Retro Synthetic Analysis, or Is Nature Perfect?


The gods of the earth and sea
Sought through nature to find this tree,
But their search was all in vain:
There grows one in the human Brain.

—From The Poems of William Blake


By Jason Socrates Bardi

Finding chemicals with biologically useful properties has become the stuff of modern day legends. Compounds that exhibit anti-tumor activity, viral replication inhibitor molecules, thermally stable enzymes, and other holy grails that achieve biological goals have been found in nature.

And so from the Pacific yew groves to the deep sea vents, we quest after these undiscovered, potentially useful compounds. Any tree, toad, or random gram of terra firma could contain the next ibuprophen or AZT. Once found, these compounds can be studied, solved, synthesized, and mass produced—to the benefit of all mankind.

Nature provides and science supplies.

But uncovering nature’s secret compounds and finding ways to synthesize them is only the beginning. A basic science laboratory is interested in not only what we can get from nature, but what we can learn from it.

“There is this impression that if it comes from nature, then we can’t do any better,” says Dale Boger, who is the Richard and Alice Cramer Professor of Chemistry at The Scripps Research Institute (TSRI). "But in fact, nature rarely makes the molecules for the reasons that we find them useful or interesting."

Because the compounds we discover may not have evolved to do exactly what we want them to do, we cannot expect to find the best agents in nature. However, what we can expect to find are lead compounds from which we can gain insight into the design of others.

Boger calls this "constructionist science"—the synthesis of function, not solely the synthesis of molecules. And he and his colleagues at TSRI seek to use the tools of organic synthesis to identify, imitate, understand, exploit, and sometimes surpass what nature provides.

A Problem in Search of its Chemistry

The science starts, of course, with synthesizing compounds that exist in nature. "We’re a group that chooses its synthetic targets based on the properties of the molecules," Boger says, "And a large proportion of our targets have a unique mechanism or properties associated with them that make them interesting in their own right."

Boger and his colleagues use the technique known as retro synthetic analysis, where a scientist looks at a structure, moves back one step to a precursor of this structure, and then thinks of a way to convert that precursor to the final product. There are usually multiple possible precursors, and each of these will have several precursors that could be used to form them.

A path must be chosen, and the selection usually reflects the personality, expertise, and the interests of the chemist. "If you have a hundred chemists, you’ll have a hundred different routes to the final molecule," says Boger.

After synthesizing an interesting natural product, further investigations can show what it is about the compound and its interaction with its biological target that makes it active, and this information can then be used to make simpler or better agents.

This type of work is time-consuming. A novel synthesis may require as many as 40 steps, each step being a reaction in which a molecule a little closer to the target is formed. Some steps may be easy to anticipate beforehand, and others may have to be invented from scratch, and each step may require 10 to 20 novel approaches before finding one that works well. Each approach may have to be repeated several times, and each time various analytical tools have to be used to determine the reaction product and yield. Then there may be 10 to 20 optimizations at each step as well.

Boger estimates his laboratory spends about 90 percent of its time tackling various syntheses. And a project may take several years from start to finish.

"It’s not something that you can make today and analyze tomorrow," says Boger.

Duocarmycins—A Classic Example

Many of the compounds that Boger and his group study and synthesize have some tumor supressor activity or are derived from anti-tumor agents.

One family of compounds that stands out in particular are the powerful cytotoxic molecules known as the duocarmycins. Naturally derived duocarmycin SA, which is produced by bacteria of the Streptomyces family, alkylates DNA and prevents replication, leading to apoptosis of the cells.

In 1982, when Boger began to synthesize a closely-related natural product, CC-1065, not much was known about the molecule, other than it had anti-tumor activity and possibly interacted with DNA. He completed its total synthesis in 1987, and has spent many years since studying the selective mechanism of the agent, looking at how it alkylates DNA through structural studies.

These structural studies have involved modifying certain atoms or moieties on the agent and its structural analogues and testing the modified molecules against the same DNA substrates that bind the original compound. In this way, the molecules can be "diced up" and their different pieces examined.

Any chemical changes to a molecule will change its structure, altering electron distributions and bond lengths. After many years of this, says Boger, you can predict to an extent how structural changes will affect the reactivity, though there are always unexpected results.

Subtle changes—even a single atom—may not look like much on paper, but can induce a million-fold difference in activity.

Florescence quenching or similar chemical assays can quantitate how much of an effect these changes will have on the molecule’s reactivity—how much and how quickly they bind to DNA, for example. Cell culture assays can be used to probe how the chemical changes affect the molecules’ biological activities—their cytotoxic efficiency, for example. And high-resolution nuclear magnetic resonance and x-ray crystal structures of the agents bound to their substrates allow unambiguous correlations between chemical, structural, and biological changes.

Structural changes can be introduced into the compound in order to probe how the compound itself exerts its biological effect. This insight, in turn, can help Boger’s group design simpler structures that have the same properties or to increase the potency or sensitivity of the natural structure.


A synthesis will yield more than just a final product. It will yield precursors, analogues, substructures, and useful chemistry along the way. Antibiotic analogues, like those of the vancomycin aglycon molecule, for instance, may be useful for treating infections with bacteria that are resistant to standard vancomycin. Other novel chemicals generated by a synthesis can be used for combinatorial chemistry screening to find compounds with biological activities against particular targets.

Also, new chemistry may be a by-product of Boger’s efforts. New synthetic methodologies and strategies can often be extended and generalized beyond any particular synthesis.

One of Boger’s well-known success stories has been his use of the hetero Diels–Alder reaction, powerful synthetic methodology which he has studied in detail for many years.

The reaction takes a compound containing a diene—conjugated four carbon chains with two doubly bonded carbons connected by a single bond—and combines them with a molecule containing a two-carbon doubly bound "–ene." Under suitable conditions, the six pi-orbital electrons in the two molecules react in such a way that the two molecules join and form a new, cyclic compound.

This type of reaction, which is called a cycloaddition, is a powerful tool for organic synthesis, since ring structures are a common feature in many target molecules and dienes are required motifs within precursor molecules.

The Diels–Alder reaction can simplify certain synthetic problems and help shortcut synthetic pathways, allowing sometimes complicated ring structures to be built in a single step. For many years, though, the reaction was limited to the all-carbon Diels–Alder reaction.

"Until we systematically explored it, the hetero Diels–Alder reaction, which contains hetero atoms in the diene, had not been applied in organic synthesis to any large extent," says Boger.

Boger has extended the scope of the reaction by using certain heterocyclic structures that naturally contain dienes, such as heteroaromatic azadienes and acyclic azadienes.

And like any good chemist, Boger spends time and energy perfecting his reactions and publishing the methodologies so that others can use it as a tool in cases where it applies.

"We do get a lot of enjoyment out of that," he says.

Go back to News & Views Index

“There are structures that we are working on today that 10 or 12 years ago would not have been considered realistic to prepare in the laboratory. And other things you can do today with 10, 20, or 1,000 times less material than it would have taken a decade ago.”















The Boger laboratory has made extensive use of the Diels–Alder reaction, which is shown in its simplest form above: the gas phase cycloaddition of Ethene and 1,3-butadiene to produce cyclohexene, for which Otto Diels and Kurt Alder shared the 1950 Nobel Prize in Chemistry.