Bioorganic Receptors for Amino Acids and Peptides: Combining Rational Design with Combinatorial Chemistry

Carsten Schmuck, Wolfgang Wienand, and Lars Geiger

2.3.1 Concept

One major research project in our group is currently the design, synthesis, and evaluation of artificial bioorganic receptors for binding amino acids and small oligopeptides in aqueous solvents. Peptides as a substrate are interesting for two reasons - their biological significance and their chemical structure. First, there are many biochemical or medicinal processes, for example enzymatic activity, bacterial infections, or neurodegenerative diseases, in which a selective molecular interaction of a peptide with another molecule or with itself (self-association) plays a decisive role [1]. Hence study of artificial receptor systems capable of selective binding to a specific biological peptide can help us understand better such peptide-molecule or peptide-peptide interactions on a detailed molecular basis. In contrast to most native proteins or natural receptors, bioorganic models are rather small and can be varied structurally deliberately. Therefore, they can be subjected to detailed physical-organic studies which cannot be performed as easily with many natural systems [2] (see also Chapter 1 for work on structural models for biological systems). The results obtained from such bioorganic model studies are not only useful for explaining and enabling better understanding of the underlying natural process, they might in the long term also facilitate the design of biosensors, the targeting of cellular processes by chemical tools, or the discovery of new medicinal therapeutics for a variety of diseases such as Alzheimer's, bacterial infections, or cancer. Second, beside this biological significance, peptides have a large variety of different potential binding sites for a receptor molecule to bind to -an amide backbone with H-bond donors and acceptors and side-chains with polar and non-polar groups. This is a great advantage if one is interested in designing peptide receptors as it enables exploitation of the whole spectrum of non-covalent interactions for complex formation [3]. In addition to the H-bond network to the peptide backbone, salt bridges to ionic amino acid residues (as in arginine, lysine, aspartic acid, etc.) or at the termini of the peptide (carboxylate or ammonium) and hydrophobic interactions with apolar side-chains (as in phenylalanine, valine, leu-cine, etc.) can further stabilize the receptor-substrate complex. The challenge of

potential binding sites Fig. 2.3.1. Potential binding sites of a peptide.

designing a peptide receptor is therefore to translate the structural features of a given peptidic substrate into a suitable receptor molecule with complementary binding sites, as shown schematically in Figure 2.3.1 [4] (for a similar problem with carbohydrates see Chapter 2.1).

We will demonstrate in this chapter our approach towards modular receptors for complexation of biologically relevant peptides in water. A new binding motif for carboxylates, the guanidiniocarbonyl pyrroles, has been designed; this enables the formation of stable ion pairs even in aqueous solvents. By a stepwise elongation of this binding motif with additional interaction sites receptors are obtained that bind not only carboxylates but also single amino acids, both side-chain- and ste-reoselectively, and even tetrapeptides.

Because many biologically important small peptides contain a free C-terminus, which under physiological conditions is an anionic carboxylate, the first step of our approach to modular peptide receptors was to develop an efficient carboxylate binding site (CBS) which also functions in highly polar solvents up to aqueous solutions (for work on artificial hosts for spherical anions see Chapter 2.2). To get an idea of what is necessary for strong complexation of carboxylates under such challenging conditions one can take a look at Nature: For example, in carboxy-peptidase A [5], an enzyme that hydrolytically cleaves an amino acid from the free C-terminus of a peptide chain, the peptidic carboxylate is essentially bound by an ion pair with the guanidinium group of arginine 145 and two additional H-bonds from asparagine 144 and tyrosine 248 (Figure 2.3.2).

The enzyme-ligand interaction is, furthermore, significantly facilitated by the overall hydrophobic character of the binding pocket which reduces competitive solvation of the binding sites by water molecules. As such ion pair formation between arginine and a carboxylate is widely found in Nature, it is not surprising that this binding motif has already been much used by supramolecular chemists over recent decades [6]. Unfortunately, without the hydrophobic shielding of an enzyme pocket the guanidinium-carboxylate ion pair is only stable in solvents of low polarity such as chloroform or acetonitrile (Box 7). Even the smallest amounts of more polar solvents such as DMSO, methanol, or even water cause immediate dissociation of these ion pairs. This is a general problem in supramolecular chem-

Fig. 2.3.2. Carboxylate binding within the active site of Carboxypeptidase A.

istry [7]. Molecular recognition is based on non-covalent interactions but, in contrast with covalent bonds, their strength is highly dependent on external conditions such as solvent composition, polarity, or even temperature. In this context water as a solvent is the most challenging. On the one hand, electrostatic interactions (H-bonds or ion pairs), which are quite well understood and have been extensively used in artificial supramolecular systems due to their complementarity and directionality, are rather weak in this solvent whereas, on the other hand, until now the stronger hydrophobic interactions [8] are much more difficult to design and use in artificial receptors.

The best solution is therefore to use not only one but several non-covalent interactions simultaneously to achieve strong ligand complexation, even in water. Although every individual contact between host and substrate by itself might be rather weak, their combined effect can still lead to high association constants (''Gulliver effect'') [9]. Our idea for the design of an efficient CBS was, therefore, to improve the binding affinity of the guanidinium cation by use of suitable additional binding sites (how efficient arginine receptors can be designed is explained in Chapter 2.4). Based on theoretical calculations we therefore introduced cationic guanidiniocarbonyl pyrroles of type 1 as a new and easily accessible binding motif for carboxylate anions [10].

0 0

Post a comment