Elsevier

Acta Histochemica

Volume 119, Issue 3, April 2017, Pages 315-326
Acta Histochemica

A simple ImageJ macro tool for analyzing mitochondrial network morphology in mammalian cell culture

https://doi.org/10.1016/j.acthis.2017.03.001Get rights and content

Abstract

Mitochondria exist in a dynamic cycle of fusion and fission whose balance directly influences the morphology of the ‘mitochondrial network’, a term that encompasses the branched, reticular structure of fused mitochondria as well as the separate, punctate individual organelles within a eukaryotic cell. Over the past decade, the significance of the mitochondrial network has been increasingly appreciated, motivating the development of various approaches to analyze it. Here, we describe the Mitochondrial Network Analysis (MiNA) toolset, a relatively simple pair of macros making use of existing ImageJ plug-ins, allowing for semi-automated analysis of mitochondrial networks in cultured mammalian cells. MiNA is freely available at https://github.com/StuartLab. The tool incorporates optional preprocessing steps to enhance the quality of images before converting the images to binary and producing a morphological skeleton for calculating nine parameters to quantitatively capture the morphology of the mitochondrial network. The efficacy of the macro toolset is demonstrated using a sample set of images from SH-SY5Y, C2C12, and mouse embryo fibroblast (MEF) cell cultures treated under different conditions and exhibiting hyperfused, fused, and fragmented mitochondrial network morphologies.

Introduction

In live cells, mitochondria undergo fusion and fission to continuously re-model a dynamic, interconnected mitochondrial network that affects cellular functions beyond ATP homeostasis. Normal mitochondrial fusion and fission is essential for mtDNA maintenance (Chen et al., 2010), cell cycle progression, and metabolic regulation (reviewed in Lee and Finkel, 2013, Salazar-Roa and Malumbres, 2016). Dysregulation of mitochondrial dynamics underlies some heritable diseases, including Charcot-Marie-Tooth Type IIA (Tufano et al., 2015). The importance of mitochondrial dynamics to normal and aberrant cell function has motivated the development of tools for its analysis.

Fluorescence confocal microscopy of live or fixed cells has been the primary approach taken for capturing ‘snapshots’ of the mitochondrial network. Initial approaches to evaluating the resultant images included qualitatively ‘binning’ mitochondrial structures into categories of ‘fused’, ‘fragmented’, and ‘intermediate’ morphologies (Mitra et al., 2009). More recently, software-based approaches allowing more quantitative and objective analyses have been developed (Ahmad et al., 2013, Koopman et al., 2006, Leonard et al., 2015, Lihavainen et al., 2012, Nikolaisen et al., 2014). The approaches take on a variety of forms, incorporating machine-learning practices (Ahmad et al., 2013, Koopman et al., 2006, Leonard et al., 2015) and more direct approaches to measuring the mitochondrial features (Lihavainen et al., 2012, Nikolaisen et al., 2014). One drawback of these methods for mitochondrial network analysis is that their implementation relies on commercial software such as Matlab, ImagePlus Pro, and GE INCell. Open source alternatives do exist, such as MitoMap and Mito-Morphology, which run in ImageJ (Dagda, 2010, Vowinckel et al., 2015). MitoMap (http://www.gurdon.cam.ac.uk/institute-life/downloadspublic/imaging-plugins) is suitable for extracting physical information about mitochondrial structures, but only from high resolution three dimensional datasets which are not always available. Mito-Morphology (http://imagejdocu.tudor.lu/doku.php?id=plugin:morphology:mitochondrial_morphology_macro_plug-in:start) is strictly for two dimensional analysis. The Mito-Morphology set of macros uses a host of parameters determined from the area, perimeter, and elliptical fitting of fluorescently labelled features to capture the morphology of the mitochondrial material within a cell. However, it does not measure branching of mitochondrial networks, and is more useful for cells exhibiting larger, rounder mitochondria as several of the measurements are based on elliptical approximation.

Cultured, adherent mammalian cells often exhibit tubular connected mitochondrial morphologies and cannot always be imaged in three dimensions using super-resolution capable devices. This motivated our development of the Mitochondrial Network Analysis tool (MiNA), which uses the freely available FIJI distribution of the ImageJ platform and amalgamates open source tools into a simple macro toolset with a user-friendly interface (for more on FIJI and ImageJ, see Schindelin et al., 2012, Schneider et al., 2012). The code is readily accessible and can therefore be modified to suit the needs of the user. The macro is largely inspired by Nikolaisen et al. (2014). The goal of MiNA is to provide simplified image analysis methods to biologists on a familiar platform that is expandable and free. MiNA was developed to evaluate the extent of mitochondrial branching and therefore to distinguish between unbranched structures, like unbranched puncta and rods (individuals), and branched structures (networks). In addition to identifying networks, MiNA will evaluate the extent of branching within individual networks, thus identifying cells in which mitochondria are ‘hyper-fused’, i.e. have very highly branched mitochondrial networks. Since both highly fragmented mitochondria (associated with specific cell cycle phases, depolarization, mitophagy) and highly fused mitochondria (associated with specific cell cycle phases, states of high aerobic metabolic activity) have important implications in cell biology (Babbar and Sheikh, 2013, Galloway et al., 2012, Mitra et al., 2009, Rossignol et al., 2004), the ability to accurately identify these states is critical. Here we present and validate MiNA as a free open source tool for analyzing the extent of mitochondrial network fusion in cultured, adherent mammalian cells.

Section snippets

Materials

C2C12 cells, Dulbecco's Modified Eagle Medium (DMEM), Modified Eagle Medium (MEM) containing non-essential amino acids, and fetal bovine serum were purchased from Sigma–Aldrich (St. Louis, MO, USA). SH-SY5Y cells were obtained from ATCC (Manassas, VA, USA). Wild-type (wt) mouse embryonic fibroblasts (MEFs) and Mfn2-null MEFs were obtained from Jackson Laboratory (Bar Harbor, Maine, USA). MitoTracker Red CMXRos was purchased from Life Technologies (Burlington, ON, Canada). The mEmerald-Mito-7

Results and discussion

Images obtained from cultured SH-SY5Y, MEFs, and C2C12 cell lines were processed using the MiNA Single Image macro and the MiNA Batch Analysis macro to investigate the efficacy of the preprocessing and accuracy of the subsequent analysis on different cell types under different treatment conditions. A variety of treatment conditions and genotypes (mitofusin-2 null cells) known to be associated with enhanced network hyper-fusion or fragmentation were chosen to allow evaluation of how well MiNA

Conclusions

MiNA successfully identified and characterized morphological features of mitochondrial networks in multiple cell lines, and cells in which mitochondria were labelled by either mito-mEFP or MitoTracker Red. Fragmentation of the networks and reductions in mitochondrial abundance in response to FCCP uncoupling, hypoxia, and the absence of Mfn-2 was detected. In addition, MiNA could detect differences in network fusion resulting from resveratrol treatment.

It is important to note that, as

Acknowledgement

This work was supported by a Natural Science and Engineering Research Council (NSERC) Discovery Grant to JAS (RGPIN-2015-05645).

References (41)

  • I. Arganda-Carreras et al.

    3D reconstruction of histological sections: application to mammary gland tissue

    Microsc. Res. Tech.

    (2010)
  • M. Babbar et al.

    Metabolic stress and disorders related to alterations in mitochondrial fission or fusion

    Mol. Cell. Pharmacol.

    (2013)
  • G. Benard et al.

    Mitochondrial bioenergetics and structural network organization

    J. Cell Sci.

    (2007)
  • G.M. Cereghetti et al.

    Inhibition of Drp1-dependent mitochondrial fragmentation and apoptosis by a polypeptide antagonist of calcineurin

    Cell Death Differ.

    (2010)
  • Class UnsharpMask, n.d. Class UnsharpMask [WWW Document]. ImageJ Dev. API Doc....
  • R.K. Dagda

    Mitochondrial Morphology [WWW Document]

    (2010)
  • T. Ferreira et al.

    ImageJ User Guide

    (2012)
  • C.a. Galloway et al.

    Mitochondrial morphology – emerging role in bioenergetics

    Free Radic. Biol. Med.

    (2012)
  • P. Giraudoux

    pgirmess: Data Analysis in Ecology

    (2016)
  • Z. Guo et al.

    Parallel thinning with two-subiteration algorithms

    Commun. ACM

    (1989)
  • Cited by (0)

    View full text