Development of a hepatic cryoinjury model to study liver regeneration

Abstract

The liver is a remarkable organ that can regenerate in response to injury. Depending on the extent of injury, the liver can undergo compensatory hyperplasia or fibrosis. Despite decades of research, the molecular mechanisms underlying these processes are poorly understood. Here, we developed a new model to study liver regeneration based on cryoinjury. To visualise liver regeneration at cellular resolution, we adapted the CUBIC tissue-clearing approach. Hepatic cryoinjury induced a localised necrotic and apoptotic lesion characterised by inflammation and infiltration of innate immune cells. After this initial phase, we observed fibrosis, which resolved as regeneration re-established homeostasis in 30 days. Importantly, this approach enables the comparison of healthy and injured parenchyma within an individual animal, providing unique advantages to previous models. In summary, the hepatic cryoinjury model provides a fast and reproducible method for studying the cellular and molecular pathways underpinning fibrosis and liver regeneration.

Keywords: Apoptosis; Cryoinjury; Fibrosis; Inflammation; Liver regeneration; Necrosis; Proliferation; Zebrafish.

Fig. 1.

Liver cell death following cryoinjury. (A-D) Simplified schematics illustrating the cryoinjury procedure in the zebrafish liver. (A) The zebrafish liver is placed ventral side up to facilitate the surgery. (B) A small incision near the midline exposes the ventral liver lobe. (C) The frozen cryoprobe is applied to the liver surface for 15 s to induce injury. (D) The damaged area in the liver appears as a blister-like structure at 1 dpci. (E,F) TUNEL-staining of sham-operated (E,E′) and injured (F,F′) liver sections at 1 dpci. IA, injured area. Yellow arrowheads indicate TUNEL⁺ cells; yellow dashed lines indicate the border zone. (G-J) Tg(fabp10a: GreenLantern-H2B; annexinV:mKate) in toto acquisitions of sham-operated (G,H) and 1 dpci (I,J) livers. (I,J) The IA (yellow dashed line) is identifiable by the absence of GreenLantern-H2B⁺ hepatocytes. Blue arrowheads indicate AnnexinV-mKate⁺ cells. Scale bars: 500 µm.

Fig. 2.

Progression of liver regeneration after cryoinjury. (A) A simplified schematic illustrating the collection of livers timeline after cryoinjury. (B-G) Whole-mount images of cleared zebrafish livers at the indicated stages of regeneration. Yellow dashed lines indicate the border zone. (H) Quantification of the IA area compared with the visible liver parenchyma area (n=16, 16, 14, 18, 7 and 12, left to right). Data are mean±s.d. P-values were calculated using one-way ANOVA followed by Tukey's multiple comparisons test (*P<0.05, ****P<0.0001). Scale bars: 500 µm.

Fig. 3.

Transient fibrotic deposition during liver regeneration after cryoinjury. (A-H) AFOG staining in sections of representative sham-operated (A,G) or injured livers (B-F,H) at the indicated stages. Blue indicates collagen; red indicates cell debris and fibrin. Anterior is towards the left; dorsal is towards the top. Outlined areas are shown at higher magnification. (G′,H′) Adjacent sections from the samples showed in G and H, immunostained using an anti-Col1a1 antibody and counterstained with DAPI. Asterisks indicate Col1a1 deposition in the liver parenchyma. (G″-H‴). Pseudo-colored Col1a1 signal (white) from G’ and H’. Magnifications of the outlined areas are shown in G‴ and H‴. Red asterisks indicate Col1a1 accumulation. (I) Quantification of the collagen area of livers from the indicated cohorts, normalized to the liver parenchyma area (n=6, 5, 6, 6 and 4, left to right). Data are mean±s.d. P-values were calculated using one-way ANOVA followed by Tukey's multiple comparisons test (**P<0.01, ****P<0.0001). bv, blood vessel; fc, fibrotic cap; IA, injured area; int, intestine; p, pancreas; vl, ventral lobe. Scale bars: 100 µm (white); 500 µm (orange and black).

Fig. 4.

Cryoinjury induces the local and transient infiltration of leukocytes. (A) Schematic representation of the experiment workflow. (B-H) Sections of livers from Tg(fabp10a:NLS-mKate) animals at the indicated stages, immunostained to detect hepatocyte nuclei (mKate) and leukocytes (Lcp1). Cyan arrowheads indicate low signal Lcp1⁺cells; dashed yellow line indicates the border zone; IA, injured area; yellow arrowheads indicate high signal Lcp1⁺ cells; asterisks indicate spared hepatocytes surrounded by injured/necrotic tissue area. (I) Quantification of the number of Lcp1⁺ cells in designated regions (n=6, 5, 6, 6, 5, 4 and 6, left to right). Solid black line indicates the mean. P-values wee calculated using one-way ANOVA followed by Tukey's multiple comparisons test. Scale bars: 25 µm.

Fig. 5.

Local hepatocyte hyperplasia upon cryoinjury. (A) Schematic representation of the experiment workflow. (B-H) Liver sections from Tg(fabp10a:NLS-mKate) animals at the indicated stages, immunostained to detect proliferation (PCNA) and hepatocyte nuclei (mKate). Yellow arrowheads indicate proliferating hepatocytes. Cyan arrowheads indicate other cell types actively cycling. Dashed yellow line indicates the separation between healthy and injured liver parenchyma; IA, injured area; asterisks indicate spared hepatocytes surrounded by injured/necrotic tissue area. (I) Hepatocyte proliferation index in the border zone at the indicated stages (n=6, 5, 6, 5, 5, 4 and 5, left to right). Solid black line indicates the mean. P-values were calculated using one-way ANOVA followed by Tukey's multiple comparisons test. Scale bars: 25 µm.

Fig. 6.

Biliary epithelial cells (BECs) and endothelial cells (ECs) recover upon cryoinjury. (A-F″) Whole-mount imaging of GreenLantern⁺ hepatocytes and Anxa4⁺ biliary epithelial cells from sham-operated (A-A″) or injured (B-F″) livers at the indicated stages. The outlined areas are shown at higher magnification. (G) Mean intensity profile of Anxa4 within the IA at the designated stages is represented by individual lines for each sample, with the IA delineated in grey (n=4 throughout). (H-M″) Whole-mount acquisitions of GFP⁺ ECs from sham-operated (H-H″) or injured (I-M″) livers at the indicated stages. The outlined areas are shown at higher magnification. (N) Mean intensity profile of fli1a:nGFP within the IA at the designated stages is represented by individual lines for each sample, with the IA delineated in grey (n=4 throughout). Dashed yellow line indicates the border zone of the injured area; IA, injured area. Scale bars: 500 µm.

Fig. 7.

Transcriptional signatures of liver regeneration upon cryoinjury. (A) Schematic representation of the experiment workflow. (B-D) Volcano plots representing the comparison of 1, 3 and 7 dpci with sham-operated adult zebrafish. DEGs: FC ≥1.5 (darker dots) or ≤−1.5 (lighter dots); P≤0.05; with top DEG annotated. (E) Bar plot representing the number of upregulated and downregulated DEGs at 1, 3 and 7 dpci. (F) Venn diagram representing DEGs at 1, 3 and 7 dpci. (G) GSEA of liver cell types during liver regeneration. (H) Dotplot representing the expression of key genes for specific liver cell populations upon cryoinjury.

Fig. 8.

Model of liver regeneration upon cryoinjury. Schematic representation of the cellular events taking place upon cryoinjury during liver repair in the zebrafish.