LiverZap: a chemoptogenetic tool for global and locally restricted hepatocyte ablation to study cellular behaviours in liver regeneration

  1. Elizabeth M. G. Ambrosio
  2. Charlotte S. L. Bailey
  3. Iris A. Unterweger
  4. Jens B. Christensen
  5. Marcel P. Bruchez
  6. Pia R. Lundegaard
  7. Elke A. Ober

Reviewed by

Read the full article

Listed in

Log in to save this article

Abstract

The liver restores its mass and architecture after injury. Yet, investigating morphogenetic cell behaviours and signals that repair tissue architecture at high spatiotemporal resolution remains challenging. We developed LiverZap, a tuneable chemoptogenetic liver injury model in zebrafish. LiverZap employs the formation of a binary FAP-TAP photosensitiser followed by brief near-infrared illumination inducing hepatocyte-specific death and recapitulating mammalian liver injury types. The tool enables local hepatocyte ablation and extended live imaging capturing regenerative cell behaviours, which is crucial for studying cellular interactions at the interface of healthy and damaged tissue. Applying LiverZap, we show that targeted hepatocyte ablation in a small region of interest is sufficient to trigger local liver progenitor-like cell (LPC)-mediated regeneration, challenging the current understanding of liver regeneration. Surprisingly, the LPC response is also elicited in adjacent uninjured tissue, at up to 100 µm distance to the injury. Moreover, dynamic biliary network rearrangement suggests active cell movements from uninjured tissue in response to substantial hepatocyte loss as an integral step of LPC-mediated liver regeneration. This precisely targetable liver cell ablation tool will enable the discovery of key molecular and morphogenetic regeneration paradigms.

Article activity feed

  4. Version published to 10.1242/dev.202217
  5. Review Commons

    Note: This rebuttal was posted by the corresponding author to Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Reply to the reviewers

    We appreciate the positive feedback from both reviewers and their critical comments, which will help us to improve the manuscript. Below, we provide a point-by-point response and how we propose to address their queries and comments.

    As the laboratory is currently undergoing a major transition, we propose essential experiments that are realistic to perform under these circumstances. We are positive that we can address all the most critical points identified by the reviewers.

    Suggestions for minor changes to the figures are already included.

    We also include responses to questions the reviewers raise.

    Reviewer #1 (Evidence, reproducibility and clarity):

    Major comments:

    1 - The authors assessed acridine orange incorporation in BECs upon LiverZap and concluded that LiverZap triggers hepatocyte-specific cell death without a bystander effect in adjacent cells (Figure 1 D-E). What happened to endothelial cells, which could also be affected either directly by ROS production in hepatocytes or indirectly by gross morphological changes in tissue organization?

    Response:

    The reviewer raises two excellent points:

    (i) Bystander effect of hepatocyte-produced ROS on endothelial cells:_ the cell death analysis included in the manuscript, shows that Acridine Orange staining overlaps with hepatocyte tg(fabp10a:dsRed) expression, but not with biliary tg(tp1:H2B-mCherry) indicating cell death is specific to hepatocytes. Moreover, singlet oxygen species as produced by LiverZap have been shown to have a very short half-life and short range of action, suggesting that neighbouring cells are unlikely affected (Liang et al., 2020).

    PLAN: Investigate potential bystander effect in endothelial cells, by activating the LiverZap tool in livers expressing transgenic tg(kdrl:mcherry) marking the vascular network followed by live staining with Acridine Orange at 8 hours post illumination.

    (ii) indirect effects of morphological tissue changes on endothelial cells:_ studying the tissue response of the vascular network to hepatocyte ablation would be very interesting. A separate and detailed study would be required to generate meaningful data and insights into said process. This could encompass for instance the use of the transgenic endothelial tg(kdrl:mcherry) line for LiverZap experiments and parallel those in Figure 4A-S. Thus, it seems beyond the scope of this work.

    NO CHANGES PROPOSED.

    2 - The evaluation criteria for distinguishing mCherry and cells in imaging experiments should be clearly described in the methods section. The authors should also provide some quantitative data regarding the level of correlation between the mCherry hepatocytes and the BEC-derived hepatocytes strictly defined based on the TP1-H2B-EGFP lineage tracing, as the former was used as a surrogate marker for the latter in some experiments.

    Response:

    Here, we believe the reviewer refers to the Tp1:H2B-mCherry-based lineage tracing, since the tg(tp1:egfp) line has not been used for this purpose. Similar to previous studies in the regeneration field (e.g. Choi et al., 2014; He at al., 2014), we have used histone inheritance of Tp1:H2B-mCherry for short-term lineage tracing. Tp1:H2B-mCherry-based lineage tracing was assessed on the whole organ level, for which we will describe the quantification pipeline. Tp1:H2B-mCherrylow cells were identified as BEC-derived hepatocytes after severe hepatocyte ablation, as shown in Fig. 2A,C, correlating with hepatocyte marker tg(fabp10a:GFP) expression. Tp1high and Tp1low cell numbers were quantified for 12, 24, 48 and 72 hpi and can be added as supplementary information.

    PLAN: Update the material and methods section and produce a more detailed description. This would include the following information: Whole-mounted livers of tg(tp1:H2B-mCherry) fish were stained for mCherry and imaged using an Leica SP8 confocal microscope. Image processing was carried out using the Imaris software. All mCherry-expressing cells in the liver were masked using the “spots” function, which allows quantification of signal intensity of all cells, represented by a sphere. Tp1high and Tp1low cells were identified using an automatically generated intensity threshold. Due to intensity differences with increasing imaging depth/z-position, segmented Tp1high cells were manually curated.

    To showcase the analysis strategy, we propose to include an example showing original image data, semi-automated quantification at the surface and deep tissue levels, as well as the overall Tp1:H2B-mCherry intensities for all positive cells and specifically Tp1high cells for all z-positions of an entire liver (see data figure below). This example could be included as supplementary data. Likewise, cell number quantification for Tp1high and Tp1low across regeneration can be added to Fig. S2.

    Fig. Quantification of Tp1:H2B-mCherryhigh cells. (A,B) 10 µm maximum intensity projections from whole mount stained tg(LiverZap);tg(tp1:H2B-mCherry) livers at 48 hpi: at the surface (A-A’) and deep in the liver (B-B’). Tp1high cells are identified by fluorescence intensity of segmented nuclei, outlined in yellow (A’ and B’). Graphs showing distribution of all Tp1:H2B-mCherry nuclei (C) and Tp1high nuclei (D) by fluorescence intensity and z-position (C). The intensity of all mCherry+ nuclei decreases with increasing z-position (C-D). The dotted line outlines the liver in A-B’.

    3 - OPTIONAL: In the locally restricted ablation model, do hepatocytes located adjacent to the ROI proliferate and/or contribute to the regeneration of the injured region?

    Response:

    An important consideration, as highlighted by the reviewer, is whether neighbouring hepatocytes also contribute to regeneration following ROI ablation.

    PLAN: To address this point, LiverZap ROI ablation will be followed by cell proliferation analysis using an EdU incorporation assay at 24 and 72 hpi. These time points are selected based on the proliferation results following global LiverZap ablation; see Fig. 2D-F. The experiment will be performed in a tg(tp1:H2B-mcherry); _tg(fabp10a:gfp)_background to distinguish proliferating GFP-positive hepatocytes, which are H2B-mCherry-negative, from LPC-derived hepatocytes that have inherited H2B-mCherry (Tp1low). The resulting insights may help to refine hypotheses regarding the process(es) stimulating the formation of new hepatocytes adjacent to the ablated region.

    4 - OPTIONAL: Figure 4, A-S. It should be of significant interest if the authors could also analyze the BEC dynamics using the locally restricted hepatocyte ablation model, comparing those in the injured region (ROI) and the outside of the ROI.

    Response:

    We agree with the reviewer that this is the exciting next question, as it likely would provide insights into the cellular mechanism by which the biliary network is de- and re-constructed, as well as the mechanism by which BECs outside the ROI may initiate the LPC response to give rise to hepatocytes in a semi-systemic response. For this, the experimental set-up introduced in Fig.4J-P, in which BECs in the ROI are distinguished from adjacent ones by photoconversion, would be followed by extended live light-sheet microscopy of the regenerating liver. Due to the complexity, extent of the experiments and current unavailability of a light-sheet microscope, we would address this optional comment in future investigations.

    NO CHANGES PROPOSED.

    5A- Figure 4, T-V'. The data shown here for the changes in E-cadherin distribution is difficult to understand and interpret. The authors should provide magnified images and better description on how to distinguish the membranous (spotted signals?) and intracellular localization. Quantitative assessment should certainly be a plus, if possible.

    Response:

    We appreciate that it may be difficult to recognize the changes in E-Cadherin localisation, in particular at BEC membranes, given that there are intracellular puncta, and that E-Cadherin is expressed both in BECs and hepatocytes. We are convinced of the related data described in Figures 4 and S4, because the first experiment allowed quantification of the staining using both Tp1:H2B-mCherry to identify BECs and intestinal E-Cadherin for normalisation, which revealed a 51% E-Cadherin reduction at BEC cell membranes following injury. Unfortunately, the signal-to-noise ratio declined in consecutive experiments precluding further quantification although we could still observe a change in localisation. We tested alternative antibodies against E-Cadherin as well as optimized staining protocols, yet without success.

    5B - OPTIONAL: In relation to the above point, it is this reviewer's candid impression that the very last part regarding the possible role of E-cadherin dynamics in regulating the biliary network remodeling is still preliminary compared to the remaining parts, thereby rather depreciating the value of the entire manuscript. Perhaps this part could be published separately, together with more functional evidence regarding the causal relationship between them (e.g., showing the effect of Ecadherin knockdown in hepatocytes on the biliary remodeling and the induction of the BECdependent regeneration program)

    Response:

    PLAN: Following this reviewer’s and reviewer 2’s comments and suggestions, we agree to remove the data on E-Cadherin. Loss of adhesion as a mechanism for adopting an LPC-state remains very exciting, future investigations with novel tools to monitor and modulate E‑Cadherin expression in BECs would thus be needed.

    6 - Do zebrafish livers possess lobular structures with the portal-to-central vein axis and the metabolic zonation as typically observed in mammalian livers? As has been described in the manuscript, the "localized" injury patters in the mammalian livers usually occur at the sub-lobular structure levels (i.e., peri-portal region-restricted vs. peri-central region-restricted). Although the "localized" injury model described in this study using the zebrafish livers was indeed localized from the viewpoint of the entire organ (or the lobe), it still seemed much more "global" when considering those situations in the mammalian livers, so that the authors' claim that the former recapitulating the latter might be too exaggerated and somehow misleading. The authors should clarify and discuss this point in the manuscript.

    Response:

    The reviewer raises an important point, and it seems that our wording might not have been clear. In mammals, boundaries between injured and healthy tissue arise, because liver injuries frequently occur at the sub-lobular level. Although zebrafish livers are composed of metabolically diverse hepatocytes, a spatial arrangement comparable to mammalian zonation has so far not been identified (Morrison et al. 2022; Oderberg and Goessling, 2023). Yet, the liver lobes in the adult zebrafish have a central vein and periportal veins at the periphery of the organ, similar to the mammalian lobular organisation (Ota et al. 2022). Therefore, the scale of injury in the mammalian setting and the ROI-ablation model introduced in the current work differs. It, nevertheless, creates boundaries of healthy and injured liver tissue relevant for uncovering dynamic cellular processes mediating tissue repair in chronic liver disease. Importantly, with its suitability for advanced live imaging and optogenetic methods (e.g. photoconversion), LiverZap, complements mammalian models, in which this is still challenging. This offers therefore the powerful opportunity to employ LiverZap to screen for dynamic repair behaviours, which subsequently can be validated in a target approach in mammalian injury models.

    PLAN: To describe the relevance of our ROI ablation paradigm for elucidating repair processes at the interface of injured and healthy tissue more precisely. We will further edit and clarify text to place the ROI ablation into the context of hepatic injuries at the sub-lobular level throughout the mammalian liver.

    Minor comments:

    7 - Figure 4. Panels D and G should correspond to the same one image and the way of labeling be changed (as in Figure 1G). Likewise, in panel J, the bars shown separately as "M" and "S" at 12 dpi should correspond to the same data, so that they should be unified as one bar.

    Response:

    Thank you for pointing this out, this is changed in the updated figures; panels Fig. 4D and I.

    8 - Figure S3L. How was the ROI border defined? Perhaps the shape of the ROI should change significantly during regeneration due to dynamic tissue remodeling processes, thereby moving the position of the border as well.

    Response:

    The ROI border was defined as the interface between photoconverted and non-converted BECs. We concur with the reviewer’s notion that cell movement and rearrangement may occur during the regeneration process (see Fig. 4A-J), and the initially straight ROI border could consequently change during the regeneration process. Nevertheless, the border between photoconverted and non-converted BECs persists, serving as a landmark for the measurements shown in Figure S3L.

    Fig.: Quantification strategy for determining the region exhibiting an LPC-response outside the ROI ablation region. The dashed line of the ROI indicates morphogenetic changes of the interface between photoconverted and nonconverted cells over time due to repair-related cell rearrangement.

    PLAN: In the revised manuscript, we propose to include the below schematic as panel J to Figure S3. Moreover, we also suggest to change the solid line of the squares indicating the ROI area in figure panels 3C,G,O,P and S3D,H,K into a dashed line at the interface between photoconverted and non-converted tissue (see below figure as an example).

    9 - The authors should comment in the manuscript as to whether the system can be applicable for induction of more restricted areas (e.g., at a single hepatocyte level; in particular metabolic zones, if existing), as well as for ablation of other hepatic cell types such as BECs and endothelial cells.

    Response:

    Indeed, the optogenetic nature of the LiverZap system allows to induce hepatocyte death at the single cell level, as well as any defined region of interest that can be generated by the light source (e.g. confocal microscope software).

    Likewise, the FAP-TAP system can be easily applied to BECs or endothelial cells, or any cell type for which a specific promoter has been identified to drive the genetic FAP component fluorogen-activating protein dL5**.

    Response:

    PLAN: Both points will be included in the discussion section of the manuscript.

    Reviewer #2 (Evidence, reproducibility and clarity):

    MAJOR COMMENTS:

    1 - The LiverZap is an elegant new tool to induce localized ablation of hepatocytes. It is not as claimed by the authors a real breakthrough: (1) While localized ablation is nice compared to NTR-MTZ model in zebrafish, mice model such as CCl4 chronic injury can also study the interaction between healthy and injured tissue. (2) Although not using MTZ, the system still requires injection or exposure to malachite green derivate dye MG-2I. A few searches suggest that this compound could induce toxicity. Can the authors study and compare the toxicity of malachite green derivate dye MG-2I to the toxicity of MTZ? This is important as this would be indeed a strong argument in favor of the presented tool.

    Response:

    Point 1 – studying interactions between healthy and injured liver tissue: The reviewer is of course correct that interactions between healthy and injured tissue can also be studied in the mouse. However, ROI ablation with the LiverZap system can be combined with live imaging, thereby enabling the observation of cellular responses of the same sample over time, at a resolution currently difficult to achieve in mammals. Moreover, the possibility to induce cell death in a defined ROI, also allows to simultaneously employ other genetic tools, including cell-type specific lineage tracing by photoconversion, which is difficult to achieve in mammalian systems. The finding that BECs beyond the ROI of hepatocyte ablation produce new hepatocytes by a LPC response, illustrates the power of this approach. The optogenetic LiverZap ablation system would therefore complement existing mammalian and zebrafish liver regeneration models.

    PLAN: to include a more detailed discussion of this point and the complementary knowledge that can be gained in the discussion section.

    Point 2 – MG-2I toxicity: Indeed, as described in the manuscript, the FAP-TAP system, underlying LiverZap hepatocyte ablation, requires MG-2I incubation for the formation of the photosensitiser. Compared to the NTR/MTZ system, incubation with MG2I is short, requiring <3 hours in contrast to more than 24hours MTZ incubation. The system, including MG-2I has also been employed in cells, as well as in the zebrafish heart and nervous system without reported adverse effects (He et al., 2016; Xie et al., 2020). Consistently, we have not observed any apparent adverse effects between 0-72 hpi following 3-18 hour MG-2I incubation (unpublished). Nevertheless, toxicity studies evaluating survival upon MG-2I incubation have not yet been carried out and may be required for comparison with MTZ.

    PLAN: To perform toxicity studies for MG-2I, similar to those previously performed for MTZ (e.g. Mathias et al. 2014), in which larval survival after 3, 24 and 48 hour MG-2I exposure starting at 4 dpf will be assessed daily until 8 days post fertilisation.

    2 -The term ablation is choose because it is anticipated that it induces heaptospecific death. However, the consequences of cell death is not shown. In particular, the inflammatory immune response is not shown nor discussed.

    Response:

    The reviewer raises an interesting point, namely the inflammatory immune response, which is not the focus of this manuscript. Acridine Orange- and TUNEL-positive cells during the ablation process indicate that the reactive oxygen species produced by the FAP-TAP system cause hepatocyte apoptosis. We predict that this would recruit and be cleared by macrophages with little or no inflammatory response, like findings for the NTR-MTZ system (Stoddard et al., 2019). However, the role of neutrophils is unclear due to a possible direct effect of MTZ on this cell type.

    PLAN: We will include this point in the discussion.

    Future in-depth live imaging of transgenic reporters will be required for detailed studies of macrophage and neutrophil recruitment and their role in efferocytosis, including transcriptome analysis of specific gene signatures to detect an inflammatory response.

    3 - The difference between mild and severe ablation is hard to grasp. Can the authors explain more clearly the differences between mild and severe: what are the criteria as there is no difference in liver volume between mild and severe ablation? How do you achieve mild or severe ablation? It appears that the severity of the ablation is judged a posteriori and not decided per the experiment.

    Response:

    Concerning the first point, there must be a misunderstanding. Mild and severe hepatocyte ablation result in clearly different liver sizes, for instance at 30 hpi, the end of ablation, liver volumes are reduced by 23 % for mild or 64 % for severe cases (Fig. 1Q). This is supported by representative image data in Figs. 1F-P and S1A-C. Nevertheless, for consistency, we had represented the 12 hpi volume data as the same two data bars, although we cannot distinguish them yet at that timepoint of the experiment, as shown by images in Fig. 1F-G.

    PLAN: Adjust Fig.1Q and represent the 12 hpi liver volume data as a shared graph for mild/severe ablation, see included figure 1Q. We propose to similarly represent all 12 hpi quantifications, as represented in Figs. S1F, 2D-F and S2A.

    For the second point, the reviewer is correct that ablation severity is evaluated and determined between 24-30 hpi, at the end of hepatocyte ablation, given there is some variability in the response. Nonetheless, both length of 660nm illumination and oxygen availability can be used to shift the proportion of mild and severe ablation, depending on the desired outcome (Figs. 1Q, S1G-H).

    NO CHANGES PLANNED.

    4 - The work supports that biliary-driven regeneration also occurs when hepatocyte ablation happens in a small area of interest. Our knowledge is that you need a large defect in hepatocyte or a chronic liver injury ro activate the BDC-driven auxiliary process for regeneration. Could this be a specificity of the fish model?

    Response:

    Like the reviewer, our understanding is that severe hepatocyte loss, senescence or chronic liver injury activate BEC-derived regeneration in mammals and in zebrafish. All these cases are characterised by substantial reduction of local hepatocyte density or loss of function (in senescence). Given the overall hepatocyte loss is only 10-20% in the ROI model, the induction of the local LPC response was very surprising, on the other hand it corresponds to a near complete local hepatocyte depletion. The hepatobiliary architecture in zebrafish is similar to that of the mammalian ductular reaction, an adaptation of the biliary network to severe hepatocyte loss. In both cases, the majority of hepatocytes connect directly via their apical canaliculi to biliary ductules to ensure physiologic transport of hepatocyte products, often preceding the LPC response (Sato et al., 2019; Caviglia et al., 2022). Therefore, we propose that the LPC response following ROI hepatocyte ablation is not specific to the zebrafish model, but a common mechanism elicited across species and related to the severity of the injury and the configuration of the hepatobiliary network at the time of injury, such as the ductular reaction.

    PLAN: To edit the text and discuss this point clearly.

    5 - Pathways revealed to control liver regeneration or BEC-driven regeneration in fish have not be found to have a similar drastic predominance in rodents. This mitigate perhaps the use of fish for this type of research?

    Response:

    On the contrary, zebrafish has been established and validated as a model to investigate and elucidate developmental hepatic programs as well as regeneration (Goessling and Sadler, 2015; Wang et al., 2017). However, we acknowledge that more comparative studies are needed to understand the molecular pathways driving regeneration both in zebrafish and mammals and their similarity.

    Specifically, zebrafish and mammals display high conservation in the parenchymal and non-parenchymal cell types of the liver as well as their developmental programs (Goessling and Sadler, 2015; Wang et al., 2017). Using different injury paradigms in zebrafish, including ethanol, acetaminophen toxicity and the pharmacogenetic NTR-MTZ model, it has been shown that cellular responses to liver injury are also remarkably conserved with mammals where hepatocyte proliferation governs repair after mild injury while severe injury repair is driven by conversion of BECs into LPCs (So, et al., 2020; Forbes and Newsome, 2019). Major pathways, such as Wnt, FGF and BMP signaling show conserved functions in restorative hepatocyte proliferation (Goessling et al., 2008; Kan et al. 2009, Böhm et al 2010). At present, only very little is known about the molecular mechanisms controlling the BEC/LPC to hepatocyte conversion particularly in rodent models (Kim et al., 2023), while a number of zebrafish studies have started to elucidate the signals governing the different steps of this process (Kim et al., 2023), due to the relative ease of using the larval zebrafish model for this work. Notably, the Notch pathway plays multiple roles in both mouse and zebrafish LPC-mediated repair (Minnis-Lyons et al., 2021; Huang et al 2014; Russel et al.,2019), however further work will be necessary to determine the detailed corresponding functions. Therefore, future work in both rodents and zebrafish will be essential to uncover the molecular mechanisms of this repair process relevant for chronic injury. Given the large conservation of developmental and repair mechanism between mammals and zebrafish observed so far, it is highly likely that this will also apply to LPC-mediated repair. Studies promise to uncover even greater similarity between zebrafish and human (e.g. Fang et al 2011), underscoring the power of using complementary vertebrate models.

    PLAN: To edit the text in the introduction and discussion to clarify and highlight the similarities, differences, and opportunities the zebrafish model offers for understanding the mechanisms of vertebrate liver regeneration in general and in particular by using the LiverZap system.

    6 - The authors show that in the case of mild ablation, hepatocytes are responsible for replenishment of the parenchyma, but in the context of severe ablation, LPC-mediated regeneration takes control. However, when the authors perform localized and controlled ablation, which is small (around 10-20%) and, to my understanding, a mild / local ablation, however the authors show that LPC mediates the regeneration. Can the authors explain the discrepancy between their results?

    Response:

    We agree with the reviewer that the LPC response in the smaller, local ROI ablation was unexpected. However, it could be explained by the following: while such ROI hepatocyte ablation represents only a 10-20% ablation of the total hepatocyte population, by sheer numbers comparable to a mild global ablation, the near-complete local hepatocyte loss however makes it more similar to a severe or chronic global injury. Notably, the zebrafish hepatobiliary architecture in zebrafish is similar to that of the mammalian ductular reaction, an adaptation of the biliary network to severe hepatocyte loss. In both cases, the majority of hepatocytes connect directly via their apical canaliculi to biliary ductules to ensure physiologic transport of hepatocyte products, often preceding the LPC response (Sato et al., 2019; Caviglia et al., 2022). We hypothesize that if a similar local, near complete hepatocyte loss would be induced in a mammalian liver exhibiting a ductular reaction, it would similarly induce local LPC-mediated repair. Since this is, to our knowledge not possible, the LiverZap model represents a unique opportunity to induce the LPC-response in a controlled manner and in addition investigate the underlying cellular and molecular processes of injured and adjacent healthy tissues at high resolution in an in vivo context.

    PLAN: We will edit the discussion to clarify this important point.

    7 - The last part of the paper about E-Cadherin expression is not convincing. I am not sure about the quality of the IF stainings of E-Cadherin, and it is not helping proving the point of the authors. Can the authors provide better stainings for this figure?

    Response:

    (Same response as to point 5A+B of reviewer 1). We appreciate that it may be difficult to recognize the changes in E-Cadherin localisation, in particular at BEC membranes, given that there are intracellular puncta and that E-Cadherin is expressed both in BECs and hepatocytes. We are convinced of the related data described in Figures 4 and S4, because the first experiment allowed quantification of the staining using both Tp1:H2B-mCherry to identify BECs and intestinal E-Cadherin for normalisation, which revealed a 51 % E-Cadherin reduction at BEC cell membranes following injury. Unfortunately, the signal to noise ratio declined in consecutive experiments, while we could still observe a change in localisation, it challenged a meaningful quantification. We tested alternative antibodies against E-Cadherin, yet without success.

    PLAN: Following both reviewers’ comments and suggestions, we agree to remove the data on E-Cadherin.

    8 - Could the authors provide a bit more information on the live imaging. Exactly how do they achieve imaging for such a long time?

    Response:

    Thank you for pointing this out, the information was not very detailed. We used relatively standard mounting conditions (low-melting point agarose and Tricaine anaesthesia, see below for details), combined with light-sheet microscopy, which was the key to achieving the long imaging. We believe that in addition to the known gentle imaging condition, the mounting set-up is critical as the fish is completely suspended in a very low-percentage, low melting point agarose within a large volume of embryo medium.

    PLAN: Update the material and methods section with the following details: Long-term live imaging was performed using a LS1 Live light sheet microscopy system (Viventis Microscopy Sàrl). Larvae were with anesthetized with 0.4% Tricaine and mounted ventrally in 0.8% low melting point agarose in E3/PTU media supplemented with 0.16% tricaine. Once the agarose solidified, the chamber was filled with E3/PTU with 0.16% Tricaine to maintain anaesthesia. A 25X objective was used and acquisition was performed every 20 minutes.

    MINOR COMMENTS:

    9 - It is hard to imagine the full-size liver in Figure 1, bad contrast. Can the authors manually delineate it?

    Response:

    The livers in this figure are now outlined in the updated figures, see new Figure 1.

    10 - "This finding is very surprising, since current understanding in the field links the generation of new hepatocytes from BECs/LPCs with global hepatocyte death." This statement lacks references.

    Response:

    PLAN: To add the following primary references to the above sentence: (Choi et al., 2014; He et al., 2014; Manco et al., 2019; Raven et al., 2017) and recent review (Kim et al_._, 2023).

    REFERENCES

    Böhm F, Köhler UA, Speicher T, Werner S. Regulation of liver regeneration by growth factors and cytokines. EMBO Mol Med. 2010 doi:10.1002/emmm.201000085.

    Caviglia S, Unterweger IA, Gasiūnaitė A, Vanoosthuyse AE, Cutrale F, Trinh LA, Fraser SE, Neuhauss SCF, Ober EA. Fraeppli: a multispectral imaging toolbox for cell tracing and dense tissue analysis in zebrafish. Development. 2022 doi:10.1242/dev.199615.

    Choi TY, Ninov N, Stainier DY, Shin D. Extensive conversion of hepatic biliary epithelial cells to hepatocytes after near total loss of hepatocytes in zebrafish. Gastroenterology. 2014 doi:10.1053/j.gastro.2013.10.019.

    Fang L, Green SR, Baek JS, Lee SH, Ellett F, Deer E, Lieschke GJ, Witztum JL, Tsimikas S, Miller YI. In vivo visualization and attenuation of oxidized lipid accumulation in hypercholesterolemic zebrafish. J Clin Invest. 2011 doi: 10.1172/JCI57755.

    Forbes SJ, Newsome PN. Liver regeneration - mechanisms and models to clinical application. Nat Rev Gastroenterol Hepatol. 2016 doi:10.1038/nrgastro.2016.97

    Goessling W, North TE, Lord AM, Ceol C, Lee S, Weidinger G, Bourque C, Strijbosch R, Haramis AP, Puder M, Clevers H, Moon RT, Zon LI. APC mutant zebrafish uncover a changing temporal requirement for wnt signaling in liver development. Dev Biol. 2008 doi:10.1016/j.ydbio.2008.05.526.

    Goessling W, Sadler KC. Zebrafish: an important tool for liver disease research. Gastroenterology. 2015 doi:10.1053/j.gastro.2015.08.034.

    He J, Lu H, Zou Q, Luo L. Regeneration of liver after extreme hepatocyte loss occurs mainly via biliary transdifferentiation in zebrafish. Gastroenterology. 2014 doi:10.1053/j.gastro.2013.11.045.

    He J, Wang Y, Missinato MA, Onuoha E, Perkins LA, Watkins SC, St Croix CM, Tsang M, Bruchez MP. A genetically targetable near-infrared photosensitizer. NatMethods. 2016 doi:10.1038/nmeth.3735.

    Huang M, Chang A, Choi M, Zhou D, Anania FA, Shin CH. Antagonistic interaction between Wnt and Notch activity modulates the regenerative capacity of a zebrafish fibrotic liver model. Hepatology. 2014 doi:10.1002/hep.27285.

    Kan NG, Junghans D, Izpisua Belmonte JC. Compensatory growth mechanisms regulated by BMP and FGF signaling mediate liver regeneration in zebrafish after partial hepatectomy. FASEB J. 2009 doi:10.1096/fj.09-131730.

    Kim M, Rizvi F, Shin D, Gouon-Evans V. Update on Hepatobiliary Plasticity. Semin Liver Dis. 2023 doi: 10.1055/s-0042-1760306.

    Liang P, Kolodieznyi D, Creeger Y, Ballou B, Bruchez MP. Subcellular Singlet Oxygen and Cell Death: Location Matters. Front Chem. 2020. doi:10.3389/fchem.2020.592941.

    Manco R, Clerbaux LA, Verhulst S, Bou Nader M, Sempoux C, Ambroise J, Bearzatto B, Gala JL, Horsmans Y, van Grunsven L, Desdouets C, Leclercq I. Reactive cholangiocytes differentiate into proliferative hepatocytes with efficient DNA repair in mice with chronic liver injury. J Hepatol. 2019

    doi: 10.1016/j.jhep.2019.02.003.

    Mathias JR, Zhang Z, Saxena MT, Mumm JS. Enhanced cell-specific ablation in zebrafish using a triple mutant of Escherichia coli nitroreductase. Zebrafish. 2014 doi: 10.1089/zeb.2013.0937.

    Minnis-Lyons SE, Ferreira-González S, Aleksieva N, Man TY, Gadd VL, Williams MJ, Guest RV, Lu WY, Dwyer BJ, Jamieson T, Nixon C, Van Hul N, Lemaigre FP, McCafferty J, Leclercq IA, Sansom OJ, Boulter L, Forbes SJ. Notch-IGF1 signaling during liver regeneration drives biliary epithelial cell expansion and inhibits hepatocyte differentiation. Sci Signal. 2021 doi:10.1126/scisignal.aay9185.

    Morrison JK, DeRossi C, Alter IL, Nayar S, Giri M, Zhang C, Cho JH, Chu J. (2022) Single-cell transcriptomics reveals conserved cell identities and fibrogenic phenotypes in zebrafish and human liver. Hepatol Commun. doi: 10.1002/hep4.1930.

    Oderberg IM, Goessling W. (2023) Biliary epithelial cells are facultative liver stem cells during liver regeneration in adult zebrafish. JCI Insight. doi: 10.1172/jci.insight.163929.

    Ota N, Shiojiri N. Comparative study on a novel lobule structure of the zebrafish liver and that of the mammalian liver. Cell Tissue Res. 2022 doi:10.1007/s00441-022-03607-y.

    Raven A, Lu WY, Man TY, Ferreira-Gonzalez S, O'Duibhir E, Dwyer BJ, Thomson JP, Meehan RR, Bogorad R, Koteliansky V, Kotelevtsev Y, Ffrench-Constant C, Boulter L, Forbes SJ. Cholangiocytes act as facultative liver stem cells during impaired hepatocyte regeneration. Nature. 2017 doi:10.1038/nature23015.

    Russell JO, Ko S, Monga SP, Shin D. Notch Inhibition Promotes Differentiation of Liver Progenitor Cells into Hepatocytes via _sox9b_Repression in Zebrafish. Stem Cells Int. 2019 doi:10.1155/2019/8451282.

    Sato K, Marzioni M, Meng F, Francis H, Glaser S, Alpini G. Ductular Reaction in Liver Diseases: Pathological Mechanisms and Translational Significances. Hepatology. 2019 doi: 10.1002/hep.30150.

    So J, Kim A, Lee SH, Shin D. Liver progenitor cell-driven liver regeneration. Exp Mol Med. 2020 doi: 10.1038/s12276-020-0483-0.

    Stoddard M, Huang C, Enyedi B, Niethammer P. Live imaging of leukocyte recruitment in a zebrafish model of chemical liver injury. Sci Rep. 2019 doi: 10.1038/s41598-018-36771-9.

    Wang S, Miller SR, Ober EA, Sadler KC. Making It New Again: Insight Into Liver Development, Regeneration, and Disease From Zebrafish Research. Curr Top Dev Biol. 2017 doi: 10.1016/bs.ctdb.2016.11.012.

    Xie W, Jiao B, Bai Q, Ilin VA, Sun M, Burton CE, Kolodieznyi D, Calderon MJ, Stolz DB, Opresko PL, St Croix CM, Watkins S, Van Houten B, Bruchez MP, Burton EA. Chemoptogenetic ablation of neuronal mitochondria in vivo with spatiotemporal precision and controllable severity. Elife. 2020 doi: 10.7554/eLife.51845.

  6. Review Commons

    Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Referee #2

    Evidence, reproducibility and clarity

    In this study, the authors present a new chemoptogenetic tool, LiverZap, to study regeneration in zebrafish. In combination with LiverZap, the authors use live imaging to study longitudinally LPC-mediated regeneration and the implication of the healthy surrounding tissue.

    Major comments

    The LiverZap is an elegant new tool to induce localized ablation of hepatocytes. It is not as claimed by the authors a real breakthrough: (1) While localized ablation is nice compared to NTR-MTZ model in zebrafish, mice model such as CCl4 chronic injury can also study the interaction between healthy and injured tissue. (2) Although not using MTZ, the system still requires injection or exposure to malachite green derivate dye MG-2I. A few searches suggest that this compound could induce toxicity. Can the authors study and compare the toxicity of malachite green derivate dye MG-2I to the toxicity of MTZ? This is important as this would be indeed a strong argument in favor of the presented tool.
    The term ablation is choose because it is anticipated that it induces heaptospecific death. However, the consequences of cell death is not shown. In particular, the inflammatory immune response is not shown nor discussed.

    • The difference between mild and severe ablation is hard to grasp. Can the authors explain more clearly the differences between mild and severe: what are the criteria as there is no difference in liver volume between mild and severe ablation? How do you achieve mild or severe ablation? It appears that the severity of the ablation is judged a posteriori and not decided per the experiment.
    • The work supports that biliary-driven regeneration also occurs when hepatocyte ablation happens in a small area of interest. Our knowledge is that you need a large defect in hepatocyte or a chronic liver injury ro activate the BDC-driven auxiliary process for regeneration. Could this be a specificity of the fish model?
    • Pathways revealed to control liver regeneration or BEC-driven regeneration in fish have not be found to have a similar drastic predominance in rodents. This mitigate perhaps the use of fish for this type of research?

    The authors show that in the case of mild ablation, hepatocytes are responsible for replenishment of the parenchyma, but in the context of severe ablation, LPC-mediated regeneration takes control.
    However, when the authors perform localized and controlled ablation, which is small (around 10-20%) and, to my understanding, a mild / local ablation, however the authors show that LPC mediates the regeneration. Can the authors explain the discrepancy between their results?
    The last part of the paper about E-Cadherin expression is not convincing. I am not sure about the quality of the IF stainings of E-Cadherin, and it is not helping proving the point of the authors. Can the authors provide better stainings for this figure?
    Could the authors provide a bit more information on the live imaging. Exactly how do they achieve imaging for such a long time?

    Minor comments

    It is hard to imagine the full-size liver in Figure 1, bad contrast. Can the authors manually delineate it?
    "This finding is very surprising, since current understanding in the field links the generation of new hepatocytes from BECs/LPCs with global hepatocyte death." This statement lacks references.

    Significance

    General assessment:

    Elegant model to ablate hepatocyte in a clean fashion and study regeneration when coupled to imaging technique.
    The work supports that biliary-driven regeneration also occurs when hepatocyte ablation happens in a small area of interest. This seems a new concept, the operation of the process needs to be ascertain in other models including humans.
    Immune/inflammatory response to the ablation as well as the way it may influence/drive or dictate a regenerative response is not investigated

    Advance:

    The advance pertains to the model because rodents offer ample possibilities to study interaction between 'intact' and 'diseased' cells. Of course the model is attractive as it is rapid, allows for 'real time' in vivo imaging, ..;

    The audience will be a specialised audience (basic research in liver regeneration, in zebrafish technologies, ...)

    Expertise

    I'm a hepatologist, devoted for the last 15 years to the experimental study of the pathophysiology of liver diseases using animal models, cell cultures models and organoids. I 'm not an expert in zebra fish. I have a large interest in regeneration and in particular I produced pioneer work in BEC-driven regeneration that is studied here.

  7. Review Commons

    Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Referee #1

    Evidence, reproducibility and clarity

    Summary:

    The manuscript by Ambrosio et al. describes a novel experimental system named "LiverZap", where spatiotemporally-regulated hepatocyte ablation can be achieved in the liver of living zebrafish for the study of the tissue/organ regeneration processes and mechanisms, particularly by in vivo live imaging. Specifically, the authors made use of a binary FAP-TAP system to achieve reactive oxygen species (ROS)-induced hepatocyte cell death upon treatment with the drug MG-2I followed by near-infrared (NIR) illumination in transgenic fish. They found that the system resulted in two categories of hepatocyte ablation, i.e., mild ablation and severe ablation. In the livers where the severe ablation was induced throughout the organ, biliary epithelial cell (BEC)-derived hepatocyte regeneration program was provoked as demonstrated by short-term lineage tracing experiments based on histone inheritance, which is quite consistent with previous studies using different methods of hepatocyte ablation in zebrafish and mice. Taking advantage of the spatial controllability of the LiverZap system, the authors further demonstrated that spatially-restricted severe hepatocyte ablation was sufficient to induce the BEC-dependent regeneration program therein. Interestingly, BECs outside the targeted region also contributed to the local hepatocyte regeneration, as revealed by using a sophisticated photoconvertible BEC labeling system. Finally, a dynamic nature of BEC aggregation and re-distribution upon liver injury was demonstrated to occur in advance of hepatocyte regeneration, reminiscent of the so-called ductular reaction in the mammalian liver. Overall, the authors' claims and the conclusions are well supported by the data presented in the manuscript, except for a few points as listed below.

    Major comments:

    • The authors assessed acridine orange incorporation in BECs upon LiverZap and concluded that LiverZap triggers hepatocyte-specific cell death without a bystander effect in adjacent cells (Figure 1 D-E). What happened to endothelial cells, which could also be affected either directly by ROS production in hepatocytes or indirectly by gross morphological changes in tissue organization?
    • The evaluation criteria for distinguishing mCherry and cells in imaging experiments should be clearly described in the methods section. The authors should also provide some quantitative data regarding the level of correlation between the mCherry hepatocytes and the BEC-derived hepatocytes strictly defined based on the TP1-H2B-EGFP lineage tracing, as the former was used as a surrogate marker for the latter in some experiments.
    • OPTIONAL: In the locally restricted ablation model, do hepatocytes located adjacent to the ROI proliferate and/or contribute to the regeneration of the injured region?
    • OPTIONAL: Figure 4, A-S. It should be of significant interest if the authors could also analyze the BEC dynamics using the locally restricted hepatocyte ablation model, comparing those in the injured region (ROI) and the outside of the ROI.
    • Figure 4, T-V'. The data shown here for the changes in E-cadherin distribution is difficult to understand and interpret. The authors should provide magnified images and better description on how to distinguish the membranous (spotted signals?) and intracellular localization. Quantitative assessment should certainly be a plus, if possible.
    • OPTIONAL: In relation to the above point, it is this reviewer's candid impression that the very last part regarding the possible role of E-cadherin dynamics in regulating the biliary network remodeling is still preliminary compared to the remaining parts, thereby rather depreciating the value of the entire manuscript. Perhaps this part could be published separately, together with more functional evidence regarding the causal relationship between them (e.g., showing the effect of E-cadherin knockdown in hepatocytes on the biliary remodeling and the induction of the BEC-dependent regeneration program)
    • Do zebrafish livers possess lobular structures with the portal-to-central vein axis and the metabolic zonation as typically observed in mammalian livers? As has been described in the manuscript, the "localized" injury patters in the mammalian livers usually occur at the sub-lobular structure levels (i.e., peri-portal region-restricted vs. peri-central region-restricted). Although the "localized" injury model described in this study using the zebrafish livers was indeed localized from the viewpoint of the entire organ (or the lobe), it still seemed much more "global" when considering those situations in the mammalian livers, so that the authors' claim that the former recapitulating the latter might be too exaggerated and somehow misleading. The authors should clarify and discuss this point in the manuscript.

    Minor comments:

    • Figure 4. Panels D and G should correspond to the same one image and the way of labeling be changed (as in Figure 1G). Likewise, in panel J, the bars shown separately as "M" and "S" at 12 dpi should correspond to the same data, so that they should be unified as one bar.
    • Figure S3L. How was the ROI border defined? Perhaps the shape of the ROI should change significantly during regeneration due to dynamic tissue remodeling processes, thereby moving the position of the border as well.
    • The authors should comment in the manuscript as to whether the system can be applicable for induction of more restricted areas (e.g., at a single hepatocyte level; in particular metabolic zones, if existing), as well as for ablation of other hepatic cell types such as BECs and endothelial cells.

    Significance

    The newly developed LiverZap system described in the present study was well designed and has multifaceted advantages compared to other "global ablation systems" that have so far been used in this research field. Indeed, the authors' original finding that the localized hepatocyte ablation provokes activation of BECs outside the injured region and their contribution to hepatocyte renewal, could have never been obtained using previous models. This finding is of considerable novelty and interest in that the localized injury model should better reflect the pathophysiological conditions in various human liver diseases. Thus, the study should make significant contribution to the field in both technological and conceptual ways, providing useful and relevant platforms for the future studies on the mechanisms of liver injury, repair and regeneration.

  8. Version published to 10.1101/2023.03.13.531762v1 on bioRxiv