Experimental Study On The Co-culture Of Hepatocytes With Bone Marrow Mesenchymal Stem Cells In Vitro

Acute liver failure remains one of the diseases with high mortality. Recently, extracorporeal bioartificial liver (BAL) support system has become an alternative to temporarily extend the lives of patients with acute liver failure and bridge to orthotopic liver transplantation. The efficacy of BAL for the individuals relies on hepatocyte biological characteristics which is the key to BAL. Therefore, it stresses a compelling need for a large number of hepatocytes with no harm to patients and more similar to human ones. The greatest challenge to the recreation of a functional engineered liver in vitro remains the short-term viability and rapid phenotypic de-differentiation of primary hepatocytes in standard monolayer culture. Over the past decades, previous attempts to provide hepatocytes with a liver-like microenvironment in vitro have been made by co-culturing hepatocytes with a range of different cell types. In diversified co-culture systems, homotypic and heterotypic cell-cell interactions play a fundamental role in the modulation of cellular behaviors such as migration, proliferation, and differentiation. Hence, co-cultivation is becoming an ideal system to maintain cells of high viability on a large scale. In bone marrow, mesenchymal stem cells (MSCS) with self-renewal and multipotent potential are typically characterized by their ability to provide a scaffold and physiological support for hematopoietic stem cells by secreting various soluble factors and endogenous extracellular matrix (ECM) proteins. Given that the stromal supporting characteristics, MSCS seem to be an alternative for co-culture with hepatocytes for liver support system. This study aims to specify the effects of MSCS on hepatocytes and uncover molecular mechanisms within this co-cultivated system, as well as identify the key molecular targets related to hepatocyte proliferation and apoptosis. It may achieve the hepatocytes with high performance on a large scale which could be a kind of ideal cellular source, and provide theoretical clues to the application of molecular targeted therapy to the large-scale hepatocyte culture. Furthermore, whether such a co-culture system could endure the cytotoxicity within sera from the patients with liver failure should be determined due to the presence of circulating regenerative inhibitors.Part I. Establishment of the optimal co-cultivation of hepatocytes with bone marrow MSCS in vitroObjective: To present an optimal randomly distributed co-culture manner with porcine hepatocytes and bone marrow MSCS in vitro, which would be an ideal cell source for BAL configuration.Methods: Mononuclear cells were isolated from bone marrow aspirate of swines (n=3) by density gradient centrifugation. MSCs of passage 1 to 3 were characterized by flow cytometry with CD29, CD44, CD45 and CD90, respectively. The viability of primary hepatocytes harvested by a two-step in situ collagenase perfusion technique was determined by trypan blue exclusion. Immunochemical staining of hepatocytes with anti-albumin and CK 18 antibodies were performed to discriminate nonparenchymal cells. A randomly distributed co-culture system composed of porcine hepatocytes and bone marrow mesenchymal stem cells was generated, and the morphological and functional changes of varying degrees of heterotypic interactions were characterized.Results: The purity of the third passage MSCS and primary hepatocytes was more than 90% and 99%, respectively. Hepatocyte viability was greater than 95%. A rapid attachment and self-organization of three-dimensional hepatocyte aggregates were encouraged. The cell ultrastructure indicating heterotypic junctions remained similar to that of hepatocytes in vivo. Confocal fluorescence microscopy further verified that MSCS served as a feeder layer for hepatocyte aggregates. Representative images recorded by live-cell microscopy indicated high hepatocyte motility in co-culture, which led to spheroid formation. Hepatic glycogen accumulation content evaluated by PAS staining was evidently increased in co-culture than that in mono-culture. Immunocytochemical staining demonstrated the increased synthesis of intracellular albumin in co-culture. Studies on hepatocyte viability showed a metabolically active, viable cell population in all co-culture configurations with occurrence of few dead cells. The maximal induction of albumin production and urea synthesis was achieved at seeding ratio of 2:1, peaking on day 2, and gradually decreased over time (P<0.05). Western blotting analysis of cultured hepatocytes separated from MSCS by fluorescence-activated cell sorting showed that albumin and CYP3A1 expressions on day 2 were significantly increased than those of homoculture (P<0.05). Further comparison within co-cultured subgroups proved hepatocyte/MSC ratio of 2:1 to be the optimal configuration (P<0.05). Cell cycle analysis indicated that larger populations of hepatocytes without MSCS treatment were accumulated in G0-G1 phase, and fewer populations of cells in G2-S phase compared with co-cultured hepatocytes (P<0.05).Conclusions: Co-cultivation of porcine hepatocytes and MSCS at a seeding ratio of 2:1 on day 2 may preserve hepatocyte morphology and functions to a great degree, which could contribute to the functional BAL configuration.Part II. Study on mechanisms of bone marrow MSCS on porcine primary hepatocyte culture in vitroObjective: To investigate the roles of endogenous ECM and soluble cytokines from bone marrow MSCS involved in the optimal co-culture system.Methods: Mononuclear cells were isolated from bone marrow aspirate of swines (n=3) by density gradient centrifugation. The optimal co-cultivation of MSCS and primary hepatocytes harvested by a two-step in situ collagenase perfusion technique was generated. Immunocytochemical analyses of fibronectin, laminin, and collagen type I, III, V were performed to reveal ECM distribution. siRNA specific for fibronectin, laminin, and collagen type I and V were designed and transfected to MSCS. Hepatic functions alteration were evaluated after inoculation with transfected MSCS. To exclude cell-cell direct contact, hepatocyte/MSC of 2:1 was co-cultured by using a semi-permeable membrane Millicell culture insert. Hepatic functions alteration were evaluated and cytokine assays were performed. Furthermore, hepatocyte specific functions were also examined upon neutralizing experiments.Results: Most of the ECM proteins were present in the cytoplasm of hepatocytes except fibronectin. As for the ECM deposits of MSCS, a few extracellular fibrils containing fibronectin, laminin, collagen type I and V but never collagen type III was observed. In the co-culture, the hepatocyte aggregates were covered with a dense network of ECM fibrils, which could be deemed as a scaffold of diversified ECM proteins similar to microenvironment in vivo. The effects of gene silencing of ECM confirmed at the protein levels by Western blotting clearly demonstrated that fibronectin, laminin, collagen type I levels in siRNA-treated MSCS were all inhibited by over 85% but siRNA for collagen type V only decreased the expression by about 50%. Both albumin and urea synthesis were significantly decreased after hepatocytes were co-cultured with siRNA-treated MSCS irrespective of any kind of the above ECM inhibition (P<0.05). The analysis of conditioned medium indicated that both albumin and urea synthesis were also elevated in separated co-culture (P<0.05). IL-6 level of hepatocyte alone was too low beyond the limit of detection while a significant increase was observed in both MSCS and co-culture (P<0.05). Statistical analysis with regard to TGF-?level showed no significant difference between co-culture and other two control groups (P>0.05). In addition, TNF-?failed to show apparent expression in any circumstances (P>0.05). In the absence of IL-6 in co-culture, the albumin secretion and urea synthesis were both reduced compared with the conditioned medium from the normal co-culture (P<0.05).Conclusions: These results demonstrate that both ECM and soluble factors have beneficial effects on the preservation of hepatic morphology and functionality in the co-culture of hepatocytes with MSCS in vitro, which could represent a promising tool for tissue engineering, cell biology, and BAL devices.Part III. Proteomic analysis of primary hepatocytes co-cultured with bone marrow MSCS in vitroObjective: To identify the differentially expressed proteins in hepatocytes within the optimal co-cultivation system, analyze the functions of specified proteins, and clarify the mechanisms underlying the functional improvement of primary hepatocytes upon co-cultivation with MSCS in protein levels.Methods: Mononuclear cells were isolated from bone marrow aspirate of swines (n=3) by density gradient centrifugation. The optimal co-cultivation of MSCS and primary hepatocytes harvested by a two-step in situ collagenase perfusion technique was generated. CD44-CD45- cell population was sorted from co-culture by flow cytometry. Hepatocyte homo-culture and MSCS alone were set as controls, respectively. Protein samples were isolated by two-dimensional (2D) gel electrophoresis, analyzed by ImageMaster 2D Elite software, and thereafter identified by mass spectrometer. Bioinformatics was employed to analyze the relevance between the altered proteins and corresponding biological events. The protein expression levels of interest were confirmed by Western Blotting.Results: CD44-CD45- (hepatocytes) and CD44+CD45- (MSCS) cell populations were successfully separated by flow cytometry. The reproducibility of the 2D gels was high. Statistical analysis revealed that 40 protein spots were differentially expressed more than 3 folds between co-culture and homo-culture. Of these hepatic protein spots, 19 were up-regulated whereas 21 were down-regulated in the co-culture. Thirty-four out of the 40 altered protein spots were unambiguously identified with high confidence, which are likely associated with cellular growth, differentiation, proliferation, apoptosis, biosynthesis, energy metabolism, and cell cycle. The results of CFTR, IL-6 and HMG-1 by Western Blotting were consistent with proteomic findings.Conclusions: Several differentially expressed functional proteins within co-cultured hepatocytes have been revealed by comparative proteomics, which may be associated with enhanced hepatic functionality, prolonged cellular survival, and immune modulation, which would provide molecular targets to optimize hepatocyte functions.Part IV. Experimental study on hepatocytes'tolerance of the cytotoxicity within sera from the patients with liver failure induced by bone marrow MSCS in vitroObjective: To clarify the hepatocytes'tolerance of the cytotoxicity within sera from the patients with liver failure induced by bone marrow MSCS, and the supporting effects on liver failure sera by co-culture.Methods: Serum samples from 18 patients with acute-on-chronic liver failure due to hepatitis B were collected with informed consent. Normal serum was obtained from 18 normal volunteers. The optimal co-cultivation of MSCS and primary hepatocytes was generated. It consisted of 4 groups as follows: hepatocyte alone with liver failure serum (Hep-F), co-culture with liver failure serum (Co-F), hepatocyte alone with normal serum (Hrp-N), and co-culture with normal serum (Co-N). The additional serum concentration was set as 10%, 20%, 40%, 60%, 80% and 100%, respectively. The albumin expression and cell cycle of hepatocytes after 24-h culture were examined to determine the optimal liver failure serum concentration. The morphology and viability were also monitored. The indices of hepatic functions in every group were performed. Subsequently, the cells were inoculated in complete medium containing 10% fetal bovine serum for another 24 h, and hepatocyte specific functions of Hep-FB, Co-FB, Hep-NB, and Co-NB were examined, respectively.Results: A well attachment and self-organization of islet-like hepatocyte aggregates were encouraged in Co-F group with 60% liver failure serum. It also achieved the best albumin secretion and largest population of hepatocytes in G2-S phase. The cell ultrastructure remained similar to that of hepatocytes in vivo. Meanwhile, hepatocyte viability assay showed a metabolically active, viable cell population in Co-F group instead of quite a few dead cells in Hep-F group. The activity of SOD, LDH, GLN and CHE from Co-F group were all significantly improved compared with those from Hep-F group (P<0.05). The similar trends were observed between Co-FB group and Hep-FB group except SOD.Conclusions: The co-cultured hepatocytes with bone marrow MSCS could counteract the cytotoxicity of mid-high concentration serum with acute-on-chronic liver failure, which would manifest a well hepatic supporting function.