Biological characterization of novel CXCR4 and CXCR7/ACKR3 receptor inhibitors

Chemokine receptors are validated therapeutic targets that have been widely studied for their involvement in cancer growth and metastasis, inflammatory disorders and infectious diseases. Chemokine receptor CXCR4 was amongst the first discovered coreceptors facilitating cellular HIV entry, and still is of major importance in this respect. Great efforts were put into the design, synthesis and in vitro characterization of CXCR4 inhibitors able to interfere with the attachment of HIV virions to the host cell membrane. The main ligand for CXCR4 is the chemokine CXCL12. However, apart from CXCR4, CXCL12 was also shown to bind to CXCR7/ACKR3, as an alternative receptor, with very high affinity. Both CXCR4 and CXCR7 are concomitantly involved in CXCL12-mediated cellular responses, whereby CXCR7 acts as a scavenging receptor for CXCL12 to regulate local and systemic CXCL12 bioavailability and thus CXCR4 activity.

In this Ph.D. work, we first performed a side-by-side comparison of in vitro CXCR4-specific cell-based assays, including a CXCL12 competition binding assay, CXCL12-induced calcium signaling and several functional assays to study CXCL12-induced receptor internalization, cell migration, and HIV-1 replication experiments with CXCR4-using virus strains. Small molecule and peptide-based CXCR4-targeting ligands were implemented within this study to validate the screening potential of these CXCR4-related assays. From our comparative analysis, we concluded that the CXCL12 competition binding assay has high value as an initial screening platform to discover potent CXCR4 inhibitors, whereas the assays based on CXCR4 effector functions are of importance for the identification of ligand-biased receptor signaling.

In contrast to CXCR4, the chemokine receptor CXCR7 cannot signal via the canonical G protein-mediated pathway. In order to study selective CXCR7 activating molecules, we developed a G protein-independent signaling assay to detect the ligand-induced recruitment of a cytoplasmic adapter protein, called ß-arrestin 2 (ARRB2), to the intracellular part of the CXCR7 membrane protein. We successfully developed an ARRB2 recruitment assay based on the NanoBiT complementation technology (Promega), which relies on the complementation of a split NanoLuc luminescent enzyme fused to CXCR7 and ARRB2. Recruitment of ARRB2 to CXCR7 resulted in the formation of a functional NanoLuc enzyme that generated a luminescent signal.

Next, CXCR4-targeting small molecules, derived from the AMD3100 bicyclam structure, were synthesized in collaboration with the group of Prof. Steve J. Archibald from the University of Hull (UK). Structural changes to the original AMD3100 bicyclam structure involved adaptations to the ring size of the cyclams, metal complexation and introduction of side/cross-bridges within each of the cyclam units. These newly synthesized bicyclams displayed high affinity CXCR4 binding and an improved interaction profile when evaluated in the above-mentioned CXCR4-specific binding and activity assays. In parallel, macrocyclic small molecules structurally derived from M40403, a validated superoxide dismutase (SOD) mimetic, were synthesized in collaboration with the group of Prof. Thomas W. Bell from the University of Nevada (USA). Evaluation of the M40403 derivatives (SH compounds) revealed an interesting dual activity for a selection of SH molecules with both CXCR4 inhibitory and SOD mimetic potency and suggested potential for the treatment of rheumatoid arthritis.

Lastly, two studies were carried out in the context of unraveling binding modes of CXCR4 and CXCR7 ligands to their cognate receptors for future development of new chemokine receptor inhibitors with either diagnostic or therapeutic potential in pathological conditions involving CXCR4 and/or CXCR7. In one study, in collaboration with the group of Dr. Andy Chevigné (LIH, Luxemburg), peptides derived from the N-terminal region of the chemokines CXCL12, CXCL11 and vCCL2 were utilized to compare binding and activation modes to CXCR4 and CXCR7, as well as the canonical CXCL11 receptor CXCR3. Our analysis showed that, in contrast to binding and activation of CXCR4 and CXCR3, neither the most proximal residues nor the N-loop are essential for binding and activation of CXCR7. These results imply a distinct ligand interaction mechanism for CXCR7 and demonstrate its strong tendency towards receptor activation. In a second study, we investigated the role of CXCR7 as coreceptor in HIV entry. To analyse CXCR7 coreceptor activity, we introduced different classes of CXCR7-targeting ligands (i.e. small molecules, monoclonal antibodies and chemokines) in our cell based HIV study. Overall, the CXCR7 ligands could all interfere with HIV entry of CXCR7-using virus, confirming previous studies on the involvement of CXCR7 in HIV replication. Nonetheless, differential activity was observed (1) between different ligand classes, which could mostly be explained by variation in molecular size and thus also in the amount of residues that participate in receptor occupancy, and (2) within each ligand class, most likely attributed to differential receptor surface binding of amino acid residues, which do or do not coincide with HIV gp120 interaction residues.