Structural Basis of the Binding Mode of the Antineoplastic Compound Motixafortide (BL-8040) in the CXCR4 Chemokine Receptor

Abstract

Modulation of the CXCL12-CXCR4 signaling axis is of the utmost importance due to its central involvement in several pathological disorders, including inflammatory diseases and cancer. Among the different currently available drugs that inhibit CXCR4 activation, motixafortide-a best-in-class antagonist of this GPCR receptor-has exhibited promising results in preclinical studies of pancreatic, breast, and lung cancers. However, detailed information on the interaction mechanism of motixafortide is still lacking. Here, we characterize the motixafortide/CXCR4 and CXCL12/CXCR4 protein complexes by using computational techniques including unbiased all-atom molecular dynamics simulations. Our microsecond-long simulations of the protein systems indicate that the agonist triggers changes associated with active-like GPCR conformations, while the antagonist favors inactive conformations of CXCR4. Detailed ligand-protein analysis indicates the importance of motixafortide's six cationic residues, all of which established charge-charge interactions with acidic CXCR4 residues. Furthermore, two synthetic bulky chemical moieties of motixafortide work in tandem to restrict the conformations of important residues associated with CXCR4 activation. Our results not only elucidate the molecular mechanism by which motixafortide interacts with the CXCR4 receptor and stabilizes its inactive states, but also provide essential information to rationally design CXCR4 inhibitors that preserve the outstanding pharmacological features of motixafortide.

Keywords: CXCR4; MD simulations; antineoplastic drugs; cancer; computational methods; motixafortide.

Figure 1

Protein complexes investigated: The tertiary structure of the CXCR4 receptor was taken from the X-ray structure with the PDB accession code 4RWS.pdb. A segment at the N-terminus was modelled as an extended structure (central panel). The same CXCR4 structure was utilized to generate the two protein complexes with either CXCL12 or motixafortide, by placing the ligands at the CXCR4 orthosteric binding site as described in the Materials and Methods section. The systems were embedded in a hydrated lipid bilayer and investigated by 1.0-microsecond-long unbiased MD simulations. A representative final structure of the motixafortide/CXCR4 complex is depicted in the left panel, where the CXCR4 structure is colored grey, while motixafortide is depicted as green spheres. In the right panel, an equivalent final structure of the CXCL12/CXCR4 complex is shown, where the CXCR4 receptor is colored teal, while the chemokine is presented as spheres using the same color code as in Figure S1.

Figure 2

Distinct CXCR4 conformations of the PIF motif: A superimposed representation of the positions adopted by the F248^6.44 aromatic residue. For the sake of clarity, the conformations adopted by the sidechain of the F248^6.44 residue during the last 200 ns are represented by the position of the most external carbon in the phenyl ring—that is, the ‘CZ’ atom. The CXCR4 structure in the motixafortide/CXCR4 complex is depicted in grey, while that of the CXCL12/CXCR4 complex is depicted in teal. Even though the initial structure of the CXCR4 receptor in both protein complexes is the same, the presence of either the agonist (CXCL12) or the antagonist (motixafortide) ligand elicits distinct conformations on this important structural motif, associated with GPCR activation. These conformations are consistent with the expected rotameric states adopted by the F248^6.44 residue (see also Figure S4).

Figure 3

Conformations at the NPxxY and toggle switch motifs: (a) Representative structures of the final stages of the MD simulation for the motixafortide/CXCR4 and CXCL12/CXCR4 protein complexes. Residues that are part of the highly conserved NPxxY motif, which has been implicated in GPCR activation, are indicated (N298^7.49, P299^7.50, and Y302^7.53); TM6 is not depicted in the figure. The position of residue Y302^7.53 changes significantly in the agonist-bound CXCR4 complex (colored teal), while in the antagonist-bound system (grey) it remains very similar to the starting conformation (see magenta arrow). (b) Comparison of the NPxxY motifs of the two CXCR4 complexes investigated here with the inactive structure of CXCR4 (4RWS.pdb) and the active-like structure of the viral chemokine receptor US28 (4XT3.pdb). While the sidechain position of residue Y302^7.53 in the motixafortide/CXCR4 complex resembles that of the inactive CXCR4 structure, the sidechain in CXCL12/CXCR4 moves towards the interior of the helical bundles, as in the case of the active-like GPCR structure. (c) Superposition of the agonist-bound system (CXCL12/CXCR4) with the active structures of the closely related CCR5 (7F1S.pdb) and CCR2 (7XA3.pdb) chemokine receptors, showing the resemblance in the conformation of Y302^7.53 from the highly conserved NPxxY motif. (d) The dihedral angle defined by the N-CA-CB-CG atoms from the toggle switch residue, W252^6.48, are shown for the antagonist-bound and agonist-bound CXCR4 systems.

Figure 4

Interactions in the motixafortide/CXCR4 protein complex: (a) Representative structure of the final stages from the atomistic MD simulation of the motixafortide/CXCR4 complex, where we can observe that all of the cationic residues of motixafortide are involved in charge–charge interactions with residues in the ligand-binding site of CXCR4 (see purple dashed lines). For the sake of clarity, TM6 is not displayed, and the sequence of motixafortide is depicted at the top of the panel, where its six basic residues (R1, R2, K7, K8, R11, and R14) are highlighted. (b) The time evolution plots of distances involving the interactions in the six basic residues from motixafortide (the moving average is colored black, while the primary data are colored gray). (c) Schematic representation of the charge–charge interactions that stabilize the molecular pose of motixafortide in the orthosteric binding site of CXCR4 (see also Video S1).

Figure 5

Interactions of the unnatural moieties of motixafortide: The two synthetic components in motixafortide are a 4-fluorobenzoyl group as an N-terminal capping group and a naphthalene derivate (NPA) sidechain at position 3 (see also Figure S7). The results from the unbiased MD simulations placed the two chemical moieties at different locations inside the orthosteric binding site (lateral view of the central panel). As shown in the left panel (extracellular view), the 4-fluorobenzoyl forms aromatic interactions with residues W94^2.60 and Y116^3.32. The presence of the 4-fluorobenzoyl group restricts the conformation of this bulky residue, facilitating the interaction of the hydroxyl group of the Y116^3.32 sidechain with E288^7.39 (purple dashed line). As shown in the right panel (extracellular view), the NPA sidechain forms aromatic interactions with H203^5.42, and it is placed in the vicinity of residues in TM6—particularly Y255^6.51.

Figure 6

Conformations of bulky residues at the orthosteric binding site linked to CXCR4 activation: (a) Motixafortide/CXCR4 superimposed conformations of residues important for ligand binding and activation at the CXCR4 orthosteric ligand binding site. The structures were taken every 50 ns for the last 500 ns of the trajectory (500 to 1000 ns) and are depicted according to the gradient color code on the right panel. (b) Similar representation as in panel (a), but for the CXCL12/CXCR4 system. In panels (a,b) we can observe more dynamic behavior in the agonist-bound system relative to the antagonist-bound CXCR4 receptor for particular residues that have been implicated in ligand binding and CXCR4 activation. (c) Representative structure of the motixafortide/CXCR4 complex, where the two synthetic chemical moieties at positions 1 and 3 and the CXCR4 residues in their close proximity are depicted as spheres. The polar interaction of residue Y116^3.32 with E288^7.39 is highlighted using a purple dashed line. (d) Similar orientation as in panel (c), but for the CXCL12/CXCR4 protein complex. Here, the polar interaction of residue Y116^3.32-E288^7.39 is replaced by the Y116^3.32-Y255^6.51 polar interaction (purple dashed line). In this case, the positively charged moiety in the K1^CXCL12 backbone establishes a charge–charge interaction with E288^7.39.