in chloroform stabilized by intramolecular hydrogen bonds between the aromatic NH and aromatic C=O groups. Because all aromatic peptide subunits point in the same direction, these conformations are closely related to the cone conformation of calixarenes. The cyclopeptide contains one aromatic subunit less than a calix[4]arene, however, and its cavity is shallower and has a slightly larger diameter.

2 3

The ability of this peptide to bind cations could be demonstrated by the upfield shift observed for the cation signals in the NMR spectrum on addition of a salt of, for example, the n-butyltrimethylammonium ion (BTMA+) to a solution of 2 in 0.2% d6-DMSO/CDCl3 [12]. This shift is a good indication of the interactions between the cation and 2, and it enabled us to determine a stability constant of 300 m—1 for the complex formed. The maximum chemical shift, A^max, observed for, for example, the N-methyl signal of the cation amounts to only —0.05 ppm, however, and is thus significantly smaller than the shift usually associated with com-

plexation of similar guests by calixarenes (typically —0.5 to —1 ppm). The effect of the aromatic subunits on the resonance of the guest protons is obviously weaker in 2, most probably because the dimensions of the cyclopeptide cavity are not optimum for complexation of BTMA+. To improve cation binding we therefore replaced the glutamic acid subunits of 2 with a more rigid amino acid, namely proline [13]. We expected the corresponding peptide 3 to be conformationally less flexible and, as a consequence, better preorganized for substrate binding. NMR spectroscopic investigations indeed confirmed this prediction. On complex formation between 3 and BTMA+ picrate a significantly larger upfield shift of the BTMA+ N-methyl signal of up to —0.70 ppm was observed; this could conveniently be used to determine a complex stability of 1260 m—1. Thus, replacement of the glutamic acids by prolines led to an approximately fourfold increase in BTMA+ complex stability, an effect that shows how sensitively the receptor properties of such peptides react toward structural modification. Currently we have no experimental evidence for participation of the peptide carbonyl groups in cation com-plexation, and therefore assume that complex formation is driven by cation-re interactions. A cooperative effect of the carbonyl groups on binding cannot be ruled out, however [14].

Further investigations revealed another interesting property of 3 [13]. On varying the anion in the BMTA+ salt we found that anions other than picrate, for example iodide, tetrafluoroborate, or tosylate, bind strongly to the NH groups of 3 by hydrogen-bond formation. This interaction stabilizes a peptide conformation in which all NH groups converge to enable simultaneous complexation of the anion. The geometry of the tosylate complex of 3 was assigned by NOESY NMR spectroscopy. Its structural relationship to the iodide complex, of which a crystal structure was obtained (Figure 2.2.3), is high.

Figure 2.2.3 nicely illustrates that 3 simultaneously interacts with the anion and the cation. As expected, the cation is included in the shallow dish-shaped cavity formed by the aromatic peptide subunits. The cyclic arrangement of peptide NH

Fig. 2.2.3. Top and side views of the crystal structure of the N-methylquinuclidinium iodide complex of 3.
Tab. 2.2.1. Stability constants Ka and maximum chemical shifts Admax of the BTMA+ complex of 3 in the presence of different anions (0.2% d6-DMSO/ CDCl3; T = 298 K; Ka in m_1; Admax maximum chemical shift of the BTMA+ N-methyl protons in ppm; error limits for Ka < 20%).





0 0

Post a comment