PET Imaging of CXCR4 Receptors in Cancer by a New Optimized Ligand

CXCR4 Imaging: Based on a high-affinity CXCR4 ligand, an imaging agent for CXCR4-positive tumors was developed through structure–activity relationship studies. The best compound was evaluated in vivo and shown to have excellent properties as a positron emission tomography (PET) tracer.

the cellular entry of the HIV, many peptidic and nonpeptidic ligands with different modes of antagonistic activity have been developed. [10][11][12][13][14][15][16][17][18] These highly CXCR4-specific agents can serve for the introduction of PET-active prosthetic groups. This approach is often complicated by loss of binding affinity, undesired alteration of biodistribution and instability in vivo. [19,20] A careful optimization of many molecular parameters is necessary to develop a suitable tracer for diagnostic application.
As a starting point for the development of the first 68 Ga-labeled, CXCR4-directed PET probe, we used cyclic pentapeptide 1 a (Figure 1) developed by Fujii et al. and the later published analogue 1 b, as it is an inverse agonist of CXCR4. [21][22][23] Small, cyclic peptides such as these should exhibit high in vivo stability towards enzymatic degradation, especially as they contain d-amino acids and N-methylated peptide bonds. [6] Although allowing first positive imaging experiments, radioiodination of the tyrosine residue increased the lipophilicity and turned out to be unsuccessful for our purpose. Consequently, we investigated the introduction of more hydrophilic groups and focused on the (radio)metal chelator 1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid (DOTA) because it can be used in combination with the corresponding radiometals for different imaging techniques like PET (e.g., 68 Ga 3 + ), single photon emission tomography (SPECT; e.g., 111 In 3 + ), or magnetic resonance imaging (MRI; e.g., Gd 3 + , Fe 3 + ) and also for radionuclide therapy (e.g., 177 Lu 3 + , 90 Y 3 + ).
Previous studies of our group and others have shown that all side chains of peptides 1 a and 1 b contribute to binding affinity. An attempt to remove the side chain of Arg 3 to introduce anchoring functions in this position resulted in a total loss of activity, whereas substitution of Arg 2 by ornithine (Orn) and its acylated derivatives gave a reduction of only one order of magnitude. Unfortunately, introduction of larger acyl or alkyl substituents on Orn 2 also strongly reduced the affinity (for details see Supporting Information). [24] Unexpectedly, ligands with benzoic acids attached to the Orn 2 side chain retained most   of the CXCR4 binding activity for example, 1 d ( Figure 1 and Table 1). [25] An important affinity improvement was achieved by starting from peptide 1 b, which differs from 1 a by an d-arginine residue instead of l-arginine in position 2 and simultaneous Nmethylation of the peptide bond between d-Tyr 1-d-Arg 2. We chose DOTA as a complexing moiety as its cyclen scaffold is also found in the CXCR4 drug AMD3100 (Mozobil TM ), and we hypothesized that we could gain receptor affinity as chelates of AMD3100 have shown to have superior affinity. [26] To attach DOTA, the d-arginine group was again substituted by d-Orn. The type and length of spacer between the peptide and DOTA was optimized in more than 25 compounds (see Supporting Information) to yield the highest affinity compounds 2 a-c ( Figure 2 and Table 1). Receptor affinities also depend on the chelation state as well as the type and radii of the metal ion in the complexing moiety. [27][28][29] Therefore, we tested gallium and indium compounds as they are relevant for imaging purposes and have different ionic radii. While the binding affinities for the free DOTA compound 2 a and its indium chelate 2 b are 150 nm and 44 AE 4 nm, respectively, the gallium complex exhibited an affinity of 5 AE 1 nm, which is virtually identical with the unmodified peptides 1 a and 1 b.
In vivo testing of [ 68 Ga]2 c was carried out in nude mice bearing OH-1 human small-cell lung cancer xenografts. The 68 Ga-labeled ligand accumulated in high levels in CXCR4-expressing tumors and allowed for a high contrasting functional imaging of the CXCR4 receptor status in vivo ( Figure 3). Co-injection of 50 mg cyclo-(-d-Tyr 1-Arg 2-Arg 3-Nal 4-Gly 5) or AMD3100 (data not shown) per mouse significantly reduced the tumor uptake, thus demonstrating specificity of CXCR4mediated tumor binding or [ 68 Ga]2 c.
Quantitative biodistribution data 60 and 120 min post-injection of [ 68 Ga]2 c alone or in the presence of 50 mg competitor are summarized in Table 2. High tumor to organ ratios were observed already 1 h post-injection. Furthermore, the results    www.chemmedchem.org from the classical biodistribution study confirm distribution and specificity of tracer accumulation as observed by the PET imaging study.
[ 68 Ga]2 c is the first 68 Ga-labeled CXCR4 imaging probe and shows excellent in vivo distribution and binding characteristics. The overexpression of CXCR4 in a variety of tumors and its role in organ-specific metastasis recommend the further clinical evaluation of [ 68 Ga]2 c. This study paves the way for molecular imaging of this important GPCR in animals and man to enable personalized medicine and individualized treatment. Furthermore, chelates of 2 a with therapeutic nuclides are an obvious choice for possible future endo-radiotherapeutic approaches.