Considering the conformational preferences of the SRSRY sequ

Considering the conformational preferences of the SRSRY sequence in the x-ray structures of soluble uPAR (SuPAR) [18], [19], [20], [50], [51], it is included in a quite flexible linker delimited by Cys76 and Cys95 residues. The SRSRY sequence adopts either α-turned or β-extended conformation. In any case, this sequence in the SuPAR is not superimposable to our antagonist structures confirming that the agonist to antagonist switch is conformationally driven [12], [25]. In Fig. 9b, as an example, the superposition of the NMR structure of peptide 1 and the segment 88–92 of a soluble uPAR crystal structure (pdb code 3BT2) [50] is shown and neither the backbone nor the side chains of the corresponding residues can be efficiently overlapped. Decrease of the anti-migratory activity of peptide 3 [SRSPRY], compared to the unphosphorylated parent 1, can be tentatively ascribed to the lower conformational stability of the first, as suggested by the CD analysis (Fig. 4). In addition to the replacements of Ser90 with Glu or Ser(P) residue, also Ser88 was replaced with a negatively charged residue both in the linear and in the cyclic peptides. Interestingly, an inhibitory effect is observed in the linear peptide 4 upon the Ser88 to Ser(P) replacement, which means that also the mutation of the first residue can switch the activity towards the anti-migratory effect. Similar Ser88 to Glu mutation in the cyclic sequence (analog 7) decreased the anti-migratory effect when compared to the lead peptide 1. The same replacement of Ser90 with Glu in the cyclic scaffold only slightly decreased the potency (analog 6vs1). Coherently, peptide 8 (Glu88 and Glu90 derivative) showed an improved activity compared to the analog 7. CD analysis indicates that Ser/Glu replacement has minor effect on the peptide conformation (Fig. S1). In particular, peptide 6 CD spectrum is almost superimposable with that of 1 while some differences are evident in the spectrum of 7. Thus, even minor conformational modification in the cyclic peptides can importantly affect the anti-migration effect. Finally, the last residue Tyr92 was also investigated by replacing it with different aromatic Ginkgolic Acid C15:1 sale (Table 1). First, Tyr92 was replaced by a Phe residue (peptide 9) to mimic the potent inhibitor RERF (AcRERFNH2) which we have previously found to inhibit cell migration [27]. Differently from the linear RERF, peptide 9 was significantly less potent than the parent peptides 1 and 6 pointing to a detrimental effect of the Tyr92 to Phe mutation. Furthermore, bulkier aromatic side chains such as Trp (10), Nal1 (11), and Nal2 (12) were considered in our study. Among those, Nal1 derivative (11) gave the same anti-migratory activity as the parent peptide 6 while peptides 10 and 12 failed or were low effective, respectively, in blocking cell migration. Since the presence of the uncoded and highly hydrophobic Nal1 residue can improve both the stability and the binding to the blood proteins of the peptide, analog 11 deserves further investigations. Although a detailed conformational analysis of Tyr92 mutated derivatives is required, it is clear from CD spectra that the preferred conformations of these peptides are different from both those of the parent peptides 1 and 6 and each other (Fig. S2), pointing to secondary structure modification as a cause of the observed different inhibitory effects on cell migration. To gain further insight into the possible binding modes of the developed peptides, a docking study was performed considering peptide 1 and the inactive analog 10. Docking results are shown in Fig. 6, Fig. 7 and Table 2. As shown, the two peptides share similar positioning within the receptor but their orientation is different. Peptide 1 has the Arg91 and Tyr92 side chains deeply inserted within the helical bundle while the corresponding side chains, Arg91 and Trp92, of peptide 10 point towards the extracellular side. Consequently, the binding energies (Table 2) indicated that FPR1/peptide 1 complex is more stable. The different observed binding modes are probably due to steric bumps of the Trp indole moiety, which hamper peptide 10 to take the same pose as peptide 1. Following MD simulations performed on the FPR1/peptide 1 complex indicate that Tyr92 side chain reorients itself after the binding thereby occupying a hydrophobic pocket formed by the residues evidenced in Fig. 8. Interestingly, a hydrogen bond between the phenolic hydroxyl group of Tyr92 and the guanidinum group of Arg84 is found in this new binding arrangement. This hydrogen bond cannot be formed if Tyr92 is replaced by the other aromatic residues of our library. The latter observation can be regarded as a further reason for the observed reduction/abolition of the activity of different analogs reported in Table 1.