Open Access Open Badges Methodology article

Toponomics method for the automated quantification of membrane protein translocation

Olga Domanova123*, Stefan Borbe1, Stefanie Mühlfeld4, Martin Becker4, Ralf Kubitz4, Dieter Häussinger4 and Thomas Berlage123

Author Affiliations

1 Fraunhofer Institute for Applied Information Technology FIT, Schloss Birlinghoven, Sankt Augustin, Germany

2 RWTH Aachen University, Information Systems Group (Informatik 5), Ahornstraße 55, Aachen, Germany

3 Bonn-Aachen International Center for Information Technology B-IT, Dahlmannstraße 2, Bonn, Germany

4 Heinrich-Heine University of Düsseldorf, Medical Faculty, Department of Gastroenterology, Hepatology and Infectiology, Moorenstr. 5, Düsseldorf, Germany

For all author emails, please log on.

BMC Bioinformatics 2011, 12:370  doi:10.1186/1471-2105-12-370

Published: 19 September 2011



Intra-cellular and inter-cellular protein translocation can be observed by microscopic imaging of tissue sections prepared immunohistochemically. A manual densitometric analysis is time-consuming, subjective and error-prone. An automated quantification is faster, more reproducible, and should yield results comparable to manual evaluation. The automated method presented here was developed on rat liver tissue sections to study the translocation of bile salt transport proteins in hepatocytes. For validation, the cholestatic liver state was compared to the normal biological state.


An automated quantification method was developed to analyze the translocation of membrane proteins and evaluated in comparison to an established manual method. Firstly, regions of interest (membrane fragments) are identified in confocal microscopy images. Further, densitometric intensity profiles are extracted orthogonally to membrane fragments, following the direction from the plasma membrane to cytoplasm. Finally, several different quantitative descriptors were derived from the densitometric profiles and were compared regarding their statistical significance with respect to the transport protein distribution. Stable performance, robustness and reproducibility were tested using several independent experimental datasets. A fully automated workflow for the information extraction and statistical evaluation has been developed and produces robust results.


New descriptors for the intensity distribution profiles were found to be more discriminative, i.e. more significant, than those used in previous research publications for the translocation quantification. The slow manual calculation can be substituted by the fast and unbiased automated method.