by ecancer reporter Clare Sansom

Tumours are characterised by the gradual accumulation of mutations that alter the function of proteins, leading to accelerated cell proliferation and growth.

Many of these are gain-of-function mutations in metabolic enzymes.

The amount of gene and protein sequence data associated with cancer has increased exponentially in recent years with improvements in sequencing technology.

This data is now finding a novel and perhaps surprising use as a source of enzyme functional diversity for the biotechnology industry.

Many processes in organic synthesis can be improved by using enzymes with novel functions. Hai Yan and his co-workers at Duke University, Durham, North Carolina, USA, first proposed the general principle that studies of cancer-causing mutations could suggest changes to enzymes used industrially that would lead them to produce novel and potentially useful metabolites.

They have now tested this hypothesis by developing a mutated enzyme to overcome a key problem in the clean biosynthesis of a dicarboxylic acid, adipic acid.

Adipic acid, a substrate for the synthesis of nylon, is one of the most widely used organic chemicals worldwide. It is therefore essential that all processes involved in its synthesis are both non-toxic and affordable; there is also much interest in developing production processes that minimise the need for fossil fuel substrates.

It has proved particularly difficult to improve the process for one important step in the synthesis of this compound, the conversion of 2-oxoadipate to (R)-2-hydroxyadipate.

The researchers observed that carcinogenic gain-of-function mutations in the human enzyme isocitrate dehydrogenase (IDH) might prove a useful guide in designing an enzyme to catalyse this reaction. Isocitrate dehydrogenases are ubiquitous enzymes that catalyse the conversion of (2R,3S)-isocitrate to 2-oxoglutarate.

Missense mutations in active site residues of IDHs have been observed in several cancer types including gliomas. These mutations convert the enzymes from dehydrogenases to oxidoreductases that catalyse the conversion of 2-oxoglutarate to (R)-2-hydroxyglutarate. An oxidoreductase with this type of activity but that favoured substrates with six-carbon, rather then five-carbon backbones would catalyse the 2-oxoadipate to (R)-2-hydroxyadipate conversion.

Yan and his co-workers hypothesised that mutating residues equivalent to those mutated in IDH in the homologous enzyme homoisocitrate dehydrogenase (HIDH) would create a synthetic enzyme with the properties required.

They compared the active site structures of human IDH with HIDH from the yeast Schizosaccharomyces pombe and identified two arginine residues that appeared analogous to residues altered in the cancer-causing mutants.

Therefore, they generated and purified mutant forms of the HIDH enzyme with point mutations equivalent to those that had been observed in IDH in tumours:  R114Q, R143C, R143H and R143K. They then tested each mutant and the wild type enzyme to see whether it would convert 2-oxoadipate to (R)-2-hydroxyadipate, assessing the reactions by first measuring the decrease in the concentration of the co-factor NADH.

The enzymes containing the R143C and R143H mutations, but not the R114QA mutant or the wild type, were found to reduce NADH concentrations significantly; the R143K mutant had some limited activity.

The substrate-binding constants for the R143H mutant were roughly equivalent to those for the wild type enzyme, although its catalytic turnover rate was only about 1% of that of the wild type. Mass spectroscopy was used to confirm the production of 2-hydroxyadipate by the mutant enzymes; in contrast, and as expected, the mass spectrum of the products of the wild type reaction was consistent with homoisocitrate.

Quantitative LC/MS/MS then showed over 99% of the 2-hydroxyadipate produced by the mutant enzyme reactions to be the R enantiomer. Similar results were observed when equivalent mutations were introduced into the active site of an HIDH from a thermophilic bacterium, Thermus thermophilus.

These results provided an important proof of the principle that it is possible to use known cancer-causing mutations in a metabolic enzyme to redesign similar enzymes so that they catalyse an industrially important reaction.

Although the reaction rates of the mutant enzymes were slow, Yan and his co-workers suggested that it should be possible to improve them further through directed evolution. The enormous amount of mutation data obtained from cancer genome sequencing is likely to yield further enzyme diversity and thus more important insights for molecular design.

Reference

Reitman, Z.J., Choi, B.D., Spasojevic, I., Bigner, D.D., Sampson, J.H. and Yan, H. (2012). Enzyme redesign guided by cancer-derived IDH1 mutations. Nature Chemical Biology, published online ahead of print 23 September 2012. doi: 10.1038/NChemBio.1065