Cytel Demonstrates Process for Difficult-to-Manufacture Carbohydrates

Only months after announcing a breakthrough that could allow the large-scale manufacture of therapeutic carbohydrates, then receiving a landmark patent for the process, Cytel Corp. (San Diego, CA) has parlayed this technology into a production-worthy process for making a human breast milk carbohydrate at a reasonable cost. Cytel and collaborators from the Institute for Biological Sciences at the National Research Council of Canada (Ottawa, Ontario), used an artificial enzyme which, when employed within Cytel's proprietary Sugar Nucleotide Cycling (SNC) process, enabled the synthesis of a difficult-to-produce carbohydrate in yields exceeding 98%. Cytel is producing the biologically active material, sialyl lactose, in 15-kg batches, quantities which are about 1000 times larger than those produced by conventional carbohydrate synthesis.

Sialyl lactose inhibits microbial attachment to human cells, especially for such bugs as helicobacter pylori (ulcers), streptococcus species (strep throat), and rotavirus. This carbohydrate also has applications in oral hygiene and consumer products.

Jim Paulson, Cytel's chief scientific officer, boasted typical chemical yields greater than 99% using SNC, since "our process allows us to run the reactions until they are complete, to the point where we cannot detect starting materials." Downstream processing results in purified yields of around 70%, but Paulson points out that purification is eminently scalable since it uses no chromatography.

Cytel has demonstrated the marriage of SNC and fusion proteins in the production of 15 kg of sialyl lactose, but Paulson points out that there are no practical limits to the scalability of this process. "We're already thinking in terms of 10 to 100 ton-per-year production of carbohydrates for foods, pharmaceuticals, and other industries, and are discussing the process with companies interested in those quantities. We do this as a batch process, so to run a larger batch we just need a bigger bucket. The limitations will come in the downstream processing, but even here all the operations are easy to scale up: ultrafiltration, microfiltration, ion exchange (batch, not chromatographic) to remove trace metal ions, and activated charcoal for decolorization."

Sugar Nucleotide Recycling and Fusion Proteins

The synthesis of carbohydrates is daunting at the bench scale, and probably impossible at production levels without using enzymes. For example, a simple carbohydrate made from four different sugar units presents 34,560 possible isomers. Pharmaceutical chemists, especially academics, take great relish in overcoming these herculean stereo- and regiochemical obstacles, but this level of uncertainty gives production engineers nightmares.

Yet the need for efficient carbohydrate production methods has grown significantly in recent years due to the increasing potential for their use in a variety of markets including pharmaceuticals, nutritional supplements, and consumer products.

Sugar nucleotide cycling is used to assemble complex, specific bioactive carbohydrates on a commercial scale and is also useful for synthetic remodeling of carbohydrates or recombinant glycoproteins. The technology imitates natural carbohydrate biosynthesis, which requires a cascade of chemical reactions initiated by specialized enzymes. The method is more time- and cost-effective than other chemical synthesis technologies.

Producing sialyl lactose involves several enzymes. One enzyme, sialyl transferase, couples sialic acid directly and specifically to the desired position on the lactose molecule. Another critical enzyme prepares the lactose for this coupling reaction by joining it to a nucleotide, which subsequently is ejected and recycled. Unfortunately, the sialyl transferase enzyme is quite unstable, especially in production settings, which is why up to now no-one has produced more than a few grams of sialyl lactose in any one batch. Through genetic engineering NRC scientists fused the transferase with the lactose-preparing enzyme to form a new enzyme, dubbed a "fusion protein," that exhibited the activities of both critical enzymes and was stable enough for large-scale batch processing.

These two benefits—stability and combining two enzymes into one—would not be useful industrially if the enzyme were difficult to generate. Most enzymes used in carbohydrate synthesis, in fact, are mammalian enzymes that must be generated either in mammalian or insect tissue culture, which is prohibitively expensive for all but the highest value-added pharmaceuticals. The NRC/Cytel fusion protein uses bacterial genes and is expressed in E. coli, which makes it inexpensive. The result, according to Paulson, is a generic manufacturing technology cheap enough for any pharmaceutical and appropriate even for food-grade carbohydrates.

Wide Industrial Applications

"We are especially pleased to see that this technology can be used effectively for the synthesis of commercially important bioactive carbohydrates," Paulson said. "As companies move into late-stage product development and commercialization of carbohydrates, the mass production of these compounds will be critical. With SNC, a new manufacturing facility in operation, and the collaborative partnership program we have established through our Glytec division, Cytel is ideally prepared to meet these needs." In addition, Cytel hopes that fusion proteins will become valuable tools for enhancing other multi-enzyme industrial processes as well.

For more information: Jim Paulson, Chief Scientific Officer, Cytel Corp., 3525 John Hopkins Ct., San Diego, CA 92121. Tel: 619-552-3000.

By Angelo DePalma