CD44 BINDING PEPTIDES

Abstract

The present invention relates to a protein which binds to the domain encoded by exon 9 of human CD44 (CD44ex9), to fusion proteins and conjugates of said protein and especially to nanoparticles conjugated to said protein. The invention further relates to a method of production for the protein and the respective conjugated nanoparticles and the use of the protein of the invention for treatment and diagnosis of cancer diseases.

Claims

1-24. (canceled)

25. A conjugate comprising: (a) protein binding to a polypeptide encoded by exon 9 of human CD44 (CD44ex9), said protein comprises or consists of (i) an amino acid sequence according SEQ ID No. 1,2,3, SEQ ID No. 39 to 50 and SEQ ID No. 52; (ii) an amino acid sequence with at least 90% identity to the amino acid sequence given in SEQ ID No. 1 to 5, SEQ ID No. 39 to 50 and SEQ ID No. 52; or (iii) an amino acid sequence with at least 97.5% identity to the amino acid sequence given in SEQ ID No. 4, and wherein said protein has a length of 100 amino acids or less, and (b) at least one of a heterologous protein, a heterologous polypeptide, or a compound conjugated, fused, or linked to the CD44ex9-binding protein.

26. The conjugate of claim 25, wherein the CD44ex9-binding protein binds to a CD44 isoform comprising the domain encoded by exon 9.

27. The conjugate of claim 26, wherein the CD44 isoform is selected from the list consisting of: a. CD44 v5, b. CD44 v5 v6, c. CD44 v3 v6, d. CD44 v3 v6, e. CD44 v2 v10, f. CD44 v3 v10, g. CD44 V4 v7, h. CD44 v4 v10,

28. The conjugate of claim 25, wherein the heterologous protein or heterologous polypeptide is an enzyme which catalyses the generation of a cytotoxic or cytostatic agent from a precursor molecule or attachment of a label to the protein.

29. The conjugate of claim 25, wherein the compound is selected from the group consisting of a nanoparticle, carbohydrate, a dye molecule, a radioactive isotope, a toxin, a cytostatic agent, a cytokine, a immunomodulatory agent, or a prodrug thereof.

30. The conjugate of claim 29, wherein the compound is linked directly or via a linker molecule to the CD44ex9-binding protein.

31. The conjugate of claim 25, wherein a nanoparticle is conjugated to the CD44ex9-binding protein.

32. The conjugate of claim 25, wherein a nanoparticle is conjugated to a conjugate or fusion protein of the CD44ex9-binding protein.

33. The conjugate of claim 31, wherein the nanoparticle is conjugated to at least one tumor antigen-binding substance and/or cytotoxic agent.

34. The conjugate of claim 33, wherein the nanoparticle the tumor antigen-binding substance is selected from the list consisting of an antibody against CEA, an antibody against CA-19-9, an adhesin, which is preferably modified, and combinations thereof.

35. The conjugate of claim 31, wherein the nanoparticle is selected from the group consisting of quantum dots, noble metal clusters, superparamagnetic iron oxide nanoparticles (IONPs), block-copolymer micelles, nanocells, dendrimers, nanotubes, polymersomes, XPclad nanoparticles, nanoparticles consisting of amorphous silica surrounded by a crystalline luminescent calcium phosphate layer, SiO.sub.2/Zn.sub.2SiO.sub.4:Mn.sup.2+ and SiO.sub.2/Ca.sub.10(PO.sub.4).sub.6OH:Eu.sup.3+ core-shell nanoparticles with diameters below 100 nm and luminescent dye labeled hybrid nanoparticles.

36. The conjugate of claim 31, wherein the nanoparticle has an inorganic core and a coating layer comprising an imidazole component, with the smallest diameter of the inorganic core including the coating layer being no more than 15 nm.

37. The conjugate of claim 36, wherein the imidazole component is a peptide comprising at least one histidyl residue, and preferably a dipeptide.

38. The conjugate of claim 36, wherein the imidazole component is a mixture of different histidyl-containing dipeptides.

39. The conjugate of claim 31, wherein the nanoparticle is non-inert.

40. A pharmaceutical composition comprising the conjugate according to claim 25 and a pharmaceutically acceptable excipient.

41. A medicament/dosimeter combination package comprising: a) a medicament to be individually dosed, and b) a diagnostic indicator system for a patient-specific property that is relevant for the action, side effect, interaction, metabolism, absorption, distribution, metabolism, and elimination of the medicament to be administered to a patient, wherein the patient-specific property is selected from the group consisting of an endogenous substance, a regulation mechanism, a gene or an indication system, and wherein the medicament or the diagnostic indicator system comprises the conjugate of claim 25.

Description

FIGURE LEGENDS

[0247] FIG. 1: Genomic structure and protein domains of CD44.

[0248] FIG. 2: (A) Schematic outline of the strategy of the yeast-two-hybrid screening. (B) Vector used for the Y2H-screen.

[0249] FIG. 3: Amino acid sequences of the positive CD44-interacting clones together with the consensus sequence derived from it.

[0250] FIG. 4: Labeling of H129 cells with the nanoparticle conjugate Q_CA19-9-3.

[0251] FIG. 5: Negative control for labeling of HT29 cells with the nanoparticle conjugate Q_CA19-9-5.

[0252] FIG. 6: Quantification of I-IT29 cell-labeling with the nanoparticle conjugate Q_CA19-9-3 or the negative control MBP-nanoparticle.

[0253] FIG. 7: Competition of HT29 cell-labeling with the nanoparticle conjugate Q_CA19-9-3 by addition of increasing amounts of the free target CA19-9.

[0254] FIG. 8: Labeling of Colo29 cells with the nanoparticle conjugate Q_CA19-9-3.

[0255] FIG. 9: Negative control for labeling of Colo29 cells with the nanoparticle conjugate Q_CA19-9-5.

[0256] FIG. 10: Labeling of SW116 cells with the nanoparticle conjugate Q_CA19-9-3.

[0257] FIG. 11: Negative control for labeling of SW116 cells with the nanoparticle conjugate Q_CA19-9-5.

[0258] FIG. 12: Labelling of mucinous adenocarcinoma tissue (sample No. 35) with a bispecific nanoparticle Q_CEA/CA19-9_12 conjugated to an anti-CEA and an anti-CA-19-9 antibody. In the four quadrants, the following conditions were analysed: [0259] Upper left: Labeling with Q_CEA/CA19-9_12 [0260] Upper right: Labeling with CEA/CA19-9_12 by addition of the free target MBP-CEA (25 M) [0261] Lower left: Labeling with Q_CEA/CA19-9_12 by addition of the free target Sialyl Lewis (250 M) [0262] Lower right: Labeling with Q_CE/CA19-9_12 by addition of the free targets MBP-CEA (25 M) and Sialyl Lewis.sub.a (250 M)

[0263] FIG. 13: Affinity ranking experiments for peptides A to E: Results of the Pseudohitpicking assay I

[0264] FIG. 14: Affinity ranking experiments for peptides A to E: Results of the FDG assay I

[0265] FIG. 15: Affinity ranking experiments for peptides A to E: Results of the Pseudohitpicking assay II

[0266] FIG. 16: Affinity ranking experiments for peptides A to E: Results of the FDG assay II

[0267] FIG. 17: Overall summary of the Affinity ranking experiments for peptides A to E

[0268] FIG. 18: Diagram showing the deletion mutants for the peptides A, C, D and E

[0269] FIG. 19: Diagram showing the deletion mutants for the peptide B (left side) and the results of their affinity analysis in the FDG assay (right side)

[0270] FIG. 20: Table showing the alanine scanning mutants for the peptide B and the results of their affinity analysis in PHP and FDG assays, both performed in twice.

[0271] FIG. 21: Diagram showing the alternative CD44 constructs that were generated for the binding analyses.

[0272] FIG. 22: Diagram giving an overview on the in vitro cytochemical experiments performed with CD44v5-binding peptides A, B or C coupled to fluorescent nanoparticles.

EXAMPLES

Example 1: Isolation of CD44ex9-Binding Proteins

[0273] For the isolation of proteins or protein fragments binding to the v5 domain of CD44 a competitive two hybrid screening was performed according to the method disclosed in EP 1 721 974 A1. As bait, the v5-fragment was fused the GAL4 binding domain. The respective ORF was cloned into a plasmid vector and designated as pGBKT7 (see FIG. 2). A human cDNA library with fragments fused to the GAL4-activation domain was used as a prey (vector pGADT7-RecAB; see FIG. 2).

[0274] The screening was performed under stringent selection conditions and resulted in the identification of 39 positive clones. After selective co-transformation of the bait and the pray plasmid, 21 of the 39 clones remained positive. The 21 clones were sequenced and analysed by BLAST search. As result the 21 sequences could be reduced to 12 different sequences since four sequences were identified twofold and one sequence was present in four clones.

[0275] Furthermore, a sequence comparison revealed that these five sequence families show extensive sequence homologies allowing the formation of a consensus sequence (see FIG. 3).

Example 2: In Vitro Detection of Antigen-Expressing Tumor Cells

2.1 Background and Objective

[0276] In order to demonstrate the ability of the protein-quantum dots conjugates to specifically detect antigen-expressing tumor cells, a tumor specimen or in vitro cultivated tumor cells were analysed by fluorescence and in parallel by immunohistochemistry for expression of the respective antigen. As a first antigen, the carcinoembryonic antigen (CEA) was analysed. CEA is a glycoprotein involved in cell adhesion. As a detection-ligand, the anti CEA antibody was conjugated to the quantum dot QDBP-655 (charge #773780). Furthermore, the cancer antigen 19-9 (CA19-9) was analysed. CA19-9 is a sialylated Lewis (a) antigen and represents a tumor marker that is used primarily in the management of pancreatic cancer.

2.2. Samples

2.2.1 Cell Culture Samples

[0277] In a first set of experiments, the in vitro cultivated human cancer cell lines HT29, Colo25 and SW116 were used. HT29 is an adherent colorectal adenocarcinoma cell line, Colo25 a human colon carcinoma cell line and SW116 a colorectal cancer cell line.

2.2.2. Tumor Sample

[0278] In a separate set of experiments, a sample taken from a human tumor was used. The tumor sample No. 35 was resected in 2011 from coecum and was diagnosed as a mucinous adenocarcinoma. It was classified according the TNM classification as pT4No(0/18)G3Ro. This sample which was not otherwise pretreated was prepared and fixed as follows.

Preparation and Fixation of Tissue Sections

[0279] The resected tumor sample, stored at 70 C. was used to prepare cryosections with a thickness of 5 m using a HM560 MV cryostate. (Thermo Scientific, Walldorf, Germany). The sections were dried for 60 to 120 min at room temperature and incubated with cold acetone (20 C.) for 10 minutes. After a further drying step for 10 min at room temperature, the sections were stored over night at 70 C.

[0280] The sections were defrosted in a closed box for 30 min at room temperature. Fixation was performed by incubation in a solution containing 20 mg/ml dimethyl suberimidate (DMS) and in 20 mM CaCl.sub.2 in 250 mM Tris-HCl buffer pH 8.0. Afterwards the sections were washed for 5 minutes in D-PBS-T buffer containing 130 mM NaCl, 7 mM Na.sub.2HPO.sub.4, 3 mM KH.sub.2PO.sub.4, and 0.1% Tween 20 by a quenching step for 20 minutes in a buffer containing 130 mM NaCl, 7 mM Na.sub.2HPO.sub.4, 3 mM KH.sub.2PO.sub.4 and 200 mM glycerine. In a final step the sections were washed for 5 minutes in D-PBS-T buffer.

Counterstaining of the Tissue Sections.

[0281] Counterstaining was performed using the Dako Autostainer Plus (DAKO Deutschland GmbH, Hamburg, Germany) according the following protocol: [0282] Rinsing with D-PBS-T buffer [0283] Blocking for 60 min with 1% BSA/5% goat serum in D-PBS buffer [0284] Blowing off [0285] Incubation with 2100 l conjugate (20 nM in 1% BSA/5% goat serum in D-PBS buffer) for 60 min [0286] Rinsing for three times with D-PBS-T [0287] Blowing off [0288] Incubation with 2100 l primary antibody (20 nM in 1% BSA/5% goat serum in D-PBS buffer) for 60 min [0289] Rinsing for three times with D-PBS-T [0290] Blowing off [0291] Incubation with 2100 l secondary antibody (20 nM in 1% BSA 5% goat serum in D-PBS buffer) for 60 min [0292] Rinsing for three times with D-PBS-T [0293] Blowing off [0294] Incubation with 3200 l Hoechst 33342 (2 g/ml in D-PBS buffer) for 10 min [0295] Rinsing for three times with D-PBS-T [0296] Embedding in Mowiol/Triethylenediamine (DABCO)

Microscopic Evaluation

[0297] The sections as prepared above were evaluated with a reverse microscope Axiovert 200 with Axiocam HrM and a Axiocam Hrc CCD camera system, HXP 120 C illuminating device. The pictures were analysed using the Axiovision software (version 4.8; Carl Zeiss, Oberkochen, Germany).

2.3 Protein-Quantum Dot Conjugates

[0298] For antigen detection the quantum dots as listed in the following table were prepared:

TABLE-US-00008 Type of QDBP- 655 QDBP/ quantum Ligand Conjugate dot Ligand ratio Q561 #773780 SI6166 N/A Q_CA19-9-20 #773780 MBP-NS-19-9-scFv-His.sub.6 + 1:50 SI6166 Q_CEA/CA19-9_9 Q561 MBP-NS-19-9-scFv-His.sub.6 + 1:10 SI6166 Q_CEA/CA19-9_10 Q561 MBP-NS-19-9-scFv-His.sub.6 + 1:20 SI6166 Q_CEA/CA19-9_11 Q561 MBP-NS-19-9-scFv-His.sub.6 + 1:30 SI6166 Q_CEA/CA19-9_12 Q561 MBP-NS-19-9-scFv-His.sub.6 + 1:40 SI6166 Q_CEA/CA19-9_13 Q561 MBP-NS-19-9-scFv-His.sub.6 + 1:50 SI6166 Q_CEA/CA19-9_1 Q561 MBP-NS-19-9-scFv-His.sub.6 + 1:50 SI6166

[0299] The conjugate Q561 is a conjugate between the quantum dot QDBP-655 (batch #773780) and the ligand SI6166. QDBP-655 has a hydrodynamic diameter of 6 nm and possesses a Ni-NTA ligand. The ligand SI6166 is an CEA-binding protein having a hydrodynamic diameter of 2.1 nm, and the final conjugate exhibits a hydrodynamic diameter of 10.1 nm. The anti CEA protein is coupled to the quantum dots.

[0300] The quantum dots Q_CA19-9-x are conjugates between the quantum dot QDBP-655 (batch #773780) and the anti-CA 19-9 antibody MBP-NS-19-9-scFv-His6. This anti-CA19-9 antibody which is a recombinant antibody mimetic consisting of the variable region of the heavy chain (VH) and the light chain (VL) and can be described as a single chain fragment of variable regions, scFv. This scFv fragment is fused to a His-tag which is used for the coupling to the quantum dot. The anti-CA 19-9 antibody MBP-NS-19-9-scFv-His6 is also denominated as SI6950.

[0301] The quantum dots Q_CEA/CA19-9_x are bispecific conjugates, whereby the quantum dot Q561 harboring the anti CEA-antibody SI6166 is further conjugated with the anti-CA 19-9 antibody SI6950.

[0302] Different batches of the conjugate Q_CEA/CA19-9_x were prepared differing by the ratio of the quantum dot to the ligand (see last column of the table).

3. Results of CA19-9 Conjugated Nanoparticles

3.1 HT29 Cells

[0303] As shown in upper left panel of FIG. 4, the conjugate Q_CA19-9-3 labels the cell membrane of the HT29 cells. A staining of the cells with the anti-CA19-9 antibody detected by Alexa Fluor 488 fluorescence (upper right panel) shows a perfect co-localization of the signals showing the specificity of the conjugate binding. In the lower right panel the cells are shown by differential interference contrast (DIC) microscopy.

[0304] In the negative control a quantum dot conjugated to MBP (MBP-His/QDP655) was used. As shown in the upper left panel of FIG. 5, no labeling could be detected. The presence of the CA19-9 antigen was verified by staining of the cells with the anti-CA19-9 antibody detected by Alexa Fluor 488 fluorescence (upper right panel).

[0305] The specific labeling of the CA19-9 antigen was also shown by a titration experiment. A shown in FIG. 6, an increasing amount of the conjugate Q_CA19-9_3 leads to an increased labeling of the HT29 cells as demonstrated by increasing relative fluorescence. In contrast, the negative control MBP-His/QDP655 leads only to a sparse fluorescent labeling which show no major increase at higher conjugate concentrations.

[0306] The specificity of the CA19-9 labeling was further verified by a competition experiment as shown in FIG. 7: An increasing amount of free CA19-9 antigen added to the labeling mixture leads to a decreased CA19-9 labeling.

3.2 Colo29 Cells

[0307] As shown in upper left panel of FIG. 8, the conjugate Q_CA19-9-3 labels the cell membrane of the Colo29 cells. A staining of the cells with the anti-CA19-9 antibody detected by Alexa Fluor 488 fluorescence (upper right panel) shows a perfect colocalization of the signals showing the specificity of the conjugate binding. In the lower right panel the cells are shown by differential interference contrast (DIC) microscopy.

[0308] In the negative control a quantum dot conjugated to MBP (MBP-His/QDP655) was used. As shown in the upper left panel of FIG. 9, no labeling could be detected. The presence of the CA19-9 antigen was verified by staining of the cells with the anti-CA19-9 antibody detected by Alexa Fluor 488 fluorescence (upper right panel).

3.3 CW116 Cells

[0309] As shown in upper left panel of FIG. 10, the conjugate Q_CA19-9-3 labels the cell membrane of the Colo29 cells. A staining of the cells with the anti-CA19-9 antibody detected by Alexa Fluor 488 fluorescence (upper right panel) shows a perfect colocalization of the signals showing the specificity of the conjugate binding. In the lower right panel the cells are shown by differential interference contrast (DIC) microscopy.

[0310] In the negative control a quantum dot conjugated to MBP (MBP-His/QDP655) was used. As shown in the upper left panel of FIG. 11, no labeling could be detected. The presence of the CA19-9 antigen was verified by staining of the cells with the anti-CA19-9 antibody detected by Alexa Fluor 488 fluorescence (upper right panel).

4. Results of Bispecific Conjugated Nanoparticles

[0311] This experiment was performed in order to test the ability of the bispecific conjugated nanoparticle to interact with both antigens in a given target. For this purpose the nanoparticle conjugated with both the anti-CA19-9 antibody and the anti CEA antibody was used to detect the respective antigens on the tumor cells of above described sample No. 35. The specificity was demonstrated by inhibition with the free targets MBP-CEA and 19-9.

[0312] As shown in the upper left part of FIG. 12, the bispecific nanoparticles labeled cell membranes of the tumor tissue. A staining of the tumor tissue with an anti-CA19-9 antibody or an anti CEA-antibody showed in each case a partial colocolization with combines to the complete labeling profile of the nanoparticle,

[0313] When the bispecific nanoparticle is used in combination with the free target MBP-CEA, a labeling can be observed that shows a perfect colocalization with the staining of the anti-CA-19-9 antibody (see FIG. 12, upper right quadrant). Hence, the free target inhibits the interaction with the cellular-bound CEA but still allows a labeling of the CA19-9.

[0314] As shown in the lower left quadrant of FIG. 12, the addition of the free target Sialyl-Lewis.sub.a inhibits the interaction with the cellular CA19-9 antigen but retains the CEA-labeling.

[0315] In a final experiment the bispecific nanoparticles were combined with both free targets, the MBP-CEA and Sialyl-Lewis.sub.a (see lower right quadrant of FIG. 12). As a result the labeling was completely inhibited (upper left corner of the quadrant). The colocalized expression of both antigens was verified by staining with the respective anti-CEA and CA19-9 antibodies.

Example 3: Characterization of CD44ex9-Binding Proteins by Affinity Ranking

Introduction:

[0316] The CD44ex9-binding proteins as identified in the competitive two hybrid screening of Example 1 were further analysed by two different affinity assays to establish an affinity ranking.

Methods:

[0317] As a first assay a pseudohitpicking assay was used, which was performed as follows: In order to compare the affinity of two different peptides, colonies expressing the first peptide or the second peptide, respectively, were plated on the first or second half of an agar plate. The threshold value for defining a colony as a hit was then determined, that a percentage of colonies of the first peptide, which is defined in advance, in the reference region of the plate are picked as hits. In the same experiment, the colonies of the second peptide are scanned and picked. The number of the colonies of the second peptide are then compared to the number of colonies of the first peptide and expressed as a percentage. Based on this percentage the relative interaction strength (i.e. the affinity to the CD44 peptide) can be derived.

[0318] The FDG assay is based on the reporter system of the two hybrid scanning system. Hereby the GAL4-binding domain with the CD44v5 peptide interacts with the respective peptides fused to the GAL4-actvation domain. After complex formation two independent reporter genes are activated, the first one encodes a fluorescent protein, the second reporter gene encodes the enzyme beta-galactosidase, whose enzymatic activity can be determined by using the fluorogenic substrate Fluorescein di-beta-D-galactopyranoside (FDG).

[0319] In the pseudohitpicking assay both reporter systems will be considered: At first the fluorescence of endogen reporter gene and after incubation with a fluorescence substrate the activity of the second enzyme-encoding reporter gene. Therewith, the results of two different reporter genes can be determined within one experiment.

[0320] As shown in FIGS. 13 to 16, ten different combinations of two of the five peptides of SEQ ID No. 1 to 5 were compared in the pseudohitpicking assay and in the FDG assay, performed in parallel. Hereby, two independent experiments were conducted as shown in FIGS. 13 to 16.

Results:

[0321] Both experiments showed that peptide B (SEQ ID No. 3) has the highest affinity towards the CD44v5 peptide followed by peptide C and peptide D. The overall summary of the affinity ranking analysis is given in FIG. 17.

Example 4: Characterization of CD44ex9-Binding Proteins by Deletion

Introduction:

[0322] For the five peptides a deletion-based epitope mapping was performed in order to identify and characterize the binding sites of these peptides. The results represent an important information which can aid in the development of CD44v5-associated therapeutics and diagnostics.

Methods:

[0323] For all five peptides A to E, deletion mutants were generated and tested in the PHP and in the FOG assay. For the peptides A, C, D and E an N- or C-terminal deleted fragment was generated and denominated as peptide A1, C1, D1 and E1. The sequences of the deletion mutants for peptides A, C; D and E are depicted in FIG. 18.

[0324] For the peptide B, altogether four deletion mutants, namely the peptides B1 to B4 were generated as shown in FIG. 19.

[0325] The peptides were comparatively tested in the PHP and the FDG assay.

Results:

[0326] For the peptide B, the N-terminal deletion of 18 amino acids (peptide B1) resulted in a peptide that slightly decreased activity (see FIG. 19, right side). Hence, this N-terminal region is not crucial for the interaction to the CD44v5 peptide and therefore peptide B1 represents a core sequence for the interaction with CD44v5.

[0327] As shown in FIG. 19, peptides with N- and C-terminal deletion (peptide B4), with C-terminal deletion only (peptide B2) and with deletion of the core sequence QLSFEVQWETS (peptide 83) do completely loss to ability to bind to CD44v5. This result is in perfect agreement with the profile of peptide B1 showing that at least the core sequence together with the C-terminal part are required for CD44v5 binding.

[0328] The peptides A1 and C1 differ from their parent peptides A and C by the lack of the C-terminal 3 or 8 amino acid long sequence including the motif AIE of the consensus sequence. In both cases the binding to CD44v5 is almost completely abolished showing the relevance of this part of the consensus sequence.

[0329] The peptides D1 and E1 differ from their parent peptides D and E by the lack of a N-terminal sequence including the motif PYYGKXLXX of the consensus sequence. In both cases the binding to CD44v5 is not diminished. Whereas peptide D1 is of similar affinity, the peptide E1 show even a stronger binding towards CD44v5 compared to the parent peptide E.

[0330] These results are in perfect accordance with the results for the peptide B as described above. The N-terminal region is not necessary for CD44v5 binding and a shortened peptide and its respective consensus sequence as a core sequence represents the preferred peptide motif for CD44v5-associated diagnosis and therapy.

Example 5: Characterization of CD44ex9-Binding Proteins by Alanine Scanning

Introduction:

[0331] For the most affine peptide B an alanine scan of the core sequence was performed in order to identify those amino acids which are crucial for CD44v5 binding. As shown by the mutated peptide B3, the presence of the core sequence QLSFEVQWETS is mandatory for CD44v5 binding.

Methods:

[0332] The peptide B was used to generate altogether 11 mutants, whereby every amino acid of the core sequence QLSFEVQWETS was independently substituted by the amino acid alanine (see FIG. 20). These mutants were tested in comparison with the peptide B in the PHP and in the FDG assay. For all five peptides A to D, deletion mutants were generated and tested in the PHP and in the FDG assay.

Results:

[0333] As shown in FIG. 20, the substitution of the first 10 amino acids of the core sequence does not led to a significant change in affinity towards CD44v5. However, the exchange of the 11th amino acid serine of the underlined core sequence by alanine resulted in a peptide with higher affinity.

[0334] Hence, there is no single amino acid within the core sequence, that is crucial for CD44v5 binding but it has to be assumed that these amino acids bind in a coordinative manner to the target.

[0335] However, based on the results an even improved peptide B and an improved consensus sequence with a Ser-Ala exchange in the 11th amino acid were identified (see consensus sequences of SEQ ID No. 58 to 62).

Example 5: Binding Analysis of CD44ex9-Binding Proteins Towards Different CD44 Variants

[0336] As shown in FIG. 21, altogether four different CD44 variants were generated that contain 20 the v5 domain in combination with different domains. As revealed by PHP and FDG assays, the CD44ex9-binding proteins of the invention were able to bind also these CD44 variants. As a negative control the CD44 variant pGKT7_CD44 was used encoding only the domains 1 to 5 and 15, 16 (not shown here). As expected, the CD44ex9-binding proteins do not show any binding to this CD44 protein.

Example 5: Binding Analysis of Nanoparticles Coupled to CD44ex9-Binding Proteins Towards In Vitro Cultivated CD44v5-Expressing Tumor Cells

[0337] In order to underscore the diagnostic and therapeutic potency of the CD44ex9-binding proteins, the peptides A, B and C were expressed as MBP-fusion proteins containing a His-tag, coupled to quantum dots and used for immunocytochemical analysis of CD44-expressing tumour cells.

Methods:

[0338] The peptides were expressed as MBP-(YTH_v5_A/B/C)-His.sub.6 fusion proteins in E. coli, purified using Ni-NTA affinity chromatography and coupled via the His.sub.6-Tag to the QDBP-655 quantum dots. These quantum dots possess a CdSe core, a ZnS shell and a passivation layer comprising an imidazole compound.

[0339] The characteristics of the QDBP-655 nanoparticles as herein can be summarized as follows:

[0340] The nanoparticles have a cadmium selenide [CdSe] core and zinc sulfide [ZnS] shell. The shell is surrounded by a dipeptide coating consisting of glycine-histidine, histidine-leucine, carnosine, and amino-PEG (polyethylene glycol) cross-linked via aminobenzophenone and 3-[Tris(hydroxymethyl)phosphonio]propionate (THPP). The dipeptide coating is coordinatively bound via its imidazole rings to the zinc ions of the shell structure. Free primary amino groups are available in the dipeptide coating for coupling reactions. Additional information concerning the dipeptide coating is disclosed in the US patent application US 200310059635 A1.

[0341] QDBP-655 is sold by Life Technologies Corporation (Eugene, Oreg., USA). It is delivered as a reddish colloidal suspension in buffer. The nanoparticles exhibit the following product characteristics:

Appearance: Reddish clear liquid
Hydrodynamic radius: 6.5 nm (determined by size exclusion chromatography)

Molecular Weight: 1000 kDa

[0342] Emission Maximum: 655 nm4 nm

Coupling:

[0343] The expressed and affinity-purified peptides A, B or C are dialysed for 2 hours against 50 mM sodium borate pH=8.3 (Slide-A-Lyzer Mini Dialysis Unit 10 KDa; (Thermo Fisher, Rockford, Ill., USA). Afterwards the protein concentration is determined and the solution diluted to the required concentration. The QDBP-655 quantum dot and the protein are mixed together to an end concentration of 0.3 M (QDBP-655) in QDBP-655/ligand ratios of 1:12.5, 1:25, 1:50, 1:75 and 1:100 in sterile borosilicate vials as summarized in the following table:

TABLE-US-00009 Q_CD44v5 QDBP/L n.sub.QDBP-655 c.sub.QDBP-655 V.sub.QDBP-655 V.sub.Ligand n.sub.Ligand c.sub.Ligand _X1 1: 78 pmol 3.9 pmol 20 l 240 l _X2 .sup.1:12.5 78 pmol 3.9 pmol 20 l 240 l 975 pmol 4.06 pmol/l _X3 1:25 78 pmol 3.9 pmol 20 l 240 l 1950 pmol 8.125 pmol/l _X4 1:50 78 pmol 3.9 pmol 20 l 240 l 3990 pmol 16.25 pmol/l _X5 1:75 78 pmol 3.9 pmol 20 l 240 l 5850 pmol 24.4 pmol/l _X6 1:100 78 pmol 3.9 pmol 20 l 240 l 7800 pmol 32.5 pmol/l

[0344] The successful coupling is shown by SDS-agarose gel-electrophoresis, whereby a ratio of 1:50 or more leads to a dramatic increase in molecular weight showing the successful coupling reaction.

Immunocytochemical Analysis

[0345] The immuncytochemical analysis of the tumor cells was performed as follows: [0346] Each 510.sup.5 tumor cells on coating glasses [0347] Cultivation for 48 hours at 37 C., 5% CO.sub.2 [0348] Washing of the cells for 31 minute with 1 mL D-PBS [0349] Fixation for 20 minutes with 1 ml dimethylsuberimidate (DMS) solution (20 mg/ml DMS in 250 ml Tris-HCl, 20 mM CaCl.sub.2, pH 8.0) [0350] Washing of the cells for 11 minute with 1 mL D-PBS [0351] Quenching by incubation for 20 min in 0.2% glycine in D-PBS [0352] Blocking by incubation for 60 min in 1% BSA/5% goat serum in D-PBS [0353] Incubation with 50 l of the preincubated Qdot-peptide conjugates (100 nM) at RT [0354] Washing of the cells for 35 minute with 1 mL D-PBS [0355] Incubation with 50 l primary antibody (anti-CD44v5: VFF-7, 1:25) [0356] Washing of the cells for 35 minute with 1 mL D-PBS [0357] Incubation with 50 l secondary antibody (GAM-A488, 1:100) [0358] Washing of the cells for 35 minute with 1 mL D-PBS [0359] Conterstaining with 300 l Hoechst 33342 (200 ng/ml in D-PBS) for 10 minutes [0360] Washing of the cells for 35 minute with 1 mL D-PBS [0361] Embedding in Mowiol 4-88 (Sigma Aldrich, St. Louis, Mich., USA)

Results:

[0362] The nanoparticles coupled to peptides A, B or C were tested with the human colon carcinoma cell line HCT-116 (see FIG. 22). When using the peptide conjugated nanoparticles a labelling of the tumour cells could be observed which colocalizes with the labeling as generated by the anti-CD44v5 antibody as positive control.