Induced Pluripotent Stem Cells Replace Liver Function in Mice


Liver transplants save lives and in the United States there is a shortage of livers for transplantation. Between July 1, 2008 and June 30, 2011, well over 14,601 adult donor livers were recovered and transplanted. Of these livers that were transplanted, many other patients died from liver failure. If there was a way to restore liver function in patients with liver failure without dependence on a liver from a liver donor, then we might be able to extend their lives.

A paper from the laboratory of Hossein Baharvand at the University of Science and Culture in Tehran, Iran provides a step towards doing just that. In this paper, Baharvand and his colleagues used human induced pluripotent stem cells to make hepatocyte-like cells or HLCs. Hepatocyte is a fancy word for a liver cell. These HLCs were then transplanted into the spleen of mice that have damaged livers, and they rescued liver function in these mice.

The liver is a vital organ. It processes molecules absorbed by the digestive system, processes foreign chemicals to make them more easily excreted. It also produces bile, which helps dispose of fat-soluble waste and solubilize fats for degradation in the small intestine during digestion. It also produces blood plasma proteins, cholesterol and special proteins to cholesterol and fat transport, converts excess glucose into glycogen for storage, regulates blood levels of amino acids (the building blocks of proteins), processes used hemoglobin to recycle its iron content, converts poisonous ammonia to urea, regulates blood clotting, and helps the body resist infections by producing immune factors and removing bacteria from the bloodstream. Thus without a functioning liver, you are in deep weeds.

Induced pluripotent stem cells or iPSCs are made from adult cells that have been genetically engineered to de-differentiate into embryonic-like stem cells. They can be grown in culture to large numbers, and can also be differentiated into, potentially, any cell type in the adult body.

In this paper, Baharvand and his colleagues grew human iPSCs in “matrigel,” and then grew them in suspension. Matrigel is gooey and the cells stick to it and grow, and they were grown in matrigel culture for 1 week. After one week, the cells were grown in liquid suspension for 1-2 weeks. The cells have better access to soluble growth factors in liquid culture and tend to grow faster. After this they were grown in a stirred culture (known as a spinner).  This expanded the cells into large numbers for further use.

 Expansion and characterization of human induced pluripotent stem cells (hiPSCs) in a dynamic suspension culture. (A) Schematic representation of the suspension culture and expansion of human pluripotent stem cells (hPSCs) from adherent to stirred bioreactor. The cells were transferred to bacterial dishes to adapt to three-dimensional (3D) environment and then transferred into the dynamic phase, stirred flask. (B) Morphology of a tested hiPSC line (hiPSC1) after passage in a stirred suspension bioreactor as monitored by dark-field microscopy. (C) Immunostaining of cross sections of spheroids for OCT4 and TRA-1-81. Scale bar: 50 μm. (D) Flow cytometry analysis and (E) normal karyotype of suspended cells in the bioreactor.
Expansion and characterization of human induced pluripotent stem cells (hiPSCs) in a dynamic suspension culture. (A) Schematic representation of the suspension culture and expansion of human pluripotent stem cells (hPSCs) from adherent to stirred bioreactor. The cells were transferred to bacterial dishes to adapt to three-dimensional (3D) environment and then transferred into the dynamic phase, stirred flask. (B) Morphology of a tested hiPSC line (hiPSC1) after passage in a stirred suspension bioreactor as monitored by dark-field microscopy. (C) Immunostaining of cross sections of spheroids for OCT4 and TRA-1-81. Scale bar: 50 μm. (D) Flow cytometry analysis and (E) normal karyotype of suspended cells in the bioreactor.

Getting cells to grow in liquid suspension tends to be a bit of an art form, but these iPSCs grew rather well. Also, the iPSCs were differentiated into definitive endoderm, which is the first step in bringing cells to the liver cell stage. The drug Rapamycin and activin (50 ng / L for those who are interested) were used to bring the growing iPSCs to the definitive endoderm.  The cells expressed all kinds of endoderm-specific genes.  Endoderm is the embryonic germ layer from which the digestive system and its accessory organs forms.

 Induction of hiPSCs into definitive endoderm. (A) Diagrammatic representation of the experimental groups for endoderm induction of hiPSCs in the bacterial dish static suspension, which include rapamycin (Rapa) “priming” and activin A “inducing” phases, and positive control groups of hiPSCs cultured in the absence of Rapa in suspension or adherent cultures in the presence of activin A. (B) Gene expression analysis of hiPSCs induced into endodermal cells. Quantitative reverse transcriptase–polymerase chain reaction (qRT-PCR) showed no significant differences in SOX17 and FOXA2 [definitive endoderm (DE) markers], and BRA (mesoendoderm marker) in all groups. SOX7 (visceral endoderm marker) was expressed at a low level. Influence of the refreshment strategy of the induction medium on DE formation in suspension cultures by qRT-PCR (C), immunostaining (D), and flow cytometry (E). We compared single (SR), double (DR), and triple (TR) refreshment of induction medium per 24 h for 4 days after Rapa administration in the static suspension of hiPSCs. Scale bar: 100 μm. The target gene expression level in qRT-PCR was normalized to GAPDH and calibrated with (presented relative to) hiPSCs. Data are presented as mean±SD. Statistical analysis as determined by one-way ANOVA with Tukey's post hoc test, n=3 for B, C, and E. *P<0.05, **P<0.01.
Induction of hiPSCs into definitive endoderm. (A) Diagrammatic representation of the experimental groups for endoderm induction of hiPSCs in the bacterial dish static suspension, which include rapamycin (Rapa) “priming” and activin A “inducing” phases, and positive control groups of hiPSCs cultured in the absence of Rapa in suspension or adherent cultures in the presence of activin A. (B) Gene expression analysis of hiPSCs induced into endodermal cells. Quantitative reverse transcriptase–polymerase chain reaction (qRT-PCR) showed no significant differences in SOX17 and FOXA2 [definitive endoderm (DE) markers], and BRA (mesoendoderm marker) in all groups. SOX7 (visceral endoderm marker) was expressed at a low level. Influence of the refreshment strategy of the induction medium on DE formation in suspension cultures by qRT-PCR (C), immunostaining (D), and flow cytometry (E). We compared single (SR), double (DR), and triple (TR) refreshment of induction medium per 24 h for 4 days after Rapa administration in the static suspension of hiPSCs. Scale bar: 100 μm. The target gene expression level in qRT-PCR was normalized to GAPDH and calibrated with (presented relative to) hiPSCs. Data are presented as mean±SD. Statistical analysis as determined by one-way ANOVA with Tukey’s post hoc test, n=3 for B, C, and E. *P<0.05, **P<0.01.
After the cells went through this culture protocol, they were grown in a stirred liquid culture called a “spinner.” The culture system contain a cocktail of growth factors that differentiated the definitive endoderm cells into HLCs.  The cells formed little spheres that expressed a host of liver-specific genes.

Differentiation of hepatocyte-like cells (HLCs) from hiPSCs in the stirred bioreactor. (A) Stepwise protocol for differentiation of hiPSCs into HLCs. (B) The morphology and cross section of spheroids at day 21. The spheroids in this step were observed as cystic and dense spheroids. Hematoxylin and eosin (H&E) staining of spheroid cross sections indicated cystic and dense epithelioid appearances in the transplant and dense spheroids, respectively. Scale bar: 100 μm. (C) Comparative relative mRNA expression in cystic and dense spheroids normalized to GAPDH and calibrated to undifferentiated hiPSCs. Transplant and dense spheroids expressed more early and late hepatic lineage markers, respectively. Data are presented as mean. n=3. (D) Transmission electron microscopy (TEM) of differentiated cells in dense spheroids at day 21. Nucleus (N), nucleoli (n), mitochondria (M), Golgi apparatus (G), lysosomes (Ly), rough endoplasmic reticuli (arrowhead), glycogen granules (GR), intermediate filaments (CK), tight junctions (TJ), gap junctions (GJ), fascia adherens (FA), junctional complex (JC), microvilli (MV), and bile-like canaliculus (BLC). Scale bar: 1 μm.
Differentiation of hepatocyte-like cells (HLCs) from hiPSCs in the stirred bioreactor. (A) Stepwise protocol for differentiation of hiPSCs into HLCs. (B) The morphology and cross section of spheroids at day 21. The spheroids in this step were observed as cystic and dense spheroids. Hematoxylin and eosin (H&E) staining of spheroid cross sections indicated cystic and dense epithelioid appearances in the transplant and dense spheroids, respectively. Scale bar: 100 μm. (C) Comparative relative mRNA expression in cystic and dense spheroids normalized to GAPDH and calibrated to undifferentiated hiPSCs. Transplant and dense spheroids expressed more early and late hepatic lineage markers, respectively. Data are presented as mean. n=3. (D) Transmission electron microscopy (TEM) of differentiated cells in dense spheroids at day 21. Nucleus (N), nucleoli (n), mitochondria (M), Golgi apparatus (G), lysosomes (Ly), rough endoplasmic reticuli (arrowhead), glycogen granules (GR), intermediate filaments (CK), tight junctions (TJ), gap junctions (GJ), fascia adherens (FA), junctional complex (JC), microvilli (MV), and bile-like canaliculus (BLC). Scale bar: 1 μm.

From the figure above, we can see that these HLCs, not only express liver-specific genes, but when they are examined in the electron microscope they look, for all intents and purposes, like liver cells.  Functional tests of these spheres of HLCs showed that they 1) took up low-density lipoprotein; 2) produced albumin (a major blood plasma protein); 3) expressed cytochrome P450s, which are the major enzymes used to process drugs; 4) produced urea from amino acids, just like real liver cells; 5) accumulated glycogen; 6) and made liver proteins (HNF4a, ALB, etc).

So it looks like liver, quacks like liver, but can it replace liver?  These HLCs were transplanted into the spleen of mice whose livers had been treated with carbon tetrachloride.  Carbon tetrachloride tends to make mincemeat of the liver, and these mice are in trouble, since their livers are toast.  Transplantation of the iPSC-derived HLCs into the spleens of these mice increased their survival rate and decreased the blood levels of liver enzymes that are usually present when there is liver damage.

This paper is significant because the procedure used provides an example of a “scalable” protocol for making large quantities of iPSCs, and their mass differentiation into definitive endoderm and then liver cells,  Because this can potentially provide enough cells to replace a nonfunctional liver, it represents a major step forward in regenerative medicine.