I. Technical aspects of the study
2. Technical improvement of tools used in Hydra for live imaging
Solution: gene knockout via CRISPR-‐Cas9 electroporation
Gene knockout could be a possible way to obtain loss-‐of-‐function data.
Currently, CRISPR-‐Cas9 is a method to achieve gene deletions well established in numerous species. Since it was used for the first time to genetically edit a eukaryotic cell (Cong et al., 2013), CRISPR has been optimized for different model system including cnidarians such as Nematostella, in the Gibson lab:
(Ikmi et al., 2014), Hydractinia by the group of Uri Frank (Gahan et al., 2017).
Lately attempts were made to establish this method via electroporation in Hydra by the group of Thomas Holstein (Lommel et al., 2017). Stable gene knock-‐in and knock-‐out would be a great asset to strengthen Hydra as model system and in the same time to open the possibility to tackle longer biological processes such as regeneration.
2. Technical improvement of tools used in Hydra for live imaging
With this study, we developed novel approaches for in vivo imaging of early ROS signals during wound healing and regeneration. These experiments could be quantified and we were able to obtain significant data that contribute to characterizing ROS signaling in injured Hydra. At the time when I joined the project available tools were not sufficient to answer this biological question. Various
improvements, especially on the level of quantification of ROS signals and size of wound during the wound healing allowed us to fill in the previous gaps in knowledge of early injury-‐induced signaling in Hydra. Hydra tissue is optically transparent, which allows an easy tracking of labeled cells using different cellular dyes. Before I became in charge for this project, there were already several tools established to follow ROS production and wound healing in Hydra. At that time, the main weakness of the ROS imaging tools was their inability to produce quantitative data in case of mitochondrial ROS imaging, and non-‐satisfactory specificity of total ROS dye that was used to address H2O2 production after bisection. I extended this live imaging expertise acquired during my PhD, and as a side-‐project could develop a live imaging tool to monitor subcellular autophagy in Hydra.
a) The MitoSOX molecular dye is used to specifically label superoxide produced exclusively by mitochondria. It shows bright red fluorescence that can be followed with any fluorescent microscope. In the past, a fluorescent stereomicroscope was used for this type of experiments, and the resolution was not sufficient for signal to be quantified. Also, animals were not imaged while mounted, but rather orientated before each imaging time point, thus making time lapse impossible to perform. There was an additional problem with this approach; since orientation of polyps during imaging often makes them extend their wound or even damage them, which can lead to artificial data.
I overcame this challenge by developing a simple system to mount bisected animals in low-‐melting agarose and then image them using the inverted spinning disk confocal microscopy. Using this method, it is now possible to perform time-‐lapse videos and follow mitochondrial ROS production in high resolution that can be quantified (Explained more in Chapter 1: Material and Methods), and since spinning disk confocal microscope has significantly faster capture time, it is possible to follow several animals in the same time. For the imaging experiments that had several conditions, such as pharmacological inhibition tests, I used a special imaging microplate that has four separate chambers, allowing following up to four different conditions. Besides high-‐
resolution imaging, I also applied a simple quantification model that counts single dots that correspond to mitochondrial ROS signal, using Imaris imaging
software. These improvements led to several experimental aspects, which were not available before: longer time-‐lapse videos of ROS production, following of several animals and several conditions (particularly useful in pharmacological inhibition studies) and full quantification of imaging results.
b) The H2DCFDA molecular dye was previously used to follow in vivo the production of H2O2. The main weakness of this method was its lack of specificity together with the fact that the obtained signals are very hard to quantify. The H2DCFDA dye is non-‐specific since it labels many different types of ROS. To precisely follow H2O2, I optimized and applied a known quantification technique previously used in human cancer cells (Fu et al., 2006), human leukocytes (Mohanty et al., 1997), E. coli (Zhang et al., 2015), plants (Xiong et al., 2007), D. melanogaster (Venkatachalam et al., 2008) and mice (Wang et al., 2015). This method relies on specificity of both components of working solution for H2O2-‐mediated oxidation: Amplex UltraRed molecule and horseradish peroxidase enzyme (HRP). While being extremely sensitive and quantitative, the drawback of this technique is that it is a purely fluorescence reading-‐based method, meaning that it cannot be used to map H2O2 producing cells, which is the case and strength of live imaging.
c) The Phalloidin dye is used to label F-‐actin in the cell, useful to follow wound closure. It has been previously used in Hydra to follow wound healing and labeling conditions were established prior to my arrival to this project. My main concern was actually the imaging condition. Previously, after the Phalloidin labeling of bisected Hydra fixed at various time points, another cut was made to produce a donut shaped tissue that represented the wounded plane. Later on, this Hydra ‘’donuts’’ were mounted on slides and imaged. The problem was exactly in this step, since after the second cut, the bisected
‘’donut’’ could easily be inverted, and actually the wrong side of the cut could be imaged, but also with this type of mounting tissue is being pressed, and wound size cannot be correctly measured. To improve this part, instead of cutting polyps again to produce ‘’donut’’ shaped tissue piece, I mounted full Hydra halves in agarose and orientated them upside down to be perpendicular to the cover slip of imaging dish. With this, several improvements can be noted: the wound is imaged in almost native 3D condition, without tissue
pressing, and wound size can be easily calculated measuring the surface of the wound with available imaging software, such as ImageJ (see details in Chapter 1: Material and Methods).
d) As previously mentioned in the Introduction, the live imaging of the autophagy flux with a tandem fluorescent biosensor is now the standardized tool to monitor and quantify autophagy in the field. However, most researchers used it on cells maintained in culture (Huang et al., 2013; Suman et al., 2014; Perez-‐
Neut et al., 2016), while its use on whole animals or organs is limited. In zebrafish, autophagy can only be followed live during embryogenesis (Schiebler et al., 2015) and in mice, which was the first organism to be used to express this biosensor, the autophagy flux can only be followed in primary culture (Li et al., 2014). Here the autophagic flux can be followed with the tandem biosensor in vivo in intact, adult animals.
When electroporated in Hydra, this complex biosensor is surprisingly well expressed with tens of positive epithelial cells and easy to follow to obtain vast experimental data. The fluorescence of these chimeric proteins is very strong, making the quantification with the Imaris software quite robust. Using this experimental setup, we obtained crucial data that contributed to the analysis of the aging project carried out in the lab. Different expression of the biosensor in the strains Ho_CS and Ho_CR contributed to the elucidation in the differentiating them on the cellular level. Our data obtained with live imaging suggests that Ho_CS animals are not equipped with fully functional autophagy machinery as evidenced in Ho_CR and Hv_Basel in the homeostatic conditions, or when challenged with MG132, an autophagic flux enhancer (Chapter 3:
Figure 3). Another key discovery that this live imaging setup contributed to is the characterization of WIPI2(RNAi) phenotype. The WIPI2 protein is known in mammalian cells as the important part of the autophagic mechanism due to its binding to Atg16L1, which is a main actor in autophagosome formation (Dooley et al., 2014). Using our experimental setup, we have confirmed that in WIPI2(RNAi) Hydra the autophagy flux is severely impaired (Chapter 3: Figure 7).
Methodological improvements presented here were a major factor in obtaining invaluable data in this PhD project that contributed to our current knowledge on immediate injury-‐induced signaling during regeneration in Hydra.
3. Chemical tools used to characterize ROS metabolism during wound