ATP7B gene therapy of autologous reprogrammed hepatocytes alleviates copper accumulation in a mouse model of Wilson's disease

Abstract

Background and aims: Wilson's disease (WD) is a rare hereditary disorder due to ATP7B gene mutation, causing pathologic copper storage mainly in the liver and neurological systems. Hepatocyte transplantation showed therapeutic potential; however, this strategy is often hindered by a shortage of quality donor cells and by allogeneic immune rejection. In this study, we aimed to evaluate the function and efficacy of autologous reprogrammed, ATP7B gene-restored hepatocytes using a mouse model of WD.

Approach and results: Sufficient liver progenitor cells (LPCs) were harvested by reprogramming hepatocytes from ATP7B^-/- mice with small molecules, which exhibited strong proliferation and hepatic differentiation capacity in vitro. After lentivirus-mediated mini ATP7B gene transfection and redifferentiation, functional LPC-ATP7B-derived hepatocytes (LPC-ATP7B-Heps) were developed. RNA sequencing data showed that, compared with LPC-green fluorescent protein-Heps (LPC-GFP-Heps) with enrichment of genes that were mainly in pathways of oxidative stress and cell apoptosis, in LPC-ATP7B-Heps under high copper stress, copper ion binding and cell proliferation pathways were enriched. LPC-ATP7B-Heps transplantation into ATP7B^-/- mice alleviated deposition of excess liver copper with its associated inflammation and fibrosis, comparable with those observed using normal primary hepatocytes at 4 months after transplantation.

Conclusions: We established a system of autologous reprogrammed WD hepatocytes and achieved ATP7B gene therapy in vitro. LPC-ATP7B-Heps transplantation demonstrated therapeutic efficacy on copper homeostasis in a mouse model of WD.

FIGURE 1

Generation and proliferative capacity of PH‐derived LPCs. (A) Schematic diagram of the reprogramming strategy on generation PH‐derived LPCs. (B) Immunostaining of proliferation marker Ki‐67 and apoptosis marker caspase‐3 in KO‐LPCs at passage 15 (n = 5 cell lines). Scale bars, 25 μm and 5 μm (zoom in). U.D.= undetectable. (C) Growth curves and doubling times of LPCs at passage 15 (n = 3 cell lines). (D) SA‐β‐gal staining of LPCs cultured in HREM with/without 5A. Scale bars, 125 μm (upper) and 25 μm (lower). (E) qPCR analyses for cell cycle and senescence‐related genes expression: CDK2, CDK4, P16, and P53 of the two groups (n = 3 cell lines). Results are presented as mean ± SD; *p < 0.05; **p < 0.01; ***p < 0.001. Student t test for (C) and (E)

FIGURE 2

Characteristics of KO‐LPCs gene expression. (A) KO‐LPCs of passage 0, 1, 3, 4, and 14 were stained with ALB (red), AAT (green), and CK19 (green). Nuclei were counterstained with DAPI. Numbers of ALB⁺, AAT⁺, and CK19⁺ cells per field were counted and their percentages were analyzed (n = 3 KO‐LPC lines and 2 fields of each line were counted). Scale bar, 25 μm. Results are shown as mean ± SD. ***p < 0.001. One‐way ANOVA with Student‐Newman‐Keuls post‐test. (B) Normalized mRNA levels on hepatic (ALB, AAT, G6PC, and HNF4α) and progenitor (CK7, CK19, SOX9, and EpCAM) markers from passage 0 to 40 of KO‐LPCs (n = 3 cell lines). Results are shown as mean ± SD

FIGURE 3

The hepatic differentiation of KO‐LPCs. (A) Representative images of KO‐LPCs (P15) and redifferentiated LPC‐Heps. Immunostaining of ALB (red) and ATP7B (green) in LPC‐Heps. Scale bars, 125 μm (bright field) and 25 μm (fluorescence). (B) ALB (red), AAT (green), and CK19 (green) staining in LPC‐Heps. A magnified picture is shown with ALB (red) and CK19 (green) double staining. Nuclei were counterstained with DAPI. Scale bars, 25 μm. The number of positive cells per field was counted (n = 3 cell lines and 2 fields of each line were counted). (C) Normalized mRNA levels for hepatocyte and progenitor‐associated markers of KO‐PHs, KO‐LPCs, and LPC‐Heps (n = 3). (D) Periodic‐Acid‐Schiff (PAS), PAS + diastase staining, and CDCFDA of KO‐PHs, KO‐LPCs, and LPC‐Heps to evaluate glycogen production and dye accumulation in the bile canaliculi. Scale bars, 25 μm (PAS) and 50 μm (CDCFDA). (E) Assays of CYP2C9, CYP1A2, and CYP3A4 activities on drug metabolism, ALB secretion, and urea synthesis of KO‐PHs, KO‐LPCs, and LPC‐Heps (n = 3). The data are expressed as mean ± SD in (B, C, and E). *p < 0.05; **p < 0.01; ***p < 0.001. One‐way ANOVA with Student‐Newman‐Keuls post‐test for (C) and (E)

FIGURE 4

Functional copper export in LPC‐ATP7B‐Heps. (A) LV transfection of KO‐LPCs and their differentiation into LPC‐Heps. Normalized mRNA levels of intracellular ATP7B in LPCs and LPC‐Heps, and flow cytometry of LPC‐ATP7B‐Heps (n = 3 cell lines). Scale bars, 50 μm and 7.5 μm (zoom in). (B) ATP7B staining and GFP expression in KO‐LPC‐GFP, KO‐LPC‐ATP7B, and LPC‐ATP7B‐Heps. Scale bars, 10 μm. (C) Representative pictures of KO‐LPC‐Heps without LV‐ATP7B transfection treated with or without CuCl₂. Upper panel: ATP7B/Golgin‐97 (Golgi marker) with no added CuCl₂. Lower panel: ATP7B/LAMP‐1 (lysosome marker) with added 50 μm CuCl₂ for 2 h. DAPI was used to counterstain nuclei. Scale bars, 10 μm. (D) Representative images of LPC‐ATP7B‐Heps treated with or without CuCl₂. The 1st line: pictures of ATP7B/Golgin‐97/GFP immunostaining without CuCl₂ addition. The 2nd line: pictures of ATP7B/LAMP1/GFP staining without CuCl₂ addition. The 3rd line: images of ATP7B/LAMP1/GFP staining with added 50 μM CuCl₂ for 2 h. The 4th line: images of ATP7B/LAMP1/GFP staining with added 50 μm CuCl₂ for 2 h followed by 50 μm BCS treatment for 1 h. The last line: images of ATP7B/Golgin‐97/GFP staining with added 50 μm CuCl₂ for 2 h followed by 50 μm BCS treatment for 1 h. Scale bars, 5 μm. Arrows point to ATP7B protein in lysosomes. (E) Copper exports of LPC‐GFP‐Heps and LPC‐ATP7B‐Heps at 3/6/9 h after CuCl₂ withdrawal (n = 3). Results are shown as mean ± SD. ***p < 0.001. Student t test for (A) and (E)

FIGURE 5

Transcriptome profiling of KO‐LPC‐ATP7B‐Heps, KO‐LPC‐GFP‐Heps, and WT‐LPC‐GFP‐Heps under copper stress. (A) Schematic of KO‐ATP7B_Cu, KO‐GFP_Cu, and WT‐GFP_Cu. Cell viabilities were presented as killing curves in KO‐ATP7B_Cu, KO‐GFP_Cu, and WT‐GFP_Cu cells treated with different copper concentrations. Venn diagram for DEGs and the heat map of microarray analysis of the 1100 DEGs in the three groups. (B) Cartoon diagram of key proteins and organelles involved in the intracellular copper transport. The heat map illustrates the copper‐trafficking‐related gene expressions of the three groups. (C) GSEA analysis of pathways on copper ion binding, cell proliferation, and apoptotic process between KO‐ATP7B_Cu and KO‐GFP_Cu cells. (D) Normalized mRNA levels of KO‐ATP7B_Cu and KO‐GFP_Cu cells on copper metabolism (ATP7B, COMMD1, and XIAP), apoptotic process (BCL2, BAX, and Caspase‐3) and cell proliferation (TGFβ2, PDGFα, IGF1, HGF, FGF18, and VEGFα) (n = 3 cell lines). Results are presented as mean ± SD. *p < 0.05; **p < 0.01; ***p < 0.001. One‐way ANOVA with Student‐Newman‐Keuls post‐test for (A). Student t test for (D). AQP, aquaporin

FIGURE 6

LPC‐ATP7B‐Heps transplantation alleviated copper deposit in ATP7B^−/− mice livers. (A) Schematic strategy of cell preparation and transplantation into ATP7B^−/− mice. (B) Representative images of ATP7B staining on liver sections of mouse PHs (mPHs), LPC‐GFP‐Heps, and LPC‐ATP7B‐Heps mice 4 months after cell transplantation. The engraftment ratios of mPHs (n = 5), LPC‐ATP7B‐Heps (n = 6), and LPC‐GFP‐Heps (n = 5) were calculated. Scale bars, 100 μm. (C) Representative images of copper staining in liver sections of mPHs, LPC‐GFP‐Heps, and LPC‐ATP7B‐Heps mice 4 months after transplantation. Scale bars, 100 μm. The levels of liver copper in WT, mPHs, LPC‐GFP‐Heps, and LPC‐ATP7B‐Heps mice were analyzed (n = 5). (D) Serum total and free copper levels, CP activities, and ALT and AST levels were assayed in WT, mPHs, LPC‐GFP‐Heps, and LPC‐ATP7B‐Heps mice (n = 4/group in free copper assay; n = 5/group in other assays). Results are median ± IQR. **p < 0.01; ***p < 0.001. One‐way ANOVA with Student‐Newman‐Keuls post‐test for (B)–(D)