atriflAMP = 3-trifluoroacetylamino-5-methylpyrazole atriflAMP = 3-trifluoroacetylamino-5-methylpyrazole ple we obtained a clear 1:1 stoichiometry, assured by means of a Job plot . NMR titrations in CDCl3 furnished association constants below 100 (Ac-Ala-Ala-OMe/Ac-Phe-Ala-OMe) . These relatively small numbers are explained by the low propensity of alanine-containing peptides to adopt the ¿-sheet conformation and were also found with the best binders developed by us earlier for dipeptides (Table 2.4.1).
The recognition module 3 with an acetyl group for polar side-chains furnished Ka values of 80 m_1 and 900 m_1 for Ac-Ala-Ala-OMe and Ac-Ser-Ala-OMe, respectively. Thus, a more than tenfold increase was observed when the additional hydrogen bond was formed between the hydroxymethyl group of serine and the acetyl group of the receptor. The reference compound 3 formed a 2:1 complex with the serine-containing dipeptide whose association constant for each step could be estimated at 180 m_1. Obviously, the introduction of an additional hydrogen-bond acceptor in the rigid framework of the receptor module resulted in a substantial increase in free binding enthalpy for a serine-containing dipeptide, but not for dialanine nor for the reference compound 3.
With our nitroarene tip in host 5, we tried to establish additional p-p interactions with electron-rich residues in aromatic amino acids. To this end, we titrated Ac-Phe-Ala-OMe first with reference compound 3 and then with host 5 and obtained 1:1-association constants of <40 m_1 and 350 m_1, respectively. Again, a ninefold increase from the steric repulsion by the reference compound and the p-stacking attraction by host 5 was found. Substantial upfield shifts in the aromatic regions of host and guest 1H NMR spectra confirmed the additional stabilizing interaction.
For electrostatic interaction with basic amino acids we investigated complex formation between ornithine-containing Ac-Orn-Ala-OMe and anionic host 6. To avoid suppression of backbone recognition by the superior electrostatics we chose acetate as the counter-ion for the ornithine ammonium ion. Preliminary NMR experiments revealed marked shifts in both host and guest signals, especially those close to the hydrogen-bond donors and acceptors. Even the ammonium signal shifted by ~1 ppm indicating the formation of the new ion pair. A high Ka value of ~2400 m_1 demonstrated the efficient recognition process. The electrostatic interaction alone was probed with 6 and n-propylammonium acetate, and gave a bind-
ing constant of 490 m-1. The difference between these values must represent the backbone recognition operating simultaneously. The reference compound 3 gave only a relatively low number of 280 m_1 in its complex with Ac-Orn-Ala-OMe. Thus, introduction of a properly placed benzoate anion into the receptor module led to an additional ion-pair-reinforced hydrogen bond with an exemplary basic amino acid (Figure 2.4.5). Again, a selectivity of @9:1 was established for the selective complexation of a representative basic dipeptide by the new anionic host 6.
Karplus analyses of the NH-a-CH coupling constants which correlate with the characteristic torsion angle 9, offer valuable information about the conformation of the peptide . We compared the 3J values of the free peptides with those in the complex with their optimized binders. Usually signals became much sharper and the coupling constants markedly increased, approaching the calculated values (MacroModel).
In detailed comparative experiments we checked the binding behavior of our new hosts for dipeptides with the inverted sequence. Ac-Ala-Phe-OMe was completely rejected, Ac-Ala-Ser-OMe was bound with only 60 m_1. Replacement of alanine with valine completely prevents complex formation, as does exchange of the acetyl protecting group for the sterically more demanding Boc group. The modification of the aminopyrazole binding site with the Kemp's triacid derivatives leads to highly selective hosts discriminating mainly as a result of steric factors.
We are currently designing special recognition tips in our new hosts for nonpolar and acidic amino acids in peptides. We are also covalently connecting two or more of the modules to achieve predictable sequence-selective recognition oflarger peptides (for another approach to peptide receptors see Chapters 2.3 and 3.1). This
should in turn enable docking to characteristic peptide sequences on protein surfaces and might ultimately be used for protein-tagging, protection, or allosteric inhibition of enzymes.
Was this article helpful?