News |  Sitemap |  Contact
PDF Export  | 

Object based colocalization analysis using ImageJ

Abstract:

Fluorescence imaging of two independently labeled proteins is commonly used to determine their co-localization in cells. Correlation between two such proteins suggests that the proteins interact, and motivates additional studies (using methods such as Foerster resonant energy transfer (FRET)), to confirm the the proteins are indeed close enough to interact. A number of approaches are in widespread use, including several popular plugins for ImageJ. Most of these approaches are based on calculation of correlation coefficients (for example, the Manders

coefficients) between two channels (generally red and green). This can be done without segmentation of the image, and generally with minimal noise reduction.

Lachanovich et al1[1] presented an object based colocalization strategy in which the images were segmented, and two collocation metrics were suggested: center of mass/particle containment and distribution of nearest neighbor distances. These metrics have two advantages:. First, they are independent of object intensity. In the case of fluorescence, intensity is often a poor parameter.[2] Second, they can be tested against a statistical model, which is required to establish their significance.

In this paper, we have updated and ported the approach of Ref. 1 to ImageJ, as a set of plugins which can be used either together or individually for colocalization analysis. These plugins consist of three sections. The first stage is point noise removal (median filtering), and background subtraction (rolling ball filter). Then, an

adaptive threshold is calculated for each object. The threshold is set at a percentage of the peak value of the object height above the noise

floor. Following segmentation, the object centers of mass are determined, and the colocalization metrics are calculated. If both classes of objects are punctate, then both the overlap and nearest

neighbor metrics may be used. If one of the two object classes is punctate, and the second is distributed, then only the overlap metric is appropriate. Finally, for images with reasonably homogeneous areas, we can create a randomized image whose stochastic properties are similar to the original image, in order to empirically verify

the significance of the colocalization results.

References:

[1] Lachmanovich, E., Shvartsman, D. E., Malka, Y., Botvin, C., Henis, Y. I., and Weiss, A. M., “Co-localization analysis of complex formation among membrane proteins by computerized fluorescence microscopy: application to immunouorescence co-patching studies,” J. Microscopy 212, 122-131 (2003).

[2] Pawley, J., “The 39 steps: a cautionary tale of quantitative 3-d

fluorescence microscopy,” Biotechniques 28, 884-886 (2000).


Keywords:
Colocalization, Fluorescence microscopy, Adaptive thresholding

Authors
Robert P. Dougherty and Aryeh M. Weiss

Organisation
School of Engineering, Bar Ilan University

Homepage

Short Biography
RPD: Robert Dougherty received his B.A. in physics from Rice University and his M.S. in physics and Ph.D. in applied mathematics from Iowa State University. He was a research engineer for The Boeing Company for 13 years. He currently develops aeroacoustic phased array techniques and systems for measurement of aircraft noise sources. His beamforming software takes the form of ImageJ plugins. He also written a number of ImageJ plugins that are used in biomedical image processing, and written and coauthored several papers in this field. A paper with coauthor K-H Kunzelmann received the 2008 Macres Award for the Best Instrumentation or Software Paper presented at Microscopy & Microanalysis 2007.

AMW:

Aryeh Weiss received his BSc, MSc, EE and PhD in electrical engineering from MIT. He was head of the Electronics Department at the Jerusalem College of Technology between 1995 and 2002. During this time he was a researcher in the the Institute for Life Sciences at the Hebrew University, where he set up the first confocal microscopy laboratory at the Givat Ram campus of Hebrew University. He is currently an associate

professor in the School of Engineering at Bar Ilan University, and he continues to work as a researcher in the Life Sciences Institute at Hebrew University.

Current projects include investigation of butyrate-induced differentiation in melanocytes and a study of the death of melanoma cells when subjected to photodynamic therapy.

Additional research interests include image processing as applied to microscopy, spectral microscopy, and superresolution microscopy.

© Luxembourg Institute of Science and Technology | Legal Notice