Clearly, the potential for identifying new drugs based on carbohydrate compounds is immense, a notion well supported by the growing number of carbohydrate-based therapeutics on the market and in clinical trials. Such compounds include acarbose and voglibose for the treatment of diabetes (and secondarily, obesity), zanamivir for influenza, acemannan for infection and wound healing, and many others.

Reactive
The great potential of carbohydrates as therapeutics has been largely underexploited, with the pool of these diverse and important compounds generally regarded as a poor source of new therapies. A major reason for this is the difficulty of synthesizing carbohydrate compounds, which has had a great impact on the discovery and production processes.

An important tool in modern drug discovery is combinatorial chemistry, with which carbohydrate synthesis has been historically thought incompatible. This perception arose from factors such as poor reactivity, difficult purification, and the necessity for regio- and stereoselective reactions. These last considerations are due to the very properties that make carbohydrates attractive members of compound libraries, namely, their rich stereochemistry and high degree of functionalization.

Because of these problems, it takes highly skilled chemists using specialized techniques to synthesize carbohydrates. This has meant that very few commercial libraries have been developed and has particularly limited the application of the automated methods required for commercial library preparations. Recently, a great deal of effort has been conducted to address these issues.

Central to many combinatorial approach-es is the use of solid-phase techniques, which have been used successfully in carbohydrate chemistry for some time (1). In a highly significant contribution, K. C. Nicolaou and colleagues of The Scripps Research Institute (LaJolla, CA) produced a dodecasaccharide on solid phase (2). The routine application of carbohydrates to solid-phase chemistry, and then to solid-phase carbohydrate combinatorial chemistry, requires innovations such as new linkers, building blocks, and coupling procedures (3). Such innovation is shown in the work of Daniel Kahne and colleagues at Princeton University, who used anomeric sulfoxide donors to synthesize a library of di- and trisaccharides and tested them on-bead (4).

Perhaps because of the pervasiveness of solid-phase methods in traditional combinatorial chemistry, and advantages such as on-bead testing, there have been only a few examples of solution-phase approaches in the carbohydrate arena (5). This is surprising since a parallel solution-phase approach to the combinatorial chemistry of carbohydrates can have clear advantages, such as the ability to produce larger quantities of compounds, release products without a cleavage step, monitor reaction progress more easily, purify intermediates where necessary, and make single-compound products without deconvolution or tagging. It is clear that both solution- and solid-phase methods need to be developed concurrently to take advantage of the benefits bestowed by each.

Proactive
Despite the elegant and innovative work mentioned above, barriers to the commercial application of carbohydrates to combinatorial chemistry remain. A few commercial examples are the exception rather than the rule (6). Michael West, Wim Meutermans, and their colleagues at Alchemia Pty. Ltd. (Brisbane, Australia), a biotechnology company specializing in large-scale custom synthesis, drug delivery application, and particularly, combinatorial synthesis of carbohydrate compounds, have devised new linkers (3), protocols, and building blocks to help overcome some of the barriers. Their aim is to help remove impediments to the discovery, manufacture, and commercialization of carbohydrates as therapeutics.

In an initial, and chiefly traditional, approach, West and Meutermans’ groups have developed a technology based on a glucosamine scaffold for the production of focused monosaccharide libraries. Initially, they prepared a library of potential anti-infective compounds, which were designed to resemble muramic acid and other substrates in the enzyme cascade that catalyzes bacterial cell wall biosynthesis. The approach used proprietary building blocks that allow for almost exclusively crystalline products during synthesis. Crystallinity was crucial to the approach because it allowed the synthesis of key intermediates of high purity without the need for chromatography. This proved to be extremely important when bulk quantities of intermediates were required for expanded second-generation libraries. The intermediates could be prepared pure on large scale without chromatography. With steps readily optimized for automation and solid-phase chemistry, the researchers used this technology to produce a small library of compounds that are currently being evaluated by collaborators.

This research highlighted the need for a series of building blocks for carbohydrate synthesis in which all the protecting groups are fully orthogonally protected and stable, with all five positions independently functionalizable. The preparation of such a “universal building block” has long been the Holy Grail of synthetic carbohydrate chemistry. Although there are academic approaches to similar compounds, such as those of Chi-Huey Wong and colleagues at The Scripps Research Institute that contain four orthogonally functionalized positions (7), they have been underused in a commercial setting. Gyula Dekany and colleagues at Alchemia have devised such a “universal building block” approach and are applying it to the 10 mammalian sugars. They term their technology Versatile Assembly on Sugar Templates (VAST).

Previous commercial applications of this approach have been further limited, not just because appropriate building blocks had not been discovered, but because any such compound could not be produced economically in sufficient quantities to allow their routine use. Alchemia has a great deal of experience in the large-scale synthesis of these building blocks, experience that has led to both a joint venture and a manufacturing alliance with The Dow Chemical Company (Midland, MI). In this collaboration, Dekany’s group has successfully transferred its technology for large-scale synthesis to Richard Wolf and colleagues in Contract Manufacturing Services at Dow, which will synthesize carbohydrates in commercial quantities. Compounds resulting from this partnership will include products of interest to third parties. Alchemia envisages the preparation and large-scale use, if required, of their five VAST building blocks. This will allow access to libraries of not only unprecedented three-dimensional diversity, but also libraries that contain natural and non-natural carbohydrate structures previously thought unfeasible for development as therapeutics.

Figure 1. In this member of the VAST library, each functional moiety is labeled as the amino acid it is intended to mimic.

To validate their approach, the Alchemia researchers are applying this technology to the production of libraries of compounds that can be decorated with diverse functional groups. Ralph Hirschmann and colleagues at the University of Pennsylvania (Philadelphia) showed previously that nonpeptidal peptidomimetics based on a glucose scaffold could act as somatostatin agonists and antagonists (8). In an approach that is not limited to any particular therapeutic area, Alchemia researchers have produced a library of compounds that mimic protein–protein interactions in a more general fashion. To accomplish this, Meutermans’ group developed a solution-phase approach to a galactose scaffold decorated with functional moieties designed to mimic amino acids. One application of such a library is the discovery of novel nonpeptidal b-turn mimetics. Concerned about the potential instability of ester and carbamate-type compounds, Meutermans and his group prepared the library so that functional moieties could be attached to the scaffold by ether linkages (Figure 1). This provides the added advantage of the greater conformational rigidity of ether linkages, which likely results in greater binding of any product compound. The major headache in this process was that any etherifications had to be regiospecific and occur in such a way as to maintain the stability of the other functional moieties. The VAST building blocks proved to be up to the challenge, and many library members have been prepared. A concurrent solid-phase approach is now being optimized.

With genomics-driven growth in the number of available targets coupled to a sense of urgency in Big Pharma to find the next blockbuster drugs, the desire for increased diversity is constantly growing. The opportunities presented by scaffolds and building blocks such as VAST should ensure that carbohydrates are not the poor cousins in that race.

References

1. Osborn, H.M.I.; Khan, T. H. Tetrahedron 1999, 55, 1807–1850.
2. Nicolaou, K. C.; Watanabe, N.; Li, J.; Pastor, J.; Winssinger, N. Angew. Chem. Int. Ed. 1998, 37, 1559–1561.
3. Pending Alchemia patents in all of these areas.
4. Liang, R., et al. Science 1996, 274, 1520–1522.
5. Boons, G.-J.; Heskamp, B.; Hout, F. Angew. Chem. Int. Ed. 1996, 35, 2845–2847.
6. Ye, X.-S.; Wong, C.-H. J. Org. Chem. 2000, 65, 2410–2431.
7. Wong, C-H; Ye, X.-S.; Zhang, Z. J. Am. Chem. Soc. 1998, 120, 7137–7138.
8. Hirschmann, R.; Nicolaou, K. C.; Pietranico, S.; Salvino, J.; Leahy, E. M., et al. J. Am. Chem. Soc. 1992, 114, 9217–9218.