BIOORTHOGONAL REPORTER GENE SYSTEM

Abstract

The present invention relates to a nucleic acid molecule encoding a fusion protein comprising (i) a secretory signal peptide; (ii) a lipocalin-derived binding protein specifically binding to an exogenous ligand; and (iii) a glycosylphosphatidylinositol (GPI) anchored and/or transmembrane domain.

Claims

1. A nucleic acid molecule encoding a fusion protein comprising (i) a secretory signal peptide; (ii) a lipocalin-derived binding protein that specifically binds to an exogenous ligand; and (iii) a glycosylphosphatidylinositol (GPI) anchored and/or transmembrane domain.

2. The nucleic acid molecule of claim 1, wherein the exogenous ligand is linked to a radionuclide.

3. The nucleic acid molecule of claim 1, wherein the encoded fusion protein further comprises a peptide affinity tag.

4. The nucleic acid molecule of claim 1, wherein the encoded fusion protein further comprises a fluorescent protein.

5. The nucleic acid molecule of claim 1, wherein the exogenous ligand comprises a small molecule, wherein the small molecule is selected from (i) a chelator, (ii) an alkaloid, (iii) an iron-chelating siderophore, (iv) a plant steroid, and (v) an organic dye.

6. The nucleic acid molecule of claim 1, wherein the secretory signal peptide is the signal peptide of a lipocalin.

7. The nucleic acid molecule of claim 1, wherein the transmembrane domain is the transmembrane domain of CD4 or CD28.

8. A vector comprising the nucleic acid molecule of claim 1, wherein the vector is a retroviral vector, an adenoviral vector or an adeno-associated vector (AAV).

9. A fusion protein encoded by the nucleic acid molecule of claim 1.

10. A cell transduced or transfected with the nucleic acid molecule of claim 1.

11. The cell of claim 10, wherein the cell is a lymphocyte.

12. The cell of claim 10, wherein the cell further comprises a chimeric antigen receptor or a transgenic T-cell receptor.

13. A kit comprising (i) the nucleic acid molecule of claim 1, and (ii) an exogenous ligand.

14. The cell of claim 10 for use in the treatment of a disease by a cell-based therapy or a gene therapy.

15. An exogenous ligand for use in an in vivo method of diagnosing the efficacy of a cell-based therapy in a subject, wherein the subject has been treated with the cell of claim 10.

16. The nucleic acid molecule of claim 2, wherein the radionuclide is selected from C-11, F-18, Sc-44, Sc-47, Cu-64, Ga-68, Y-86, Y-90, Zr-89, Tc-99m, In-111, I-123, I-124, I-131, Tb-152 and Lu-177.

17. The nucleic acid molecule of claim 3, wherein the peptide affinity tag is selected from a V5-, Strep-II, Flag-, c-myc-, HA-, Spot-, T7-, and NE-epitope tag.

18. The nucleic acid molecule of claim 4, wherein the fluorescent protein is an autofluorescent protein selected from mRuby3, GFP, eGFP, sfGFP, UnaG, miRFP703 and miRFP720.

19. The nucleic acid molecule of claim 1, wherein the encoded fusion protein lacks a fluorescent protein.

20. The nucleic acid molecule of claim 1, wherein the exogenous ligand is a peptide with at least 2 amino acid residues and less than 10 amino acid residues.

Description

[0133] The figures show:

[0134] FIG. 1: Crystal structure of two Anticalin⋅ligand pairs for use in the invention. (A) Crystal structure of the Anticalin D6.2 (M69Q) in complex with colchicine (PDB ID: 5NKN). (B) Crystal structure of the Anticalin CL31 in complex with Tris-CHX-A″-DTPA⋅.sup.89Y (PDB ID: 4IAX). In both protein structures the C-terminus of the lipocalin-based binding protein, which is fused with the V5-tag and/or the transmembrane domain, is highlighted. Furthermore, the unique primary amino-group within the ligand moiety, which can be conjugated to a linker and a functional moiety, is highlighted.

[0135] FIG. 2: Schematic representation of the membrane-associated PET-reporter proteins. (A) The reporter protein intended for preclinical development comprising an Anticalin featuring a binding activity for DTPA, a V5-tag, a CD4 transmembrane domain and a cytoplasmic mRuby3 fluorescent protein. (B) Schematic representation of a similar reporter protein without a fluorescent protein (intended for clinical use).

[0136] FIG. 3: Modular DNA- and Protein-design of the reporter protein. The various expression cassettes are composed of regulatory and coding regions which are both assembled from different elements. Upstream regulatory elements are promoters that initiate transcription of the open reading frame (ORF), such as the Chicken Actin Promoter (CAG), the 5′ Long Terminal Repeats (5′LTR) or the Cytomegalovirus promoter (CMV) together with introns for increased expression strength. Downstream of the ORF, polyadenylation signals are incorporated for optimal protein expression. Furthermore, with viral vector systems genetic elements for amplification and packaging are incorporated such as the 5′LTR/3′LTR of Lentivirus or the Inverted Terminal Repeats (ITRs) of the Adeno-associated Virus (AAV). The reporter protein is composed of a secretory signal sequence/signal peptide for the incorporation into the membrane, one or more lipocalin-derived binding proteins that serve as binding domain for the radioligand, an optional V5-epitope tag and a transmembrane domain derived from human CD4 followed by an optional fluorescent protein. Furthermore, some expression constructs featured two different membrane proteins that are both expressed from a single mRNA, each having a secretory signal peptide for insertion into the membrane, just separated by a 2A self-cleaving peptide. A set of exemplary expression constructs that were constructed: (A) Full reporter gene with red fluorescent protein (“preclinical DTPA-R”), (B) without fluorescent protein (“clinical DTPA-R”), (C) without V5-tag, (D) with sequence for the GPI-anchor amidation of e.g. a serine residue instead of a transmembrane domain, (E) with a colchicine binding Anticalin (called “Colchi-R”), (F) a homo-bivalent reporter gene, (G) a hetero-duovalent reporter gene, (H) a reporter gene expression cassette for the use of lentiviral transduction technique, (I) a T2A-separated expression cassette for a Chimeric Antigen Receptor (CAR) and a reporter gene, (J) the same construct as before with inverted orientation, (K-L) a reporter protein with miRFP720 as fluorescent protein that features ITRs for gene transfer of the reporter gene using recombinant AAV viral vectors. (M) Example of a reporter protein that features an intracellular fluorescent protein which is not connected to the cell membrane by a polypeptide chain but separated by a self-cleaving 2A sequence. Intracellular self-cleavage and separation of the fluorescent protein leads to much higher cell surface expression of the membrane-anchored reporter protein compared to the situation where the fluorescent protein is permanently fused to the reporter protein (see, e.g., FIG. 4B).

[0137] FIG. 4: Determination of expression levels obtained for different reporter proteins and reporter gene expression cassettes by flow cytometry. Lentiviral vectors were produced and used for the infection of the human T-cell line Jurkat. Cells were analyzed after infection using flow cytometry on a LSRFortessa (BD Biosciences) for intrinsic mRuby3 fluorescence or the V5-tag presented on the cell surface. (A) Expression levels of different reporter genes. Sodium/iodide symporter fused to mRuby3 (NIS-mRuby3) and a soluble mRuby3 protein were compared to the reporter proteins of the invention, that is DTPA-R and Colchi-R, both fused to mRuby3. (B) Comparison of different fluorescent proteins fused to the C-terminus of the reporter protein DTPA-R and located in the cytoplasm. Cells were stained with an Alexa Fluor 488-conjugated antibody (SV5-PK1) directed against the extracellular V5-tag. (C) Comparison of reporter proteins without or together with a chimeric antigen receptor (CAR) fused via a C-terminal 2A-sequence on a single mRNA. (D) Comparison of inverted order of reporter protein and CAR for DTPA-R and Colchi-R.

[0138] FIG. 5: Western blot analysis of cells expressing the reporter proteins of the invention. The human T-cell line Jurkat was modified for expression of the reporter proteins of the invention using lentiviral transduction or CRISPR/Cas9 mediated gene transfer. Corresponding cells were cultivated, harvested, and proteins were extracted using RIPA-buffer (Thermo Scientific). Samples were adjusted for the same total protein amount and separated using SDS-PAGE (4-20% under reducing conditions), followed by blotting onto a PVDF membrane. The membrane was incubated with the monoclonal anti-V5-antibody SV5-PK1 and appropriate fluorescent secondary reagents (800 nm channel), together with a control antibody recognizing beta-actin (DyLight680-AbD12141, Bio-Rad). Signals were detected using an Odyssey scanner (LI-COR) (A) and signals were analyzed using LI-COR software, which allowed quantitative comparison of the individual reporter proteins (B). The following samples were analyzed: 1) Jurkat cell line without transgene, 2) Jurkat DTPA-V5-TMD-mRuby3, 3) Jurkat Colchi-V5-TMD-mRuby3, 4) Jurkat DTPA-V5-TMD, 5) Jurkat Colchi-V5-TMD, 6) Jurkat DTPA-V5-TMD-mRuby3 “clone Jurkat 7-14”. While sample 1) was a non-transgenic mock sample/control, 2-5) were lentivirally transduced cell lines and 6) was stably modified using CRISPR/Cas 9 plasmid transfection and subsequent selection using appropriate antibiotics. Besides the intact protein corresponding to the complete reporter protein with processed signal peptide, a truncated protein species is present, in which the reporter protein has been proteolytically cleaved, most probably within the intracellularly located fluorescent protein domain. The detection of cell lysates using the B1 antibody also results in an unspecific band produced by a cross-reactivity of the antibody with an unknown host protein, which is also present in the HEK cells that were not genetically modified.

[0139] FIG. 6: Fluorescence microscopy of cells expressing the reporter proteins of the invention. (A) HEK (humand embryonic kidney) or (B) PC3 (human prostate carcinoma) cells were transduced with lentiviral vectors to express different versions of the DTPA-R or the Colchi-R, subjected to FACS cell sorting and subsequently grown in 96-well plates, then stained with Hoechst 33342 and the anti-V5-tag antibody SV5-PK1 as well as the appropriate secondary reagent, and finally imaged using an Evos M7000 fluorescence microscope (Thermo Scientific) with appropriate filters at 40-fold magnification.

[0140] FIG. 7: Magnetic-activated Cell Separation (MACS) using the reporter protein of the invention. Jurkat cells expressing the DTPA-R (A) or the Colchi-R (B) were mixed with un-transduced Jurkat cells at a 5:95 ratio. A total of 10×10.sup.6 cells were incubated for 30 min on ice with 2 μg anti-V5-tag antibody SV5-PK1 (Bio-Rad Laboratories). Subsequently, 20 μl anti-mouse IgG (H+L)-microbeads were added and binding was allowed for 15 min at 4° C. Then, the cell suspension was loaded onto a MiniMACS separation column (type MS) which was magnetized using a MiniMACS separator magnet (all from Miltenyi Biotec). After washing with MACS buffer, the column was removed from the magnetic field and the cells of interest were eluted together with the MACS buffer. Finally, a sample of the initial cell suspension the flow through and the eluted cell fraction were analyzed using flow cytometry.

[0141] FIG. 8: Detection of cells tagged with the invention on a cellular level using immunohistochemistry.

[0142] Jurkat cells, either transfected with DTPA-R or Colchi-R were injected s.c. into CD1 nude mice (Charles River Laboratories) and explanted after mice had been sacrificed. Tumor tissue was fixed in 10% (w/v) neutral-buffered formalin solution (Otto Fischar, Saarbrücken, Germany) for 48 h and stored in PBS at 4° C. Tissue samples were dehydrated using an automated system (ASP300S; Leica Biosystems) and subsequently embedded in paraffin. Serial 2 μm sections were prepared with a rotary microtome (HM355S; ThermoFisher Scientific) and subjected to histological and immunohistochemical analysis. Hematoxylin and eosin (H&E) staining was performed on deparaffinized sections with Eosin and Mayer's Haemalaun (Morphisto, Frankfurt am Main, Germany). Immunohistochemistry was performed using a Bond RXm system (Leica Biosystems) with primary antibodies against V5-tag (SV5-Pk1; Biorad) using 1:250 dilutions. Briefly, slides were deparaffinized using deparaffinization solution (Leica Biosystems), pretreated with Epitope retrieval solution 1 (corresponding to citrate buffer, pH 6) for 30 min. Bound antibody was detected with a polymer refine detection kit without post primary reagent and with an intermediate Rabbit anti-mouse IgG secondary antibody (diluted 1:400; Leica Biosystems) and signals were developed with 3,3′-diaminobenzidine (DAB). Representative images were collected on an Aperio AT2 digital pathology slide scanner using ImageScope (ver.12.3) software (both from Leica Biosystems).

[0143] FIG. 9: Determination of expression levels of the reporter proteins in T-cells and the influence on the cell division rate.

[0144] The human T-cell line Jurkat was lentivirally transduced with expression cassettes for different reporter proteins and after FACS sorting of the 10% highest expressing clones, individual stable sub-cell lines were created. For these cell lines the doubling time was assessed using the CFSE assay (A) and the number of receptors on the cell surface was quantified using flow cytometry measurements (B).

[0145] To this end, cells were cultivated in RPMI medium with 10% FCS and penicillin and streptomycin. In order to further assess the influence of the T-cell activation on cell division rate and reporter protein expression, the cells were unspecifically activated with Phorbol myristate acetate (PMA) and lonomycin (both from InvivoGen, Toulouse, France). Cells were seeded at a density of 0.5×10{circumflex over ( )}6 cells/ml in medium with or without 1 μg/ml PMA and 10 μg/ml lonomycin for 24 h and were then transferred into medium without chemical activators and measurements were performed after 2 days. (A) Exponentially dividing Jurkat cells were counted and a number of 15×10{circumflex over ( )}6 cells was dissolved in 500 μl PBS and labeled with the CFSE Cell Division Tracker Kit (BioLegend, CA, USA) for 20 min at 37° C. The fluorescent probe 5-(and 6)-Carboxyfluorescein diacetate succinimidyl ester (CFSE) conjugates to primary amine groups on the cells which subsequently can be quantified using flow cytometry. As the CFSE-signal of cells is reduced by 50% by every cell division, the quantification of the signal after some days allows calculation of the number of cell division that had occurred. In order to quantify the loss of signal, in addition to the stained cells also beads with a known concentration of the same fluorescent dye (Quantum FITC-5 MESF, Bangs Laboratories) were used for flow cytometry on a LSRFortessa (BD Biosciences) instrument. The reduction of the fluorescence signal was analyzed following 4 days of cell growth and the MESF kit was then used to calculate the number of corresponding fluorescein dyes that was then used to calculate the doubling time. (B) The same cells, with and without chemical activation using PMA/lonomycin were counted and a number of 0.5*10{circumflex over ( )}6 cells was stained with 2 μl of a 147 μg/ml SV5-PK1 antibody solution (Bio-Rad Laboratories, Hercules, CA) that had been conjugated with Alexa Fluor488-NHS (Lumiprobe, Hannover, Germany) at a degree of labeling of 3.85 dyes per antibody. Here the Quantum Alexa Fluor 488 MESF kit (Bangs Laboratories, Fishers, In) was used to calculate the number of Molecules of Soluble Fluorochrome (MESF) based on mean fluorescence intensities and it was assumed that one antibody binds with its two paratopes (Fab arms) to two reporter proteins. Error bars indicate standard deviation and represent the variance of the cell populations recorded in flow cytometry.

[0146] FIG. 10: Absence of negative influence of the reporter protein on the cellular function of a CAR T-cell

[0147] In order to confirm the absence of negative impact of the reporter protein of the invention on the cellular function of a CAR T-cell, two different expression constructs were compared, both featuring an anti-CD19 chimeric antigen receptor, a self-cleaving 2A peptide followed by either a truncated version of the epidermal growth factor receptor (EGFRt) or the reporter protein DTPA-R. In short, peripheral blood mononuclear cell (PBMC) were isolated from a healthy donor and subsequently transduced with lentiviral vectors encoding the described expression cassettes (CAR/EGFRt or CAR/DTPA-R). After isolation of the cell population expressing the transgene, cells were expanded in RPMI medium with 200 U/mI IL-2, 12.5 ng/ml IL-7 and IL-15, each (all from PeproTech, Rocky Hill, NJ, USA) and maintained at a cell density of 0.25×10.sup.6 cells/ml. The ability of the CAR T-cells to kill CD19 expressing target cells was assessed by Chrome-51 (.sup.51Cr) release assay that works by loading the target cells with the radioactive isotope and the subsequent quantification of the Cr-51 that has been released into the supernatant of the culture. For this assay, the CD19-positive target cell lines Nalm-6 (A) or Raji (B) were incubated together with the respective CD19-CAR T-cells at different ratios of T-cells over target cells. A number of 10 000 target cells was incubated with the respective number of PBMCs in a V-bottom 96-well plate in 150 μl culture medium. The plate was incubated for 4 h with occasional agitation at 37° C. Afterwards, the radioactivity in the supernatant was quantified using a Wizard2 gamma counter (PerkinElmer) and the Percent Specific Lysis was calculated using the following formula: [(experimental release−spontaneous release)/(maximum release−spontaneous release)]*100. In addition, the ability of the CAR T-cells to kill CD19 expressing target cells was assessed by a FACS killing assay (C-F). To this end, GFP expressing NALM-6 target cells were incubated with 1:1 or 5:1 excess of CAR/DTPA-R or CAR/EGFRt CAR T-cells. After 48 h a comparable fraction of the cultured cells (as measured by FACS quantification beads) was analyzed and the number of GFP-positive target cells that remained after the killing assay was quantified.

[0148] FIG. 11: Composition of an exemplary .sup.18F-labeled radioligand composed of a ligand moiety (here colchicine, alternatively Bn-CHX-A″-DTPA), a PEG linker and a radiolabel moiety (e.g. radio-fluorinated pyridine for PET-detection). Hydrophilicity optimization of a fluorine-18 labeled radioligand used in conjunction with the Colchi-R. In this example, the radioligand composed of the colchicine ligand moiety, a PEG-linker and a fluorinated pyridine was analyzed (A), resulting in an unfavorable LogP value that indicated low hydrophilicity. Of note, hydrophilicity is crucial to prevent unspecific accumulation in cells or cell membranes and to ensure favorable distribution in the body as well as subsequent elimination by the renal route. Thus, two derivatives with increased hydrophilicity were designed featuring one (B) or two (C) interspersed D-Glu residues carrying negatively charged side chains. The incorporation of one D-Glu residue shifted the LogP value to −1.92 (meaning 98.8% in the PBS phase) and the additional incorporation of a second D-Glu residue further increased the LogP value to −2.97 (which means 99.9% of the radioligand is distributed in the aqueous PBS phase) rendering it a promising radioligand candidate.

[0149] FIG. 12: Labeling reaction, quality control and in vivo biodistribution for .sup.18F.sup.− for .sup.18F-Py-Glu.sub.2-PEG(4)-Colchicine.

[0150] The precursor .sup.+TMA-Py-D-Glu-D-Glu-PEG(4)-Colchicine was used for synthesis of .sup.18F-Py-D-Glu-D-Glu-PEG(4)-Colchicine (called .sup.18F-Colchicine). (A) In short, the radioactive .sup.18F.sup.− was produced by a cyclotron in house. Fluorine was eluted from a Sep-Pak Accell Plus QMA Carbonate (Waters) cartridge using 75 mM Tetrabutylammonium (TBA HCO.sub.3) solution and was subsequently dried in a Wheaton V-Vial under a continuous flow of dry argon gas at 95° C. by azeotropic distillation. An amount of 0.5 to 2 mg of the precursor was dissolved in DMSO and added into the V-Viral. The exchange reaction of .sup.18F for the Trimethyl ammonium (TMA) leaving group also occurred at 95° C. for 10 minutes. Subsequently, the reaction mixture was purified by a Chromafix PS-H.sup.+ (Macherey-Nagel) and a Sep-Pak C18 Classic (Waters) cartridge, which allows the quantitative removal of the unlabeled precursor by its positive charge. Subsequently, the activity was eluted from the C18 cartridge by EtOH and in some cases the radioligand was further purified by HPLC with a Multospher 100 RP 18−5μ 5μ 250×10 mm column with an isocratic elution at 25% Acetonitrile with 0.1% TFA.

[0151] (B) Quality control by HPLC of the radiosynthesis of .sup.18F-Py-D-Glu-D-Glu-PEG(4)-Colchicine: i) after radiolabeling reaction, ii) after cartridge purification and iii) after preparative HPLC purification, which were performed on a 100×4.6 mm Chromolith performance RP-C18 column (Merck Millipore) with a gradient of 5 to 55% acetonitrile with 0.1% TFA at a flow rate of 3 ml/min. A single peak was obtained as for both other precursors shown in FIG. 9 indicating a pure radioligand.

[0152] FIG. 13: Modular chemical design of radioligands for the DTPA-R

[0153] Based on the required ligand moiety for the DTPA-R (here NH.sub.2-Bn-CHX-A″-DTPA) there are different variants of radioligands that can be synthesized and compared in order to find the lead candidate which is expected to exhibit high affinity for the DTPA-R, in vivo stability and the absence of quantitative structural modifications in the patient and a favorable excretion pattern (ideally only via the renal route).

[0154] (A) The ligand moiety alone can serve as the radioligand when used together with an appropriate radiometal. (B) A variant where the amino-group within the NH.sub.2-Bn-CHX-A″-DTPA was replaced by the leaving group for radiofluorination (trimethylamine leaving group according to the “minimalist approach for radiofluorination” (Richarz et al., 2014)). (C) A design with a radiofluorination moiety, two D-Glu residues and a PEG(4) linker (called .sup.18F-DTPA) as in FIG. 11-12 for .sup.18F-Colchicine. (D-E) A bivalent radioligand with two ligand moieties which allows the loading with radioactive and non-radioactive metals. By the bivalent binding of the radioligand to the cell of interest via two different Anticalins, the overall affinity of the radioligand is drastically increased compared to a monovalent radioligand (comparable to the bivalent binding of full-length antibodies to cells). Dimerization was accomplished by a homo-bifunctional PEG(13)-NHS.sub.2 (D) or a aliphatic linker with interspaced peptide bonds (E).

[0155] FIG. 14: Confirmation of radioligand binding to cells tagged with the invention using radioligand binding assays.

[0156] The binding of different radioligands to cells that express the DTPA-R or Colchi-R was shown in binding assays. To this end, Jurkat or PC3 cells that stably express the respective reporter protein of the invention were cultivated, counted and a defined concentration was incubated together with the radioligand in PBS with 2% BSA for 1 h at 37° C. Subsequently, the cells were washed twice with PBS with 2% BSA and finally in the cell-bound fraction and the supernatant fraction the respective radioactivity was detected. (A) NH.sub.2-Bn-CHX-A″-DTPA⋅Y-90 binding was competed with a non-radioactive competitor NH.sub.2-Bn-CHX-A″-DTPA⋅Y-89. (B) Specific binding of .sup.18F-Py-D-Glu-D-Glu-PEG(4)-Colchicine to Colchi-R Jurkat and Colchi-R PC3 cells was quantified and compared to the reporter gene with another specificity (DTPA-R). Furthermore, the specific binding was blocked by 200 μM Colchicine solution to prove the specificity of the binding. (C) A comparative binding assay with .sup.18F-Py-Ahx-DTPA and .sup.18F-Py-Glu-Glu-PEG(4)-Colchicine to Jurkat cells expressing either DTPA-R or Colchi-R proving the orthogonal functionality of the two reporter protein & reporter probe pairs.

[0157] FIG. 15: In vivo PET-imaging of subcutaneous PC3 xenograft tumors tagged with the invention.

[0158] CD1 nude mice were subcutaneously injected with a Colchi-R PC3 xenograft in the right flank and a DTPA-R PC3 xenograft in the left flank. After xenograft tumors had developed the fluorine labeled radioligand was injected i.v. and a dynamic PET/MR was recorded for 90 minutes using a Mediso nanoScan PET/MR scanner. (A) Maximal intensity projections after .sup.18F-Py-D-Glu-D-Glu-PEG(4)-Colchicine i.v. injections are shown for the 5-10 min and 75-90 min time frame and exhibit a hepato-biliary excretion pattern. (B) In contrast, the i.v. injection of .sup.18F-Py-D-Glu-D-Glu-PEG(4)-DTPA⋅Tb (.sup.18F-DTPA) resulted in the specific accumulation of the radioligand in the tumor at the opposite shoulder (PC3 DTPA-R) together with a favorable, exclusively renal excretion pattern. Here, the mouse was awake for 60 min ahead of the PET-scan that was recorded von 60-90 min post injection. For the successful detection of this small xenograft tumor, also axial image sections are depicted. (C) Furthermore, Jurkat cells expressing DTPA-R were incubated in vitro with an excess of .sup.18F-DTPA and washed twice. Subsequently, a dilution series was prepared and the cells were scanned within PCR tubes for 60 min to determine the detection limit of the reporter gene system, which was found to be around 5.000 cells.

[0159] FIG. 16: Dynamic PET-imaging of DTPA-R xenograft tumor with .sup.18F-DTPA⋅Tb

[0160] (A) Comparable to the PET-imaging study in FIG. 15 a mouse with larger PC3 DTPA-R and PC3 Colchi-R xenograft tumors on the right and left shoulder were dynamically scanned for 90 min after injection of .sup.18F-DTPA. (B) at the time frame 75-90 min p.i. a remarkably low background signal and a pronounced tumor signal are visible. Additionally, there are elimination related signal in gall bladder, kidneys, ureter and the urine bladder. (C) an axial section through the tumor reveals a signal for the DTPA-R tumor and no signal above the background for the Colchi-R tumor. Furthermore, a necrotic core was visible in the sectional view of the PET which is characteristic for PC3 xenografts. (DE) The distribution of the radioligand at the border of the tumor and the absence of radioligand within the necrotic core was confirmed by autoradiography experiments.

[0161] FIG. 17: Use of DTPA-R for the detection and quantification of AAV9 transduction events in vivo

[0162] (A) For the detection of viral transduction events based on the adeno-associated virus, viral vectors were constructed by flanking an expression cassette for the DTPA-R or the Colchi-R with AAV2 inverted terminal repeats (ITRs). This genetic construct was used to produce and purify AAV9 viral vectors. (B) A dose dependent transduction was confirmed for DTPA-R and Colchi-R reporter genes after transduction of HEK cells with respective AAV viral vectors. (C) Anti-V5 immunohistochemistry staining of an axial cryosection of a mouse heart that had been i.v. injected with AAV before. (D) Maximal intensity projection of a PET-scan of a mouse that had been injected 7 days before with AAV9 viral vectors encoding DTPA. Data was recorded 70-90 min post injection of the .sup.18F-DTPA⋅Tb radioligand. Axial sections are depicted showing specific signal in shoulder and dorsal muscle tissue as well as the myocardium.

[0163] FIG. 18: Direct comparison of expression levels of an Anticalin in comparison with other binding proteins on the cell membrane

[0164] (A-C) Design of expression constructs for the comparison of Anticalin-based reporter proteins according to this invention versus scFv-based reporter proteins from the state of the art. (A) Construct for CL31d-V5-CD4TMD, (B) construct for scFv(muC825)-V5-CD4TMD and (C) construct for scFv(huC825)-V5-CD4TMD. Jurkat T-cells after stable retroviral transduction with one of the constructs were stained with an anti-VS-tag antibody conjugated with AlexaFluor488, which was detected using flow cytometry. (D) Histogram of the transduced cells analyzed by flow cytometry along with (E) the conversion to absolute numbers of receptors per cells (as explained for FIG. 9 herein above). From these stably transduced cell lines, the 10% highest expressing cells were isolated using Fluorescence Activated Cell Sorting (FACS). (F) Histogram of the FACS-sorted cells again analyzed by flow cytometry along with (G) the conversion into absolute receptor densities.

[0165] The following Examples illustrate the invention.

EXAMPLE 1

Development of a Novel Reporter Gene/Protein System

[0166] A reporter system was developed comprising a genetically encoded, membrane anchored reporter protein and a cognate small-molecule radioactive probe (radioligand) which can be injected intravenously. The radioligand is distributed in the blood compartment and selectively bound by the reporter protein, which is expressed on the genetically modified cell type. After rapid excretion of excess unbound radioligand via the kidneys or the hepato-biliary system, its radioactive decay in the body can be registered and by this way the distribution and location of the reporter gene-tagged cell type can be analyzed. This concept resembles the use of radiopharmaceuticals in nuclear medicine to detect various tumor targets (such as somatostatin, PSMA or integrins). The difference here is that the molecular target in the patient is a synthetic membrane-associated protein, whose coding gene has been incorporated specifically into the tagged cell type, which may serve for cell or gene therapy.

[0167] For an ideal reporter gene system, its components need to be bioorthogonal: neither the reporter (membrane) protein nor the radioligand should interfere with the healthy organs or tissues. This allows the monitoring of biological processes with minimal functional impairment and, on the other hand, with a minimal background signal. Furthermore, also the radioligand should be inert and not be modified or cleaved within the organism. A further requirement for clinical application is the lack of immunogenicity. For signal normalization, or for the simultaneous analysis of different biological processes, it is also desirable to have several independent reporter systems at hand that can be used in parallel without mutual interference (called multiplexing), similar to the well-established dual-luciferase systems or fluorescent proteins with distinct spectral properties used in biomedical research.

[0168] Anticalins are a class of engineered binding proteins based on the natural lipocalins. Most often, the human lipocalin 2 (Lcn2), also known as neutrophil gelatinase-associated lipocalin, NGAL, is used as scaffold for the selection of specific binding proteins (Richter et al., 2014, Schiefner and Skerra, 2015). The natural ligand of Lcn2 is the iron-chelating siderophore enterobactin that is employed by bacteria to sequester iron ions which are essential for their growth. The plasma protein Lcn2 binds the iron-siderophore-complex (Fe⋅enterobactin) before it can be taken up by the bacterium and, thus, restricts the growth of bacteria in the human body by deprivation of this essential metal. Lcn2 is part of the innate immune system and, in an engineered form, ideally suited for biomedical applications. Furthermore, the calyx-shaped ligand pocket of the lipocalins favors the tight binding of small-molecule ligands in correspondingly engineered Anticalins (FIG. 1).

[0169] The human origin of many Anticalins lowers immunogenicity after expression on a given cell of interest, e.g. an immune cell. As in current clinical investigations the immunogenicity of CAR receptors limits the persistence of some CAR T-cells in patients and hampers a positive therapy outcome, this is one of the most critical factors. Apart from protein or peptide targets, Anticalins have been selected to recognize several small molecules with high affinity (Table 1), for example petrobactin, a specific siderophore of certain Bacillus species (Dauner et al., 2018), colchicine (Barkovskiy et al., 2018) (FIG. 1A) and Bn-CHX-A″-DTPA (Kim et al., 2009) (Eggenstein et al., 2013) (FIG. 1B). These artificial ligands represent chemical structures foreign to human physiology (in contrast to the natural metabilites biotin, thymidine and analogues or iodine, which often have been used for cell tagging and, thus, may serve (after labelling) as bioorthogonal radioligands in the human body that are specifically recognized by the membrane-anchored Anticalin. The extraordinary affinity in the picomolar range of these Anticalins allows the imaging of the reporter construct and/or the transformed cell even at later time points after injection with a low background signal (due to the elimination of the free ligand). The high affinity of the Anticalin expressed on the cell surface can even be increased when a radioligand is used comprising two different ligand moieties, conjugated by an appropriate linker, that can be bound by two individual membrane-anchored Anticalins and, thus, result in an increased overall affinity (avidity effect, see FIG. 13 D-E for exemplary radioligands).

[0170] Generally, the reporter gene system of the invention features a modular design in a way that (i) different radionuclides can be incorporated into the radioligand via suitable chelator groups or selective conjugation chemistry to allow PET and/or SPECT imaging and (ii) the membrane-bound reporter protein can be expressed in diverse cellular context driven by different promoters (constitutive, chemically inducible or inducible by biological processes). Furthermore, iii) the reporter protein itself is modular in a way that it features further to mandatory protein domains (secretory signal peptide, Anticalin, membrane anchor) also optional protein domains such as the extracellular V5-epitope tag and the intracellular fluorescent proteins (FIG. 2). The gene transfer of the reporter gene of the invention into the cell of interest has been established with three different methods: (i) targeted CRISPR/Cas9-mediated genome integration after plasmid transfection or nucleofection and subsequent selection using appropriate antibiotics, (ii) lentiviral or retroviral gene transfer with stable random integration into the genome and (iii) adeno-associated viral vectors that lead to transient gene expression without genomic integration.

TABLE-US-00001 TABLE 1 Prior art Anticalins that bind small-molecule ligands CL31d D6.4(Q77E) (also (also abbreviated abbreviated herein as herein Anticalin DTPA) as Colchi) M2 DigA16(H86N) FluA(R95K/A45I/S114R) Ligand Bn-CHX-A″- Colchicine Petrobactin Dig(it)oxigenin Fluorescein DTPA•Me Literature (Kim et al., 2009) (Barkovskiy et al., 2018) (Dauner et al., 2018) (Schlehuber et al., 2000) (Vopel et al., 2005) (Eggenstein et al., 2013) Patent WO2009156456 WO2011069992 WO2011069992 WO2000075308 WO1999016873 Applicatio No. Scaffold human Lcn2 human Lcn2 human Lcn2 bilin-binding bilin-binding used protein protein Affinity ~500 pM ~450 pM ~20 pM for Fe.sup.III 350 pM ~1 nM (K.sub.D) ~50 PM for Ga.sup.III Radioligand CHX-A″- .sup.18F-Py-PEG(4)-Colchicine Petrobactin•.sup.68Ga.sup.III .sup.18F-Py-PEG(4)-Digoxigenin .sup.18F-Py-PEG(4)-Fluorescein DTPA•.sup.152Tb (or dimers thereof) or .sup.18F- DTPA•Tb Advantages Radiometal- Perfectly suitable .sup.18F- Gallium-68 labeled No more patent Patent protection charged radiolabeling due to the radioligand allows easy protection by by Pieris Pharma. CHX″- lack of acidic protons radiolabeling Pieris Pharma. DTPA can be used will expire soon

[0171] Anticalins binding suitable small-molecule radioligands with high affinity (see Table 1) need to be expressed in sufficient density on the surface of the engineered/transformed cell in order to allow complex formation between the Anticalin and its cognate radioligand in reasonable amounts. This design is somewhat similar to a membrane-based reporter gene that binds a DOTA⋅Me complex (Wei et al., 2008, Krebs et al., 2018). This reporter gene was made of the following protein moieties: (i) a signal peptide, (ii) a murine antibody scFv fragment (2D12.5/G54C), (iii) a transmembrane domain from CD4 and (iv) a P2A-separated green fluorescent protein (GFP) (Krebs et al., 2018). Disadvantages of this design, which is based on a single chain variable fragment (scFv) binding protein include: 1) oligomerisation of the antibody fragments (Hudson and Kortt, 1999), 2) low surface expression, 3) murine origin and, hence, potential immunogenicity in humans, 4) the cross-reactive radioligand lanthanoid(S)-2-(4-acrylamidobenzyl)-DOTA (AABD) which, as acrylamide derivative, unspecifically binds also to biological structures other than the DAbR1 reporter protein. In contrast, advantages of the lipocalin-based reporter proteins according to the invention include, but are not limited to: 1) human origin of the protein scaffold, 2) smaller size of the reporter protein, 3) single-chain protein architecture by nature, 4) high stability of the folded protein and 5) the absence of oligomerisation tendency. While antibodies exist in nature both in soluble and in a membrane-anchored form, lipocalins are exclusively found as soluble secretory proteins in mammals. ApoD may be seen as a rare exception as it is found associated with high density lipoprotein particles; however, this lipocalin does not carry a transmembrane domain or GPI anchor (Schiefner and Skerra, 2015).

[0172] In one embodiment, the reporter protein of the invention (FIG. 2) is composed of (i) the natural Lcn2 signal peptide [SEQ ID NO:1/2], (ii) a small-molecule-binding Anticalin (see Table 1 and [SEQ ID NO:3/4 or 5/6]), (iii) e.g. the V5-epitope-tag [SEQ ID NO:7/8], (iv) the transmembrane domain of human CD4 [SEQ ID NO:9/10] or CD28 [SEQ ID NO:11/12] and/or a GPI-anchor sequence [SEQ ID NO:13/14 or 15/16] and, optionally, (v) a fluorescent protein (such as mRuby3 [SEQ ID NO:17/18], miRFP703 and miRFP720 [SEQ ID NO:19/20]). In one embodiment, the fluorescent protein is separated from the reporter protein using a 2A self-cleaving peptide sequence. The reporter protein without a fluorescent protein (which may be suitable for method development and preclinical studies) only contains 259 amino acid residues (FIG. 3), which offers the advantage of a minimal genetic design with 777 base pairs (bp) [SEQ ID NO:23/24], thus allowing incorporation together with other functional elements into viral vectors with limited packaging capacity, such as the AAV (with a maximal packaging size only 4700 bp).

[0173] Various different expression vectors for the different reporter proteins have been constructed and tested (FIG. 3). These vectors comprise variants of the reporter protein for different forms of gene transfer, such as transfection (FIG. 3A-G), lentiviral or retroviral gene transfer (FIG. 3 H-J) or recombinant AAV gene transfer (FIG. 3 K,L).

[0174] Some applications as well as the beneficial properties of the reporter proteins according to this invention will be described and illustrated in the following Examples.

EXAMPLE 2

Demonstration of High Expression Levels, Correct Intracellular Transport to the Cell Membrane and Absence of Proteolytic Cleavage

[0175] High surface expression of exemplary constructs on transfected cells was verified by flow cytofluorometry (FIG. 4) and western blot analysis (FIG. 5). Flow cytometry analysis was performed and confirmed a much higher surface expression of the invention compared to the Sodium-Iodide Symporter (NIS) which has been often proposed as a reporter protein (Ravera et al., 2017). The surface expression level of the reporter protein of the invention was higher compared to the NIS-mRuby3 but lower than the fluorescent protein mRuby3 expressed as a soluble protein in the cytoplasm (FIG. 4A). Surprisingly, those reporter proteins lacking a fluorescent protein revealed 5-10 fold higher expression of the reporter protein on the cell surface, thus omitting the fluorescent protein for a given application of the reporter protein can be used to increase the expression level of the reporter protein and thus the signal that can be generated (FIG. 4B). Furthermore, the integrity of the reporter protein within T-cells and the absence of shedding of the extracellular domain comprising the Anticalin has been confirmed by western blot analysis of different stably transduced Jurkat cell lines using the V5-tag (FIG. 5).

[0176] Finally, the sub-cellular distribution pattern of the fluorescent protein mRuby3 as part of the longer reporter gene construct was investigated in transgenic HEK-(FIG. 6A) and PC3-cell (FIG. 6A) lines by fluorescence microscopy. The V5-tag as well as the intrinsic fluorescence of the mRuby3 were both detected at the cellular membrane of the cells, confirming the efficient secretion and membrane incorporation of the reporter proteins. This membrane localization was independent of the used lipocalin-derived binding proteins (Bn-CHX-A″-DTPA- or Colchicine-binding) and also independent of whether a fluorescent protein was incorporated into the design (FIG. 5).

EXAMPLE 3

Exploiting the V5-Epitope Tag for MACS and IHC

[0177] The incorporation of an epitope-tag into the reporter protein enables the use of antibodies against this tag for the detection of the protein on the cell surface (e.g. flow cytometry, masscytometry or MACS) or in an isolated form (e.g. western blot). For these applications, the V5-epitope tag (Southern et al., 1991, Dunn et al., 1999) was selected because of its high affinity (K.sub.D=24 pM for SV5-PK1), the strong denaturing conditions that are required to break this interaction (9 M urea and 1% Tween-20) as well as the hydrophilic and only slightly charged amino acid sequence (GKPIPNPLLGLDST, see SEQ ID NO:7/8) of the tag (Dunn et al., 1999).

[0178] The possibility to bind an accessible epitope on the cell surface with a high affine antibody was used for magnetic-activated cell sorting (FIG. 7). For this technique, the transgenic cells were stained with the SV5-PK1 antibody and subsequently incubated with an antibody against the murine Fc-domain that was conjugated with (super)paramagnetic beads that can be magnetized in a magnetic field. The decoration of the cell surface of the transduced cells was then utilized in a second step to retain the transgenic cells within a magnetic field, while at the same time the cells that are not linked to magnetic particles were washed from the column by the gravity flow of the MACS buffer. Using this method, a purity of ˜99% transgenic cells (FIG. 7) could be achieved which represents an interesting additional benefit of the invention, given the fact that MACS is often used in the production and purification process of patient-derived, autologous CAR T-cells.

[0179] Furthermore, the presence of the V5-epitope tag on the cell surface was used to identify the transgenic cells on a cellular level in a histological context (FIG. 8). For this application, the SV5-PK1 antibody was used for immunohistochemical staining of tissue sections which allowed the easy and clear identification of transgenic cells with a high contrast (FIG. 8) which is a big advantage compared to the IHC staining of a cell type of interest with antibodies against naturally occurring antigens. In the Jurkat xenograft tumor model that was investigated, it was possible to clearly distinguish between DTPA-R positive T-cells and murine cells in which the tumor cells had invaded (FIG. 8B) while this is not possible in consecutive HE stained tissue sections (FIG. 8C). This was also true for tissue that was mainly made up of tumor cells with a strong DAB-staining, where the non-stained blood vessel and the nerve fiber in the left part of the tissue section are clearly visible and distinguishable from the labelled cells (FIG. 8D).

EXAMPLE 4

The Reporter Protein is well Tolerated in Cells

[0180] When the reporter gene system of the invention is used for PET-detection of a cell population of interest, such as infused CAR T-cells or cells transduced with an AAV viral vector, an important prerequisite for the application of the reporter protein is the absence of negative impact on the therapeutic function of this cell population.

[0181] In order to check some of the major aspects that may be impaired, stable Jurkat cell lines with different reporter proteins were established and subsequently the doubling time of these cell lines (FIG. 9A) as well as the reporter protein surface expression levels of these cell lines (FIG. 9B) were determined.

[0182] The determination of the doubling time for these Jurkat cell lines using the CFSE cell proliferation assay confirmed the absence of a negative effect of the expression of the reporter gene on the cell division rate, as there were no significant differences between the doubling times determined for these cell lines. Although, the unspecific activation of the T-cell using PMA/ionomycin slightly increased the doubling time of all cell lines, there was no detectable difference between all evaluated reporter gene constructs and the untransduced control cells (FIG. 9A).

[0183] The determination of the total cell surface expression of the reporter gene was conducted with anti-V5 antibody that had been labeled with a fluorescent dye for which a kit with calibration beads was available. The results for the total number of receptors on the cell surface reflected the results earlier determined for different reporter gene constructs when quantifying the intrinsic fluorescence of the fluorescent protein (FIG. 4). The highest expression was measured for Jurkat transduced with DTPA-R with a mean of 870.000 copies on the cell surface (FIG. 9B). For all the cell lines measure the activation with PMA/ionomycin increased the surface expression of the reporter protein slightly, as T-cell activation is known to increase size and metabolic activity. For the Jurkat DTPA-R cell line, the surface expression of 870.000 copies was increased upon T-cell activation by ˜10% to 960.000 (FIG. 9B).

[0184] Thus, it is not only confirmed that the expression of different reporter proteins of the invention does not change the proliferation rate, but also that major events in the physiology of the T-cell only have a minor effect on the surface expression of the reporter gene.

[0185] Furthermore, anti-CD19 CAR T-cells created from peripheral blood mononuclear cell (PBMC) of a healthy donor were evaluated for their potential to kill CD19-positive target cells. The cellular toxicity to the target cells was measured by a radioactive Chrome-51 release assay at a 4 h end point (FIG. 10). The assay was conducted with B-cell precursor leukemia cell line NALM-6 (FIG. 10A) and the Burkitt lymphoma cell line Raji (FIG. 10B). With both target cell lines it was confirmed that the PBMCs transduced with CAR/DTPA-R showed equal killing efficacy compared to the PBMCs transduced with CAR/EGFRt, which is a successful CAR construct known to literature and serves as a reference standard (Wang et al., 2011, Paszkiewicz et al., 2016).

[0186] This result was furthermore confirmed by a FACS based killing assay in which the successful killing of GFP-expressing NALM-6 target cells was confirmed by both CAR T-cells, CAR/EGFRt and CAR/DTPA-R, respectively (FIG. 10 C-F).

EXAMPLE 5

Design of Radiopharmaceuticals for the Use with the Reporter Proteins of the Invention

[0187] It is crucial for the development of both a preclinically and a clinically useful reporter system to select the best affinity pair composed of a genetically encodable binding protein (here an Anticalin) and its ligand. Relevant criteria for this choice include: (i) the affinity of the binding protein to the small molecule ligand (see Table 1), (ii) the availability of different radiolabeling strategies to generate radioligands with different properties (FIG. 13), (iii) the in vivo distribution, clearance and metabolization properties of the radioligand, (iv) the pharmacological knowledge about the ligand and its previous use in medicine, (v) a high molar/specific activity and (vi) technical aspects of the radioisotope and the radiosynthesis itself.

[0188] Given these criteria, different ligands were synthesized (FIG. 11-13), their chemical properties were tested and their binding to transformed cells, followed by investigation of their biodistribution in vivo in mice. The result was that the incorporation of one or two D-Glu amino acid residues into the radioligand for use together with the Colchi-R increased the hydrophilicity drastically, as expected (FIG. 11). This increase in hydrophilicity is an intended characteristic of radiopharmaceuticals as it facilitates the clearance of the radioligand from the tissue and, thus, improves the signal to background ratio, which is necessary to detect even weak signals in PET, which are expected from small cell populations. While the .sup.18F-Py-PEG(4)-colchicine radioligand had a LogP value of 0.89 and, thus, only 11.4% was found in the PBS phase, the incorporation of one D-Glu residue improved the LogP to a more hydrophilic value of −1.92 (98.8% in PBS phase) and a second D-Glu residue even brought the LogP to −2.97 (99.9% in the PBS phase). The D-Glu residues were chosen over the proteinogenic L-Glu residues as they are known to limit the susceptibility of the resulting radioligand for proteolytic cleavage and, thus, improve the stability of the ligand in vivo (Grishin et al., 2020). Furthermore, a binding motif for a serum or plasma protein known in the art may be included into the linker, as it has been shown to modify the tissue distribution profile and extend the plasma half-life of small molecule ligands (Benesova et al., 2018). Such binding motifs could for example be selected from known albumin binding domains (Dumelin et al., 2008). The fluorination was accomplished by the incorporation of a nicotinic acid building block that featured a trimethyl-ammonium (TMA) leaving group that allowed the one-step radiofluorination of the precursor molecule according to the “minimalist approach” (Richarz et al., 2014). The advantage of this new method lies in the robust fluorination reaction that is based on a simple nucleophilic substitution reaction in which the positively charged TMA leaving group is exchanged for the radioactive Fluorid-18 (FIG. 12A). Furthermore, the positive charge within the leaving group allows the separation of the non-radioactive precursor and the radioactive radioligand by cartridges or chromatography in order to obtain an optimal specific (molar) activity of the radioligand. The different steps in the radiolabeling and purification process are depicted and illustrate the fact that cartridge purification with a QMA and C18 cartridge is sufficient to purify the product of the radiolabeling reaction to an acceptable purity, while this can even be improved by a final preparative HPLC run (FIG. 12B).

[0189] Finally, the radioligands for the Colchi-R were injected into nude mice and a dynamic PET-scan was recorded. While the biodistribution of the .sup.18F-Py-PEG(4)-colchicine radioligand is dominated by its lipophilic nature (positive LogP value) and, thus, is only eliminated insufficiently from the organism (FIG. 12C), the incorporation of the D-Glu residues improves hydrophilicity and, thus, improves the clearance from the tissue via the kidneys (FIG. 15A). Nevertheless, the compound is not sufficiently hydrophilic to be exclusively cleared via the kidneys.

[0190] Besides the radioligands for the Colchi-R (FIG. 11-12), also various radioligand for the DTPA-R were designed and synthesized (FIG. 13). These radioligands are all based on the binding motive that is tightly bound by the respective Anticalin, the NH.sub.2-Bn-CHX-A″-DTPA (FIG. 13A). Based on this ligand moiety, which can be used to chelate radio-metals suitable for diagnostic imaging or therapy, novel precursors for radiofluorination (FIG. 13BC) were constructed. Finally, also bivalent DTPA radioligands were constructed that feature two interconnected ligand moieties that can simultaneously bind to two different Anticalins on the cell surface and, thus, improve the overall affinity of the radioligand in vitro and in vivo (FIG. 13DE). These radioligand feature an improved affinity and are able to also deliver radionuclides for therapeutic purposes, such as Y-90, Tb-149, Tb-161, Bi-213 or a combination of such a therapeutic radioisotope together with a cold metal ion.

EXAMPLE 6

Specific Binding of the Radioligand to the Reporter Protein In Vitro and In Vivo

[0191] For the sensitive and specific detection of cells that are labeled with the reporter gene of the invention, the binding between the Anticalin binding protein and the radioligand is of utmost importance. For this reason, the binding of different radioligands to cells that were stably modified with the reporter genes was studies in vitro (FIG. 14). At first, the binding of the ligand domain for the DTPA-R (NH.sub.2-CHX-A″-DTPA) to Jurkat cells expressing the reporter protein DTPA-V5-TMD-mRuby3 was assessed. For this binding study, the radioactive Yttrium isotope Y-90 was chelated by the ligand domain and a fixed amount of the radioligand was applied to the cells that had been previously incubated with a dilution series of the same chelator molecule that had been charged with the cold Yttrium isotope Y-89. Curve fitting of the obtained binding curve resulted in a IC.sub.50-value of 1.4 nM, demonstrating the high-affine binding of the radioligand to the Anticalin on the cell surface (FIG. 14A). In order to test the specificity of the reporter protein—Colchicine radioligand interaction, the binding was blocked by a high molar excess of the ligand domain (here 200 μM colchicine) prior to incubation with the radioligand (FIG. 14B). Independent of the cell type in which the Colchi-R has been expressed (Jurkat or PC3), there was high accumulation of the radioligand in the cell fraction after washing them twice, while there was no signal detectable for the DTPA-R expressing cells. Comparable low levels of bound radioligand were measured for the cells that have been blocked with the colchicine prior to the application of the radioligand, again proving the specificity of the radioligand binding. Finally, the orthogonality of the DTPA-R and the Colchi-R have been studied in a single cell binding experiment (FIG. 14C). To this end, Jurkat cells expressing the respective reporter genes or untransduced Jurkat cells were incubated with the two radioligands, each. The result clearly shows that the DTPA radioligand only binds to the DTPA-R cells and the Colchi radioligand is specific for the Colchi-R modified cells. At the same time, there was no detectable binding of any radioligand to the cells of the second Anticalin-ligand pair or to the wild type Jurkat cells (FIG. 14C). Taken together, these experiments clearly demonstrate the high affinity of the membrane-bound Anticalin for the respective radioligand in an affinity-range comparable to the affinity determined for the recombinant Anticalin. Furthermore, the binding is specific to the respective reporter protein and there is no unspecific binding detectable.

[0192] The ability to use the reportergene system for in vivo PET-imaging was assessed (FIG. 15-17). For this experiment xenograft tumors were inoculated in nude mice and the mice where then used for the PET-imaging experiments. The intravenous injection of the radioligand .sup.18F-Py-D-Glu-D-Glu-PEG(4)-Colchicine resulted in a systemic distribution of the radioligand and a rapid clearance via the hepato-biliary system into the colon (FIG. 15A). Within the xenograft tumor that had been grown from Colchi-R positive cells, there was a clear and strong accumulation of the radioligand. In contrast, the xenograft tumor that had been grown from DTPA-R expressing cells showed no accumulation of .sup.18F-Colchicine above the level of the surrounding tissue. Upon an injection of the radioligand the PC3 xenograft showed an accumulation of the radioligand. This experiment clearly demonstrates the ability of the invention to allow the specific and sensitive detection of transgenic cells in vivo by means of Colchi-R PET-imaging. Furthermore, the radioligand of DTPA-R was tested in the same animal model comprising two PC3 xenograft tumors expressing both reporter proteins. The .sup.18F-DTPA radioligand .sup.18F-Py-D-Glu-D-Glu-PEG(4)-DTPA⋅Tb was synthesized and injected into a tumor-bearing nude mouse and after 60 min a static PET-scan was recorded (FIG. 15B). Even the very small PC3 DTPA-R xenograft tumor of only 11 mm.sup.3 was clearly visible in the MIP of the PET-scan as well as in the axial sections (shown below). Besides the desired signal of the tumor cells that were labeled with the DTPA-R reporter protein, only minor elimination related signal could be detected. These signals in gall bladder, kidneys, ureter and urine bladder are located in well defined tissues and organs of the animal and, thus, allow the differentiation between reporter protein related and elimination related PET-signals. Using later image time frames for the PET-scan, it will obviously be possible to allow the .sup.18F-DTPA to even further eliminate from the organism and to obtain an even cleaner background for the detection of DTPA-R tagged cells. Besides the exquisite specificity of the reporter gene PET-imaging system of the invention also the sensitivity of the system is of importance. Besides the clear detection of very small xenograft tumors (FIG. 15A), also the detection limit was determined with Jurkat cells expression the DTPA-R. These cells were incubated with an excess of the radioligand and were then washed twice. Subsequently, a dilution series in PCR tubes was prepared and scanned for 1 h in the Mediso nano-PET/MR scanner (FIG. 15 C). Here the detection limit was at ˜5.000 cells that showed enough signal to be discriminated from the background.

[0193] To further analyze the dynamics of the .sup.18F-DTPA⋅Tb radioligand in vivo, a dynamic PET-scan was conducted for 90 min post injection. An equal distribution of the radioligand in the whole animal can be seen in the first 5 minutes, which is then followed by a rapid elimination of the radioligand via the kidneys. Already after 30 minutes, a clear contrast is visible that allows the identification of the tumor that was DTPA-R modified, while the Colchi-R labeled tumor xenograft remained invisible. Again, elimination related PET-signals were visible and somewhat more pronounced because the lack of physical activity of the mouse during anesthesia led to a slower clearance of the radioligand. At the same time, the signal of the radioligand within the tumor remained constant over the whole 90 minutes of the dynamic PET-scan, which is caused by the high affinity of the Anticalin for the DTPA ligand-moiety that efficiently limits the dissociation and subsequent elimination of the radioligand. In addition, it must be pointed out, that the radioligand only accumulated in the vital border regions of the xenograft tumor, while the core region containing secrete and necrotic cells showed no PET-signal above background level (FIG. 16C-D). This clearly proves that the signal obtained from a .sup.18F-DTPA scan is limited to living cells and, thus, no false-positive signal from dead cells is measured. Finally, it was confirmed that the reporter gene can also be delivered to host cells via a genetic vector. For this purpose plasmids were constructed featuring inverted terminal repeats (ITRs) of the adeno-associated virus serotype 2 and respective expression cassette for two reporter proteins of the invention (FIG. 17A). These plasmids where then subsequently used to produce AAV viral vector of the serotype 9 encoding the reporter proteins. In an in vitro experiment different titers of these viral preparations were used to transduce HEK cells and resulting fluorescence intensities as well as the fraction of the DTPA-R-miRFP720 or Colchi-R-miRFP720 positive cells was determined (FIG. 17B). These viral vectors were also used in in vivo experiments in nude mice, which received an i.v. injection of the viral vectors. The expression of the Colchi-R after viral transduction events in different cells within the myocardial muscle of the mouse was detected via IHC staining. For the DTPA-R, also a PET-scan was used to detect viral transduction events in a quantitative and sensitive manner. This PET-scan resulted in the detection of a pronounced signal within the heart of the mouse but also muscles in the neck and dorsal region where clearly visible. Thus, it can be claimed that the reporter proteins of the invention are suitable for the quantification of AAV transduction and transgene expression in a longitudinal in vivo study.

EXAMPLE 7

Advantage of Using Anticalins Instead of Other Binding Proteins

[0194] As explained herein above in Example 1, a murine antibody scFv fragment (2D12.5/G54C) was previously employed as a reporter protein for PET imaging (Krebs et al., 2018). This scFv fragment, binds DOTA:metal complexes and was derived from the murine monoclonal antibody 2D12.5/G54C—It was used to construct the synthetic reporter gene DabR1 (DOTA Antibody Reporter 1) (Krebs et al., 2018). Based on the Antibody 2D12.5, also scFv fragments with improved affinity were developed and proposed as binding proteins which can be used as part of a membrane-anchored reporter protein (improved version of DAbR1). The corresponding muC825 scFv as well as its humanized version, huC825 scFv, were described for example in EP3256164B1 and WO 2019/060750A2. To best knowledge of the inventors, the scFv huC825 represents the most advanced and most appropriate candidate of an antibody-derived binding protein for the construction of a PET reporter gene according to the art (Dacek et al., 2021)(Krebs et al., 2018). Due to the increased affinity of C825 for DOTA:metal complexes, the complicated formation of a covalent bond between the engineered Cys residue G54C and the acrylamidobenzyl group as part of the radioligand (Krebs et al., 2018) was no longer necessary. The free Cys residue in the oxidizing extracellular milieu is prone to chemical modification while the chemical cross-reactivity of the acrylamidobenzyl group leads to undesired background signals. Apart from that, the humanization can be expected to reduce the potential immunogenicity of the murine protein in immunocompetent patients, which is of high importance for clinical translation (Dacek et al., 2021).

[0195] Both the murine and the humanized scFv C825 were compared with the Anticalin-based reporter protein according to this invention side by side. To this end, three membrane-anchored reporter proteins were constructed which only differ in the binding protein domain specific for the radioactive chelator:metal complex. These were (a) the DTPA-binding Anticalin CL31d (SEQ ID NO 23/24, FIG. 18A), (b) the murine C825 scFv (SEQ ID NO: 30, FIG. 18B) and (c) the humanized C825 scFv (SEQ ID NO: 31, FIG. 18C).

[0196] The amino acid sequences for the murine and human scFv, including the signal peptide (MGWSCIILFLVATATG; SEQ ID NO: 32), were taken from WO 2019/060750A2 and codon-optimized for Homo sapiens. The synthesized genes were cloned using Pacl and Clal restriction sites on pMP71-Lcn2_SignalP-CL31d-V5-V5-CD4TMD, that is the plasm id used for the retroviral transduction with Lcn2_SignalP-CL31d-V5-CD4TMD. Sanger sequencing confirmed correctness of the coding regions and, subsequently, the three plasmids were used for the production of retroviral vectors (Engels et al., 2003), followed by transduction of the human T-cell line Jurkat. The amount of virus was chosen to achieve a low transduction rate (20-33%) in order to avoid events of multiple transduction and genomic integration.

[0197] Transduced Jurkat cells were then stained with the AlexaFluor488-conjugated V5-binding IgG antibody SV5-PK1. Surface density of the reporter constructs in the three stably transduced cell lines was analyzed under mutually identical conditions by detection of the V5-tag via flow cytometry in two independent experiments (FIG. 18D-G). The histogram of the fluorescence signals is shown for all three constructs (FIG. 18D), demonstrating a huge difference in the surface expression levels as assessed from the Median Fluorescence Intensity (MFI). Both scFv-based reporter proteins showed only a very low expression (MFI=1'134 and 1'078 for murine and humanized scFv (C825), respectively). In contrast, the Anticalin-based reporter protein exhibited a very high expression level (MFI=6'559 for CL31d). This corresponds to a 5.8/6.1-fold difference in the MFI between the Anticalin-based reporter construct according to this invention and the corresponding constructs employing scFv binding moieties as known in the art.

[0198] Furthermore, the MFI values were converted into absolute numbers of detected fluorescent dye molecules (as described in the legend of FIG. 9) to assess the absolute number of receptors per cell (FIG. 18E). While mu_scFv(C825) revealed a median of 31'600 and hu_scFv (C825) revealed a median of 28'600 receptors per cell, the Anticalin-based reporter construct resulted in a median of 260'700 receptors per cell. This corresponds to a 8.2/9.1-fold higher cell surface expression level for the Anticalin-based reporter protein design according to this invention. This strong effect is somewhat reduced after fluorescence activated cell sorting (FACS) of the 10% highest expressing clones (FIG. 18F), after which the muC825 cell line had 292'000 receptors, huC825 expressed 237'000 receptors, while CL31d displaying cells showed an expression of 886'000 receptors per cell (FIG. 18G). While the difference was reduced to a 3.0/3.7-fold higher expression for the Anticalin-based design after this more stringent selection process, the use of the Anticalin-based design still offers a significant advantage over reporter gene systems known in the art. As all three mRNAs were expressed from the same promoter, with the same 5′ UTR, and the three reporter constructs that were compared here feature the same membrane anchor domain, the difference in expression levels must be attributed to the inferior properties of the antibody-derived scFvs. The use of the (Lcn2) Anticalin scaffold, which naturally constitutes a single polypeptide chain with stable fold that is not prone to aggregation, offers a clear improvement over scFv fragments, for which poor folding efficiency and a tendency towards oligomer formation and aggregation is known in the art. In fact, the ˜9-fold higher expression level on the cell surface, as demonstrated in this example, should lead to the binding of a ˜9-fold larger number of radioligands and allow for the much more sensitive detection of transduced or transfected cells in PET imaging studies.

REFERENCES

[0199] BAJAR, B. T., WANG, E. S., LAM, A. J., KIM, B. B., JACOBS, C. L., HOWE, E. S., DAVIDSON, M. W., LIN, M. Z. & CHU, J. 2016. Improving brightness and photostability of green and red fluorescent proteins for live cell imaging and FRET reporting. Sci Rep, 6, 20889. [0200] BARKOVSKIY, M., ILYUKHINA, E., DAUNER, M., EICHINGER, A. & SKERRA, A. 2018. An engineered lipocalin that tightly complexes the plant poison colchicine for use as antidote as well as bioanalytical applications. Biol Chem. [0201] BENESOVA, M., UMBRICHT, C. A., SCHIBLI, R. & MULLER, C. 2018. Albumin-Binding PSMA Ligands: Optimization of the Tissue Distribution Profile. Mol Pharm, 15, 934-946. [0202] BRAASCH, D. A. & COREY, D. R. 2001. Locked nucleic acid (LNA): fine-tuning the recognition of DNA and RNA. Chem Biol, 8, 1-7. [0203] BRADER, P., SERGANOVA, I. & BLASBERG, R. G. 2013. Noninvasive molecular imaging using reporter genes. J Nucl Med, 54, 167-72. [0204] DACEK, M. M., VEACH, D. R., CHEAL, S. M., CARTER, L. M., MCDEVITT, M. R., PUNZALAN, B., BURNES VARGAS, D., KUBIK, T. Z., MONETTE, S., SANTICH, B. H., YANG, G., OUERFELLI, O., KESNER, A. L., CHEUNG, N. V., SCHEINBERG, D. A., LARSON, S. M. & KREBS, S. 2021. Engineered Cells as a Test Platform for Radiohaptens in Pretargeted Imaging and Radioimmunotherapy Applications. Bioconjug Chem, 32, 649-654. [0205] DAUNER, M., EICHINGER, A., LUCKING, G., SCHERER, S. & SKERRA, A. 2018. Reprogramming Human Siderocalin To Neutralize Petrobactin, the Essential Iron Scavenger of Anthrax Bacillus. Angew Chem Int Ed Engl, 57, 14619-14623. [0206] DERENZO, C. & GOTTSCHALK, S. 2019. Genetic Modification Strategies to Enhance CAR T Cell Persistence for Patients With Solid Tumors. Front Immunol, 10, 218. [0207] DUMELIN, C. E., TRUSSEL, S., BULLER, F., TRACHSEL, E., BOOTZ, F., ZHANG, Y., MANNOCCI, L., BECK, S. C., DRUMEA-MIRANCEA, M., SEELIGER, M. W., BALTES, C., MUGGLER, T., KRANZ, F., RUDIN, M., MELKKO, S., SCHEUERMANN, J. & NERI, D. 2008. A portable albumin binder from a DNA-encoded chemical library. Angew Chem Int Ed Engl, 47, 3196-201. [0208] DUNN, C., O'DOWD, A. & RANDALL, R. E. 1999. Fine mapping of the binding sites of monoclonal antibodies raised against the Pk tag. J Immunol Methods, 224, 141-50. [0209] EGGENSTEIN, E., EICHINGER, A., KIM, H. J. & SKERRA, A. 2013. Structure-guided engineering of Anticalins with improved binding behavior and biochemical characteristics for application in radio-immuno imaging and/or therapy. J Struct Biol. [0210] ENGELS, B., CAM, H., SCHULER, T., INDRACCOLO, S., GLADOW, M., BAUM, C., BLANKENSTEIN, T. & UCKERT, W. 2003. Retroviral vectors for high-level transgene expression in T lymphocytes. Hum Gene Ther, 14, 1155-68. [0211] GEBAUER, M. & SKERRA, A. 2012. Anticalins small engineered binding proteins based on the lipocalin scaffold. Methods Enzymol, 503, 157-88. [0212] GHIM, C. M., LEE, S. K., TAKAYAMA, S. & MITCHELL, R. J. 2010. The art of reporter proteins in science: past, present and future applications. BMB Rep, 43, 451-60. [0213] GRISHIN, D., ZHDANOV, D., POKROVSKAYA, M. & SOKOLOV, N. 2020. D-amino acids in nature, agriculture and biomedicine. All Life, 13, 11-22. [0214] GRUTZKAU, A. & RADBRUCH, A. 2010. Small but mighty: how the MACS-technology based on nanosized superparamagnetic particles has helped to analyze the immune system within the last 20 years. Cytometry A, 77, 643-7. [0215] HUDSON, P. J. & KORTT, A. A. 1999. High avidity scFv multimers; diabodies and triabodies. J Immunol Methods, 231, 177-89. [0216] KIM, H. J., EICHINGER, A. & SKERRA, A. 2009. High-affinity recognition of lanthanide(III) chelate complexes by a reprogrammed human lipocalin 2. J Am Chem Soc, 131, 3565-76. [0217] KIM, J. H., LEE, S. R., LI, L. H., PARK, H. J., PARK, J. H., LEE, K. Y., KIM, M. K., SHIN, B. A. & CHOI, S. Y. 2011. High cleavage efficiency of a 2A peptide derived from porcine teschovirus-1 in human cell lines, zebrafish and mice. PLoS One, 6, e18556. [0218] KREBS, S., AHAD, A., CARTER, L., EYQUEM, J., BRAND, C., BELL, M., PONOMAREV, V., REINER, T., MEARES, C. F., GOTTSCHALK, S., SADELAIN, M., LARSON, S. M. & WEBER, W. A. 2018. Antibody with infinite affinity for in vivo tracking of genetically engineered lymphocytes. J Nucl Med, 59, 1894-1900. [0219] KUMAGAI, A., ANDO, R., MIYATAKE, H., GREIMEL, P., KOBAYASHI, T., HIRABAYASHI, Y., SHIMOGORI, T. & MIYAWAKI, A. 2013. A bilirubin-inducible fluorescent protein from eel muscle. Cell, 153, 1602-11. [0220] LABANIEH, L., MAJZNER, R. G. & MACKALL, C. L. 2018. Programming CAR-T cells to kill cancer. Nat Biomed Eng, 2, 377-391. [0221] LI, C. & SAMULSKI, R. J. 2020. Engineering adeno-associated virus vectors for gene therapy. Nat Rev Genet, 21, 255-272. [0222] LIU Z, CHEN O, WALL J B J, ZHENG M, ZHOU Y, WANG L, VASEGHI H R, QIAN L, LIU J. 2017. Systematic comparison of 2A peptides for cloning multi-genes in a polycistronic vector. Sci Rep, 7, 2193. [0223] LUNDSTROM, K. 2018. Viral Vectors in Gene Therapy. Diseases, 6. [0224] MILIOTOU, A. N. & PAPADOPOULOU, L. C. 2018. CAR T-cell Therapy: A New Era in Cancer Immunotherapy. Curr Pharm Biotechnol, 19, 5-18. [0225] MILTENYI, S. & SCHMITZ, J. 2000. High Gradient Magnetic Cell Sorting. In: RADBRUCH, A. (ed.) Flow Cytometry and Cell Sorting. Berlin, Heidelberg: Springer Berlin Heidelberg. [0226] OWENS, G. C., CHAPPELL, S. A., MAURO, V. P. & EDELMAN, G. M. 2001. Identification of two short internal ribosome entry sites selected from libraries of random oligonucleotides. Proc Natl Acad Sci USA, 98, 1471-6. [0227] PASZKIEWICZ, P. J., FRASSLE, S. P., SRIVASTAVA, S., SOMMERMEYER, D., HUDECEK,

[0228] M., DREXLER, I., SADELAIN, M., LIU, L., JENSEN, M. C., RIDDELL, S. R. & BUSCH, D. H. 2016. Targeted antibody-mediated depletion of murine CD19 CAR T cells permanently reverses B cell aplasia. J Clin Invest, 126, 4262-4272. [0229] PAULICK, M. G. & BERTOZZI, C. R. 2008. The glycosylphosphatidylinositol anchor: a complex membrane-anchoring structure for proteins. Biochemistry, 47, 6991-7000. [0230] PRICE, E. W., EDWARDS, K. J., CARNAZZA, K. E., CARLIN, S. D., ZEGLIS, B. M., ADAM, M. J., ORVIG, C. & LEWIS, J. S. 2016. A comparative evaluation of the chelators H4octapa and CHX-A″-DTPA with the therapeutic radiometal (90)Y. Nucl Med Biol, 43, 566-576. [0231] PRICE, E. W. & ORVIG, C. 2014. Matching chelators to radiometals for radiopharmaceuticals. Chem Soc Rev, 43, 260-90. [0232] RAVERA, S., REYNA-NEYRA, A., FERRANDINO, G., AMZEL, L. M. & CARRASCO, N. 2017. The Sodium/Iodide Symporter (NIS): Molecular Physiology and Preclinical and Clinical Applications. Annu Rev Physiol, 79, 261-289. [0233] RICHARZ, R., KRAPF, P., ZARRAD, F., URUSOVA, E. A., NEUMAIER, B. & ZLATOPOLSKIY, B. D. 2014. Neither azeotropic drying, nor base nor other additives: a minimalist approach to (18)F-labeling. Org Biomol Chem, 12, 8094-9. [0234] RICHTER, A., EGGENSTEIN, E. & SKERRA, A. 2014. Anticalins: exploiting a non-Ig scaffold with hypervariable loops for the engineering of binding proteins. FEBS Lett, 588, 213-8. [0235] SCHIEFNER, A. & SKERRA, A. 2015. The menagerie of human lipocalins: a natural protein scaffold for molecular recognition of physiological compounds. Acc Chem Res, 48, 976-85. [0236] SCHLEHUBER, S., BESTE, G. & SKERRA, A. 2000. A novel type of receptor protein, based on the lipocalin scaffold, with specificity for digoxigenin. J Mol Biol, 297, 1105-20. [0237] SERGANOVA, I. & BLASBERG, R. 2005. Reporter gene imaging: potential impact on therapy. Nucl Med Biol, 32, 763-80. [0238] SERGANOVA, I., MAYER-KUKUCK, P., HUANG, R. & BLASBERG, R. 2008. Molecular imaging: reporter gene imaging. Handb Exp Pharmacol, 167-223. [0239] SHAH, N. N. & FRY, T. J. 2019. Mechanisms of resistance to CAR T cell therapy. Nat Rev Clin Oncol, 16, 372-385. [0240] SHCHERBAKOVA, D. M., BALOBAN, M., EMELYANOV, A. V., BRENOWITZ, M., GUO, P. & VERKHUSHA, V. V. 2016. Bright monomeric near-infrared fluorescent proteins as tags and biosensors for multiscale imaging. Nat Commun, 7, 12405. [0241] SHCHERBAKOVA, D. M., COX CAMMER, N., HUISMAN, T. M., VERKHUSHA, V. V. & HODGSON, L. 2018. Direct multiplex imaging and optogenetics of Rho GTPases enabled by near-infrared FRET. Nat Chem Biol, 14, 591-600. [0242] SOUTHERN, J. A., YOUNG, D. F., HEANEY, F., BAUMGARTNER, W. K. & RANDALL, R. E. 1991. Identification of an epitope on the P and V proteins of simian virus 5 that distinguishes between two isolates with different biological characteristics. J Gen Virol, 72 (Pt 7), 1551-7. [0243] VARGAS, J. E., CHICAYBAM, L., STEIN, R. T., TANURI, A., DELGADO-CANEDO, A. & BONAMINO, M. H. 2016. Retroviral vectors and transposons for stable gene therapy: advances, current challenges and perspectives. J Transl Med, 14, 288. [0244] VOPEL, S., MUHLBACH, H. & SKERRA, A. 2005. Rational engineering of a fluorescein-binding anticalin for improved ligand affinity. Biol Chem, 386, 1097-104. [0245] WANG, X., CHANG, W. C., WONG, C. W., COLCHER, D., SHERMAN, M., OSTBERG, J. R., FORMAN, S. J., RIDDELL, S. R. & JENSEN, M. C. 2011. A transgene-encoded cell surface polypeptide for selection, in vivo tracking, and ablation of engineered cells. Blood, 118, 1255-63. [0246] WEI, L. H., OLAFSEN, T., RADU, C., HILDEBRANDT, I. J., MCCOY, M. R., PHELPS, M. E., MEARES, C., WU, A. M., CZERNIN, J. & WEBER, W. A. 2008. Engineered antibody fragments with infinite affinity as reporter genes for PET imaging. J Nucl Med, 49, 1828-35.