Gel filtration chromatography should not be relied upon in isolation.

Gel filtration chromatography should not be relied upon in isolation.

A method of molecular weight (MW) determination used by thousands of researchers worldwide could be responsible for some alarmingly inaccurate data, report researchers in Taiwan and the US.

The controversy surrounds Superose 12, a gel filtration medium designed primarily for high resolution chromatography separation and purification of proteins across a wide MW-range. Although not specifically designed for MW determination, says manufacturer GE Healthcare, it can be used for that purpose when properly calibrated.

Researchers at the University of California, Davis (UCD), US, and Da-Yeh University, Taiwan, found that an inhibitor of the enzyme α-amylase had a MW of about 50 kDa when measured by laser-assisted time-of-flight mass spectrometry, polyacrylamide gel electrophoresis, or another gel-filtration system produced by GE Healthcare, Sephadex G-100. The same protein had an MW of only around 20 kDa using Superose 12 gel filtration.

’Since Superose 12 is widely used, there could be many mistakes in determining molecular weights,’ said John Whitaker, professor emeritus at the UCD Food Science and Technology Department. ’We worked to determine the problem so that others would not rely entirely on the use of Superose 12,’ he told Chemistry World.

It is ’highly atypical’ that Superose 12 would yield a result deviating as much as 50 per cent from other methods, says Allan Simpson, vice president, research & analysis protein separations at GE Healthcare. He is disappointed that Whitaker’s conclusions were drawn from the performance of just one Superose 12 column.

’Superose columns are very durable and may be used for hundreds of runs without effect,’ he said. ’However, there is no reference in the paper to how many runs this particular column had been subjected to, or the conditions applied to the column during any previous runs.’

It is even possible, though unlikely given the magnitude of deviation, that the proteins being investigated interact ’in a hitherto unknown way’ with the agarose matrix in the column, suggests Simpson.

’In the absence of a detailed review of the data with the researchers, our comments are of course hypothetical,’ he conceded.

Bea Perks