BOSTON, MASSACHUSETTS

When biologists need a particular peptide or strand of DNA, they ask a machine to make it for them. Automation gives fast, inexpensive access to these molecules, and it has helped turbocharge biology.

But constructing complex carbohydrates from sugar building blocks still needs expert chemists, with researchers spending weeks — and tens of thousands of dollars — forging molecules by hand.

Two groups of chemists now say they have independently developed automated carbohydrate synthesizers with the aim of making the molecules on demand. One machine is already churning out carbohydrates for clients, and the other will be available for labs to buy next year. Advocates say the machines could unplug a major bottleneck in the field of glycobiology, which studies the sugar chains (glycans) present on many proteins and lipids.

Click for a larger version.

The main problem with carbohydrate synthesis is that sugar building blocks attach to each other in myriad ways to create a dazzling variety of branched structures — in contrast to peptides, which are made from a linear chain of amino acids linked by identical chemical bonds. Making the right carbohydrate means avoiding all the unwanted possibilities.Click here for larger image

In 2001, chemist Peter Seeberger, who now directs the Max Planck Institute of Colloids and Interfaces in Potsdam, Germany, published the outline of a carbohydrate synthesizer that attached sugars one by one to a resin-linked chain (O. J. Plante, E. R. Palmacci and P. H. Seeberger Science 291, 1523–1527; 2001). The idea was based on solid-phase peptide synthesis, which won biochemist Bruce Merrifield the Nobel Prize in Chemistry in 1984; the resin holds on to the growing peptide or carbohydrate chain, with unreacted building blocks being washed away. But many researchers believed the chemistry behind Seeberger's proposal would be too complicated to be reliable, and the number of different sugars needed too great to manufacture. So Seeberger has spent the past decade completely remodelling his machine — overhauling the chemistry that links the sugars to the resin, for example.

Commercial machines will be available next year, he says. The device is being tested at Leiden University in the Netherlands and the University of Alberta in Edmonton, Canada. It is also being used by Ancora Pharmaceuticals, based in Medford, Massachusetts, a company co-founded by Seeberger to design and test carbohydrate-based vaccines. Seeberger says that 90% of known mammalian carbohydrates can be constructed using just 35 building blocks, which he plans to supply in bulk.

Seeberger presented the early fruits of research using his machines at this week's American Chemical Society meeting in Boston, Massachusetts. For example, he is imaging sugar-covered bacterial cells — and targeting drugs at them — by attaching a variety of glycans to nanoparticles.

Not everyone thinks Seeberger's technology is currently versatile enough to make a wide range of carbohydrates on demand. "I don't think the chemistry is sufficiently well developed," says Geert-Jan Boons at the University of Georgia in Athens. But it could build libraries of carbohydrates using well-established synthetic routes, he adds.

Meanwhile, a complementary system is already delivering shorter carbohydrates than those targeted by Seeberger. Developed by Nicola Pohl of Iowa State University in Ames, it uses a hydrophobic fluorocarbon tag, rather than a resin, to anchor growing sugar chains in solution. Pohl's company, LuCella Biosciences in Ames, has been filling orders to build carbohydrates since November 2009. Pohl hopes that automated machines will persuade more biologists to study glycans rather than being deterred by their complexity. "A lot of our early work is about educating the biologists that these carbohydrates are now readily available," she says.