Solid Support Combinatorial Chemistry in
Lead Discovery and SAR Optimization

Adnan M. M. Mjalli, Ph.D. and Barry E. Toyonaga, Ph.D.

Ontogen Corporation
2325 Camino Vide Roble
Carlsbad, California 92009

horizontal line

http://www.netsci.org/Science/Combichem/feature03.html

horizontal line

The widespread acceptance and use of high throughput screening technologies for the purposes of drug discovery and development has created an unprecedented demand for small organic molecules. The requirements for (i) large numbers of diverse and novel chemical entities and (ii) methods to rapidly optimize the compounds or 'hits' found by screening may not be met by medicinal chemistry teams employing traditional synthetic methods. Alternatively, combinatorial chemistry in solution or on solid support, is being developed to increase the efficiency of organic syntheses. Furthermore, successful applications of such methods leading to the discovery of therapeutic candidates have been reported [1]. A review of this emerging field is presented by M. R. Pavia elsewhere in this issue of NetSci .

THE ONTOGEN APPROACH

Hardware and software platforms have been designed and developed to significantly increase the number of compounds that a synthetic organic/ medicinal chemist can prepare in a given period of time. Thus, we are able to create libraries of compounds for biological screening and perform medicinal chemistry optimization strategies ultimately leading to compounds for human clinical trials (Figure A). A full description of the OntoBlock and OntoCODE systems and chemistry optimized on solid supports will be published elsewhere. Brief summaries follow.

Figure A
Preclinical Drug Discovery



The synthesis of complex small molecules on solid support using different organic reactions such as multi-step sequential substitution reactions, multi-component condensation reactions and pharmacophore modifying reactions has been accomplished (Figure B).

Figure B
a) Sequential substitution,
b) Multi-component condensation array (MCCA),
c) Pharmacophore transformation



In this fashion complex, diverse, non-peptide, chemical compound libraries such as:

  • beta-lactams;
  • hydantoin imides and thioimides;
  • imidazoles;
  • N-acyl-alpha-amino amides, esters, acids;
  • oxazoles;
  • phosphonates (alpha-hydroxy, alpha-amino, alpha-acylamino);
  • phosphinates;
  • pyrroles;
  • tetra-substituted 5 membered ring lactams;
  • tetra-substituted 6 membered ring lactams and
  • tetrazoles.

are synthesized on solid support using a wide range of organic transformations including:

  • acylations;
  • aldol condensations;
  • alkylations;
  • Claisen couplings;
  • Heck reactions;
  • heterocycle forming reactions such as condensations, dipolar cycloadditions, annulations, etc.;
  • Michael additions;
  • Mitsunobu couplings;
  • multicomponent condensation reactions and
  • reductions.

The final products are cleaved into a standard 96 well microtiter plate, one compound per well. Each plate can be directly submitted for high throughput screening as well as quantitative and semi-quantitative analysis in order to assess purity, identity and yield of each compound synthesized.

Ontogen has adapted the method of synthesis of non-peptide small molecules on solid support aiming for compound synthesis in a 96 well formatted reaction vessel [2], spatially dispersed one compound per well.

HARDWARE AND SOFTWARE

The OntoBLOCK system was developed to create spatially dispersed combinatorial libraries (SDCL) [3]. SDCL allows for the production of milligram quantities of compounds in a 96 well format, and the structure of each library component is inferred from its Cartesian coordinates on the microtiter plate. At the heart of the system is a pair of septa sealed reaction blocks which, together, contain 96 reaction chambers. The reaction block can maintain temperatures between -80 degrees C and +100 degrees C as well as inert atmospheres (nitrogen or argon) and contain internal pressures (and volatiles) above 15 psi. Chemically inert (polypropylene or teflon) reaction chambers are easily replaced. In general, the reaction block has been economically designed allowing for relatively inexpensive production and widespread use. In addition, the blocks were designed to be compatible with gantry robots capable of moving them between different workstations or platforms.

The independent workstation concept was adopted to fully optimize certain operations and achieve production line efficiencies. Specialized workstations carry out delivery of (i) reagents, (ii) rinse solutions and (iii) linker cleavage reagents as outlined in Figure C.

Figure C
Workstation schematic showing compound synthesis and testing



The Reagent Delivery Station delivers reagents, including the appropriate functionalized resin, into the reaction block from septa-sealed bottles under inert atmosphere conditions using a coaxial liquid/gas handling pipeting needle. Reaction mixture agitation within each vessel is achieved by vortexing the reaction block on the "Oscillating Mixing Stations". Resin in each vessel of the reaction block is washed between individual reaction steps and at the end of the synthesis using a dedicated "Wash Station". Until the development of such a high speed Wash Station, resin rinsing was a rate limiting activity in the overall synthetic process. The final products are cleaved from the resin within each reaction vessel of the reaction block into corresponding wells of a standard 96 well microtiter plate. Thus, a plate of compounds in solution, one compound per well, is produced.

Central to the design of the OntoBLOCK system was the creation of a docking station which provided a universal platform onto which the reaction block could be mounted. Each workstation is equipped with a docking station which securely holds the reaction block and provides, among other functions, the delivery of circulating gases and liquids used for temperature and inert atmosphere control.

Typically, the eluant(s) used to release the product from the resin are removed in a vacuum oven. Individual samples could be weighed at this point and then redissolved (Master Plates). Compounds can be produced on a 10 to 25 umol scale (2 to 10 mg) at rate of 1,000 to 2,000 compounds per day per production line.

Copies of the 96-well Master Plates are replicated using 96 well pipeting devices. These Daughter Plates are ready for both high throughput screening and automated electrospray mass spectrometric analysis. Contents of the Master Plates are vacuum dried, vacuum packed under inert atmosphere and archived at -78 degrees C for future use.

An alternative to the SDCL approach called OntoCODE has also been developed. The OntoCODE system allows the synthesis of a combinatorial library via radio frequency (RF) tag encoded solid phase resin capsules. Each unique synthesis capsule can be tagged with a unique identifier code. The inert nature of the RF tag, a glass encased microchip, avoids potential conflicts between the chemistries related to a chemical tag and the associated compound undergoing synthesis. Furthermore, the transmission or retrieval of information from any capsule is instantaneous, avoiding the added reaction times associated with synthesizing chemical tags. A divide, track and apportion procedure is utilized in this synthetic strategy. Initially, a large quantity of RF tagged capsules are manipulated chemically to prepare the encapsulated solid phase resin for the synthesis. The capsules are then divided evenly into the number of reaction vessels required to couple each monomer in the library repertoire. The reaction vessel location of each capsule is recorded via scanning of its RF transponder code. Redistribution of the capsules is then guided by a computer database whereby the past and future reaction vessel locations of each capsule is recorded and determined, via scanning of the capsule's RF transponder code. By repeating this procedure throughout each step of the library synthesis, a histogram of the synthesis is developed in the database. The advantages of having compounds cleaved in a spatially dispersed manner can be achieved in practice for chemical libraries of approximately 2X104 or less, using the OntoBLOCK system described above. Thus, in the mode routinely used by Ontogen, individual compounds, not mixtures, are made available for screening. Obviously, OntoCODE is fully compatible with the more conventional output of tagged bead synthesis, i.e. the preparation of mixtures of compounds.

Finally, software systems have been designed to enable the facile entry, manipulation, retrieval and analysis of data. For example, measured mass spectrometric data is automatically compared to the predicted mass of every compound to determine the success of the synthesis. The entire spectrum of information is tracked and is accessible: from the purchase of raw materials, to synthesis planning, to physical analysis (NMR, mass spec, etc.),to biological analysis, through to structure-activity relationship analysis (Figure D).

Figure D
Software Systems



LIBRARY DESIGN

The design of the pharmacophore basis of a particular library is driven by the nature of the biological target of interest. The following types of information are considered, if available:

  1. the biology of the target enzyme or receptor;
  2. the nature of substrate;
  3. the mechanism of target-substrate interaction;
  4. related literature information;
  5. 3-D structural information;

To date, several libraries requiring diverse reaction conditions have been synthesized. These libraries are highly functionalized to achieve maximum structural and functional diversity. For example, the imidazole based library (Figure. E) contains up to seven different substituents around the pharmacophore of interest. In general, the method of synthesis is designed to allow full control over each of the individual substituents. This is accomplished through the selection of the starting materials or inputs (charge, electron withdrawing/donating, hydrogen bond donor/acceptor, hydrophobicity, steric bulk, etc.). In general the inputs are chosen to be commercially available. On occasion, inputs are synthesized for specific cases, fully aware that input synthesis has the potential to dramatically reduce the efficiencies of the combinatorial approach.

Figure E
7 Input imidazole library



PROTEIN TYROSINE PHOSPHATASE (PTPASE) INHIBITION: AN EXAMPLE OF LEAD DISCOVERY AND SAR OPTIMIZATION

The importance of protein tyrosine phosphorylation by protein tyrosine kinases (PTKase) and dephosphorylation by PTPases in the signal transduction pathways which control the cell growth and differentiation has been reviewed [4, 5]. PTPases have been implicated in several disease areas [6, 7, 8, 9, 10].

A wide variety of PTPases has been reported and, interestingly, each has at least one highly conserved catalytic domain of eleven amino acids: (I/V)HCXAGXXR(S/T)G. Mutation experiments indicate that the consensus sequence cystine and arginine residues are required for PTPase catalytic activity [11]. Furthermore, the natural substrates for most of these tyrosine phosphatases have not been identified. Thus, the discovery of competitive, phosphorous containing, potent, reversible and selective small molecule inhibitors is expected to be quite challenging.

A library of potential substrate based inhibitors was designed. The library consists of three different inputs such that R1, R2, and R3 represent a wide range of functional groups selected on the basis of their potential to mimic the phosphate, phenyl and amino acid (AA) groups respectively (Figure F).

Figure F
Design of a PTPase substrate-based mimic



A novel method of synthesis has been established on solid support using three different sequential steps (overall yield > 80%). This approach would allow the creation of over 200,000 compounds using commercially available starting materials as inputs. More than 350 plates, over 28,000 compounds, were synthesized using the OntoBLOCK system. These compounds were screened against a particular PTPase, PTP-1B. The details of this project will be published elsewhere.

The first active compounds (IC50 values between 5 and 100 µM) were discovered within the first thirty plates synthesized from that library (Figure G, Box B). SAR information derived from the analysis of these compounds was worked into the design of compounds of plate number 103. The synthesis and screening of plate 103 gave rise to active compounds with IC50 values around 1uM (Box D). Additional SAR data was generated in the subsequent round of design, synthesis and assay (Boxes E and F). Finally, sufficient knowledge was compiled to design plate number 163 (Box G) from which potent inhibitors of our target enzyme were discovered (Box H). Representatives from plate 163 showed selectivity towards PTP-1b by a factor of 100 when tested against similar PTPases.

Figure G
PTPase discovery flow chart



We believe this example demonstrates the strength of our approach to the discovery of active compounds and SAR optimization. The work described was accomplished by a group of 5 to 6 scientists, chemists and biologists, over a 3 to 4 month period. That approach lead to the discovery of novel, phosphorous containing PTP-1b inhibitors with an IC50 of 20nM. Detailed kinetic analysis showed that these compounds are competitive and reversible inhibitors. Studies with two other PTPases show selectivity of greater than 100-fold.

REFERENCES

General material:

  1. Gordon et al., J. Med. Chem., 1994, 37, 1385-401
  2. Gallop et al., J. Med. Chem., 1994, 37, 1233-51

Combinatorial libraries used:

  1. to identify nucleic acid aptamers: Latham et al., Nucl. Acids. Res., 1994, 22, 2817-2822
  2. to identify RNA ligands to reverse transcriptase: Chen and Gold, Biochemistry, 1994, 33, 8746-56
  3. to identify catalytic antibody specific to a particular transition state: Posner, Trends. Biochem. Sci., 1994, 19, 145-150
  4. to identify ligands for 7-transmembrane G-protein-coupled receptors: Zucherman, R. N., Martin, E. J., Spellmeyer, D. C, et al., J. Med. Chem., 1994, 37, 2678-85
  5. to identify aminimide as mimetics: Hogan, J. C., WO 9401102
  6. to identify inhibitors of carbonic anhydrase enzyme: Baldwin, J. J., Burbaum, J. J., Henderson, I., Ohlmeyer, M. H. J., J. Am. Chem. Soc., 1995, 117, 5588-5589.
  7. Geysen, H. M. , Meleon, R. H., Barteling, S. J., Proc. Nat. Acad. Sci. USA, 1984, 81, 3998-4002.
  8. Armstrong R., PCT 95/02566
  9. Hunter, T., Curr. Cell Biol., 1989, 1, 1168.
  10. Chan, A. C., Desai, D. M., Weiss, A., Annu. Rev. Immun., 1994, 12, 555-92
  11. Flint, A. J., Gebbink, M. F. B. G., Franza, B. R. J, Hill D. E., Tonks, N. K., EMBO J, 1993, 12, 1937.
  12. Woodford-Thomas, T. A., Rhodes, J. D., Dixon, J. E., J. Cell. Biol., 1992, 117, 401.
  13. Brown-Shimmer, S., Johnson, K. A., Hill, D. E., Olefsky, J. M., Bruskin, A. M., Cancer Res., 1992, 52, 478.
  14. Kenner, K. A., Hill, D. E., Olefsky, J. M., Kusari, J., J. Biol. Chem., 1993, 268, 25455.
  15. Frangioni, J. V., Smith, M., Oda A., Salzman, E. W., Neel, B. G., EMBO J., 1993, 12, 4843.
  16. Bradford, D., Flint, A., Tonks, N. K., Science, 1994, 263, 1397.


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