Normalisation genes for expression analyses in the brown alga model Ectocarpus siliculosus

Aude Le Bail^*¹^,2 , Simon M Dittami^*¹^,2 , Pierre-Olivier de Franco¹^,2 , Sylvie Rousvoal¹^,2 , Mark J Cock¹^,2 , Thierry Tonon¹^,2 and Bénédicte Charrier¹^,2

¹UPMC Univ Paris 6, UMR 7139 Végétaux marins et Biomolécules, Station Biologique, F 29682, Roscoff, France

²CNRS, UMR 7139 Végétaux marins et Biomolécules, Station Biologique, F 29682, Roscoff, France

author email corresponding author email* Contributed equally

BMC Molecular Biology 2008, 9:75doi:10.1186/1471-2199-9-75


Published:	18 August 2008

Abstract

Background

Brown algae are plant multi-cellular organisms occupying most of the world coasts and are essential actors in the constitution of ecological niches at the shoreline. Ectocarpus siliculosus is an emerging model for brown algal research. Its genome has been sequenced, and several tools are being developed to perform analyses at different levels of cell organization, including transcriptomic expression analyses. Several topics, including physiological responses to osmotic stress and to exposure to contaminants and solvents are being studied in order to better understand the adaptive capacity of brown algae to pollution and environmental changes. A series of genes that can be used to normalise expression analyses is required for these studies.

Results

We monitored the expression of 13 genes under 21 different culture conditions. These included genes encoding proteins and factors involved in protein translation (ribosomal protein 26S, EF1alpha, IF2A, IF4E) and protein degradation (ubiquitin, ubiquitin conjugating enzyme) or folding (cyclophilin), and proteins involved in both the structure of the cytoskeleton (tubulin alpha, actin, actin-related proteins) and its trafficking function (dynein), as well as a protein implicated in carbon metabolism (glucose 6-phosphate dehydrogenase). The stability of their expression level was assessed using the Ct range, and by applying both the geNorm and the Normfinder principles of calculation.

Conclusion

Comparisons of the data obtained with the three methods of calculation indicated that EF1alpha (EF1a) was the best reference gene for normalisation. The normalisation factor should be calculated with at least two genes, alpha tubulin, ubiquitin-conjugating enzyme or actin-related proteins being good partners of EF1a. Our results exclude actin as a good normalisation gene, and, in this, are in agreement with previous studies in other organisms.