Application of a concept drawn from two areas of macromolecular chemistry shows how artificial binding sites that resemble those found in globular proteins can be made.
Chemists have often looked with awe at the dazzling array of biochemical functions and chemical transformations in nature. So, if nature can do it, why can't we? Part of the answer is, of course, the many millions of years over which natural selection has had time to optimize an enzyme active site or a protein–DNA interface, compared to the usual term of a research grant or the four or five years available to PhD students. Synthetic chemists are, nonetheless, on the case, and a clever way of creating molecular-recognition systems is described by Zimmerman et al.1 on page 399 of this issue. They have taken an important step in establishing a family of synthetic compounds that can be easily prepared and show recognition properties that bear comparison to those of antibodies.
The exquisite recognition of an antigen by its antibody is just one instance of the remarkable chemical control achieved in biology. Regulation of gene expression by transcription factors and the seemingly effortless execution of difficult reactions by enzymes are other examples. Yet these transformations are carried out with a very limited set of building-blocks — just four nucleotide bases and 20 amino acids. Nature can fashion the correct chemical microenvironments for binding a substrate or cleaving a bond by simply folding a string of amino acids, but synthetic chemists have yet to design and create an equivalent receptor or catalyst. There are essentially no examples of synthetic molecules that bind to an antigen with the strength or selectivity of an antibody, or that split the bonds between amino acids at physiological pH and temperature like the digestive enzyme trypsin. But although chemists don't have the time that nature has had, they do have a large array of synthetic reagents and potential building-block designs for constructing functional mimics of proteins or nucleotide chains.
Using entirely synthetic components, Zimmerman et al.1 have created a globular molecule containing an interior recognition site that can bind to certain modified porphyrins and differentiate between closely related analogues. Porphyrins are a family of large organic molecules that occur naturally (for instance, as a component of haemoglobin) but are commonly used in synthetic chemistry. The key element in Zimmerman and colleagues' approach is the merging of two areas of macromolecular chemistry, dendrimers and molecularly imprinted polymers.
Dendrimers have long been seen as a discrete branch of polymer chemistry, where the macromolecule is grown out from a central core to create a roughly spherical compound with a single molecular weight2. But dendrimers suffer from their design. They tend to be unduly flexible and have little interior space in which to form an active site with well-positioned binding groups. On the other hand, molecularly imprinted polymers often contain effective binding pockets within the confines of a stable polymer network. The pockets are formed by carrying out the polymerization around a molecular template which is then removed from the matrix3. The resulting cavity possesses the shape (and complementary chemical characteristics) of the template and can be used to bind related chemical structures. But molecularly imprinted polymers can be limited by the chemical heterogeneity of the polymer in terms of size and binding-site structure, as well as restricted solubility.
Zimmerman et al.1 have designed a single molecular template from which they construct a dendrimer by attaching functional groups to create a divergent framework. The template is a porphyrin with four phenyl groups, each of which bears two hydroxyl groups (Fig. 1a). All eight hydroxyls are linked through easily cleaved ester bonds to a wedge-shaped dendrimer segment composed of three generations of benzene derivatives4. The third-generation component contains terminal alkenes. The result is a classical, almost spherical dendrimer with 64 alkene groups on its surface.
The significance of the alkenes is that they contain carbon–carbon double bonds, which can be broken and re-formed as new double bonds between neighbouring alkene groups. It is here that Zimmerman and colleagues' approach comes into its own. Using a catalyst — Grubbs' olefin metathesis catalyst5 — the peripheral alkenes are crosslinked to stabilize the shape of the dendrimer that has been formed around the porphyrin template. Analysis with nuclear magnetic resonance, mass spectrometry and chromatography confirmed that virtually every alkene had reacted with a neighbour. The whole superstructure is then sufficiently stable to allow the porphyrin template to be removed by cleaving the ester bonds. This leaves behind a cavity, lined with eight carboxylic-acid groups, within the interior of the globular molecule (Fig. 1b). The size and shape of the cavity betray its origins from the porphyrin template.
The authors show that the dendrimer host binds strongly to a test substrate — a porphyrin with four pyrimidine groups, with a total of eight basic nitrogen atoms in appropriate positions to form hydrogen bonds to the carboxylic-acid groups in the cavity. There is subtlety in the recognition, suggesting a well-defined binding pocket. The original porphyrin template with its eight hydroxyls does not itself bind because the combined sizes of a carboxylic acid and a hydroxyl group make too tight a fit.
However, the design is not perfect. There is flexibility in the dendrimer host structure that leads to intramolecular hydrogen bonds between carboxylic acids in the binding pocket. This results in an affinity that is similar for substrates irrespective of whether four or eight hydrogen bonds are formed. Also, the lack of discrimination among different isomers of a tetrapyridyl porphyrin suggests that the carboxylic-acid groups can move within the binding cavity and contact the nitrogen atoms in different positions.
Nonetheless, the overall concept outlined by Zimmerman et al. is a promising one. The idea of using a substrate (or at least a close analogue) as the template for the synthesis of its own host will find many future applications, for example in drug delivery and catalyst design, and in devising novel separation strategies. Nature still has an edge in terms of affinity and selectivity, but Zimmerman and colleagues' approach will now permit the construction of artificial binding sites that resemble those in globular proteins.
Zimmerman, S. C., Wendland, M. S., Rakow, N. A., Zharov, I. & Suslick, K. S. Nature 418, 399–403 (2002).
Newkome, G. R., Moorefield, C. N. & Vögtle, F. Dendritic Macromolecules: Concepts, Syntheses, Perspectives (VCH, Weinheim, 1996).
Wulff, G. Angew. Chem. Int. Edn Engl. 34, 1812–1832 (1995).
Hawker, C. J. & Frechet, J. M. J. J. Am. Chem. Soc. 112, 7638–7647 (1990).
Trnka, T. M. & Grubbs, R. H. Acc. Chem. Res. 34, 18–29 (2001).
About this article
Comptes Rendus Chimie (2003)