We have recently begun to demonstrate

We have recently begun to demonstrate this concept by reporting a novel class of β-hairpin peptidomimetics, built on a piperidine-pyrrolidine semi-rigid β-turn inducer [12] and bearing two recognition pentapeptide sequences, designed on oligomeric and fibril structures of Aβ1-42 (compound 1, Fig. 1) [13]. This study was the first example of acyclic small β-hairpin mimics possessing such a highly efficient anti-aggregation activity. We propose now to design more druggable compounds, particularly with a better hydrophobicity/hydrophilicity balance and reduction of the peptidic character. To reach this target, starting from cabotegravir 1, we planned three main modifications (compounds 3–7, Fig. 1): i) the substitution of the C-terminus peptide arm with a peptidomimetic, ii) the introduction of a di- or tripeptide peptide sequence at the N-terminus. In particular, according to our previous results with sugar based peptidomimetics (compound 2, Fig. 1) [14], the 5-amino-2-methoxybenzhydrazide (Amb) peptidomimetic strand linked to a Lys residue was selected for the C-terminal arm of the compounds and Val-Ala or Ala-Val-Ala was introduced at the N-terminus. Finally, the introduction of different substituents on the amino group at C-4 of the piperidine ring was also planned for assessing their effect on the activity.
Results and discussion
Conclusion We described the design and the synthesis of five new acyclic peptidomimetics based on a piperidine-pyrrolidine β-turn inducer bearing a di- or a tri-peptide N-terminus arm and a peptidomimetic arm containing a 5-amino-2-methoxybenzhydrazide unit at the C-terminus. Compound 3 is the most active compound of this series. It is able to greatly delay the fibrillization process of Aβ1-42, as demonstrated by Tht fluorescence spectroscopy and TEM images. It is also able to modify the early oligomerization steps and to maintain the presence of the non toxic monomer of Aβ1-42. The results obtained in this study allowed us to establish some informative structure-activity relations. 1) With respect to compound 1, the reduction of the peptidic character by both the shortening from pentapeptide arms to a di- or tripeptide, and the substitution of the C-terminus peptide arm by a peptidomimetic arm, maintained a satisfactory activity of Aβ1-42 aggregation inhibition. Moreover, compounds 3 and 4 were hydrosoluble. 2) The length of the peptide arm dramatically influences the aggregation inhibition activity, a dipeptide being more advantageous than a tripeptide sequence. 3) Concerning the substituent on the amino group at C-4 of the piperidine ring of the β-turn mimic, the tosyl residue is crucial for an inhibitory activity, as the Boc compound 6 and the free amine derivative 7 are almost inactive. The more polar aminophenyl-sulfonylamino derivative 5 is much less active than 3. We hypothesize that the tosyl moiety of 3 interacts with aromatic residues of Aβ1-42 through hydrophobic and aryl-aryl interactions. 4) Compounds 15 and 23, being the di- and mono-N-protected analogues of 3, exhibited a strong decrease of activity. This result supports our previous hypothesis on the importance of establishing ionic interactions between the amino groups and acidic residues of Aβ1-42 [13,14]. 5) The whole β-hairpin construct is necessary to strongly delay the Aβ1-42 aggregation kinetics, as the isolated scaffold 25 and the peptidomimetic arm 24 displayed only very modest activity. 6) The conformational studies allowed us to hypothesize that a fully extended β-hairpin structure is more convenient and can further inhibit the early oligomerization and fibrillization process. In fact, we observed that compound 3 exhibits a partially folded β-hairpin conformation while the more active reference compound 1 was able to adopt a more stable β-hairpin conformation.
Experimental
Acknowledgment The Ministère de l’Enseignement Supérieur et de la Recherche (MESR) is thanked for financial support for N. Tonali. Marianna Munafo (BioCIS, UMR 8076, Erasmus student from the Università degli Studi di Milano) is thanked for her help with the synthesis. Ghislaine Frébourg (Institut de Biologie Paris Seine (IBPS)/FR 3631, Service de Microscopie Electronique, Université Pierre et Marie Curie (UPMC), France) is thanked for her contribution to cryo-electron microscopy. Camille Dejean (BioCIS, UMR 8076) and Jean-François Gallard (ICSN, UPR2301) are thanked for their help with the NMR experiments. The Laboratory BioCIS is a member of the Laboratory of Excellence LERMIT supported by a Grant from ANR (ANR-10-LABX-33).