Mitochondria-Targeted Green Fluorescent Protein for Quantitative Monitoring of Mitochondrial Morphology in Living Cells

Abstract

Mitochondrial morphology within a cell appears to be correlated with cellular function. Quantitative measurement of its morphology is therefore essential in order to understand the coupling in detail. In addition to mitochondria‐selective dye such as rhodamine 123 (RH123), a recent progress in genetic engineering allowed the microscopic observation of mitochondria with a matrix‐targeted green fluorescent protein (GFP). Here we evaluated these labeling procedures for the ability to analyze mitochondrial morphology quantitatively. In contrast to no cytosolic GFP, fluorescence of RH123 in cytosol was intense and peripheral mitochondria with weak fluorescence were masked by thresholding without any filters. Although applying a top hat filter improved binarized images for RH123 remarkably, the mitochondrial structures differed slightly from those in raw images. When form factor, aspect ratio, mean area and number of mitochondrion per cell were analyzed from the binarized images, these parameters for GFP were stable to 16 times of repetitive recordings but mean area for RH123 decreased by 30% probably due to photo‐bleaching. We concluded that expression of matrix‐targeted GFP was the optimal procedure to monitor mitochondrial morphology quantitatively because of photo‐stability and precise localization.

Live-Imaging Readouts and Cell Models for Phenotypic Profiling of Mitochondrial Function

“…Protein-based probes targeted to the mitochondria, such as mito-GFP, are also a valid tool to study mitochondrial morphology and dynamics (Rizzuto et al, 1995; Nomura et al, 2009), however, these do not allow the simultaneous study of Δψ.…”

Image Processing Methods for Quantitative Analysis of Mitochondrial DNA Dynamics

“…In this review, theoretical background of the typical methods and several results of the pilot experiment are introduced. fluorescence imaging systems (e.g., laser scanning microscope, LSM) are adopted [22]. As shown in Figure 2A, the high quality image data were obtained with FluoView FV1000 Olympus and were used in this paper.…”