Chemically Generated Screening Libraries:
Present and Future

Michael R. Pavia

Sphinx Pharmaceuticals
A Division of Eli Lilly & Co.
840 Memorial Drive
Cambridge, MA 02139



http://www.netsci.org/Science/Combichem/feature11.html

Introduction

The pharmaceutical industry is rapidly changing. The days where industrial scientists had the luxury of exhaustively studying the function and structure of a novel biological target then spending years to identify and optimize a lead, only to find that the compound failed in development are gone. Emerging new technologies such as genomics, gene sequencing, transgenic animals, and molecular biology are affording the industry a huge number of novel, clinically relevant biological targets. High throughput screening technologies coupled with compound libraries obtained through combinatorial chemistry and/or high throughput synthesis methods are being utilized to rapidly identify and optimize ligands for these targets [1,2,3,4].

In this brief article I review chemically generated screening libraries, describing the current state of the art, as well as what advances are still required to make this technology optimally useful within the pharmaceutical industry.

The field of combinatorial chemistry began only a decade ago. For the majority of its existence, the field was devoted to preparing large libraries of peptides. Next came the progression to peptide-like oligomeric compounds and most recently, the preparation of small organic molecules. Today, the vast majority of efforts in the field are devoted toward the preparation of these small, non-oligomeric molecules.

The first examples of non-polymeric small molecule diversity generation was reported only three years ago [5, 6]. Since that time we have seen the steady development of new chemistries and equipment applied to library generation. It is now possible to synthesize a broad structural range of compounds using these methodologies.

What trends are we observing in this field of research? What are the future improvements necessary to make this technology even more useful? What breakthroughs can we hope for? I will attempt to answer these questions for two major issues in library generation: "How to Make A Library" and "Which Library to Make".

How to Make a Library?

Three key decisions must be made before starting library synthesis:

  • Mixtures or single compounds
  • Solid phase or solution synthesis
  • Fully integrated synthesis station or selected automation

Mixtures or single compounds: For the generation of compound mixtures, solid phase synthesis is the preferred methodology. The most widely used technique is the split-pool method which assures that each component of the mixture is present in approximately equimolar concentrations. The structure of the bound ligands are determined either through an iterative, or recursive, deconvolution strategy or through the use of encoded libraries. An exciting recent advance is the use of Rf transponders for encoding [7, 8].

Examples exist where deconvolution methods have been successful in identifying active compounds amid a large mixture of candidates. However, there remains a number of issues in screening large compound mixtures, all related to the question: can a single active component be consistently identified from a mixture?

Screening of single, structurally defined molecules has a proven track record in the industry. A number of laboratories have recently developed methods for the rapid, simultaneous synthesis of large numbers of compounds using array synthesis that results in one single and well defined compound being prepared at each site. Compounds prepared by these methods are being used for both lead generation and lead optimization [9]. For example, at Sphinx, we have pursued parallel array synthesis using 96-well commercial microtiter plates, and have developed a low-cost, low-tech approach easily transferred to traditional medicinal chemistry labs [10].

It appears now that most practitioners of combinatorial chemistry are testing their libraries as single compounds or as very small mixtures (3-10 compounds), probably for the reasons discussed earlier. However, synthesis methods appear to be nearly equally divided between those who prefer to prepare mixtures versus those who prefer to synthesize individual compounds. There may indeed be some time advantage in synthesizing large numbers of compounds using the split pool methodology if one wishes to prepare very large numbers of compounds. However, the preparation of single compounds is now quite fast with automation and I believe this technology is particularly well suited to lead optimization studies.

Solid phase or solution synthesis: Solid phase synthesis has many advantages over solution-based methodologies which have been discussed in detail elsewhere [11]. These include the ability to force reactions to completion using large excesses of reagents and simple purification through washing of excess reagents and impurities away from the solid support. Furthermore, the range of organic reactions which can be successfully carried out on solid phase is rapidly expanding and is now frequently utilized for multi-step synthetic sequences [11].

However, solid phase chemistry is not without its problems such as difficulties in analysis of resin bound products, and low loading capacities of many resins. Reactions frequently proceed slower on solid phase, and heterogeneous reagents cannot be employed. Many resins display significant swelling and shrinking properties which can severely affect reaction rates and site accessibility. In addition, it is generally quite labor intensive to adapt solution chemistry to the solid support.

A number of groups have also generated libraries in which parallel reactions occur in solution [12]. Solution-generated libraries do not have the advantage seen in solid phase synthesis of being able to drive the reactions to completion by adding large excesses of reagents. For this reason, solution based syntheses were initially utilized for the preparation of simple structures. However, the introduction of parallel purification schemes is allowing more complex chemistries to be performed in parallel solution format.

Recently, the use of resin bound reagents [13], scavenger resins, as well as soluble resins which can be made insoluble for washing steps have been reported [14]. These techniques will certainly add to the scope of reactions that can be employed for solution based library synthesis.

In conclusion, both solid phase and solution techniques have been shown to be useful for library generation. To give the scientist the broadest selection of available chemistries it is advisable to have the ability to carry out both types of syntheses.

Fully integrated synthesis station or selected automation: The use of automation is an important component in library generation. Considerable efforts to develop automated systems for organic chemistry are underway. These efforts not only free the chemist for more productive endeavors but also assure consistency in repetitive procedures. Among the procedures that are currently being automated on a routine basis are liquid transfer (solvents and reagents) solid transfer (reagents and resins), filtration, solid phase extraction, and several analytical procedures such as TLC spotting and injection of samples into an autosampler attached to a MS or HPLC system.

There are however, differences of opinion as to whether a fully-integrated synthesis station is preferred to a work station approach where only select procedures are carried out by the robot with the remainder being carried out by the chemist. In my opinion, the latter approach offers a less capital intensive and more flexible approach to automation.

Which Library to Make?

Progress to date in library generation has primarily focused on methodologies for generating compounds, with less emphasis placed on library design. A number of groups are developing novel methods to analyze diversity space, although it is safe to say that no one has yet provided a completely satisfactory computational solution to effectively analyzing biological target diversity space. As we achieve a better understanding of diversity space analysis, a reasonable goal will be to attempt to prepare "universal libraries" designed to systematically explore diversity space with a limited number of structural motifs and effectively identify a chemical lead for any biological target of interest. Such an approach is among the goals being pursued at Sphinx Pharmaceuticals [15].

An alternate approach is to design biased sets of compounds which contain specific pharmacophores expected to be important for activity, or which are of a particular size or shape. Others have maximized diversity around a scaffold known to be prevalent in therapeutic agents.

The Future

The field of combinatorial chemistry has moved at an incredibly fast pace in the past several years. What can we expect in the future?

Advances in adapting solution chemistry to solid phase organic synthesis will continue. It can be expected that nearly all of the standard reactions in organic chemistry will be successfully carried out on solid phase. The creation of novel cleavable linkers for tethering reaction components to supports and improvements aimed at identifying novel solid supports to improve reaction yields and product loadings will all contribute to this effort.

An expanded scope of reactions that can occur with solution techniques using increasingly effective parallel purification schemes is expected.

In short, soon we will have the ability to make nearly any class of molecule we wish using these rapid techniques.

Improved generations of automated instrumentation built to perform organic synthesis is expected. Furthermore, progress is likely to continue to drive combinatorial technologies and instrumentation towards miniaturization.

With the creation of so much chemical and biological data, new data handling tools to process, track, and interpret results will be required as well.

Finally, new computational techniques will afford us a more complete understanding of what we mean by diversity space. It will be possible to design a library, or collection of libraries, to effectively explore biological target space and allow for the discovery of novel leads for any novel biological target.

For this new paradigm to achieve optimal success, chemistry, molecular biology, genomics, screening technology, engineering, computational technology, and data processing expertise must all work together.

Successes resulting from combinatorial synthesis have already been reported for both lead generation and optimization and we can expect that these successes will be reported at an ever increasing pace.

References

  1. Moos WH, Green GD, Pavia MR, Ann Rep Med Chem, 1993, 28 315-324.
  2. Pavia MR, Sawyer TK, Moos WH, Bioorg Med Chem Lett, 1993, 3, 387-96
  3. Gallop MA et al, J Med Chem, 1994, 37, 1233-1251.
  4. Gordon EM et al, J Med Chem, 1994, 37, 1385-1401.
  5. Bunin BA, Ellman, J Am Chem Soc, 1992, 114, 10997-10998
  6. DeWitt SH et al, Proc Natl Acad Sci USA, 1993, 90, 6909-6913.
  7. Nicolaou KC et al, Angew Chem, 1995, 34, 2289-91.
  8. Moran EJ et al, J Amer Chem Soc, 1995, 17, 10787-8.
  9. Terrett NK et al, Tetrahedron 51: 8135-8173.
  10. Meyers HV et al, Molecular Diversity, 1995, 1, 13-20.
  11. Fruchtel JS, Jung G, Angew Chem, 1996, 35, 17-42.
  12. Storer R, Drug Discovery Today, 1996, 1, 248-54.
  13. Parlow JJ, Tet Lett, 1995, 36, 1395-6.
  14. Han H et al, Proc Nat Acad Sci USA, 1995, 92, 6419-23.
  15. Pavia MR et al, Bioorg Med Chem, 1996, in press.


NetSci, ISSN 1092-7360, is published by Network Science Corporation. Except where expressly stated, content at this site is copyright (© 1995 - 2010) by Network Science Corporation and is for your personal use only. No redistribution is allowed without written permission from Network Science Corporation. This web site is managed by:

Network Science Corporation
4411 Connecticut Avenue NW, STE 514
Washington, DC 20008
Tel: (828) 817-9811
E-mail: TheEditors@netsci.org
Website Hosted by Total Choice