Compartmentation

Compartmentation is the great bug-bear of current metabolite profiling, and is a major barrier to linking the proteome with the metabolome. Most methods of measuring metabolites see only the sum of metabolites present in all the compartments of the tissue that is analysed.

Databases such as KEGG can draw very nice metabolic pathways linking metabolites with arrows and marking the enzymes that catalyse the interconversions, and the genes that encode them. These are wonderful tools. However, many (most?) enzymes exist as multiple isoforms in different cell-types, and in different parts of the cell. Glucose may be a substrate of hexokinase in the databases, but glucose in a stomatal cell is not a substrate of hexokinase in a mesophyl cell. Sucrose in the vacuole is not a substrate of invertase in the cytoplasm. Translocators are also a problem. They are just enzymes catalysing the reaction:

Thingy_inside ↔ Thingy_outside

The removal of the translocator by antisense, mutation, or whatever, is highly likely to affect the metabolite profile, even if only the sum of [Thingy] is measured, but all the same, wouldn't it be nice if we could measure the substrates of this enzyme, Thingy_inside, and Thingy_outside directly?

Unfortunately current methods for measuring compartmentation of metabolites are hard work. The current favourite method is non-aqueous fractionation, about which I can provide more information if necessary, but so far as I know no one at the John Innes Centre currently carries out this time-consuming and difficult technique.

Like many localisation methods it is based on separating things down a gradient and comparing the profile of the Thingy you are interested in with the profiles of marker enzymes marking each of the compartments Thingy might be found in.

Further information is available on the mathematics of gradient separations.