_ η _

Therapeutic compositions

The present invention relates to methods and means to diagnose and treat leukemia.

Acute lymphoblastic leukemia (ALL) of the T cell immunophen- otype accounts for about 15% of childhood ALL. While with contemporary treatment protocols relapse free survival of T-ALL has been achieved at similar rates as for B cell precursor (BCP) ALL, it has been considered for a long time a malignancy with a dismal prognosis. In fact, this type of leukemia is still at a higher risk for treatment failure and early relapses than BCP ALL. Stem cell transplantation has increased the cure rate for some but not all subtypes of ALL and further intensification of existing treatment is unlikely to raise cure rates substantially but may only increase treatment-related mortality and morbidity (Pui CH et al. N Engl J Med. (2004) 350:1535-1548).

For more than a decade, clinical and immunological studies have addressed the elimination of leukemic cells by alloimmune effector cells. While donor lymphocyte infusions (DLI) have induced sustained remissions in patients with a relapse of chronic myeloid leukemias (CML) after stem cell transplantation, they have resulted in complete remission only in a small percentage of patients with ALL. Studies in BCP ALL have demonstrated that these leukemias are poor antigen presenting cells and induce T cell tolerance due to the absence of efficient B7 family-mediated costimulation. Thus, restoring the presentation of tumour antigens as well as the reversal of tolerance seems to be necessary to elicit a specific immune response.

Although many efforts were undertaken to develop new methods for the treatment of leukaemia, chemotherapy and bone marrow transplantation are still the therapies of choice. Chemothera- peutics used in the course of the treatment of leukemia show many side effects exerting a negative influence on the overall conditions of the patient to be treated (e.g. infertility, cataracts, secondary cancers, damages to the liver, kidneys, lungs and heart) . In many cases chemotherapy is followed by bone marrow transplantation. The major risk of bone marrow transplantation is an increased susceptibility to infection and bleeding as a result of the high-dose cancer treatment. To prevent or treat infections resulting from transplantation, antibi-

otics are administered to the patients. However, also transfusions of platelets may be given to the patient to prevent bleeding and red blood cells to treat anemia. Side effects resulting from transplantation may be nausea, vomiting, fatigue, loss of appetite, mouth sores, hair loss and skin reactions. With allogeneic transplants sometimes a complication known as graft- versus-host disease may develop. This disease occurs when white blood cells from the donor (the graft) identify cells in the patient's body (the host) as foreign and attack them. The most commonly damaged organs are the skin, liver and intestines. This complication can develop within a few weeks of the transplant or much later. To prevent this complication, the patient may receive medications that suppress the immune system such as steroids or other immunosuppressive agents.

In order to overcome these drawbacks several alternative or supplementary means and methods for leukemia therapy were developed and applied. Some of these methods involve the use of vaccines against specific leukemia antigens (see e.g. Gokbuget N and Hoelzer D (2004) Ann. Hematol. 83:201-205).

WO 03/072068 A2 discloses methods and compositions for the treatment of cancer, in particular for acute lymphoblastic leukemia .

WO 03/087315 A2 discloses methods for identifying biological targets for improving therapies like cancer therapies.

US 2001/0044103 Al relates to the diagnosis of the distinction between acute lymphoblastic leukemia and acute myeloid leukemia and prognosis of acute myeloid leukemia.

It is an object of the present invention to provide means and methods to treat patients suffering from leukemia, in particular of acute lymphoblastic leukemia, as a supplementary or exclusive therapy. Another object of the present invention is to provide a vaccination for leukemia treatment or prevention.

Therefore the present invention provides a composition for the treatment of leukemia, in particular of acute lymphoblastic leukemia, comprising antigen-presenting cells loaded with a protein or an immunogenic isoform-specific fragment thereof selected from the group consisting of HECTD1δ (SEQ ID No. 1), CX-OR- F15δ (SEQ ID No. 2), hCAP-Eδ (SEQ ID No. 3) and combinations thereof.

These proteins are novel isoforms of known genes. These iso-

forms are expressed specifically in leukemia cells, especially in lymphoblastic leukemia cells. The proteins according to the present invention are - in contrast to their known "wild type" counterpart - immunogenic and are able to induce cellular and humoral immune responses . In order to provoke a cellular immune response against these antigenic proteins and fragments therof and against these leukemia cells and consequently to reduce the number of said cells or even to eliminate said cells from an individual, antigen presenting cells loaded with HECTD1δ, CX-OR- F15δ, hCAP-Eδ or isoform-specific fragments thereof are also provided with the present invention.

The isoforms according to the present invention have been identified by a specific screening approach using SEREX-techno- logy and Jurkat cell lines. In this specific approach, 13 different antigens were detected that are differentially expressed in primary T-cell leukemias. However, only the three isoforms (HECTD1δ (SEQ ID No. 1), CX-ORF15δ (SEQ ID No. 2) and hCAP-Eδ (SEQ ID No. 3)) were shown to induce a leukemia specific humoral and/or cellular immune response. These proteins therefore turned out to be valuable for the treatment of leukemia, because these isoforms are specifically expressed in leukemia cells and patients, preferably children up to the age of 18 suffering from leukemia are able to produce antibodies against said proteins.

The term "fragments", as used herein, relates to parts of said proteins containing less amino acids than the protein from which they are derived from. "Fragments" relate also to polypeptides (having 11 to 100 amino acid residues) and peptides (having 2 to 10 amino acid residues) being part of a protein. The fragments according to the present invention may comprise at least 4, 5, 6, 7, 8, 9, 10, 15, 20, 30, 50 or 100 consecutive amino acids of HECTD1δ, CX-ORF15δ and hCAP-Eδ. All fragments according to the present invention comprise at least the portion specific for the isoform (e.g. for HECTD1δ the deletion of E1749 (i.e. direct peptide connection between Q1748 and E1750 of the HECTD1δ) , alternative splicing resulting in the deletion of 'F35- L166 for CX-ORF15δ and usage of an alternative exon comprising a stop codon leading to the insertion of K754-A760 for hCAP-Eδ) ("isoform-specific") . Preferred fragments are designed as T cell antigens and therefore have a length of 6 to 20, preferably 7 to 15, more preferred of 8 to 12, especially 9 to 11, amino acids.

Longer fragments with at least 20, 30 or 40 amino acids up to the whole length of the isoforms may be used for other purposes e.g. for eliciting B cell response.

The term "immunogen" has the well known meaning in the art and refers to a molecule that can elicit an adaptive immune response upon injection or delivery by another mode into a human being or animal. Such molecules may be peptides, polypeptides as well as proteins. Hence, an "immunogenic fragment" is a fragment (protein, polypeptide or peptide) that is an immunogen. It is noted that an immunogenic peptide may preferably comprise at least 6, preferably at least 7, more preferably at least 8, (8-12 AA in the MHCI context and 14-22 AA in the MHCII context) amino acids in order to be recognized and bound to a T cell receptor (antibody) and to elicit a T-cell mediated immune response. Protein fragments comprising 8-12, preferably 9-11, amino acids are suitably employed in the MHCI and protein fragments comprising 14-22, preferably 15-20, amino acids in the MHCII context.

As used herein, "antigen presenting cells" comprise naturally occurring as well as artificial antigen presenting cells and acellular antigen presenting systems. A typical representative of naturally occurring antigen presenting cells are dendritic cells. Artificial cells can be for instance insect cells, mouse fibroblasts or K32 cells (see e.g. Kim JV et al. (2004) Nature Biotechnol. 22:403-410). Since the stimulation of T lymphocytes is mediated through the T-cell receptor, which recognizes a specific major histocompatibility complex/peptide- polypeptide complex, and co-stimulation it is important that these cells express also co-stimulatory molecules which are required in this process (Kim JV et al . (2004)). If acellular antigen presenting systems (e.g. magnetic beads, liposomes) are used, the co-stimulatory molecules required for T lymphocyte stimulation have to be immobilised on the surface of said cells (Kim JV et al. (2004) ) .

However, the antigen presenting cells according to the present invention are preferably dendritic cells.

The main requirements for generating a tumor specific immune response are the adequate presentation and recognition of tumor antigens by the immune system. Especially dendritic cells turned out to be suited to fulfil this task. Dendritic cells are considered to be the most potent antigen presenting cells involved

in host defence. All immunogenic and tolerogenic responses are initiated upon the recognition of antigen presented by mature dendritic cells, playing an important role in the immune response e.g. in cancer. Dendritic cells can be generated in vitro, for instance, from peripheral blood monocytes or from CD34+ haematopoietic precursor cells in culture medium containing human granulocyte macro-phage-colony stimulating factor (GM- CSF), IL-4 (e.g. WO 01/39600, WO 01/09288, WO 02/34887, WO 2004/24900, WO 2005/012509), or some other cytokines such as IL-13 (US Patent 6399372; Salcedo et al., 2005, DOI 10.1007/s00262-005-0078-6, Cancer Immunol. Immunother. Dec. 2005, p. 1-11) .

Immature dendritic cells are characterized by high endocytic activity and low T-cell activation potential. They constantly sample the surroundings for pathogens such as viruses and bacteria. This is done through pattern recognition receptors such as the toll-like receptors. Latter receptors recognize specifically chemical structures of, for instance, bacterial pathogens like lipopolysaccharides . Once they have come into contact with such a pathogen or chemical structure, they become activated into mature dendritic cells. Therefore, in order to obtain dendritic cells loaded with an antigen said cells are exposed to HECTD1δ, CX-ORF15δ, hCAP-Eδ or isoform-specific fragments thereof, RNA or DNA molecules coding for said proteins, viral and non-viral expression vectors, cellular lysates containing said proteins (reviewed in Felzmann et al . 2002, Onkologie 25: 456) according to the present invention and to, e.g., a lipopolysac- charide. The dendritic cells phagocytose said proteins and fragments and degrade them into small pieces and upon maturation present those fragments at their cell surface using major histocompatibility complex molecules. Simultaneously, they upregulate cell-surface receptors that act as co-receptors in T-cell activation such as CD80 and CD86, greatly enhancing their ability to activate T-cells. They also upregulate CCR7 , a chemotactic receptor that induces the dendritic cell to travel through the blood stream to the spleen or through the lymphatic system to a lymph node. Here they act as antigen presenting cells and activate helper T-cells and killer T-cells by presenting them with the leukemia specific antigens derived from (or being) HECTD1δ, CX-ORFl5δ and hCAP-Eδ.

Therefore the dendritic cells are preferably activated. According to another preferred embodiment of the present invention the antigen presenting cells are pulsed with said protein or isoform-specific fragments thereof or modified with a polynucleotide, preferably RNA or DNA, encoding said protein or fragment thereof.

The antigen presenting cells, in particular the dendritic cells, may be exposed to the HECTD1δ, CX-ORF15δ, hCAP-Eδ or iso- form-specific fragments thereof according to the present invention or to polynucleotides encoding said proteins and fragments. Especially the introduction of mRNA into antigen presenting cells is a valuable alternative (see. e.g. Jarnjak-Jankovic S et al. (2005) BMC Cancer 5:20).

The fragment or fragments of the proteins HECTD1δ, CX-ORF15δ and hCAP-Eδ loaded on the antigen presenting cells have to comprise those regions of the protein which enables the cells to present an antigen to the immune system and hence to provoke the formation (priming) of HECTD1δ-, CX-ORF15δ- and hCAP-Eδ-specific antibodies (effector cells) . Therefore the fragment of HECTDlδ (SEQ ID No. 1) comprises preferably amino acids 1388 to 2611 (more preferably amino acids 1707 to 2611, amino acids 1726 to 1774 or amino acids 1747 to 1769) of SEQ ID No. 1.

According to a preferred embodiment of the present invention the fragment of (full-size) CX-ORF15δ (SEQ ID No. 2) comprises amino acids Ml to D396 of SEQ ID No. 2.

According to another preferred embodiment of the present invention the fragment (full-size) of hCAP-Eδ (SEQ ID No. 3) comprises amino acids Ml to A760, of SEQ ID No. 3.

According to a preferred embodiment of the present invention the composition comprises further the proteins or fragments thereof alone or in combination with immune stimulatory agents (adjuvants) such as antigen presenting cells including dendritic cells, immune stimulatory microbial products including membrane components such as LPS, proteins such as toxins, DNA such as synthetic CpG oligodeoxynucleotides or double stranded RNA molecules, other molecules such as alum or peptide and synthetic polycations, or combinations thereof, or adjuvants such as Fre- und' s adjuvans or microbes such as BCG. (Singh et al . Nat. Bio- technol. (1999) 17:1075-81)

Another aspect of the present invention relates to the use

of a composition according to the present invention for the manufacture of a medicament for the treatment of patients suffering from leukemia, in particular of acute lymphoblastic leukemia.

The composition according to the present invention may be used for the manufacture of medicaments like vaccines. Since said composition is able to provoke an immune response against leukemia cells in an individual, the medicament may be administered to treat and/or prevent the formation of leukemia. The medicament may be applied by subcutaneous or intradermal injection, intravenous infusion, injection into lymph nodes, into muscle, by oral application or combinations thereof.

The medicament is applied preferably administered prior bone marrow transplantation. Depending on the route of administration or the adjuvant the dose may range from pg to mg amounts applied weekly or monthly or on a continuous basis or combinations thereof .

According to the present invention the medicament comprising a composition as outlined above may be employed at various stages of leukemia treatment. For instance, it may be administered instead of chemotherapeutic substances, in combination with such substances and prior or after bone marrow transplantation. However, in order to avoid bone marrow transplantation and hence to reduce further risks leading to irreversible defects in the patient it is preferably used prior said transplantation, preferably after the initial round of chemotherapy.

Another aspect of the present invention relates to a protein comprising the sequence SEQ ID No. 1 (HECTD1δ) , protein comprising the sequence SEQ ID No. 2 (CX-ORF15δ) , protein comprising the sequence SEQ ID No. 3 (hCAP-Eδ) and nucleic acid molecules coding for said proteins comprising the sequence SEQ ID No. 4 (HECTD1δ) , nucleic acid molecule comprising the sequence SEQ ID No. 2 (CX-ORF15δ) and nucleic acid molecule comprising the sequence SEQ ID No. 3 (hCAP-Eδ) .

The aforementioned proteins and nucleic acid molecules may be used in various applications. For instance, antibodies directed to the said proteins or isoform-specific fragments thereof may be manufactured which in turn are used as part of vaccines against leukemia or of diagnostic kits for the diagnosis of leukemia. The nucleic acid molecules can serve as templates for the recombinant production of the corresponding proteins or

fragments thereof. Additionally said nucleic acid molecules may also serve as template for the construction of siRNA.

The term "fragments", as used herein, relates to parts of said proteins and nucleic acid molecules containing less amino acids and nucleotides, respectively, than the protein from which they are derived from. The fragments according to the present invention may comprise at least 4, 5, 6, 7, 8, 9, 10, 15, 20, 30, 50 or 100 consecutive amino acids or nucleotides in the case of siRNA of HECTD1δ, CX-ORF15δ and hCAP-Eδ. All fragments according to the present invention comprise at least the portion specific for the isoform (e.g. entire isoform specific cDNA sequence) ("isoform-specific") .

Another aspect of the present invention relates to the use of HECTD1δ (SEQ ID No. 1), CX-ORF15δ (SEQ ID No. 2), hCAP-Eδ (SEQ ID No. 3) or isoform-specific fragments thereof as described above for the manufacture of a vaccine for the treatment or prevention of leukemia, in particular of acute lymphoblastic leukemia.

It turned out that patients suffering from leukemia, in particular from acute lymphoblastic leukemia, produce HECTD1δ-, CX- ORF15δ- and hCAP-Eδ-specific antibodies. Therefore, isoform-spe- cific fragments of HECTD1δ, CX-ORF15δ and hCAP-Eδ are preferably used for the manufacture of a vaccine for the treatment or prevention of leukemia. The vaccine according to the present invention may be used for active or passive immunisation of an individual, especially an indivdual having or being at risk of developing leukemia, preferably ALL, especially leukemia or ALL which has been proven to be characterised by the presence of HECTD1δ, CX-ORF15δ and/or hCAP-Eδ. Active immunisation will require the administration of HECTDlδ, CX-ORF15δ, hCAP-Eδ or isoform-specific fragments thereof (optionally with an appropriate adjuvant, e.g. KLH, alum) to an individual, whereas in the course of a passive immunisation antibodies binding to said proteins or fragments are administered which will be typically applied by intravenous infusion but also by injection into skin or muscle at a wide dose range.

Yet another aspect of the present invention relates to an antibody directed to HECTDlδ (SEQ ID No. 1), CX-ORF15δ (SEQ ID No. 2), hCAP-Eδ (SEQ ID No. 3) or fragments thereof.

Antibodies directed to said proteins or fragments thereof

may foe employed to diagnose, prevent or treat all kind of tumors, in particular leukemia tumors, expressing HECTD1δ, CX-OR- F15δ and hCAP-Eδ . These antibodies preferably bind specifically to said proteins or fragments thereof and substantially not the corresponding isoforms (HECTDl, CX-ORF15 and hCAP-E) .

As used herein, the term "antibody" refers to amino acid containing molecules which are capable to bind specifically to a target molecule. Hence, said term refers to antibodies as well as to fragments thereof, also to parts of an antibody which are capable to bind to the target molecule (e.g. variable region of an antibody) . The antibody according to the present invention may be monoclonal or polyclonal or of recombinant origin. Of course also multispecific (e.g. bispecific) , humanized and chimeric antibodies fall under this definition. The methods for producing such antibodies are well known in the art.

Preferred fragments of HECTD1δ (SEQ ID No. 1) comprise amino acids 1388 to 2611, amino acids 1707 to 2611, amino acids 1726 to 1774 or amino acids 1747 to 1769 of SEQ ID No. 1.

According to a preferred embodiment of the present invention a fragment of CX-ORF15δ (SEQ ID No. 2) comprises amino acids 1 to 396, of SEQ ID No. 2.

According to another preferred embodiment of the present invention a fragment of hCAP-Eδ (SEQ ID No. 3) comprises amino acids 1 to 760, of SEQ ID No. 3.

Another aspect of the present invention relates to a method for monitoring the progress in leukemia therapy, preferably in acute lymphoblastic leukemia therapy, comprising the steps:

- providing a sample of a patient suffering from leukemia,

- determining an amount of HECTD1δ (SEQ ID No. 1, SEQ ID No. 4), CX-ORF15δ (SEQ ID No. 2, SEQ ID No. 5) and/or hCAP-Eδ (SEQ ID No. 3, SEQ ID No. 6) in said sample,

- comparing said amount with an amount of HECTD1δ (SEQ ID No. 1, SEQ ID No. 4), CX-ORF15δ (SEQ ID No. 2, SEQ ID No. 5) and/or hCAP-Eδ (SEQ ID No. 3, SEQ ID No. 6) determined prior to the therapy.

Since leukemia cells specifically express HECTD1δ, CX-ORF15δ and hCAP-Eδ these proteins represent marker proteins whose detection and quantification may help to monitor the progress of a leukemia therapy and to change or to adapt the treatment strategy. For instance, the decrease of a marker protein in the

samples provided in the course of the treatment indicates that the therapy shows positive effects, because the amount of leukemia cells in an individual could be reduced.

The sample is preferably blood or bone marrow.

According to a preferred embodiment of the present invention the amount of the proteins or fragments thereof is determined by an immunochemical method or by a nucleic acid detection method.

Antibodies directed to HECTD1δ, CX-ORF15δ, hCAP-Eδ or fragments thereof may be used to detect and quantify these proteins in a sample. Since the mRNA levels correlate directly with the amount of proteins present in such a sample, also nucleic acid detection methods can be used to quantify indirectly the proteins of interest.

The immunochemical method is preferably selected from the group consisting of enzyme-linked immunosorbent assay (ELISA) , radioimmuno assay (RIA), Dot Blot and Western Blot.

All known methods requiring antigen specific antibodies may be used according to the present invention.

Although methods like Northern Blot analysis involving chemically, enzymatically or radioactively probes may be used to detect specifically the mRNA levels in a sample the nucleic acid detection method is preferably a polymerase chain reaction method, more preferably a reverse transcriptase polymerase chain reaction method, in particular a quantitative reverse transcriptase polymerase chain reaction method. The polymerase chain reaction may be performed in situ or by extraction of the mRNA from the sample followed by the amplification reaction.

The nucleic acid detection method is preferably performed with a primer or primer pair derived from SEQ ID No. 4, when the amount of HECTD1δ is determined, from SEQ ID No. 5, when the amount of CX-ORF15δ is determined, and from SEQ ID No. 6, when the amount of hCAP-Eδ is determined.

As used herein, "derived from" means that the primers of the present invention are directly deduced from the nucleotide sequences SEQ ID No. 4 to 6 or their complementary strands. The primers are selected as such to comprise or to include characteristic regions of HECTD1δ, CX-ORF15δ and hCAP-Eδ which allows to distinguish them from their isoforms HECTDl, CX-ORF15 and hCAP-E.

A "primer pair" according to the present invention comprises

at least one forward and at least one reverse primer. The primer pair is used in methods which require the presence of two primers . The primers disclosed herein may also be used in methods which require a minimum of one primer (e.g. Northern Blot) .

The primer or primer pair to determine the amount of HECTDlδ is preferably 5'-GGG AGC AGG AAG AAG AGT ACG-3 ' (SEQ ID No. 7) and/or 5'-CAA GAG CTC TGA ATG AGG GGT-3 ' (SEQ ID No. 8) .

According to a preferred embodiment of the present invention the primer or primer pair to determine the amount of CX-0RF15δ is 5'-GCG GCA GAA GCT GGA GGA GA-3' (SEQ ID No. 9) and/or 5'-GTT TTC CTG TCG GAG TTT GGC G-3' (SEQ ID No. 10) .

The primer or primer pair to determine the amount of hCAP-Eδ is preferably 5'-GGA CAT TGA GTG GAG GTG CTC GAT C-3' (SEQ ID No. 11) and/or 5'-GCA TGG ACT GTT TTC ATC TTA CCA ATG G-3' (SEQ ID No. 12) .

Another aspect of the present invention relates to a method for diagnosing leukemia, preferably acute lymphoblastic leukemia, especially a leukemia or ALL being characterised by the presence of HECTDl, CX-ORF15 and/or hCAP-E, in a patient:

- providing a sample of a patient,

- determining the presence of HECTDlδ (SEQ ID No. 1), CX-OR- F15δ (SEQ ID No. 2), hCAP-Eδ (SEQ ID No. 3) or fragments thereof or antibodies directed to said proteins in said sample and

- diagnosing leukemia (or a specific form thereof) if HECTDlδ (SEQ ID No. 1), CX-0RF15δ (SEQ ID No. 2), hCAP-Eδ (SEQ ID No. 3) or fragments thereof or antibodies directed to said proteins are determined in said sample.

For the diagnosis of leukaemia, the proteins or peptide fragments or antibodies against these molecules to detect the antigen or the antibody immune response against them, respectively, or DNA or RNA coding for the proteins or fragments of the proteins, may be used in micro arrays (commonly called DNA chips but also protein chips) alone or in combination.

Due to the specific expression of the isoforms of HECTDl, CX-ORF15 and hCAP-E in leukemia cells, these proteins may also be used as markers for diagnosing leukemia, in particular lymphoblastic leukemia, more precisely acute lymphoblastic leukemia.

When diagnosing leukemia it is not necessary to quantify the amount of the marker proteins present in a sample. The simple presence of the proteins or fragments thereof indicates that the

individual from whom the sample was obtained is suffering from leukemia.

The sample is preferably blood or a fraction thereof, preferably plasma or serum.

According to a preferred embodiment of the present invention the amount of the proteins or fragments thereof is determined by an immunochemical method or by a nucleic acid detection method.

The immunochemical method is preferably selected from the group consisting of enzyme-linked immunosorbent assay (ELISA) , radioimmuno assay (RIA) , Dot Blot, micro arrays and Western Blot.

According to another preferred embodiment of the present invention the nucleic acid detection method is a polymerase chain reaction method, preferably a reverse transcriptase polymerase chain reaction method, in particular a quantitative reverse transcriptase polymerase chain reaction method.

The nucleic acid detection method is preferably performed with a primer or primer pair derived from SEQ ID No. 4, when the amount of HECTD1δ is determined, from SEQ ID No. 5, when the amount of CX-ORF15δ is determined, and from SEQ ID No. 6, when the amount of hCAP-Eδ is determined.

According to a preferred embodiment of the present invention the primer or primer pair to determine the amount of HECTDlδ is 5'-GGG AGC AGG AAG AAG AGT ACG-3 ¹ (SEQ ID No. 7) and/or 5'-CAA GAG CTC TGA ATG AGG GGT-3 ' (SEQ ID No. 8) .

According to another preferred embodiment of the present invention the primer or primer pair to determine the amount of CX- ORF15δ is 5'-GCG GCA GAA GCT GGA GGA GA-3 ¹ (SEQ ID No. 9) and/or 5'-GTT TTC CTG TCG GAG TTT GGC G-3' (SEQ ID No. 10).

The primer or primer pair to determine the amount of hCAP-Eδ is preferably 5'-GGA CAT TGA GTG GAG GTG CTC GAT C-3' (SEQ ID No. 11) and/or 5'-GCA TGG ACT GTT TTC ATC TTA CCA ATG G-3' (SEQ ID No. 12) .

Yet another aspect of the present invention relates to a kit comprising:

- HECTDlδ (SEQ ID No. 1, SEQ ID No. 4), CX-ORF15δ (SEQ ID No. 2, SEQ ID No. 5), hCAP-Eδ (SEQ ID No. 3, SEQ ID No. 6) or isoform specific fragments thereof, and

- antigen presenting cells, preferably dendritic cells. Such a kit may be used, for instance, to manufacture a com-

position and a medicament according to the present invention.

Another aspect of the present invention relates to a composition for inhibiting the growth of tumor cells, in particular of leukemia cells, comprising an effective amount of an antisense polynucleotide specifically hybridizing with HECTDlδ (SEQ ID No. 4), CX-ORF15δ (SEQ ID No. 5), hCAP-Eδ (SEQ ID No. 6) or isoform- specific fragments thereof and inhibiting the expression of HECTDlδ, CX-ORF15δ and hCAP-Eδ.

The knowledge of leukemia specific isoforms allows also the identification of antisense oligonucleotides which, when admin ^¬ istered to a patient, are able to inhibit the expression of a distinct protein or enzyme and consequently influence the growth of leukemia cells. Such influence may result in the reduction of leukemia cells in a patient. The oligonucleotides have to be isoform-specific and hence comprise those regions of HECTDlδ, CX-ORF15δ and hCAP-Eδ, which allows the specific distinction from HECTDl, CX-ORF15 and hCAP-E.

Given the known sequence of HECTDlδ, CX-ORF15δ and hCAP-Eδ, specific inhibitors of expression may be rationally designed. Most commonly, these inhibitors will be relatively small RNA or DNA molecules because they can be designed to be highly specific. In general, so-called "antisense" molecules will have a sequence which is complementary to a portion of HECTDlδ, CX-ORF15δ and hCAP-Eδ DNA or mRNA.

As indicated, the antisense molecules can have a variety of chemical constitutions, so long as they retain the ability specifically to bind at the indicated control elements. Thus, especially preferred molecules are oligo-DNA, RNA and protein nucleic acids (PNAs) . The oligonucleotides of the present invention can be based, for example, upon ribonucleotide or deoxyribonuc- leotide monomers linked by phosphodiester bonds, or by analogues linked by methyl phosphonate, phosphorothioate, or other bonds.

Phosphodiester-linked oligonucleotides are particularly susceptible to the action of nucleases in serum or inside cells, and therefore in a preferred embodiment the oligonucleotides of the present invention are phosphorothioate or methyl phosphon- ate-linked analogues, which have been shown to be nuclease-res- istant. See Stein et al . , Phosphorothioate Oligodeoxynucleotide Analogues in "Oligodeoxynucleotides-Antisense Inhibitors of Gene Expression" Cohen, Ed. McMillan Press, London (1988) .

"Hybridising", as used herein, is the ability of primers consisting of nucleic acids to bind specifically to a nucleic acid sequence under stringent conditions. Stringent conditions are sequence-dependent and will be different in different circumstances. Longer sequences hybridise specifically at higher temperatures. An extensive guide to the hybridisation of nucleic acids is found in Tijssen, Techniques in Biochemistry and Molecular Biology-Hybridisation with Nucleic Probes, "Overview of principles of hybridisation and the strategy of nucleic acid assays" (1993). Generally, stringent conditions are selected to be about 5 to 10°C lower than the thermal melting point (Tm) for the specific sequence at a defined ionic strength and pH. The Tm is the temperature (under defined ionic strength, pH, and nucleic acid concentration) at which 50% of the probes complementary to the target hybridise to the target sequence at equilibrium (as the target sequences are present in excess, at Tm, 50% of the probes are occupied at equilibrium) . Stringent conditions will be those in which the salt concentration is less than about 1.0 M sodium ion, typically about 0.01 to 1.0 M sodium ion concentration (or other salts) at pH 7.0 to 8.3 and the temperature is at least about 30 ⁰ C for short probes (e.g. 10 to 50 nucleotides) and at least about 60 ⁰ C for long probes (e.g. greater than 50 nucleotides) .

Said antisense polynucleotide is preferably a cDNA, and/or siRNA molecule.

According to a preferred embodiment of the present invention the antisense polynucleotide is encapsulated in a liposome.

Antisense molecules can be delivered in a variety of ways . They may be synthesized and delivered as a typical pharmaceutical. A preferred formulation involves encapsulation or association with cationic liposomes. They may be modified with a targeting sequence, and optionally linked to a polyamine, such as polylysine (Bachmann et al . , J. MoI. Med. (1998) 76:126-32). Alternatively, antisense molecules may be delivered using gene therapy methods. Using gene therapy vectors, single or multiple tandem copies of antisense molecules can be used. Administration of an antisense oligonucleotide to a patient can be effected orally or by subcutaneous, intramuscular, intraperitoneal or intravenous injection. Compositions according to the present invention, however, are advantageously administered in the form of

injectable compositions. A typical composition for such purpose comprises a pharmaceutically acceptable solvent or diluent and other suitable, physiologic compounds.

The composition according to the present invention comprises further a pharmaceutically acceptable carrier.

The polynucleotide is preferably 8 to 50, preferably 10 to 40, more preferably 12 to 30, in particular 14 to 25, nucleotides in length.

According to a preferred embodiment of the present invention the polynucleotide is S'-TTCGTACTCTTCTTCCTGCTCCCCAATAGG -3', when the expression of HECTDlδ is inhibited, 5 ' -TCCT- GCTCTCCTCCAGCTTCTGCCGCGGGCTG-3' , when the expression of CX-OR- F15δ is inhibited, and 5'-TTAAGCATGGACTGTTTTCATCTTACCAATG- GTTTTTTTAAGGGCATCT-3 ' , if the expression of hCAP-Eδ is inhibited.

Another aspect of the present invention relates to the use of a composition according to the present invention for the manufacture of a medicament for the treatment of patients suffering from leukemia, in particular of acute lymphoblastic leukemia.

The present invention is further illustrated in the following figures and examples without being restricted thereto.

Fig. 1 shows exon usage of SEREX clones and novel isoforms isolated from the Jurkat cDNA expression library. A: HECTDl, B: CX-ORF15, C: MPP-I and D: hCAP-E gene loci. Exons are depicted in white rectangles, alternative exons and different boundaries of exons in black rectangles. Newly identified exons are indicated by a δ. Complete cds mRNA of clones 28C, 13C and IB is indicated by a closed bar, the 5' incomplete mRNA of clones 24A, 10D, 8C and HA by an open bar. Illustrated clones represent the longest of homologous clones. Arrowheads indicate the primer annealing site for clone-specific RT-PCR.

Fig. 2 shows mRNA expression profile of the novel isoforms in T-cell leukemias and normal tissues.

RT-PCR expression patterns are shown for CX-ORF15δ, HECTDlδ, MPP-1δ and hCAPEδ. Samples are indicated at the top of each lane and include 16 primary leukemic samples, 5 T-ALL cell lines, normal hematopoieitic cells and other normal tissues. cDNA quality control is shown for G3PDH. Note that in leukemias with low control gene amplification (T-ALL 11, 12 and 15) novel isoforms were highly expressed. *, not tested.

Fig. 3 shows reactivity of novel isoforms with plasma from

T-ALL patients. ELISA sample identity is shown at the bottom of the Fig. : patient samples are indicated by black bars, controls (Cl-13) by white bars; na, not analyzed. The reactivity of each plasma sample derived from two independent experiments was calculated as the ratio: OD value test antigen/ OD value β- galactosidase control antigen. For each antigen the threshold level was calculated as the mean+2SD from healthy control samples with an OD ratio smaller than 2. A. Reactivity of full size CX-0RF15δ (clone 13C), mean+2SD = 1.56. B. Reactivity of full size hCAP-Eδ (clone IB), mean+2SD = 1.82. C. Reactivity of E1747-K1769 peptide of HECTD1δ (clone 24A), mean+2SD = 1.42. D. Reactivity of amino acid S559-D617 reflecting the COOH-terminus of MPP-1δ (clone 8C), mean+2SD = 2.02.

Fig. 4 shows INF-γ production of HECTD1δ specific T cells. Data from one out of three representative experiments using PBM- NCs from the former T-ALL patient are shown. Cells were stimulated twice with autologous ThI polarizing DCs, charged with recombinant (rec) , synthetic (synth) HECTDlδ or HECTDl proteins, as indicated on the X-axis. After restimulation of the T cell cultures using the respective DCs or β-galactosidase (β-gal) DCs as control, specific reactivity was measured by INF-γ production of CD4+ (top) and CD8+ (bottom) T cells.

Fig. 5 shows HECTDlδ protein expression in T-ALL. Western blotting was performed using T-ALL cell lines, primary leukemic samples and hematopoieitic cells as indicated at the top of each lane. The HECTDlδ (290 kD) was detected with an anti-HECTDlδ antibody exclusively in leukemic cells (top) . Protein loading was controlled with an anti-tubulin antibody (bottom) .

Fig. 6 shows in vitro ubiquitination assay using HECTDlδ E3 ligase. A: Thioester formation was performed in reactions containing Ubal El and UbcHδb E2 ligase with and without purified His-tagged HECTDlδ (lanes 1, 3 and 2, 4, respectively). Reactions were quenched with SDS-PAGE loading buffer under reducing conditions (+βME, lanes 1 and 2) and non-reducing conditions (-βME, lanes 3 and 4) . Ubiquitinated Ubal~Ub and UbcH5b~Ub was detected by Myc-tagged ubiquitin with an anti-Myc antibody (lanes 3 and 4, bottom), ubiquitinated His-HECTDlδ~Ub (lane 3, top) and His-HETCDIδ after dissolving the ubiquitin thioester linkage (lane 1, top) was detected with an anti-HECTDlδ antibody. The difference in protein size between lane 1 and 3 indie-

ates thioester complexes of ubiquitin with HECTDlδ, UbcH5b and Ubal. B: The specificity of the anti-HECTDlδ antibody was tested by inhibition with HECTDlδ peptide. The protein ladder below the His-HECTDIδ (arrow) represents degradation products of the purified protein. C: Schematic presentation of the proposed full size HECTDlδ. Position of the enzymatic HECT domain, the PEST domain with the isoform specific glutamic acid deletion and the two putative protein interaction domains (mib/herc2 and NH2-ter- minal ankyrin repeats) indicated as black boxes. HECTDlδ SEREX clone: isolated in this study; recombinant and synthetic HETCD1δ proteins: for ubiquitin thioester formation and cellular immune reactions; HECTDlδ peptide: for antibody generation and HECTDlδ specific ELISA.

Fig. 7 shows a protein alignment of HECTDl (SEQ ID No. 13) with HECTDlδ.

Fig. 8 shoes a protein alignment of CX-ORF15 (SEQ ID No. 14) with CX-ORF15δ.

Fig. 9 shows a protein alignment of hCAP-E (SEQ ID No. 15) with hCAP-Eδ.

Fig. 10 shows a cDNA alignment of HECTDl (SEQ ID No. 16) with HECTDlδ.

Fig. 11 shows a cDNA alignment of CX-ORF15 (SEQ ID No. 17) with CX-ORF15δ.

Fig. 12 shows a cDNA alignment of hCAP-E (SEQ ID No. 18) with hCAP-Eδ.

EXlUyIPLES :

Patient and control samples:

Bone marrow samples were obtained from 18 children (median age at diagnosis 10.7 years, range 1.5 - 16.5) with T-ALL during routine diagnostic procedures. In addition, peripheral blood samples from 13 age-matched healthy individuals without infections or chronic disease were collected. Blood samples from healthy donors were obtained from three adult males. Thymus was obtained from children, aged 1 - 4 years, undergoing cardiac surgery. Table 1 includes the clinical data of the patients as well as the differentiation stage of T-ALLs assessed by flow cytometric analysis according to EGIL standards (Bene MC et al. Leukemia (1995) 9:1783-1786).

Table 1- Clinical data from patients with T-ALL.

T-ALL _Seχ Age at Imniuiio- _CR

Patient ID ^{"" '} diagnosis plieiiotype

I ^* M 14.3 immature 43

M 5.0 immature 40

3 ^* M 2.4 immature 40

4' M 8.S immature 35

5 ^* M 11.9 immature 30

6 M 3.3 immature 25

7 F 15.3 intermediate 4S

S F 13.8 intermediate SRel ^T

9 F 6.4 intermediate 42

10 M 10.7 intermediate 39

11 F 13.5 intermediate 33

12 M 16.5 intermediate 30

13 M 5.3 mature 22

14 F 11.6 mature 47

15 F 10.S mature 33

16 M 8.2 mamre 46

17 M 1.5 mature 33

18 F 93 mature SRei ^{T * Indicates patients from whom plasma was used for library screening: CR. complete remission in month: τ These patients relapsed both 8 months after attaining first CR.}

Plasma and bone marrow mononuclear cells (MNCs) containing more than 96% leukemic cells from patients as well as plasma and peripheral blood MNC from controls were isolated by Ficoll- Hypaque (Amersham Bioscience, Buckinghamshire, UK) density- gradient centrifugation. Plasma was further sterile-filtered and stored at -20 ⁰ C; MNCs were vitally frozen and stored in liquid nitrogen until use.,

Example 1: Immunoscreeninq - Isolation of immunoreactive clones from the Jurkat cell line by SEREX

A Uni-ZAP XR custom cDNA expression library prepared from the Jurkat cell line (Stratagene, USA) was screened with a plasma pool from four patients that were reactive with T-ALL cell lines by Western blotting. The library consisted of 2 x 10 ⁵ primary recombinants with an insert size larger than 400 bp and was amplified to 6 x 10 ⁹ plaque forming units. E. coli transfec- ted with a total of 1.5 x 10 ⁶ recombinant Uni-ZAP XR phages were plated on 30 agar plates containing 30 mM Isopropyl beta-Dith- iogalactopyranoside (IPTG) for induction of protein expression.

After emergence of visible plaques nitrocellulose membranes were placed on the plates and cultured for 4h. The plasma pool used for the screening procedure was 1:10 diluted and incubated for 5h at room temperature with filters that were precoated with proteins derived from E. coli/phage lysates in order to remove antibodies reactive to vector-related antigens. The preabsorbed plasma was diluted 1:400 and used for an overnight incubation of the nitrocellulose membranes at 4 ⁰ C. Plasma-IgG, bound to recombinant proteins expressed in the lytic plaques, was detected by a 1 hour incubation with alkaline phosphatase conjugated goat anti-human IgG antibody (Sigma, USA), 1:1000 and visualized with 5-bromo-4-chloro-3-indolylphosphate/nitro blue tetrazolium (Sigma) .

In this example the reactivity of plasma-IgG from children with TALL to the allogeneic leukemic T-cell lines Jurkat, HSB-2, MOLT-3, Loucy and Peer by Western blot analysis was tested. Since plasma samples reacted predominantly with Jurkat cell derived proteins, a cDNA expression library derived from the Jurkat cell line for the screening with a plasma pool from four highly reactive children was selected (T-ALLl, T-ALL3, T-ALL4 and T-ALL5) . After three rounds of library screening 18 positive clones with high homology to previously described genes were isolated (Table 3). They encoded 13 different proteins. Strikingly, while most of the genes (MPP-I, hCAP-E, LGALS-9, OP18, PP4R1, TNKS-2, ZNF-292 and FTH) were detected only once, CX-OR- F15 and HECTDl were represented by 8 and 3 clones, respectively. MPP-I, hCAP-E and OP18 encode proteins involved in chromosome organization and microtubule dynamics during mitosis; TNKS-2, PP4R1 and LGALS-9 gene products bias cell senescence and apop- tosis. LGALS-9 is an eosinophil chemo-attractant produced by activated T-cells and serves also as a transmembrane urate transporter. The function of CX-ORF15 is unknown and ZNF292 and HECTDl are assumed to be a DNA interaction protein and an E3 Ub ligase based on its zinc finger and HECT-domains, respectively. With the exception of LGALS-9 all proteins are located exclusively in the cytoplasm.

In order to identify the proteins according to the present invention SEREX, the serological screening of cDNA expression libraries from tumors with sera from cancer patients has been used. SEREX is a valuable tool to demonstrate that patients may

mount a potent humoral immune response to a wide variety of antigens, including mutational antigens, splice variant antigens and with regard to their expression pattern cancer-testis antigens as well as differentiation and amplified antigens (Preuss KD et al. (2002) 188:43-50). Even though these leukemia-associated antigens (LAA) were identified in the context of a humoral immune response by serum-IgG from patients, many of such SEREX- defined antigens were found to be the target of a cellular immune response as well. For instance, the targets of an anti-leukemia immune reaction in chronic myelogenous leukemia (CML) and Chronic Lymphocytic Leukemia (CLL) patients were recently characterized by a SEREX-based analysis.

Example 2: Sequence analysis of reactive clones Positive clones were subcloned for isolation and excised to pBluescript phagemids with an ExAssist helper phage system (Stratagene) . Sequencing reactions of cDNA inserts were performed using an ABI Prism 3100 Genetic Analyzer (Applied Biosys- tems, USA) . Sequences were further analyzed for homology with known genes using public databases (BLAT, BLAST, SEREX) , exon usage (AceView/NCBI database) and functional domains (PESTfind Analysis) . New isoforms of genes were submitted to gene bank data bases and have the accession numbers as listed in Table 2.

Table 2. Identity of alternatively spliced genes, accession numbers, primers and conditions for RT-PCR.

Antigen. Accession No. Forward prirner/ Reverse pπrner TA [ ⁰ C]Z t TE [ ⁵ C]/ 1 eye!

CX-ORF15δ AY739713 ₅ , 5 _G ' G _T C _T G _ττ G _c CA _CT G _G A _T A _C G _G C^TG ₀ G _τ A _r G _r G , 61/4Os 72/ 9Os 38 _G A _GC GA _G 3 ₃ '

HECTDU AY739714 5' GGGAGCAGGAAGAAGAGT ACG3' 68/ 30s 72/ 60s 34 5' CAAGAGCTCTGAATGAGGGGT3'

MPP-U AY7397I5 5' CACACCAGTGAC AGTTAAGATTCC 3 ¹ 59/ 30s 72 SOs 38 5' CTTTCTCCCAGTTGA CTGCATC 3' hCAP-Eλ AY739716 5' GGACATTGAGTGGAGGTG CTC GAT C 3' 64' 40s 72' 90s 38 5' GCATGGACTGTTTTC ATC TTACCAATGG 3' house-keepmg gene G3PDH 5' TGAAGG TCG GAGTCAACG GAT TTGGT 3 ¹ TA=TE 68 / 180s 34 5' CAT GTG GGC CAT GAG GTC CAC CAC 3'

Sequence analysis of clones 24A, the shorter clone of 13D, 13C, 8C and IB revealed novel splice variants from the HECTDl, CX-ORF15, MPP-I and hCAP-E genes, respectively (Fig. 1, Table 2 and 3) .

Table 3. Identified clones and homology to published sequences.

Homology to pnblished sequences in GeuBaiik Database

No. of _size

Clone* Gene Accession Chromosomal Description Clones ¹ <bp> No. Location

2SC S ¹ 4343 CX-ORFl 5/ LSR-S f AK002071 Xp22.22 putative nuclear protein

13C 1 3945 CX-ORFl 5/ LSR-5 § na Xp22.22 unknown isofomi of CX-ORF15

BD 3 3693 HECTDI/ KLiAlIiI na 14ql2 unknown isoforiu of a potential E3 ubiquitin ligase

1OD 1 2635 MPP-L' MPHQSPH-I 'KRMPl § AB033337 10q23.31 mitotic kinesin-related protein

SC 1 3347 MPP-I/ MPHOSPH-I/ KRMPl § na 10q23.31 unknown isoforni of MPP-I

HA 1 3S41 hCAP-E/ SMC3L1 AK0014S5 9q31.1 structural maintenance of chromosome protein

IB 1 2474 hCAP-E/ SMC2L1 na 9q31.1 unknown isoform of hCAP-E

12D 1 1533 LGλLS-9 § AB006782 17qll.2 β-galactostde binding protein

2A 1 914 OPlSfSTMNl Zl 1566 Ip36.11 niicrorubule-destabilizing pho spho-proteiii

ID 1 268O PP4R1 § AF111106 lSpll.22 subunitof a serine/ threonine phosphatase

1OB 1 3S21 TNKS-2/ TλNK-2 AF43S201 10q23.2 telomere-associated (ADP-ribose) polymerase

HB 1 5739 ZNF-292 ABOl 1102 6ql4.3 zinc finger protein

23A 1 895 FTHl BC015946 ^" Ilql2.3 iron transporter protein

*, Longest of homologous clones; \ Total number of homologous clones in this study; ~. Sequence of four clones was too small to differentiate between and NGO-Br-62, the two isofomis of CX-ORF15: na, not available; -'These genes have homology to the SEREX database clones NGO-Br-30. NGO-St-12. Hom-HDl-21. respectively.

In the novel isoform of HECTDl, designated HECTD1δ, the 5' splice site of the commonly used exon 62 is shifted 3 nucleotides downstream (Fig. IA) . This leads to the deletion of one glutamic acid (E1749) located in a putative PEST domain and thereby decreasing the PEST score from +11.5 to +7.7. Such a change is assumed to result in a higher stability of the novel isoform.

The novel isoform CX-ORF-15δ lacks exon 2 and 3 compared to the published sequences of CX-ORF15 (Fig. IB) . MPP-1δ uses the alternative exon δ with an in frame stop codon leading to the COOH-terminal truncation of the protein (Fig. 1C) . hCAP-Eδ uses the commonly involved exon 36, but has retained an intronic region of 26 nucleotides downstream to exon 36, which results also in the COOH-terminal truncation of the protein (Fig. ID) isolated from the Jurkat cDNA expression library.

Example 3: RT-PCR and detection of mRNA expression of novel isoforms in leukemic T-cells and normal tissue mRNA expression of newly identified isoforms of CX-ORF15, hCAP-E, HECTDl and MPP-I was evaluated in T-ALL cell lines, in primary leukemic cells and in different normal tissues using normalized standard tissue panels (Multiple tissue cDNA panel I, II and human blood fractions cDNA panel, Clontech, USA) . Total RNA was extracted from cell lines and primary T-ALL cells using

RNA preparation kit (Qiagen, Germany) . Two μg of each sample were reverse transcribed with random hexamers and M-MLV reverse transcriptase (Promega, USA) . For PCR-amplification of alternatively spliced genes isoform-specific primers were designed. Sequences of primers and PCR conditions are listed in Table 2. G3PDH was used as a control gene amplification. The PCR reaction mix contained an equivalent of 0.1 μg total RNA, 2.5 mM MgCl ₂ and 0.02 U/μl GoldTaq polymerase (Applied Biosystems) . Amplification was performed in a GenAmp PCR System 9700 thermal cycler (Applied Biosystems) with a 1 min denaturation at 95 ⁰ C at the beginning and a 30 s denaturation at 95 ⁰ C for each cycle. PCR products were analyzed on 2% agarose gels.

The expression of the novel isoforms CX-ORF15δ, HECTD1δ, MPP-1δ and hCAP-Eδ by RT-PCR in leukemic T-cells from 16 patients, T-ALL cell lines (Jurkat, Molt-3, HSB-2, Loucy and Peer) as well as in normal tissues was analysed. In contrast to the ubiquitous expression of the known isoforms of CX-ORF15, MPP-I and hCAP-E (derived from the EST data base) , the novel isoforms CX-ORF15δ, HECTD1δ, MPP-1δ and hCAPEδ demonstrated a restricted expression pattern (Fig. 2).

CX-ORF15δ and HECTD1δ showed high expression in all T-ALL cell lines and in 75% (12/16) and 69% (11/16) of primary T-ALL samples, respectively. CX-ORF15δ was also weakly expressed in resting hematopoietic cells, and HECTD1δ was present in all resting hematopoietic T and B cells. MPP-1δ was strongly expressed in all T-ALL cell lines and primary leukemic samples, but also in hematopoietic cells, spleen and thymus. There was also weak expression of MPP-1δ in lung, liver, pancreas, prostate, testis and ovary. High expression of hCAP-Eδ mRNA was found in all T-ALL cell lines and in 63% (10/16) of primary T- ALL samples.

Example 4: Presence of antibodies against the novel isoforms in T-ALL patients and healthy controls:

Bacterially expressed NH2-terminally His-tagged fusion pro- ^• teins of newly identified isoforms were purified. The expression plasmids were constructed by PCR-based subcloning of the 24A, 13C, 8C and IB phagemids into vector pET100/D-TOPO (Invitrogen, USA) . Sequencing of inserts confirmed the identity with the original clone. Recombinant proteins were then purified using Pro- Bond columns (Invitrogen) . Correct size and purity of the fusion

proteins were confirmed by Coomassie staining of SDS-PAGE and Western blotting using a HRP conjugated anti-His-tagged antibody (Invitrogen) . Recombinant Histagged beta-galactosidase was expressed and purified in the same way as described above for the new isoforms and used as control antigen.

A peptide homologous to the deduced amino acid sequence of HECTD1δ (clone 24A) (E1747-K1769) was generated by fluorenyl- methoxycarbonyl (Fmoc) synthesis for rabbit immunization and specificity testing of antibodies (Eurogentec, Belgium) . For cellular immune reactions two further 48-mer peptides (H1726- F1774, H1726-S1774) were designed spanning the spliced region of the known and new isoform (Thermo Electron, USA) . The location of these synthetic peptides is indicated in Fig. 6C. Identity and purity of the petide and proteins was determined by reverse phase high-performance liquid chromatography (HPLC) and confirmed by mass spectral analysis. The peptide and proteins were dissolved in PBS at a concentration of 3 mg/ml and stored at -70 ⁰ C until use.

ELISA plates (Nunc, Roskilde, Denmark) were coated with 50 μl of purified recombinant His-tagged fusion protein or with the peptide at a concentration of 5 μg/ml in coating buffer (PBS + 0.05% sodium azide) over night at 4°C. Plates were washed with PBST (PBS + 0.05% Tween 20) and blocked with 250 μl/well 2% BSA coating buffer over night at 4 ⁰ C. Plasma from patients and healthy individuals was added at a final dilution of 1:500 in PBS-T (50 μl/well) and incubated for 3h at room temperature. After several washing steps with PBS-T, the plates were incubated with 50 μl/well of alkaline phosphatase-conjugated goat anti-human IgG antibody (Sigma) for Ih at room temperature. Finally, IgG binding was visualized with 100 μl/well of PNPP substrate (Sigma) and measured at OD _40S using a micro plate reader (Anthos Labtec Instruments, Austria) . The OD ratio of the test antigens, based on His-tagged beta-galactosidase control antigen, was used to determine the degree of specificity above background. A sample was considered to be positive, when its OD ratio exceeded the mean value plus 2-fold standard deviation (mean+2SD) from healthy control samples with an OD ratio less than 2.

To examine whether the novel isoforms have induced leukemia- specific immune reactions, plasma from children with T-ALL and

from age-matched controls was used for antigen ELISA. Full size CX-ORF15δ and hCAP-Eδ antigens as well as the proposed antigenic region of MPP-1δ, represented by the isoform-specific exon δ (amino acid S559-D617), were purified. As an antigen for HECTD1δ a peptide, which was designed homologous to the region with the single glutamic acid deletion (Fig. 1 and 6C) , was used.

As shown in Fig. 3, a leukemia-associated immune reaction was detectable for CXORF15δ, hCAP-Eδ and HECTDlδ in two, three and six of 16 patients, respectively. No reactivity was observed in plasma samples from 13 controls against CX-ORF15δ and hCAPEδ.

However, plasma from one of 10 controls reacted with HECTDlδ. Reactivity with MPP-1δ was detected in four of 16 T-ALL patients and in five of 13 controls and therefore considered non-specific. Plasma-IgG from at least one of those 4 T-ALL patients, which were initially used for screening the Jurkat expression library, reacted with the respective antigenic isoform of the isolated SEREX clones thereby validating the specificity of antibody binding.

The specificity of the antibody binding was further confirmed by a competitive ELISA.

Example 5: Induction of an isotype-specific T cell response against HECTDl by employing T helper type 1 (ThI) polarizing dendritic cells (DC)

Peripheral blood mononuclear cells (PBMNCs) from healthy individuals were collected and frozen in liquid nitrogen. For the generation of monocyte derived ThI polarizing DCs from PBMNCs the previously established protocol was followed (Felzmann T et al. (2005) 54:769-780). Briefly, 10 ⁶ DCs were exposed to 5 μg of recombinant His-tagged HECTDlδ or beta-galactosidase protein or one of the two synthetic isoform specific HECTDl proteins (H1726-F1774, H1726-S1774) for 2 hours at 37°C. The DC culture was then supplemented to a final concentration of 200 U/ml LPS (Calbiochem, Germany) , 50 ng/ml INF-γ (Boehringer-Ingelheim, Austria), 400 U/ml IL-4, 1000 U/ml GM-CSF and 2% human plasma and further cultured for 6 hours at 37 ⁰ C.

Screening for HECTDlδ specific immunoreactive T cells PBMNCs were stimulated every week and restimulated thereafter (day 14) for 12 hours with DCs in a 5:1 ratio. Four hours later, CD3 ⁺ 4 ⁺ as well as CD3 ⁺ 8 ⁺ cells were analyzed for their INF-γ production by flow cytometry according to the BD Bioscience standard protocol

for intracellular protein staining and evaluated on a FACS ARIA (BD Bioscience, USA) . The following directly conjugated antibodies were used: anti-CD3-PE (DAKO, Denmark), anti-CD45~FITC (DAKO), anti-CD4-PE-Cy7 (BD Bioscience), anti-CD8-PerCP (BD Bioscience) and anti-INF-γ-APC (BD Pharmingen, USA) .

Since the specific antibodies in the patients' plasma were directed against cytoplasmic proteins they are unable to target the antigen expressing cells. Exemplarily for all LAA' s, the most attractive antigen for in vitro studies, the HECTD1δ protein based on its ability to induce humoral immune response in about a third of children with T-ALL, was used to provoke a cellular immune response. For this purpose two 48-mer synthetic and one recombinant protein, covering the spliced region of HECTDl, were used for loading ThI polarizing DCs. In addition, recombinant beta-galactosidase charged DCs were used as control DCs. After three rounds of stimulation, T-cells from three healthy individuals were tested for reactivity by INF-γ production. As exemplified for one donor, the former patient (no 12), a specific response of both CD4+ and CD8+ T-cells could be induced. Fig. 4 represents data from one of three independent experiments that have revealed similar results . While an immune response was raised against the novel isoform HECTDlδ that was used as a recombinant as well as a synthetic protein, no reactivity was induced upon stimulation with the known HECTDl isoform presented by a synthetic protein. All three healthy individuals reacted in the same manner. These data show that naturally processed epitopes of the synthetic protein as well as of the much larger recombinant HECTDlδ protein induce a helper as well as a cytotoxic T-cell reactivity. Moreover, it also demonstrates that the novel splice site had created immunogenic epitope (s).

Example 6: Presence of the novel isoform of HECTDl, a new E3 liqase, in T-ALL

An anti-HECTDl antibody was generated by rabbit immunization with peptide E1747-K1769 and its specificity confirmed by ELISA. The antibody was directly used for Western blotting at a 1:1000 dilution. Cultured T-ALL cell lines, vitally frozen primary T- ALL samples, fresh MNCs from peripheral blood as well as thymocytes were lysed in RIPA buffer and separated by SDS-polyacrylamide electrophoresis. HECTDlδ expression was detected with the rabbit

anti-HECTDlδ antibody and visualized with HRP-conjugated goat anti-rabbit IgG antibody 1:10000 (Biorad Laboratories, USA). As loading control tubulin was detected with mouse anti-tubulin antibody (Calbiochem, Germany) and visualized with HRP-conjugated goat anti-mouse IgG antibody 1:10000 (Pierce, USA).

Thioester-Ubiquitination was induced under physiological conditions (50 mM Tris/HCl pH 7.5, 150 mM KCl, 0.1 itiM DTT, 0.1% Tween 20, 2 mM ATP, and 10 mM MgCl ₂ ) with 7 μg purified His- tagged HECTD1δ, 4 μg His-tagged UbcH5b, 0.1 μg Myctagged ubi- quitin and 1.6 μg Myc-tagged Ubal enzyme. A control assay without HECTD1δ protein was also performed. The 2x reactions of 60 μl were split after 10 min incubation at 37 ⁰ C for reducing and non-reducing conditions. For reducing conditions 15 μl 5x SDS loading buffer containing β-mercaptoethanol (βME) was added to a 30 μl reaction, for thioester-conserving, non-reducing conditions 15 μl 5x SDS-loading buffer without βME was added per 30 μl reaction. Samples with reducing conditions were further incubated at 95 ⁰ C for 5 min, those with non-reducing conditions at 37 ⁰ C for 5 min and subsequently analyzed by SDS polyacrylamide electrophoresis and Western blotting using an anti-Myc 9E10 antibody, 1:5000 (Abeam limited, UK) and anti-HECTDlδ, 1:1000 (Eurogentec) . Proteins were visualized with HRP-conjugated goat anti-mouse IgG antibody and HRP-conjugated goat anti-rabbit IgG antibody, 1:2000 (Biorad).

Finally, after having demonstrated specific immunogenicity of the novel HECTDl isoform also its proposed function was assessed. While the HECTD1δ protein was expressed at its predicted size of 290 kD at high abundance in all T-ALLs, it was not detectable in normal hematopoietic cells by Western blotting (Fig. 5) . These data not only endorse the leukaemia restricted presence of this protein, but also correlate well with mRNA expression and potentially also with the proposed increased stability due to alternative splicing (Fig. 2) .

Finally, it was assessed whether HECTD1δ possesses the proposed E3 ligase activity by its ability to form thioester complexes with ubiquitin in the presence of the Ubal El and UbcH5b E2 (Fig. 6A) . The purified COOH-terminus of HECTD1δ (M1707- N2611, Fig. 5C) , comprising the HECT-domain, accepted ubiquitin from UbcH5b, which was in turn activated by Ubal. Binding of ubiquitin to HECTD1δ as well as to Ubal and UbcH5b was revers-

ible under reducing conditions, thereby demonstrating that the process of transferring ubiquitin chains by the E1-E2-E3 system is mediated via thioester formation with cysteine residues . The specificity of this reaction was further confirmed by inhibiting the antibody binding by the HECTD1δ peptide (Fig. 6B) .

Discussion :

In these examples it was demonstrated that children with T- ALLs develop a spontaneous immune response against LAA. Thirteen antigens of known genes including four novel isoforms of HECTDl, CX-ORF-15, MPP-I and hCAP-E were isolated by a SEREX-based approach. Three of these isoforms exhibited a restricted mRNA expression in T-ALLs and had induced a specific humoral immune response in a subgroup of children. The mounted antibodies, however, cannot exert their specific function since the target proteins are exclusively intracytoplasmic. The potential to induce a T cell response was therefore evaluated in vitro for one of the novel isoforms, the HECTD1δ. A specific reactivity of autologous CD4+ and CD8+ T cells from three healthy individuals was induced after priming against naturally processed recombinant or synthetic HECTD1δ, but not against the known splice variant. These data suggest to further evaluate the potential of LAA for immunotherpeutic strategies in children with T-ALL that might be particularly required for those children with a treatment resistant recurrence of the disease.

Most of the identified proteins in this study were intracytoplasmic, a subcellular localization that is common to antigens defined by SEREX (Sahin U et al . PNAS (1995) 92:11810- 11813) . Therefore, the specific antibodies in the patients' plasma cannot target the leukemic cells. Hence, the ability of these LAAs to induce a specific T cell reaction was evaluated, which appeared as a likely scenario since a helper T-cell response is considered to be needed to induce humoral immune responses. Even though T and B lymphocytes react against the same protein, they target different epitopes and it also explains why the SEREX approach serves as an antigen screening method not only for the humoral but also the cellular immune system. Further evidence for a B-T cell collaboration in the context of tumor immunology is provided by a murine model that demonstrates the necessity of both B and T cells for T cell-mediated tumor rejection.

To exemplify the potential to induce a specific T cell reaction HECTD1δ was selected and a recombinant as well as a synthetic protein was used, spanning the differentially spliced region, for loading ThI polarizing autologous DC from three healthy donors. Among them was a former patient (no 12), who was in complete continuous remission and already six months off chemotherapy when he donated blood for the in vitro experiments. In all three individuals an antigen specific reaction of CD4+ and CD8+ T cells could be demonstrated against the novel isoform by INF-γ production. Thus, here a link between the proposed B and T cell interaction by the detection of specific antibodies produced in the patients and a CD4+ T cell stimulation in vitro by one of the leukemia associated antigenic isoforms was provided.

By the SEREX approach according to the present invention 13 antigens were isolated with open reading frames of known gene loci. Some were recognized here for the first time, i.e. HECTDl, hCAP-E and ZNF-292, while CX-ORF15, MPP-I, PP4R1, LGALS-9, and TNKS-2 have been recently detected also in other malignancies including Hodgkin' s disease and adult T-ALL (SEREX database of the Ludwig Institute for Cancer Research) . Some of theses proteins, LGALS-9 (galectin-9) and TNKS-2 (tankyrase-2) , as well as STMNl (stathmin) have an already confirmed or proposed etiological association with cancer as deduced from their cancer-restricted expression pattern in Hodgkin' s disease, lymphomas and leukemic T-cells, respectively. These proteins have functions that are associated with the mitotic apparatus and chromosome condensation as well as with transcriptional control, DNA replication, post-transcriptional regulation and apoptosis.

In addition to the already described isoforms, four novel splice variants of T-ALL associated antigens were detected, a finding reported also by others for hematopoietic malignancies and carcinomas . A further focus was drawn on the newly identified isoforms as they are often specific for the respective malignancy. The four novel splice variants were highly expressed at the mRNA level in more than 60% of T-ALL and - with the exception of MPP-1δ - only weakly in some subgroups of hematopoietic cells. While CX-ORF15δ and hCAP-Eδ were recognized only by the patients' plasma, MPP-1δ specific antibodies were detected also in plasma from healthy controls, which is in concordance with

its ubiquitous expression. An autoimmune reaction against MPP-I, the known isoform, has been reported recently in patients with autoimmune diseases. HECTD1δ reactivity was observed in about 30% of children with T-ALL, but also with one control sample. In the latter case the reactivity could have resulted from a clinically silent viral infection or inflammation, conditions that have the potential to lead also to immune reactions with autoantigens . Of note, each of the four novel splice variants from this study suggests alterations of protein function. The specific impact of CX-ORF15δ on the leukemic phenotype of T-ALL is unclear, as is the function of CXORF15 in general. As the truncated hCAP-Eδ lacks the functional SMC-C domain it may lead to disturbed mitotic chromosome condensation.

HECTDl proteins contain 10 putative PEST domains, the signals for protein degradation. Hence, it can be assumed that this protein has a high turn over. In fact, a single glutamic acid deletion, which was identified in one of the PEST domains of the novel HECTD1δ isoform, leads to a decreased PEST score thereby suggesting higher protein stability. This notion is supported by a recently described block in GCN4 protein degradation due to the exchange of a single amino acid in a PEST domain of this transcription factor.

The proposed function of HECTD1δ as an E3 ubiquitin ligase was confirmed by an in vitro ubiquitin ligase assay in this study. It further demonstrated that the COOH-terminus comprising the HECT domain is both necessary and sufficient for the ligase activity. This adds the HECTD1δ to other E3 ligases with a cancer association in different types of malignancies and thus to a potentially interesting target for alternative treatment options.

Substrate specificity of HECTD1δ is presumably defined by the four NH ₂ -terminal ankyrin repeats and a mib/herc2 domain potentially involving HECTDl proteins in Notch activation and oncogenic pathways.