Conjecturing about Small-Molecule Agonists and Antagonists of α4β1 Integrin: From Mechanistic Insight to Potential Therapeutic Applications

Abstract

Integrins are heterodimeric cell-surface receptors that regulate cell-cell adhesion and cellular functions through bidirectional signaling. On the other hand, anomalous trafficking of integrins is also implicated in severe pathologies as cancer, thrombosis, inflammation, allergies, and multiple sclerosis. For this reason, they are attractive candidates as drug targets. However, despite promising preclinical data, several anti-integrin drugs failed in late-stage clinical trials for chronic indications, with paradoxical side effects. One possible reason is that, at low concentration, ligands proposed as antagonists may also act as partial agonists. Hence, the comprehension of the specific structural features for ligands' agonism or antagonism is currently of the utmost interest. For α4β1 integrin, the situation is particularly obscure because neither the crystallographic nor the cryo-EM structures are known. In addition, very few potent and selective agonists are available for investigating the mechanism at the basis of the receptor activation. In this account, we discuss the physiological role of α4β1 integrin and the related pathologies, and review the few agonists. Finally, we speculate on plausible models to explain agonism vs. antagonism by comparison with RGD-binding integrins and by analysis of computational simulations performed with homology or hybrid receptor structures.

Keywords: agonist; crystal structure; inflammation; molecular docking; α4β1 integrin.

Figure 1

(A) Closed structure of full αIIbβ3 integrin (8T2V) [8]; densities are not visible for the αIIb and β3 short cytoplasmic tails. (B) Open-extended conformation of αIIbβ3 ectodomain (four domains) in the presence of LGGAKQRGDV (2VDR) [9]; blue arrows indicate the directions of the main displacements underwent by elements of the βI- and hybrid domains during conformational reorganization; PSI, plexin-semaphorin-integrin. No density is detected for thigh, leg, and TM domains. (C) Inset of carboxylate binding site in the β-subunit and the metal ion-dependent adhesion sites (green spheres) (2VDR).

Figure 2

Simplified representation of leukocyte adhesion cascade involved in immune cell infiltration. The process consists of leukocyte rolling, capture, arrest (firm adhesion), and transmigration (diapedesys) from the circulatory system to the infected site. This cascade is mediated by various molecules: selectins and their ligands (PSGL), integrins and the cell adhesion molecules (CAM). Extravasation is driven by junctional adhesion molecules (JAM) and platelet endothelial cell adhesion molecules (PECAM).

Figure 3

Structures of α4β1 integrin antagonists discussed in this paper: (A) N-acylphenylalanine derivatives RO0505376, firategrast, carotegrast-methyl; LDV peptide BIO1211; cyclic peptide 1 containing N-terminal Tyr; peptidomimetics containing a β-residue: DS-70, 2, and 3; (B) peptidomimetics Ds13g (partial agonist) and Ds13d (pure antagonist).

Figure 4

(A) Fluorescent zeolite L microcrystal biofunctionalized with an integrin ligand derived from 2, utilized for coating a leukocyte-responsive nanostructured surface. (B) Collagen-based scaffold functionalized with integrin ligand peptides LLP2A or LXW7; SILY (RRANAALKAGELYKSILY-NH₂) is a high-affinity collagen-binding peptide.

Figure 5

Structures of urea-based integrin agonists: THI0019, 4 and 5; c[Amp(MPUPA)LEV] (6); LDV cyclopeptides designed as mimetics of BIO1211, containing a phenylalanine-urea (Phu) residue (in red), and isoAsp (in blue), (7).

Figure 6

Simplified sketch of the binding site of wild-type FN/αvβ3 (4MMX) and of the binding site of αvβ3 integrin hosting the 10th type III domain of mutant FN (4MMZ), showing the stacking between Trp1496 of FN and Tyr122 of the b-subunit which seems to prevent the dislocation of β1-α1-loop and subsequent receptor extension. Red spheres represent water molecules, grey spheres represent divalent cations.

Figure 7

Sketch of the binding site of RO0505376 in α4β7integrin/Fab ACT-1 (3V4V). Red spheres represent water molecules, grey spheres represent divalent cations.

Figure 8

Sketches of the metal-ion-dependent adhesion sites for the complexes α5β1 with a linear RGD peptide (4WK0), and with an RGD cyclopeptide (4WK4). The latter differs particularly in the repositioning of Ser134 to displace water1 (W1) at MIDAS. Red spheres represent water molecules, grey spheres represent divalent cations.

Figure 9

Simplified sketches of the MIDAS and ADMIDAS region of complexes αIIbβ3/UR-2922 and αIIbβ3/tirofiban. The “closing” antagonist UR-2922 stabilizes water1 between the MIDAS metal ion, the nitrogen N2 of the 1H-pyrazole ring, and Ser123 sidechain. With the opening ligand tirofiban, β1-α1 loop movement brings Ser123 to displaced water1 (W1) to coordinate directly to the MIDAS. Red spheres represent water molecules, grey spheres represent divalent cations.

Figure 10

(A) Ribbon image of the chimeric α4β1 model from Ref. [41], and (B) details of the binding site rendered by the solvent accessible surface, showing subpockets A–E, U, L; the positions of relevant residues are also shown.

Figure 11

Comparison of the predicted binding poses obtained by molecular docking for the MPUPA antagonists DS-70 and 2 [41], 3, [57], BIO1211 as reported in Ref. [94], and the agonist THI001 [54]. A–E, U, L refer to binding sites subpockets as shown in Figure 10.

Figure 12

Predicted complexes between composite α4β7/α5β1 receptor and the agonist 7a, c[Phu-LDV-isoAsp] (left), and the antagonist 7c, c[(R)-Phu-LDV-isoAsp] (right) [58]. A–E, U, L refer to binding sites subpockets as shown in Figure 10. Red spheres represent water molecules, grey spheres represent divalent cations.