Introduction

Chemistry research at the Broad Institute’s Chemical Biology (BCB) Program is focused on bringing the power of modern organic synthesis to bear on the study of complex biological systems. We have adopted the terms chemical genetics and chemical genomics to describe this approach because our aim is to emulate the success of classical genetics by providing tools that are generally useful in exploring complex biological systems. The specific goals of the chemistry groups are to use diversity-oriented split-and-pool chemical synthesis as an engine to efficiently synthesize large collections of complex and diverse small molecules and test their ability to induce specific and novel biological phenotypes. Our research interests also include the elaboration of hit compounds to increase potency and specificity. We are interested in tools for a wide range of purposes, ranging from in vitro biochemical analyses to the systematic dissection of patterning processes in whole organisms.

We have built a flexible synthesis chemical platform that harnesses the power of split-and-pool synthesis to build large collections of both natural-product like molecules and small molecules based on novel core structures that are amenable for use in any type of chemical genetic or chemical genomic experiment. Our objective in synthesizing structural diverse libraries of compounds is to find out whether such compounds have qualitatively different properties than traditional “drug-like” compounds that form the bulk of existing chemical collections at pharmaceutical companies and elsewhere. We believe that compounds that do not strictly follow Lipinski’s rules will have different biorelevant properties from small, achiral chemicals, enabling us to explore previously unprobed chemical space. The experience gained from the multitude of biological screens run at the BCB screening facility will teach us how to elaborate rules and metrics specifically for the identification of novel biological probes.

Because of the iterative nature of this experiment, we collaborate very closely with biologists. As we identify areas of chemical space that are particularly important to biology, we will apply this insight to the synthesis of new small molecule libraries to increase the rate of future scientific discoveries using chemical genetics.

Split/pool synthesis of DOS libraries

In contrast to target-oriented synthesis (TOS) and medicinal or combinatorial chemistry, which aim to access precise or dense regions of chemistry space, diversity-oriented synthesis (DOS) populates chemical space broadly with small-molecules having diverse structures. The goals of DOS include the development of pathways leading to the efficient (three- to five-step) synthesis of collections of small molecules having skeletal and stereochemical diversity with defined coordinates in chemical space. Ideally, these pathways also yield compounds having the potential to append appendages site- and stereoselectively to a variety of attachment sites during a post-screening, maturation stage. The diverse skeletons and stereochemistries ensure that the appendages can be positioned in multiple orientations about the surface of the molecules.

The preparation of compound libraries using diversity-oriented synthesis has provided a crucial compliment to the compounds that BCB has acquired from commercial sources. DOS exploits the power of split/pool combinatorial chemistry to produce compounds that possess skeletal and stereochemical diversity that far exceed compounds that are commercially available. DOS efforts at BCB have produced approximately 130,000 compounds for screening.

Screening of molecules synthesized at the Broad Institute using our DOS chemical technology platform offers several advantages over their commercially available counterparts, including a high level of preliminary SAR that typically emerges from the primary screen. Each DOS library offers between 2 and 6 different core structures, each represented by 200 to 1000 members each. If 10 DOS libraries appear in the screening collection, 20-60 different core structures will be represented, so that any hit that emerges from this collection will already possess 100-300 structurally related compounds that will provide SAR information without any additional chemical synthesis.  More importantly, each compound in the DOS library results from synthetic pathway developed at the facility, so the subsequent synthesis of analogs in Stage 2 will not be plagued by unreliable or unreported syntheses, as has frequently been the case with commercially available compounds.

Split/pool synthesis of focused libraries

Exemplary for the workflow is a research project based on Harvard Medical School professor Tim Mitchison’s interest in the mitotic spindle kinesin Eg5. He initiated a search for new structural classes which would compliment the activity of monastrol – a compound derived from a commercial library that was the first molecule to be identified as selective inhibitor of the human kinesin KSP (Eg5). A preliminary screen of the available DOS compounds revealed that a dihydropyrancarboxamide (later named “monastramide”) was a potent inhibitor of KSP, and that the activity was similar to that of monastrol despite the obvious structural differences.

Discovery of probe from screening of DOS compound collection and subsequent rapid optimization from the synthesis of a focused library

The synthetic routes used in DOS and executed on small scale for the production of libraries are easily applied to a large-scale technology platform for the preparation of “focused” libraries. In the case of KSP, monastramide was produced by an enantioselective Diels-Alder reaction which was a new method first reported in 1998, and used only three years later in the preparation of a DOS library. The chemistry developed for the preparation of this library on 500uM “macrobeads,” which provide roughly 0.1 mmol of each compound, was adapted to our lantern platform which provides up to 20 mmol of each target structure. Since the technique exploits split/pool chemistry, 180 molecules were synthesized in only 3 synthetic steps. This highly efficient synthesis strategy allowed for the rapid discovery of two new structures with increased affinity.  More importantly, the 180 structurally related analogs are being used to probe the active sites of mutated forms of the Eg5 protein, thus revealing key structural requirements for binding in the wild-type Eg5.  This study would not have been possible without the large collection of structural analogs which were made available as a result of the DOS pathways developed for the original library.

Our previous experience with the synthesis of focused libraries has revealed that 192 compounds, ie two full plates, can be prepared from an established DOS pathway in approximately 2 months.  The ability to synthesize far more compounds when compared to traditional synthesis stems from the split/pool technique that provides an exponentially increasing number of compounds relative to the number of steps executed.  In total, one chemist using the DOS technology platform can produce nearly 200 compounds in the time required to produce 20 using traditional techniques.

Target Identification for Phenotypic screens:  Preparation of Affinity Reagents

Once a potent molecule is discovered in a phenotypic screen, the elucidation of affected pathways and target is often the next challenge. The synthetic chemistry group is working closely with the screening investigator in the design of reagents for affinity chromatography. This process involves the synthesis of derivatives of the optimized probe which maintain activity in the screen.  Since this technique often involves trial and error, as well as the synthesis of analogs of greater complexity than the parent compound, we have allotted the effort equivalent to the synthesis of 5 small molecules in Stage 2.  In the event that affinity chromatography is prevented by the nature of the organism under investigation, the synthetic chemistry group will work with the screening investigator to identify other way in which synthetic chemistry can help elucidate the mode of action of the probe molecule.