WO2013117581A1

WO2013117581A1 - Metabolically stable variants of chemerin 9

Info

Publication number: WO2013117581A1
Application number: PCT/EP2013/052308
Authority: WO
Inventors: Carsten Groetzinger; Sebastian BANDHOLTZ; Bertram Wiedenmann
Original assignee: Charite - Universitätsmedizin Berlin
Priority date: 2012-02-10
Filing date: 2013-02-06
Publication date: 2013-08-15

Abstract

The invention relates to a chemerin-9 variant of the sequence Y¹ F² P³ G Q⁵ F⁶ A⁷ F⁸ S⁹, wherein a position is substituted, characterized by at least one of the following substitutions: - F² is one of the group comprised of Y, W, J, V, M, an (S)-2-aminopropionic acid substituted at carbon 3 by a C4-, C5- or C6 linear alkyl or by a C5- or C6 circular alkyl moiety, or an (S)-2-amino-3-(nitrophenyl)-propionic acid; - Q⁵ is selected from the group comprised of M, T, S, Cit and Hnv; and/or - A⁷ is selected from the group comprised of T, S, H, G, C, and w, and to methods of diagnosis and therapy of oesophageal and pancreatic cancer, wherein the chemerin-9 variant of the invention is provided attached to a detectable label or therapeutically active agent.

Description

Metabolically stable variants of chemerin 9

Description

Pancreatic and oesophageal tumours are comparatively rare but characterised by high mortality rates. One reason being that such tumours are commonly detected only at advanced stages of cancerous disease. Often the disease is detected by chance, for instance during the course of other medical examinations. The five-year survival rate is approximately 5%. A labelled ligand to a tumour specific cell surface marker (receptor) may have so called theranostic applications, utility in both disease diagnosis and treatment. Currently, no receptor-ligand systems are available for oesophageal squamous cell carcinoma and pancreatic adenocarcinoma.

Chemerin (UniProt ID Q99969) is a 14 kDa protein, highly expressed in adipose tissue, liver and lung. It is secreted as the inactive, 143 amino acid (AA) long, prochemerin and activated by proteolytic cleavage of a 6 AA C-terminal sequence by serine proteases involved in coagulation-, inflammation or fibrinolysis cascades. Chemerin is a chemoattractant with an important role in innate and adaptive immunity. It acts as a ligand for CMKLR1 (ChemR23, UniProt ID Q99788), a G protein-coupled receptor.

Wittamer et al. (J. Biol. Chem. 279, 9956-9962 (2004)) identified a nonapeptide corresponding to the C terminus of processed chemerin, ⁴⁹YFPGQFAFS¹⁵⁷ (Chemerin-9) that retains most of the activity of the full length protein. Shimamura et al. (Peptides 30(8), 2009) show a metabolically more stable analogue of the murine Chemerin-9.

Starting from the above listed state of the art, the invention presented here discloses improved means and methods for both therapeutic and diagnostic utilisation of the interaction of Chemerin with its receptor.

Terms and definitions

Amino acid sequences are given from amino to carboxyl terminus. Capital letters for sequence positions refer to L-amino acids in the one-letter code (Stryer, Biochemistry, 3^rd ed. p. 21 ). Lower case letters for amino acid sequence positions refer to the corresponding D- or (2R)-amino acids.

Citrulline (Cit) is (2S)-2-amino-5-(carbamoylamino)pentanoic acid (CAS No. 372-75-8).

4-Nitrophenylalanine (Fni) is (2S)-2-amino-3-(4-nitrophenyl)-propionic acid; (CAS No. 949-99-5) Hydroxynorvaline (Hnv) is (2S,3S)-2-amino-3-hydroxypentanoic acid (CAS No. 10148-66-0).

Hydroxyproline (Hyp) is (2S,4R)-4-hydroxypyrrolidine-2-carboxylic acid (CAS No. 51-35-4).

Tic is L-1 ,2,3,4-tetrahydroisoquinoline-3-carboxylic acid (CAS No. 74163-81-8).

J is isoleucine or leucine.

In the context of the present description, the term alkyl means saturated hydrocarbons comprising one to ten carbon atoms, which may be present in a branched, linear, or cyclic structure, or a cyclic structure with side chains. The term alkyl also encompasses partially unsaturated hydrocarbons, e.g., propenyl. Examples of alkyls are methyl, ethyl, isobutyl, pentyl, n-hexyl or cyclohexyl. The term alkyl also encompasses heteroalkyi groups. Linear or circular n-alkyls are alkyls in which n denominates the number of carbon atoms in the longest linear or circular chain. For example, C4-alkyl in the present context can be, by way of non-limiting example, an n-butyl, isopentyl or 2,2- or 2,3 dimethylbutane. C3-alkyl can be an n-propyl or isobutyl; a C5-alkyl can be an n-pentyl or a dimethylpentyl group. A circular C6-alkyl can be, by way of non-limiting example, cyclohexyl, methylcyclohexyl or

dimethylcyclohexyl.

According to a first aspect of the invention, a chemerin-9 variant of the sequence Y F² P³ G Q⁵ F⁶ A⁷ F⁸ S⁹ is provided, wherein one or more positions are substituted by exchange of an amino acid of the native sequence for a different amino acid. This chemerin-9 variant according to the invention has at least one of the following substitutions:

the position F² is taken by one of the group comprised of Y, W, J, V, M, an (S)-2-aminopropionic acid substituted at carbon 3 by a C3-, C4-, C5- or C6 linear alkyl or by a C5- or C6 circular alkyl moiety, or an (S)-2-amino-3-(nitrophenyl)-propionic acid;

Q⁵ is selected from the group comprised of M, T, S, Cit and Hnv; and/or

A⁷ is selected from the group comprised of T, S, H, G, C, and w.

Even one substitution of the alternatives provided in this first aspect of the invention provides chemerin-9 variants with remarkably increased potency (see Table 1 ).

According to one embodiment, a chemerin-9 variant comprises more than one substitution, and at least one of the following substitutions is additionally present:

- Y is f or N-acetyl-Y;

- P³ is Hyp;

- F⁶ is Y, W or w;

- A⁷ is a;

- F⁸ is Tic;

S⁹ is s or f.

According to one embodiment of this aspect of the invention, at least two of F², Q⁵ and A⁷ are substituted, for example F² is W or L, Q⁵ is M or Q and A⁷ is S, G, H or C.

According to one embodiment of this aspect of the invention, Y is N-acetyl-Y; F² is Cha or Fni; and Q⁵ is selected from the group comprised of Q, M, Cit and Hnv.

According to one embodiment of this aspect of the invention the chemerin-9 variant has a sequence set forth in Table 3.

According to one embodiment of this aspect of the invention, chemerin-9 variants are provided which have a metabolic stability, as measured in half-life in human serum, of at least five times that of native chemerin-9, or an EC₅₀ for CMKLR1 , of <50% that of native chemerin-9.

Particularly preferred chemerin-9 variants are AcY Cha P G M Y A F f (C080) and

AcY Cha Hyp G Cit F a Tic S (C091 ). These variants show a higher potency than native chemerin-9, and vastly increased half-lives. Not all applications, however, will necessitate such long half-lives. While high protease resistance as characterized by long half lives in the serum or liver extract stability assays may be favourable, other factors such as plasma protein binding, uptake by transporters, unspecific binding and aggregation will influence biodistribution, tracer kinetics and accumulation in the tumor. Variants with less than the highest protease resistance may turn out to be the better tracer molecule in the animal experiment or in the human situation. Table 3 provides a broad range of half- lives to select from for whatever value an application may require.

The compounds of the invention are useful for diagnosis and therapy of cancer, particularly in the diagnosis of tumour extension and distribution, in post-operative or post-therapeutic control of tumour removal, and in control of tumour remission. For this to be achieved, according to one aspect of the present invention, a variant of chemerin-9 as set forth above, preferably a variant capable of binding to CMKLR1 with an EC₅₀ of below 50 nmol/l is provided, wherein the chemerin-9 variant carries a detectable label.

According to observations of the present inventors, tumours originating from oesophageal squamous cell carcinoma and pancreatic adenocarcinoma show an elevated expression of CMKLR1 (>90% and >50% of samples, respectively) (see Example 6 below).

According to one embodiment of the invention, a chemerin-9 variant as set forth above is provided, which comprises or is covalently linked to a detectable label. It comprises the label -without being attached to it- if the detectable label is part of the chemerin-9 variant as set forth above, for example by having a radioisotope forming part of an amino acid structure incorporated into the chemerin-9 peptide chain. Alternatively, a radioisotope or dye molecule, by way of non-limiting examples, is attached to the chemerin-9 variant covalently or via a covalently attached chelator molecule such as DOTA (1 ,4,7, 10-tetraazacyclododecane-1 ,4,7, 10-tetraacetic acid), NOTA (1 ,4,7-triazacyclononane- Ν,Ν',Ν''-triacetic acid), HYNIC (6-Hydrazinopyridine-3-carboxylic acid) or others. A radioactive label may also be attached in the form of a nanoparticle.

According to one embodiment of this aspect of the invention, the detectable label is a radiolabel for PET (positron emission tomography) or SPECT (Single-photon emission computed tomography), such as one of the radioisotopes carbon-1 1 , nitrogen-13, oxygen-15, fluorine-18, gallium-68, technetium- 99m, indium-1 1 1 or iodine-123, iodine-124. The radiolabel may be comprised in the chemerin-9 or its variant by incorporation or covalent coupling to the peptide itself, or a carrier molecule attached thereto.

According to one embodiment of the invention, the detectable label is a near infrared fluorescent dye. One example for such dye is an indotricarbocyanine dye. Near infrared dyes have been described, inter alia, by Umezawa (J. Am. Chem. Soc. (2008) 130, 1550-1551 ) and are available, for example, from Amersham /GE Healthcare (Little Chalfont, GB), mivenion (Berlin, DE), and LI-COR Biosciences (Lincoln, Nebraska, USA).

According to one embodiment of the invention, a chemerin-9 variant as set forth above is provided, which comprises or is covalently linked to a radioisotope emitting beta or gamma radiation. A radioisotope such as carbon-1 1 for example may form part of the chemerin-9 variant backbone. Alternatively, a radioisotope is attached to the side chain of an amino acid constituting the chemerin-9 variant, such as may be, by way of non-limiting example, a tyrosine, phenylalanine or histidine having an iodine radioisotope attached to its aromatic ring.

Similarly, chemerin-9 and its variants described herein are useful for the therapy of oesophageal squamous cell carcinoma and pancreatic adenocarcinoma. According to one aspect of the present invention, a variant of chemerin-9 capable of binding to CMKLR1 with an EC₅₀ of below 50 nmol/l is provided, wherein the chemerin-9 or chemerin-9 variant carries covalently attached to it a therapeutic radioisotope or a cancer drug or toxin. Non-limiting examples for therapeutic radioisotopes are phosphorus-32, strontium-89, ytrrium-90, iodine-125 and -131 , samarium-153, erbium-169, lutetium- 177 and rhenium-186/188. Non-limiting examples for toxins and cancer drugs are ricin toxin, diphtheria toxin, anthrax toxin, pro-aerolysin, pseudomonas exotoxin, shigella toxin, cone snail neurotoxin, auristatin, doxorubicin, daunorubicin, taxol, irinotecan, vincristine, vinblastine, cisplatin, carboplatin, oxaliplatin, ifosfamideetoposide.

According to another aspect of the invention, a method for detecting the presence of a tumour cell expressing CMKLR1 is provided, wherein a peptide is covalently linked to or comprises a detectable label, said peptide being selected from the group comprising chemerin-9 and a chemerin-9 variant according to any of aspects and embodiments described herein. Said peptide is brought into contact with said tumour cell under conditions allowing for selective binding of said peptide to CMKLR1 , and binding is detected.

According to one embodiment, the tumour cell is selected from an oesophageal squamous cell carcinoma cell and pancreatic adenocarcinoma cell.

According to one embodiment, the method of the invention is practiced in-vitro and the tumour cell is comprised in a sample obtained from a patient. A diagnostic method may for example comprise the steps of

providing a histological sample for analysis obtained from a patient;

bringing into contact with said histological sample a solution of a chemerin-9 variant according to any of the above embodiments, said chemerin-9 variant further being covalently attached to a visually detectable label such as a fluorescent dye molecule;

incubating the sample under conditions that allow for the selective binding of chemerin-9 to a cell- surface-bound CMKLR1 molecule, subsequently washing the sample to remove non-bound chemerin-9, and

detecting the presence of the detectable label in the sample.

Alternatively, the visually detectable label may be an enzyme having an activity resulting in an optically detectable precipitate, which may give rise to further signal by amplification through precipitate-specific secondary signalling systems. One example is the Tyramide Signal Amplification System

commercialized by Perkin Elmer.

According to one embodiment, the method of the invention is practiced in-vivo. According to a preferred embodiment, the method of the invention is practiced in-vivo and detection of the detectable label is performed by positron emission tomography (PET), single photon emission computed tomography (SPECT) or near-infrared imaging.

According to one embodiment of this aspect of the invention, the detectable label is a radiolabel for PET or SPECT, such as one of the radioisotopes carbon-1 1 , nitrogen-13, oxygen-15, fluorine-18, gallium-68, technetium-99m, indium-1 1 1 or iodine-123, iodine-124. The radiolabel may be comprised in the chemerin-9 or its variant by incorporation or covalent coupling to the peptide itself, or a carrier molecule attached thereto.

Similarly, chemerin-9 and its variants described herein are useful for the therapy of oesophageal squamous cell carcinoma and pancreatic adenocarcinoma. According to one aspect of the present invention, a variant of chemerin-9 capable of binding to ChemR23 with an EC₅₀ of below 50 nmol/l is provided, wherein the chemerin-9 or chemerin-9 variant carries covalently attached to it a therapeutic radioisotope or a toxin. Non-limiting examples for therapeutic radioisotopes are phosphorus-32, strontium-89, yttrium-90, iodine-125 and -131 , samarium-153, erbium-169, lutetium-177 and rhenium- 186/188.

According to one embodiment of this aspect of the invention, an isolated chemerin-9 peptide or a variant of chemerin-9 capable of binding to ChemR23 with an EC₅₀ of below 50 nmol/l is provided, wherein the chemerin-9 or its variant is coupled covalently to a near-infrared fluorescent dye. This embodiment has particular utility in localisation of primary tumours and metastasis in patients with advanced stage disease.

According to yet another aspect of the invention, a chemerin-9 variant as set forth above is provided for therapy of oesophageal squamous cell carcinoma or pancreatic carcinoma, said chemerin-9 variant comprising or having covalently attached to it a radionuclide or a toxin. Framed differently, the present invention encompasses a method for treating a patient suffering from oesophageal squamous cell carcinoma or pancreatic carcinoma, said method comprising delivering to said patient a therapeutically active amount of a chemerin-9 variant according to the present invention, attached to a radionuclide or toxin.

Preferred radionuclides according to this aspect of the invention are therapeutic beta- and gamma radiation emitting therapeutic nuclides such as exemplified above, particularly yttrium-90, lutetium-177 and bismuth-213.

Wherever alternatives for single features such as positions Y F² P³ G Q⁵ F⁶ A⁷ F⁸ S⁹ or labels are laid out herein as "embodiments", it is to be understood that such alternatives may be combined freely to form discrete embodiments of the entire molecule provided as such or for use in a method or medical indication herein. Thus, any of the alternative embodiments for F² may be combined with any of the alternative embodiments of Q⁵ and these combinations may be combined with any label mentioned herein.

The following examples and figures, from which further advantages and embodiments can be drawn, illustrate the invention without limiting its scope. Brief description of the figures.

shows receptor internalisation measured by vesicle surface area in U20S cells depending on dose of chemerin-9 and two variants.

shows near-infrared imaging of tumours expressing CMRLK1 in mice. shows single substitution variants of chemerin-9. EC₅₀ values were determined by Ca2+ mobilisation in HEK293A cells.

shows multiple substitution variants of chemerin-9. Data were acquired as shown for table 1.

shows non-proteinogenic amino acid substitution variants of chemerin-9, their EC₅₀ values and their half life values.

shows the cAMP content of HEK293A-cells stimulated with different chemerin-9 variants.

shows a summary of results of competitive binding studies of ²⁵l-chemerin-9 to CMKLR1 expressing HEK293A membranes in the presence of increasing concentrations of different ligands.

shows expression of CMKLR1 in non-cancerous tissue samples,

shows expression of CMKLR1 in cancerous tissue samples.

Examples

Methods

Standard techniques are employed for cell culture and molecular biological manipulations (Sambrook et al., Molecular Cloning: A Laboratory Manual, Third Edition, Cold Spring Harbour Press; Freshney: Culture of Animal Cells, Wiley-Blackwell).

Peptides are purchased from Peptides & Elephants (Potsdam, Germany) or Bachem (Bubendorf, Switzerland).

Ca²⁺-Mobilisation

Ca²⁺-lmaging is used for functional analysis of CMKLR1 ; the heterologously expressed receptor is artificially tethered to the IP₃- signal transduction pathway using Ga₁₆. Multiple Ga, and Ga_s binding GPCRs can be tethered to the IP₃- signal transduction pathway using Ga₁₆ (Offermanns and Simon, J. Biol. Chem. 270(25): 15175-15180 (1995)).

Fluorometric detection (Cellux, Perkin Elmer) of calcium sensitive fluorescent dyes, such as Fluo-4AM, is used to measure intracellular calcium flux.

Cells are passaged, seeded in black, clear-bottomed 96 well plates (BD Falcon, Microtest, Optilux) and, if applicable, transfected. Ca²⁺-lmaging is performed 18-24 h after transfection (Cellux, Perkin Elmer). For imaging cells are loaded with calcium sensitive fluorescent dye Fluo4-AM (2 μΜ Fluo4- AM, 2.5 μΜ Probenecid in serum-free RPMI cell culture medium) by 45-60 min incubation at 37°C. Excess dye is subsequently removed by washing three times at 15 minute intervals in C1-solution (130 mmol/l NaCI, 5 mmol/l KCI, 10 mmol/l Na-HEPES, 2 mmol/l CaCI₂, 10 mmol/l Glucose). Ligands are diluted as appropriate in C1 -solution. To avoid peptides adhering to plastic surfaces, C1- solution is supplemented with 0.1-0.5% BSA, depending on peptide hydrophobicity.

Plates are quantified (Cellux, Perkin Elmer) immediately after the final wash (excitation filter: 485nm; emission filter: 535nm; integration time: 500 ms and high sensitivity detection). Ligands are applied 30 seconds after commencing measurement; total duration of the measurement is 120 seconds per plate.

For analysis of antagonistic ligands, a second measurement is made 5 minutes after the first, adding a complete agonist at a concentration equivalent to its EC₈o value.

Data are analysed using AssayPro software. Signals are scored in duplicate and compared to the appropriate controls. Maximal fluorescence scores upon ligand application (F) are normalised to the respective basic fluorescence value (F₀), with AF/F₀ = (F-F₀)/F₀.

Metabolic stability of peptides

Metabolic stability of peptides is determined using human serum or liver extract [Janssen and Aspmo 2008, Methods in Molecular Biology 494: 177-186]. RPMI is supplemented with 25% (v/v) human serum or 25% (v/v) liver extract and pre-warmed to 37°C. Peptides are added to a final concentration of 100 μg/ml, mixed, incubated at 37°C and samples taken at time 0 and other appropriate time points; at each sampling a 50 μΙ aliquot of the reaction is precipitated on ice with 100 μΙ 6% v/v trichloroacetic acid for 15 minutes. Precipitates are pelleted by centrifugation for 2 minutes at 18 000 g and supernatants retained for analysis.

Samples are analysed by HPLC. The complete reaction mix is applied to a ZORBAX 300Extend-C18- column (Agilent) and separated on a concentration gradient from 10% to 40% acetonitrile with 0.1 % TFA over 15 minutes (flow rate 1 ml/min). Peptides are detected by absorption at wavelengths of 214 nm and 280 nm as well as fluorescence with an excitation of 280nm and emission at 340nm.

Areas under the curve are integrated for the given times using ChemStation (Agilent) and half-lives calculated as y = y₀ + ax using SigmaPlot 1 1.

Internalisation of EGFP labelled proteins

Cells are cultured for at least 1 h in serum free RPMI to avoid artefacts through possible serum contained ligands. Then, between 22 and 32 h prior to ligand exposure, U20S-cells expressing either EGFP labelled receptor or the receptor and GFP/EGFP labelled Arrestin are seeded onto cover-slips and cultured. Ligands are diluted to the desired concentration in C1-solution. To avoid adherence of peptides to plastic surfaces, C1-solution is supplemented with between 0.1 and 0.5% BSA, depending on peptide hydrophobicity. After aspiration of the growth medium, 500 μΙ ligand solution is added per well and cells incubated at 37°C for 20 minutes. Cells are then fixed with 4% formaldehyde solution (10 minutes, room temperature). Coverslips are washed twice in PBS, and fixed on microscope slides with mounting medium (Fluorescent Mounting Medium, Dakocytomation, Hamburg). Fluorescent cells are analysed and digitised using an inverted confocal microscope (Zeiss Axiovert 100) with laser light source and appropriate filters and software.

Dyes are detected individually, at a resolution of 1024 X 960 pixels, superimposed and adjusted according to brightness and contrast.

Example 1 : Functional analysis of single and multiple L-AA substitutions in chemerin-9

47 variants having only one substitution in the native sequence were synthesized and characterized functionally.

Substitution variants were analysed using intracellular Ca²⁺-mobilisation following receptor stimulation. A HEK293A cell line stably expressing CMKLR1 und Ga₁₆ was produced. Ga₁₆ binds to a range of GPCRs and so functions as universal adaptor for the Ca²⁺ signalling pathway. Increasing peptides concentrations were applied. Duplicate measurements were made in between three and five independent experiments; mean values ± standard deviations are given. EC₅₀- values of different chemerin-9 single substitution variants are shown in Table 1 . These point toward a high degree of flexibility in Position 2 of chemerin-9. This is surprising, given that Wittamer et al. (ibid.) found this position to be important in interaction between peptide and receptor. Positions 5 and 7 are similarly flexible.

No. sequence ECso [nM] No. sequence EC so [nM]

YFPGQFAFS 1.25 ± 0.54 C024 YFPGAFAFS 6.56 ± 3.04

C001 WFPGQFAFS 34.47 ± 10.94 C025 YFPGVFAFS 6.41 ± 3.96

C002 FFPGQFAFS 9.04 ± 3.60 C026 YFPGKFAFS 9.56 ± 5.25

C003 DFPGQFAFS 96.68 ± 44.10 C027 YFPGNFAFS 9.69 ± 4.42

C004* YMPGQFAFS 4.90 ± 1.04 C028 YFPGFFAFS 162.17 ± 54.76

C005* YLPGQFAFS 0.78 ± 0.51 C029* YFPGQYAFS 0.46 ± 0.21

C006* YWPGQFAFS 1.75 ± 0.80 C030 YFPGQEAFS 396.62 ± 105.71

C007* YYPGQFAFS 5.32 ± 2.33 C031* YFPGQWAFS 3.39 ± 2.30

C008 YRPGQFAFS 9.53 ± 0.74 C032 YFPGQFDFS 70.60 ± 5.31

C009 YPPGQFAFS 95.46 ± 36.82 C033 YFPGQFYFS 85.52 ± 34.43

C010 YDPGQFAFS 189.42 ± 77.04 C034* YFPGQFTFS 0.39 ± 0.19

C01 1 YFGGQFAFS 928.85 ± 185.20 C035* YFPGQFSFS 0.33 ± 0.16

C012 YFDGQFAFS 221.25 ± 71.97 C036* YFPGQFHFS 0.56 ± 0.28

C013 YFIGQFAFS 92.95 ± 13.75 C037* YFPGQFGFS 0.93 ± 0.43

C014 YFLGQFAFS 18.14 ± 13.43 C038* YFPGQFCFS 1.24 ± 0.59

C015 YFAGQFAFS 16.30 ± 6.80 C039 YFPGQFEFS 191.56 ± 13.88

C016 YFVGQFAFS 28.56 ± 6.24 C040 YFPGQFAQS 1079.23 ± 374.51 C017 YFTGQFAFS 7.03 ± 2.27 C041* YFPGQFAIS 4.23 ± 2.00

C018 YFPPQFAFS 264.81 ± 41.68 C042 YFPGQFALS 64.86 ± 23.44

C019 YFPYQFAFS 3141.75 ± 1000.39 C043* YFPGQFAFG 0.79 ± 0.54

C020* YFPGMFAFS 0.46 ± 0.21 C044* YFPGQFAFA 3.52 ± 1.71

C021* YFPGTFAFS 3.30 ± 1.46 C045 YFPGQFAFC 23.84 ± 10.61

C022* YFPGSFAFS 2.81 ± 1.35 C046 YFPGQFAFH 1798.08 ± 1030.88

C023 YFPGHFAFS 6.63 ± 2.91 C047* YFPGQFAFM 3.1 1 ± 2.27

Table 1 (substituted AA given in bold face)

Table 2 shows an analysis of chemerin-9 variants having multiple substitutions. The measurements were conducted as described for table 1. It is remarkable that all variants analysed show the same or higher potency as the original chemerin-9 sequence.

Nr. sequence EC 50 [nM] Nr. sequence EC so [nM]

YFPGQFAFS 1.25 ± 0.54 C062 YFPGMYSFM 0.36 ± 0.10

C048 YLPGMYSFG 0.43 ± 0.25 C063 YFPGMYSFS 0.14 ± 0.05

C049 YLPGMFAFS 0.52 ± 0.05 C064 YFPGQYSFM 0.63 ± 0.06

C050 YWPGMFAFS 0.89 ± 0.16 C065 YFPGMYAFS 0.18 ± 0.03

C051 YLPGMFAFG 0.75 ± 0.83 C066 YLPGQFTFS 0.38 ± 0.09

C052 YWPGMFAFG 1.67 ± 1.96 C067 YLPGQFSFS 0.26 ± 0.14

C053 YLPGQFSFG 0.30 ± 0.02 C068 YLPGQFHFS 0.65 ± 0.05

C054 YWPGQFFFG 0.47 ± 0.13 C069 YLPGQFSFG 0.60 ± 0.55

C056 YLPGMFSFG 0.29 ± 0.10 C070 YLPGQFHFG 0.37 ± 0.19

C057 YLPGQYSFG 0.1 1 ± 0.07 C071 YLPGQFGFG 1.32 ± 0.64

C058 YLPGQFAFG 0.39 ± 0.1 1 C072 YLPGMYAFS 0.14 ± 0.06

C059 YLPGMYGFF 0.30 ± 0.08 C073 YFPGMFAFG 0.27 ± 0.14

C060 YLPGMYHFG 0.79 ± 0.56 C074 YFPGMYAFS 0.45 ± 0.09

C061 YLPGMYCFG 0.60 ± 0.71 C075 YLPGQYAFG 0.24 ± 0.06

Table 2 (substituted AA given in bold face)

Example 2: metabolic stability of chemerin-9 variants

Single substitutions of L-amino acids by their respective D enantiomer had little if any influence on half-life (data not shown), with those variants having the greatest half-life extension (positions 2 and 3) also showing the greatest reduction in potency. Consequently, non-natural amino acid substitutions were introduced to provide additional stability. The sequences and results of their analysis are shown in Table 3. The determination of EC₅₀ values was performed as indicated for Example 1.

The half-life (t1/2 [min]) in 25% human serum was determined by linear regression of the area under the signal curve as a function of time. Duplicate measurements were made in two to three independent experiments; mean values ± standard deviations are given.

Nr. Sequence ECso [nM] i/₂ [min]

YF P GQ FAF S 1.24 24

C076 yChaP GM YAF s 2.49 199

C077 fChaP GM YAF f 2.14 280

C078 fChaP GM YAF s 4.10 284

C079 AcYChaP GM YAF s 0.70 908

C080 AcYChaP GM YAF f 0.30 17427

C081 AcYChaP GM YAF S 2.12 663

C083 fCh HypGCitFAF s 54.34 496

C084 AcY Cha Hyp G Cit F A F f 1.33 153

C085 Acy Cha Hyp G Cit F A F S 0.33 1195

C086 AcYFniP GQ FAF S 1.69 295

C087 AcY Fni P G Hnv F A F S 1.09 412 C088 AcY Fni P G Hnv F A F s 6.20 797

C089 AcY Fni P G Hnv F A F f 21.85 2 2

C090 AcY Fni Hyp G Hnv F A F S 0.80 142

C091 AcY Cha Hyp G Cit F a Tic S 0.29 14994

C092 AcY Cha P G Q Y w F f 0.39 1230

C093 AcY Cha P G Y A Tic S 3.96 69

C094 AcY Cha P G M Y a Tic S 3.04 1560

C095 AcY Cha P G M Y A F S 0.12 405

C096 AcY Cha P G M Y a F S 0.69 1532

Table 3

All sequences in Table 3 show a significant half-life increase. The sequences with longest half-life show an increase of almost 4 orders of magnitude, while keeping potency of the peptide -within the error of measurement- at that of the native chemerin-9.

Example 3: Receptor internalisation

Cell lines stably expressing CMKLR1 as well one of C-terminally labelled n32-Arrestin-GFP, ήβ1- Arrestin-EGFP or h 2-Arrestin-EGFP were generated. Labelling of Arrestins with green fluorescent protein (GFP) enabled localisation and observation of cells before and after ligand stimulation.

Reduction of membrane fluorescence and formation of vesicles in the cytoplasm following stimulation with chemerin-9 was detected by fluorescence microscopy. EGFP-CMKLR1 was found to be internalised into U20S-cells upon chemerin-9 stimulation. Internalisation can take place in conjunction with either h 2-Arrestin-EGFP or r 2-Arrestin-GFP.

The variants of the invention were analysed with this system. All showed internalisation similar in order of magnitude to that of the native chemerin-9 peptide (data not shown).

To elucidate differences observed in these experiments, internalisation of EGFP-CMKLR1 and 2- Arrestin-GFP was quantified by automated microscopy. Vesicle area per cell was determined after application of increasing peptide concentrations. Fig. 1 shows the dose response of chemerin-9 (EC₅₀ = 25.4 nM), Ac-Y-Cha-Hyp-G-Cit-FAFf (EC₅₀ = 2.15 nM) and Ac-Y-Cha-PGMYa-Tic-S (EC₅₀ = 52.3 nM) plotted against the vesicle surface area per cell. U20S-cells stably expressing CMKLRI und n32- Arrestin-GFP were used to generate data shown. Application of increasing agonist concentrations resulted in increased vesicle numbers within the cell and the resulting total vesicle surface area increased with agonist concentration.

Example 4: Quantitative analysis of cAMP-concentration after receptor stimulation cAMP GloSensor (Promega), a modified glow worm luciferase linked to a cAMP binding domain, was used for cAMP-analysis. Increased cytoplasmic concentrations of cAMP lead to heightened activation of the cAMP-GloSensors and an increase in the detected luminescent signal.

A HEK293A cell line stably expressing CMKLR1 and cAMP-GloSensor was generated. To confirm functional expression of CMKLR1 , cells were transiently transfected with Ga₁₆ plasmid and analysed (not shown). Stimulation with increasing chemerin-9-concentrations resulted in an EC₅₀-value of 29 nM, confirming stable expression of CMKLR1. Functional expression of cAMP-GloSensor was confirmed by applying 10 μΜ Forskolin and 1 μΜ Isoproterenol, both of which either directly or indirectly activate the 2-adrenergic receptor and adenylate cyclase respectively, resulting in an elevated intracellular cAMP concentration. Application of either agent gave a luminescent signal, confirming functional expression of the cAMP-GloSensor. No luminescence signal was detectable following stimulation with 1 μΜ chemerin-9.

It was subsequently tested whether stimulation with either chemerin-9 or Ac-Y-Cha-PGMYaFS led to adenylate cyclase inhibition (not shown). Incubation of cells in buffer alone and application of 1 μΜ Forskolin were used as negative and positive controls respectively. As expected, application of buffer, 100 nM chemerin-9 and 100 nM Ac-Y-Cha-PGMYaFS gave no luminescent signal. Initial addition of 1 μΜ Forskolin yielded a measurement of approximately 6000 relative light units (LU).

A second application of 5 μΜ Forskolin after initial buffer incubation yielded a luminescent signal of approximately 4000 LU. Application of 5 μΜ Forskolin yielded luminescence of only approximately 1000 LU in cells previously stimulated with chemerin-9 or Ac-Y-Cha-PGMYaFS. Adenylate cyclase was inhibited 4-fold by addition of either of the ligands. Different chemerin-9 variants were analysed quantitatively and compared to each other using this approach. First increasing peptide concentrations were applied; then cells were stimulated with Forskolin and IC₅₀-values calculated. Results are summarised in Table 4. Duplicate measurements were made in three independent experiments; mean values ± standard deviations are given.

Nr. Sequence ICso LE*10³

YFPGQFAFS 5.57 ± 0.02 3.63 ± 1.88

C080 Ac-Y-Cha-PGMYAFf 0.47 ± 0.05 4.24 ± 2.56

C081 Ac-Y-Cha-PGMYAFS 0.67 ± 0.32 4.29 ± 2.77

C084 Ac-Y-Cha-Hyp-G-Cit-FAFf 37.92 ± 3.86 3.85 ± 0.29

C085 Ac-y-Cha-Hyp-G-Cit-FAFS 0.90 ± 0.13 0.89 ± 0.16

C086 Ac-Y-Phe(4-N0₂)-PGQFAFS 4.55 ± 1.48 3.02 ± 1.02

C087 Ac-Y-Phe(4-N0₂)-PG-Hnv-FAFS 37.50 ± 3.26 3.84 ± 0.27

C090 Ac-Y-Phe(4-N0₂)-Hyp-G-Hnv-FAF 2.68 ± 0.39 3.94 ± 1.96

C091 Ac-Y-Cha-Hyp-G-Cit-Fa-Tic-S 35.65 ± 5.86 4.27 ± 0.34

C092 Ac-Y-Cha-PGQYwFf 1.13 ± 0.18 0.96 ± 0.25 C094 Ac-Y-Cha-PGMYa-Tic-S 0.96 ± 0.04 0.94 ± 0.23

C096 Ac-Y-Cha-PGMYaFS 10.08 ± 0.38 4.34 ± 2.70

Table 4

Example 5: Competitive binding studies with CMKLR1 and ²⁵lodine-labelled chemerin-9

Peptide-receptor binding was analysed using ²⁵lodine-labelled chemerin-9 binding to CMKLR1- expressing HEK293 membranes. IC₅₀-values were determined for chemerin-9 and the three stabilised variants C080, C091 und C092. Results are summarised in Table 5; duplicate measurements were made in two independent experiments; mean values ± standard deviations are given.

Table 5

Example 6: CMKLR1 expression in vivo

a) Validation of Polyclonal Antibodies against CMKLR1

Polyclonal sera against CMKLR1 were raised and analysed by western blot and antibodies were affinity purified. Affinity purified antibodies not only showed similar western blot signals compared to CMKLR1 N-terminally tagged with the epitope 7.12E3, also the western blot signals and the immunofluorescence staining of the two antibodies recognizing either peptide 86 OR peptide 87 were identical. To further validate these antibodies, HEK293A cells, stably expressing CMKLR1 , were analysed by flow cytometry. This offers the possibility to discriminate between intracellular and extracellular epitopes, since cells with permeabilized membranes can be excluded from the analysis by simultaneous staining with propidium iodide, a DNA intercalating, membrane impermeant fluorescent dye. Indeed, the antibodies purified against the extracellular epitope of CMKLR1 (peptide 86) showed strong staining of CMKLR1 expressing cells compared to untransfected cells, but not the antibodies purified against the intracellular epitope (peptide 87).

b) Expression of CMKLR1 in Normal Human Tissue

Published CMKLR1 expression analyses focus on mRNA level, with cartilage tissue being the only histological sample investigated for CMKLR1 protein expression. A panel of human cryo-tissues, mainly from the gastrointestinal tract, was analysed for CMKLR1 expression with

immunofluorescence. CMKLR1 expression was found mainly in epithelial regions of gastrointestinal tissues and in pancreatic islets, whereas thyroid, liver and lung samples showed no expression compared to negative controls. Strong signals were observed in the squamous epithelium of the tonsil, in the basement membrane of the oesophagus and in single cells in the epithelium of the antrum and the small and the large bowel. Also single cells in pancreatic islets showed strong staining. Moderate staining was observed in single cells of the epithelium of the prostate and in the vascular epithelium of the kidney (Table 6: Cryo-sections of eleven non-neoplastic human tissues (n=2) were analysed by immunofluorescence with polyclonal antibody 20/86).

Table 6

c) Expression of CMKLRI in Gastrointestinal Tumours

A panel of eight different gastrointestinal tumour entities were analysed by immunofluorescence on cryo-sections with the polyclonal antibody 20/86 for expression of CMKLR1. Among these tumour entities, 50% (5 of 10) of the pancreas tumours and 100% (3 of 3) of the oesophageal squamous cell carcinomas (ESCC) showed areas with strong membranous expression of CMKLR1 , whereas the others did not (Table 7. Cryo-sections of eight different gastrointestinal tumour entities were analyzed in immunofluorescence with polyclonal antibody 20/86. Samples were classified as positive if a clear membrane-staining was observed in more than 5% of the cells).

Tumour Type Positive/T otal Samples

pancreatic carcinoma 5/10

insulinoma 0/5

islet cell carcinoma 0/5

colorectal carcinoma 0/5

hepatocellular carcinoma 0/5

oesophageal squamous cell carcinoma 14/16

oesophageal adenocarcinoma 4/16

neuroendocrine tumour 0/1 1 d) Expression of CMKLR1 in Oesophageal Cancer

The screening of gastrointestinal tumour entities (c) showed strong membranous expression of CMKLR1. Expression of CMKLR1 in oesophageal squamous cell carcinoma was more stable, since all three samples showed strong membranous expression in about 25-60% of the tumour area. For this reason, we decided to analyse a set of 32 oesophageal cancer samples, comprising 16 squamous cell carcinomas and 16 adenocarcinomas (one of the adenocarcinoma was classified as Barrett's oesophagus). For thirteen samples of each group, non-neoplastic adjacent tissue samples were also available, and analysed in parallel. This set was analysed for CMKLR1 expression not only by immunofluorescence on cryo-sections, but also in western blot of consecutive cryo-sections solubilized in a Tris-buffered solution containing 1 % SDS prior to SDS-PAGE.

To further characterize the tumour regions expressing CMKLR1 , double immunofluorescence staining was performed with the polyclonal antibody 20/86 and monoclonal antibodies for cytokeratin 19, an intermediate filament and marker for epithelial cells, and a monoclonal antibody for vimentin, another intermediate filament and marker for mesenchymal cells.

The staining of these samples showed that 88% (14/16) of the oesophageal squamous cell carcinomas (ESCC) and 25% (4/16) of the oesophageal adenocarcinomas (EAC) strongly expressed CMKLR1.

A quantifiable indicator for overexpression was the intensity of the western blot signal. All samples of pure non-neoplastic tissue (19 samples) showed no signal at all. From the EAC tumour samples, only the one classified as Barrett's Oesophagus showed a weak signal. From the ESCC samples, eight of 16 samples showed a signal.

e) Expression of CMKLR1 in Esophageal Cancer Cell Lines

Seven oesophageal cancer cell lines were analysed for the expression of CMKLR1. The cell lines Kyse 140, Kyse 180, Kyse 410, Kyse 520 and OE21 originate from squamous cell carcinomas, OE19 from an adenocarcinoma of the esophageal gastric junction and OE33 from an adenocarcinoma further classified as Barrett's metaplasia.

In immunofluorescence, Kyse 410 and OE33 showed no staining, Kyse 140 and OE19 showed a rather diffuse, not clearly membraneous staining. Kyse 180, Kyse 520 and OE21 showed membrane staining, but only in about 10%, 40% and 20% of the cells. In western blot, Kyse 140 and Kyse 410 showed no band at all, whereas Kyse 180, Kyse 520 and OE21 showed a band at the characteristic size of about 55 kDa also found ESCC samples. Kyse 520 also showed a signal of about 65 kDa, a signal not observed in cancer samples, but in stably transfected HEK293A cells. OE19 showed a band corresponding to a slightly smaller molecular weight than Kyse 180, Kyse 520 and OE21. OE33 finally showed a signal corresponding to a molecular size of about 80 and also 100 kDa. These signals were characteristic for OE33 and not observed in any other cell line.

Example 7: Tumour detection in-vivo

Using solid phase peptide synthesis, a peptide derived from C080 (Acetyl-Tyr-Cha-Pro-Gly-Met-Tyr- Ala-Phe-ser-OH) was synthesized without the N-terminal acetyl group. Instead, the pre-activated near- infrared fluorescent dye indotricarbocyanine (ITCC) was coupled to the amino group of the N-terminal tyrosine via an aminohexanoic acid linker. The resulting labelled peptide (CG37) was used in tumour model imaging experiments. Nude mice carrying subcutaneous xenograft tumours of CMKLR1- expressing wild-type Kyse180 cells (a cell line derived from a patient with squamous cell carcinoma of the oesophagus) or U20S cells (a cell line isolated from a patient with osteosarcoma) stably expressing CMKLR1 from a transfected plasmid were used in near-infrared fluorescent optical planar imaging. The fluorescent peptide conjugate CG37 was dissolved in 50% DMSO at a concentration of 1 mmol/l. For the imaging experiment, from this stock solution 0.6 nmol of the fluorescent peptide CG37 were diluted into 150 μΙ of 0.9% NaCI. This solution was injected into a lateral tail vein of a tumour-carrying nude mouse. Under gas anaesthesia, animals were imaged in a near-infrared fluorescence small animal imaging device with dedicated laser excitation at 785 nm (FWHM < 2 nm) and fluorescence emission detection at 820 nm (FWHM = 20 nm). Imaging was performed at 24 hours post injection. The results are shown in Fig. 2, where subcutaneous tumours at the flanks of the animals are clearly visible.

Claims

1. A chemerin-9 variant of the sequence Y F² P³ G Q⁵ F⁶ A⁷ F⁸ S⁹, wherein a position is

substituted, characterized by at least one of the following substitutions:

F² is one of the group comprised of Y, W, J, V, M, an (S)-2-aminopropionic acid substituted at carbon 3 by a C4-, C5- or C6 linear alkyl or by a C5- or C6 circular alkyl moiety, or an (S)-2-amino-3-(nitrophenyl)-propionic acid;

Q⁵ is selected from the group comprised of M, T, S, Cit and Hnv; and/or

A⁷ is selected from the group comprised of T, S, H, G, C, and w.

2. A chemerin-9 variant according to claim 1 , wherein at least one of the following substitutions is additionally present:

- Y is f or N-acetyl-Y;

- P³ is Hyp;

- F⁶ is Y, W or w;

- A⁷ is a;

- F⁸ is Tic;

S⁹ is s or f.

3. A chemerin-9 variant according to any of the above claims, wherein at least two of F², Q⁵ and A⁷ are substituted.

4. A chemerin-9 variant according to any of the above claims, wherein

- Y is N-acetyl-Y;

F² is Cha or Fni; and

Q⁵ is selected from the group comprised of Q, M, Cit and Hnv.

5. A chemerin-9 variant according to any of the above claims, having a sequence set forth in Table 3.

6. A chemerin-9 variant according to any of the above claims, having a metabolic stability, as measured in half-life in human serum, at least five times that of native chemerin-9.

7. A chemerin-9 variant according to any of the above claims, having an EC₅₀ for CMKLR1 , of <50% that of native chemerin-9.

8. A chemerin-9 variant according to any of the above claims, which comprises or is covalently linked to a cancer drug.

9. A chemerin-9 variant according to any of the above claims, which comprises or is covalently linked to a radioisotope emitting beta or gamma radiation.

10. A chemerin-9 variant according to any of the above claims, which is covalently linked to a detectable label.

1 1. A chemerin-9 variant according to claim 10, wherein the label is a near infrared fluorescent dye or a radionuclide for PET or SPECT.

12. A method to detect the presence of a tumour cell expressing CMKLR1 , wherein a peptide is covalently linked to or comprises a detectable label, said peptide being selected from the group comprising chemerin-9 and a chemerin-9 variant according to any of the claims 1 to 10, and wherein said peptide is brought into contact with said tumour cell under conditions allowing for selective binding of said peptide to CMKLR1 , and binding is detected.

13. A method according to claim 12, wherein the tumour cell is selected from an oesophageal squamous cell carcinoma cell and pancreatic adenocarcinoma cell.

14. A method according to any of claims 12 to 13, wherein the method is practiced in-vitro method and the tumour cell is comprised in a sample obtained from a patient.

15. A method according to any of 12 to 13, wherein the method is practiced in-vivo.

16. The method of claim 15, wherein detection of the detectable label is performed by PET, SPECT or near-infrared spectroscopy.

17. A chemerin-9 variant according to any of claims 1 to 1 1 for therapy of oesophageal squamous cell carcinoma or pancreatic carcinoma, said chemerin-9 variant comprising or having covalently attached to it a radionuclide, a cancer drug or a toxin.