An automated quantitative image analysis pipeline of in vivo oxidative stress and macrophage kinetics

Abstract

Macrophage behavior is of great interest in response to tissue injury and promotion of regeneration. With increasing numbers of zebrafish reporter-based assays, new capabilities now exist to characterize macrophage migration, and their responses to biochemical cues, such as reactive oxygen species. Real time detection of macrophage behavior in response to oxidative stress using quantitative measures is currently beyond the scope of commercially available software solutions, presenting a gap in understanding macrophage behavior. To address this gap, we developed an image analysis pipeline solution to provide real time quantitative measures of cellular kinetics and reactive oxygen species content in vivo after tissue injury. This approach, termed Zirmi, differs from current software solutions that may only provide qualitative, single image analysis, or cell tracking solutions. Zirmi is equipped with user-defined algorithm parameters to customize quantitative data measures with visualization checks for an analysis pipeline of time-based changes. Moreover, this pipeline leverages open-source PhagoSight, as an automated keyhole cell tracking solution, to avoid parallel developments and build upon readily available tools. This approach demonstrated standardized space- and time-based quantitative measures of (1) fluorescent probe based oxidative stress and (2) macrophage recruitment kinetic based changes after tissue injury. Zirmi image analysis pipeline performed at execution speeds up to 10-times faster than manual image-based approaches. Automated segmentation methods were comparable to manual methods with a DICE Similarity coefficient > 0.70. Zirmi provides an open-source, quantitative, and non-generic image analysis pipeline. This strategy complements current wide-spread zebrafish strategies, for automated standardizations of analysis and data measures.

The Effect of Fluence on Macrophage Kinetics, Oxidative Stress, and Wound Closure Using Real-Time In Vivo Imaging

“…All time-lapse images were acquired using ZEN Ò software with a Zeiss LSM 510 confocal microscope, as described previously. 33 Fluorescent excitations 488 nm for FIG. 1.…”

“…All time-lapse images were analyzed using open source Zirmi software in MATLAB R2015a, as described previously. 33 Quantification of cell kinetics was based on centroid positions, not whole cell morphology. Macrophage centroid positions were tracked using PhagoSight software's keyhole algorithm and visualization tools for proofing.…”

“…Time intervals were defined as T1 [30- To analyze the extracted data more efficiently, wound regions were normalized, orienting positions with respect to the epicenter or zero position (S o ) of the wound margin for discrete distance to wound determinations. 33 In unwounded fish and for baseline measurements, the most distal position of the caudal fin was selected as the zero position. Zirmi software was used to manually trace wound region to isolate background subtracted DHE mean fluorescent intensity values *65 lm radially extended from the wound margin.…”

“…Zirmi software was used to manually trace wound region to isolate background subtracted DHE mean fluorescent intensity values *65 lm radially extended from the wound margin. 33…”

“…The zebrafish model presents powerful transgenic phenotypes to facilitate in vivo real-time characterization of macrophage recruitment and oxidative stress after injury. [26][27][28][29][30][31][32][33] Insights into leukocyte migration behavior and reactive oxygen species generation at zebrafish wound sites have yielded high impact discoveries, providing novel insights translatable to mammalian systems. 31,32,[34][35][36] To simultaneously track individual macrophage trajectories and tissue oxidative stress levels, zebrafish transgenic for a fluorescent macrophage marker, Tg(mpeg-dendra2), underwent dihydroethidium (DHE) loading.…”