Abstract
The question whether two proteins interact with each other or whether a protein localizes to a certain region of the cell is often addressed with fluorescence microscopy and analysis of a potential colocalization of fluorescence markers. Since a mere visual estimation does not allow quantification of the degree of colocalization, different statistical methods of pixel‐intensity correlation are commonly used to score it. We observed that these correlation coefficients are prone to false positive results and tend to show high values even for molecules that reside in different organelles. Our aim was to improve this type of analysis and we developed a novel method combining object‐recognition based colocalization analysis with pixel‐intensity correlation to calculate an object‐corrected Pearson coefficient. We designed a macro for the Fiji‐version of the software ImageJ and tested the performance systematically with various organelle markers revealing an improved robustness of our approach over classical methods. In order to prove that colocalization does not necessarily mean a physical interaction, we performed FRET (fluorescence resonance energy transfer) microscopy. This confirmed that non‐interacting molecules can exhibit a nearly complete colocalization, but that they do not show any significant FRET signal in contrast to proteins that are bound to each other.
Highlights
In the last two decades, we have seen an incredible progress in fluorescence microscopy techniques and in parallel a significant advance in methods to label proteinsAbbreviations: ANOVA, analysis of variances; DMEM, Dulbecco’s modified essential medium; ECFP, enhanced cyan fluorescent protein; ER, endoplasmic reticulum; EYFP, enhanced yellow fluorescent protein; FBS, fetal bovine serum; fluorescence energy transfer (FRET), fluorescence resonance energy transfer; HEK, human embryonic kidney cells; ICQ, intensity correlation quotient; JACOP, Just Another Colocalization Plugin; NA, numerical aperture genetically by fusing them to spectrally distinct fluorescent proteins
In a previous project [23], we had noticed that molecules, which exhibited clearly different localizations in the cell as assessed by fluorescence microscopy, still gave high numerical colocalization values with intensity-based coefficients such as Pearson’s or Manders’ correlation parameters
Human cells transfected with ECFP- and EYFP-tagged markers for cytoplasmic membranes, mitochondria, ER or transfected with EYFP alone, which spreads out through cytosol and nucleus, revealed the expected localization pattern in high-resolution confocal laser-scanning microscopy (Fig. 1A) with a clear separation of the two spectrally different markers
Summary
In the last two decades, we have seen an incredible progress in fluorescence microscopy techniques and in parallel a significant advance in methods to label proteins. Www.biotechnology-journal.com www.advancedsciencenews.com subcellular regions does not necessarily mean that they are physically binding to each other. Sophisticated novel techniques such as superresolution microscopy address these questions with better precision [1], they can still not discriminate unambiguously between functional interaction and incidental colocalization. These techniques require expensive equipment that is not available and rather complex to use
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
