Hepatogenic differentiation of human mesenchymal stem cells from peritoneal adipose tissue

BACKGROUNDS/AIMS: It has been reported that functional hepatogenic differentiation has the possibility to occur in subcutaneous adipose tissue-derived stem cells. However, no studies have investigated whether the adipose tissue-driven stem cells present in various body parts differ according to hepatogenic differentiations. In this study, stem cells were separated from body visceral fat and abdominal subcutaneous adipose tissue, and cultured, and then hepatogenic differentiation was induced. We aim to investigate the possibilities and aspects of hepatogenic differentiations within the two types of fat cells. METHODS: Omental fat tissues were obtained as visceral fat and abdominal subcutaneous adipose tissues were obtained from patients who had suction-assisted lipectomy. Stem cells were separated from the obtained fat tissues, and then, hepatogenic differentiation was carried out by utilizing 2-step differentiation protocols. RESULTS: After the differentiation, two types of cultured cells that showed the similar neuron-like shapes were changed to cuboidal shapes and included several binucleated cells which could be characteristics of mature hepatocytes. We confirmed that hepatocyte specific genes and proteins such as albumin and CYP3A4 were being expressed. By utilizing the ELISA test, we were able to observe that the albumin was secreted into the culture fluids in both cells. After completing the differentiation, we observed the presence of the hepatocyte specific properties by confirming glycogen storage within the cells and the ICG reagent uptake. CONCLUSIONS: We confirmed that hepatogenic differentiation was possible to occur in the omental fat as well as subcutaneous adipose tissue.

Diferenciação de células-tronco em hepatócitos e desenvolvimento de modelo pré-clínico de fibrose hepática para ensaios de terapia celular / Mesenchymal stem cell differentiation in hepatocytes and development of pre-clinic model of hepatic fibrosis for cellular therapy assays

Este trabalho teve como objetivo desenvolver um protocolo para a diferenciação in vitro de células-tronco mesenquimais (CTM) em hepatócitos e a padronização de um modelo animal de fibrose hepática induzida por dimetilnitrosamina (DMN) para ensaios pré-clínicos de transplante de CTM. CTM isoladas de fontes variadas apresentaram morfologia fibroblastóide e aderência ao plástico e o padrão de marcadores de superfície celular esperado na análise por citometria de fluxo. A capacidade de diferenciação osteogênica e adipogênica dessas células foi comprovada pelas colorações de vermelho de alizarina, oil red e azul de toluidina, respectivamente, confirmando, que as células isoladas para este estudo se comportaram como CTM conforme proposto pela Sociedade Internacional de Pesquisa em Células-tronco. A diferenciação hepática foi avaliada quanto à morfologia e capacidade das células diferenciadas de estocar glicogênio confirmada por PAS (ácido periódico-Schiff), de sintetizar albumina confirmada por imunofluorescência, além da capacidade de expressar genes hepato-específicos verificada por ensaios de PCR em tempo real. Com base na literatura para diferenciação hepática, diferentes protocolos de um, dois e três passos foram testados. CTM humanas mostraram capacidade de produzir e estocar glicogênio e de sintetizar albumina, apenas quando diferenciadas com protocolos de três etapas, porém sem uma expressão aumentada dos genes hepato-específicos albumina, α-fetoproteína e c-Met. Uma etapa de diferenciação endodérmica, previamente aplicada à diferenciação hepática, aumentou a capacidade de produzir e estocar glicogênio das CTM diferenciadas. Para a padronização do modelo de fibrose hepática induzida por DMN, foram realizados experimentos de dose-resposta e foi verificado o efeito da hepatectomia em modelos mistos DMN/hepatectomia. A injúria hepática e o efeito do transplante de CTM foram avaliados por análise macroscópica dos fígados, histologia das biópsias de fígados corados com HE e tricromo de Masson e parâmetros bioquímicos séricos. Alterações macroscópicas, histológicas e nos níveis séricos de fosfatase alcalina indicam a indução da fibrose hepática nos ratos Wistar tratados com DMN na dose de 10 µg/g de peso animal por três dias consecutivos durante quatro semanas, mas não observamos nenhum efeito induzido pela hepatectomia. Porém, este modelo com DMN se mostra semelhante a estágios iniciais de uma fibrose hepática. O transplante de 1 x 107 CTM de veia de cordão umbilical humano (VCUH) no modelo de injúria hepática induzida por DMN não resultou em melhora da fibrose, diminuição dos níveis séricos de fosfatase alcalina e nem em ganho de peso dos animais quando comparados aos animais tratados com PBSA após a injúria hepática (grupo placebo). Em conjunto, esses resultados sugerem que CTM humanas se diferenciam após tratamentos mais complexos, onde os indutores hepatogênicos são sequencialmente adicionados ao meio de modo a mimetizar a sinalização durante o desenvolvimento embrionário. O transplante de CTM de VCUH parece não ter efeito positivo em um modelo pré-clínico de injúria hepática similar a estágios iniciais de fibrose. Financiado por CNPq (573578/2008-7) e FAPESP (2007/54260-2)

This study aimed to develop an in vitro differentiation protocol of mesenchymal (MSC) stem cells to hepatocytes and to standardize an animal model for hepatic fibrosis induced by dimethylnitrosamine (DMN) for preclinical transplant assays of MSC. MSC isolated from various sources presented fibroblastoid morphology, plastic adherence, and the expected pattern of cell surface markers by flow cytometry analysis. The capacity of osteogenic, adipogenic and chondrogenic differentiation of these cells was confirmed by alizarin red, oil red and toluidine blue staining, respectively, confirming that the cells isolated for this study behave as MSC, as proposed by the International Society for Stem Cell Research. Hepatogenic differentiation was evaluated by analysis of cell morphology, capacity to store glycogen confirmed by PAS (periodic acid-Schiff), albumin synthesis confirmed by immunofluorescence, as well as hepatic-specific gene expression verified by real time PCR assays. Based on the published literature on hepatic differentiation, several protocols of one, two, and three steps were tested. Human MSC differentiated solely when treated in a three step-protocol, showing the ability to produce and store glycogen and synthesize albumin; however the expression of hepatic-specific genes such as albumin, α-fetoprotein and c-Met was not increased. An endoderm differentiation stage, added to the hepatic differentiation protocol, increased the capacity to produce and store glycogen of differentiated MSC. In order to standardize the model of liver fibrosis induced by DMN, dose-response experiments were performed and the effect of hepatectomy in mixed models DMN/hepatectomy was observed. Severity of liver injury and the effect of cell transplantation were evaluated by macroscopic analysis of the livers, histology of liver biopsies stained with HE and Masson's trichrome, and evaluation of serum biochemical parameters. The macroscopic and histological observations, and altered alkaline phosphatase serum levels indicated the success in inducing liver fibrosis in DMN-treated rats at a dose of 10 µg/g of animal weight for three consecutive days, during four weeks, without any additional effect upon hepatectomy. Transplanting 1 x 107 umbilical cord MSC in the model of liver injury induced by DMN did not result in improvement of the fibrosis, decrease of alkaline phosphatase serum levels, or in weight gain of the treated animals compared to animals treated with PBSA after liver injury (placebo group). Together, these results suggest that human MSC are capable of differentiating to hepatocyte-like cells after more complex protocols, where hepatogenic inducers are sequentially added to the medium in order to mimic signaling that occurs during fetal development. Transplantation of undifferentiated umbilical cord MSC did not have any positive effect in a preclinical liver injury model characterized by an early stage of fibrosis. Supported by CNPq (573578/2008-7) and FAPESP (2007/54260-2)