Not so cold storage

One of the principal proteomics techniques for identifying and measuring human disease biomarkers is two-dimensional differential gel electrophoresis (2D-DIGE) which is commonly used for the analysis of plasma and serum and has been used extensively in clinical research. The protein abundances from two different conditions, such as diseased and healthy states, are compared to try and discover variations which are diagnostic of the disease.

When current standard operating procedures are followed, serum and plasma collected from patients should be stored in a freezer at -80°C, or preferentially in liquid nitrogen at -196°C. However, in practice, this eliminates many plasma samples from legitimate study because they are often routinely stored in freezers that operate at the higher temperature of -30°C.

Aware of the number of "lost" samples that this SOP creates, a team of Spanish researchers set about to see if the storage temperature made any difference in practice. Hector Escobar-Morreale and colleagues from the University Hospital Ramon y Cajal, IRYCIS, the University of Alacal, Madrid, and the Institute of Health Carlos III, Madrid, compared the proteomics profiles of plasma stored at -30°C and -80°C.

Portions of each plasma sample collected from eight women were stored at both temperatures for at least 18 months. After thawing out, the two major proteins, albumin and immunoglobulin G, were removed with a commercial ultrafiltration kit to prevent them from masking less abundant proteins during electrophoresis.

The proteins remaining in the plasma were labelled with the fluorescent CyDyes and separated by 2D gel electrophoresis. One set was labelled with the Cy3 dye and another with Cy5 to allow comparison of the protein abundances. By cross-labelling samples stored at each temperature with either of the dyes, the degree of experimental variability was determined. Only those protein abundance variations of more than 1.5-fold were regarded as significant.

An average of 1300 protein spots were detected in all of the images and an average of 770 of these were matched between the gels using the commercial software. The differences were attributed to spots not present in all gels, faint spots, noisy spots and dust particles. The research team considered an overlap of 60-80% to be adequate.

Protein spots that were present in at least 75% of the images were then compared to identify the significant variations in protein abundance between plasma stored at the two temperatures. Surprisingly few proteins were affected.

Four protein spots were more abundant in plasma stored at -30°C and five were less abundant. Subsequent mass spectrometric analysis identified eight of the nine spots and they all corresponded to the complement C3 precursor protein. The relative instability of this protein during freezing was tentatively attributed to the presence of three intrachain disulphide bonds, the presence of an internal thioester group and the possibility of autolytic cleavage.

A similar analysis was carried out for pooled plasma samples from the individuals. In this case, 15 spots were more abundant in the sample stored at -30°C and seven were less abundant. Again, nine of these corresponded to complement C3 precursor protein and 10 of the remaining spots corresponded to alpha-2-macroglobulin. One other spot was albumin and a further spot was a mixture of alpha-2-macroglobulin and complement factor H.

However, the researchers declared that the results in the pooled samples should not be regarded as reliable. Although pooling reduces individual variation, the total sample size is not increased, so the likelihood of false positives is increased. However, they did confirm the results from the individual samples.

So, the team concluded that there is very little to be gained by storing plasma at -80°C compared with -30°C, as long as the instability of complement C3 precursor protein is taken into consideration.

The study is not without its limitations, notably the fact that low-abundant proteins might not be detected by 2D-DIGE and proteins outside the pI and molecular weight ranges of the technique are overlooked.

Nevertheless, the researchers declared that "long-term storage at -30°C is appropriate for 2D-DIGE analysis of albumin- and IgG-depleted human plasma without losing the essential information about highly abundant proteins." This recommendation opens up the way for the analysis of clinical samples stored at -30 to -80°C, providing a wider range of samples for future proteomics studies of disease biomarkers.

The views represented in this article are solely those of the author and do not necessarily represent those of John Wiley and Sons, Ltd.