Synthesis Takes the Shuttle

In a development that could point the way toward the drug discovery laboratory of the future, a chemistry group at an industrial pharmaceutical research facility has designed, constructed, and demonstrated a solution-phase organic synthesis system with a level of automation that may be unprecedented. Whether it's worth the investment is not clear.

The system, called SynCar, uses a set of four shuttles to transport glass reagent vessels to a series of workstations where individualized synthesis, cleanup, and analysis steps are carried out. The shuttles resemble railroad tracks, except that the reagent vials hang below the tracks rather than ride on top. Workstation functions include temperature control and liquid-reagent handling for chemical synthesis; filtration, liquid-liquid extraction, solid-phase extraction, and evaporation for purification; and weighing and high-performance liquid chromatography/mass spectrometry for product analysis. The system generates purified and characterized products that can be screened for desired properties, such as affinity to molecular targets of interest.

Angelika Weber, Erich von Roedern, and Hans Ulrich Stilz of the High Throughput Parallel Synthesis Unit at Sanofi-Aventis, in Frankfurt am Main, Germany, developed SynCar, in collaboration with Accelab GmbH, an engineering firm in Kusterdingen, Germany.

The Sanofi-Aventis group reports on the design of the system and its use to automate amide-coupling reactions in the latest issue of the Journal of Combinatorial Chemistry (2005, 7, 178). "It has tremendously improved our productivity," says Stilz, head of chemical sciences at the research facility.

Researchers in the field who were asked to comment on SynCar say it is innovative and perhaps groundbreaking, but some point out that the reported level of efficiency improvement is so far modest rather than spectacular and that SynCar's apparent high cost may impede broader use of similar systems by others. The Sanofi-Aventis group hasn't revealed SynCar's price tag, but photos of the room-sized system suggest that it's anything but cheap.

WHEN THE field of combinatorial chemistry began to pick up steam in the 1980s, the first generation of commercial automated synthesizers that emerged to meet the needs of synthetic chemists were primarily solid-phase synthesis systems, which closely resembled earlier solid-phase peptide and nucleic acid synthesizers. In solid-phase synthesis, reagents are bound to a solid support, making it easier to carry out multistep reactions and to isolate pure products from reagents and by-products. Well-established solution-phase chemistry often has to be adapted for use in solid-phase synthesis protocols, and this process can be difficult and laborious.

Because it is so much simpler to use traditional solution-phase protocols than it is to adapt them to solid-phase synthesis, combinatorial and high-throughput labs have gravitated toward the predominant use of solution-phase synthesis in the past few years. Instrument manufacturers tried to adapt to this trend by modifying automated solid-phase synthesizers for solution-phase synthesis. Most of these adapted systems ultimately proved unpopular and commercially unsuccessful. Chemspeed, in Augst, Switzerland, is "the only remaining vendor" for solution-phase automation systems, the Sanofi-Aventis researchers note in their paper.

The dearth of commercial options for automated solution-phase synthesis spurred the Sanofi-Aventis scientists to design and develop their own systems. The first of these was a robotic synthesizer they created in the late 1990s, which had inadequate productivity and versatility, in their view. Seeing some ways it could be improved, they went on to design and develop SynCar, with its multiple-track system and modular workstations. SynCar isn't available commercially, and there are no current plans to make it available.

An innovative aspect of SynCar is that "software tracks the whole system," Stilz says. "Each flask can be programmed, so you can fine-tune reaction conditions for each individual vessel. It's a modular system, so you can always expand it to add new operations."

SynCar was designed to move routine chemical synthesis out of the medicinal chemistry lab so medicinal chemists could concentrate on more sophisticated research, such as developing new synthetic schemes and new molecular scaffolds, Stilz says. Currently, SynCar can make about 20,000 compounds a year, whereas Sanofi-Aventis' earlier robotic system worked at about one-third that rate.

"It was, however, not primarily the throughput in terms of number of compounds a year that attracted us to build this system," Stilz says. "The major value of the system is that it allows us to produce individual medicinal chemistry compounds of greater than 90% average purity and in 50- to 100-mg quantities to enable broader biological profiling without a need for resynthesis."

John C. Reader, founder and vice president of chemistry at Sareum, Cambridge, England, calls SynCar "a step forward. It takes away a lot of manual and tedious steps that other groups have to do." Sareum is a structure-based drug discovery company that uses automated solution-phase chemistry extensively.

"The software side of it sounds particularly impressive. It makes the most efficient use of the modules in the system, which sounds like a fairly complex task," Reader says. He notes, however, that "the standard amide chemistry the Sanofi-Aventis group describes in their paper is not particularly interesting. For the system to be really useful, I'd want to see some examples of the other kinds of chemistry they've done with it, which they allude to" in their paper.

"NOT MANY people are going to be able to reproduce the system because a huge amount of resources has gone into generating it," Reader says. "And the actual throughput the Sanofi-Aventis researchers quote is not massively impressive, so I wonder if they're getting the value from their investment."

Daryl R. Sauer, senior group leader in charge of high-throughput organic synthesis at Abbott Laboratories, Abbott Park, Ill., says: "The SynCar system is definitely a significant step in automated synthesis. It's probably the most comprehensive and automated solution-phase synthesis platform that's ever been built."

But he has reservations as well. "I'm not sure if the system is a lot more productive than what four full-time employees could do using commercial off-the-shelf automation equipment in a modular approach," Sauer says. "Typically, in a high-throughput lab you can buy equipment from Chemspeed that will give you the same level of productivity [as that reported in the paper] with four people"--the number of people the Sanofi-Aventis group is using to operate the SynCar system.

He also notes that Chemspeed's equipment can handle solid additions to reaction mixtures, whereas the SynCar system apparently cannot. "That's kind of important, as there are a lot of useful heterogeneous reagents," he says. Weber, von Roedern, and Stilz point out, however, that SynCar is modular and totally reconfigurable, so modules with different functions can always be added.

Research fellow Ronald N. Zuckermann of Chiron Corp., Emeryville, Calif., who specializes in combinatorial discovery and peptoid research, calls SynCar "a tremendous engineering achievement. Having built and run our own custom automated synthesizers at Chiron, I am very impressed with the planning and thought that went into this. It should have a great long-term impact on producing a consistent stream of high-quality compounds."

On the other hand, "the success of such a complex, expensive, and centralized resource will be driven by economic factors, not science or engineering ones," he adds. "The cost to manage, maintain, and operate the system will be high and will have to be weighed against the advantages of distributing smaller personal workstations to more of the workforce."

Andreas Termin, director of chemistry at Vertex Pharmaceuticals, San Diego, points out that other companies are working on similar industrialized library production facilities but that the SynCar paper "is the first publication that describes in detail the efforts that went into the setup of a fully integrated system and the lessons learned. The system is definitely an engineering feat and shows a great deal of collaboration between Sanofi-Aventis and its engineering partner. Also, the modularity of the system is a plus, and I believe that some components will find broader application in the industry."