A liver-specific gene expression panel predicts the differentiation status of in vitro hepatocyte models

Abstract

Alternative cell sources, such as three-dimensional organoids and induced pluripotent stem cell-derived cells, might provide a potentially effective approach for both drug development applications and clinical transplantation. For example, the development of cell sources for liver cell-based therapy has been increasingly needed, and liver transplantation is performed for the treatment for patients with severe end-stage liver disease. Differentiated liver cells and three-dimensional organoids are expected to provide new cell sources for tissue models and revolutionary clinical therapies. However, conventional experimental methods confirming the expression levels of liver-specific lineage markers cannot provide complete information regarding the differentiation status or degree of similarity between liver and differentiated cell sources. Therefore, in this study, to overcome several issues associated with the assessment of differentiated liver cells and organoids, we developed a liver-specific gene expression panel (LiGEP) algorithm that presents the degree of liver similarity as a "percentage." We demonstrated that the percentage calculated using the LiGEP algorithm was correlated with the developmental stages of in vivo liver tissues in mice, suggesting that LiGEP can correctly predict developmental stages. Moreover, three-dimensional cultured HepaRG cells and human pluripotent stem cell-derived hepatocyte-like cells showed liver similarity scores of 59.14% and 32%, respectively, although general liver-specific markers were detected.

Conclusion: Our study describes a quantitative and predictive model for differentiated samples, particularly liver-specific cells or organoids; and this model can be further expanded to various tissue-specific organoids; our LiGEP can provide useful information and insights regarding the differentiation status of in vitro liver models. (Hepatology 2017;66:1662-1674).

Figure 1

A workflow for developing the LiGEP. (1) Standard RNA‐seq analysis and selection of liver‐specific genes. RNA‐seq analysis with total RNA from 20 tissues purchased from Clontech and screening liver‐specific genes compared with 18 nonliver tissues through TopHat‐cufflinks pipelines. (2) Validation. Validation of candidate genes using the public database of Human Protein Atlas and quantitative RT‐PCR analysis. (3) LiGEP algorithm. Construction of LiGEP algorithm to measure the liver similarity of liver organoids and 3D culture and validation of LiGEP algorithm with normal tissues (730 normal tissues).

Figure 2

Construction and characterization of LiGEP. (A) A multidimensional scaling plot analysis was performed with RNA‐Seq results of LiGEP in 20 human tissues. The liver and other tissues are clearly separated in two major sections. The x axis represents the human tissues, and the y axis shows the 93 LiGEP genes. (B,C) A heat map representing the gene expression level of LiGEP in 20 (in‐house) and 32 (Human Protein Atlas) tissues. Tissues with similar FPKM values (log₁₀[FPKM +1]) were clustered. (D) Correlation plot of LiGEP in the Human Protein Atlas data. The x and y axes describe the names of human tissues.

Figure 3

Functional enrichment analysis of LiGEP from the Ingenuity Pathway Knowledge Base. (A) Canonical pathway analysis. Representative canonical pathways are depicted as a histogram. The P value associated with the pathways is a measure of the significance with respect to the pathways for the imported data set and a reference set of molecules involved in a given pathway. The orange line indicates the threshold of significance (P < 0.05). (B) Disease and biofunction analysis. The top regulated disease and biofunction are selected, and the significance is displayed with a heat map. (C) Gene networking analysis. Each gene was matched to human homolog proteins, and the human genes are represented in the networking map. The expression intensity of the focused genes is colored in red. The predicted diseases and functions are displayed in blue and purple, respectively. Abbreviations: FXR, farnesoid X receptor; LXR, liver X receptor; RXR, retinoid X receptor.

Figure 4

The LiGEP algorithm represents the developmental stage of liver. (A) Distribution of LiGEP algorithm scores in TCGA normal data. Healthy human organ RNA‐seq data were downloaded from the TCGA cohort. A box plot shows the interquartile range of LiGEP algorithm scores in 15 tissue types. Liver samples (n = 50) and other tissue samples (n = 680) are used to assess the accuracy of the LiGEP algorithm. (B) LiGEP algorithm analysis between liver and fetal liver. Red dots represent the LiGEP expression of fetal liver, and the red line represents the criteria for distinguishing between liver and nonliver. (C) Liver similarity by the LiGEP algorithm of clinical liver samples (livers 1‐3), ATLAS liver, and fetal liver. (D) Results of the LiGEP algorithm with each liver developmental stage. The dot means the genes passed over LiGEP criteria for distinguishing liver. (E) Liver similarity using the LiGEP algorithm for each liver developmental stage in mice.

Figure 5

LiGEP indicates liver similarity of 3D culture and hPSC‐derived HLCs. (A) Morphology of 2D and 3D HepaRG cells and spheroids. HepaRG cells were dropped onto the culture plates. Spheroids were formed by the hanging drop method for 48 hours and then transferred to suspension culture tubes, and spheroid formation was monitored by light microscopy. Scale bar, 200 μm. (B) CYP3A4 enzyme activity in 2D (n = 3, pooled) and 3D (n = 6, pooled) cultures. The data, reported as relative light units, were normalized to the DNA content of hepatocytes. (C) Albumin expression between 2D and 3D cultures. The mRNA level of the albumin gene was measured by real‐time PCR. Gene expression was normalized to that of glyceraldehyde 3‐phosphate dehydrogenase and compared to that of the control (2D monolayer culture). (D) The result of the LiGEP algorithm with 2D and 3D cultures. Liver similarity of 2D and 3D culture cells by the LiGEP algorithm. (E) Quantitative RT‐PCR analysis of expression of hepatic maturation markers, including alpha‐1 antitrypsin, CYP3A4, and glutathione S‐transferase A1 and A2, was performed in undifferentiated hPSCs and hPSC‐HLCs. (F) Immunocytochemical analysis of hepatocyte‐specific markers (hepatocyte nuclear factor 4 alpha, alpha‐1 antitrypsin, and cytokeratin‐18) in HLCs derived from hESCs and human‐induced pluripotent stem cells. (G,H) LiGEP algorithm with hPSC‐derived HLCs. Expressed genes of LiGEP (G) and liver similarity (H). Abbreviations: AAT, alpha‐1 antitrypsin; CK‐18, cytokeratin 18; DAPI, 4′,6‐diamidino‐2‐phenylindole; GSTA1/2, glutathione S‐transferases A1 and A2; hiPSC, human‐induced PSC; HNF4A, hepatocyte nuclear factor 4 alpha; Undiff., Undifferentiation.

Figure 6

Graphic summary of LiGEP algorithm to assess differentiation of 3D culture and hPSC‐derived HLCs. The score of the LiGEP algorithm according to each stage of generation can be used as important information to generate high‐quality liver organoids. Low percentage scores of the LiGEP algorithm mean that the quality of the 3D culture/hPSC‐derived HLCs is not high. To increase the quality of the liver organoid, the pipeline (organoid generation–LiGEP algorithm) must be performed.