Shin-Huei Fu

Loofa Sponge as a Scaffold for the Culture of Human Hepatocyte Cell Line

Loofa sponge was investigated as a three‐dimensional scaffold for stationary and perfusion culture of human hepatoblastoma cell line C3A/HepG2. In stationary culture, C3A/HepG2 cells in loofa cubes showed higher α‐fetoprotein and albumin secretion rates than those in polyurethane foam (PU). To use loofa cylinders in a packed‐bed reactor, immobilization of C3A/HepG2 cells by recirculating medium at 26 mL/min (superficial velocity = 51.7 cm/min) resulted in a cell loading density of 5.15 × 107 cells/cm3‐loofa. This cell loading density is higher than values reported in the literature for packed‐bed reactor intended for bioartificial liver. During 9 days of perfusion culture in the reactor, immobilized C3A/HepG2 showed steady synthesis of albumin with an average synthesis rate at 42.2 μg/106 cells/day. These experimental results and observations by SEM suggested that loofa sponge is a suitable scaffold for high‐density culture of human hepatocyte cell line and the immobilized cells could express high levels of liver‐specific functions.

Microencapsulation of islets in PEG-amine modified alginate-poly(l-lysine)-alginate microcapsules for constructing bioartificial pancreas

Abstract Two positively charged derivatives of poly(ethylene glycol) (PEG) were coated onto alginate-poly( l -lysine)-alginate (A-P-A) microcapsules by allowing them to interact them with the negatively charged alginate on the capsule surface. The polymers are methoxypolyoxyethylene amine (PEGA1) and polyoxyethylene bis(amine) (PEGA2), which contain charged amine groups at one or both ends, respectively, with PEG as the backbone. The coating of the microcapsules with PEG-amine resulted in a much smoother capsule surface than A-P-A microcapsule surfaces as examined under a scanning electron microscope. The diffusivity of bovine serum albumin into the microcapsules remained the same after PEGA1 coating. But the diffusivity decreased to less than one-fifth that in A-P-A microcapsules coated with PEGA2. The biocompatibility of the microcapsules also improved as investigated by an in vivo study. Microcapsules were implanted in the peritoneal cavity of BALB c mice and retrieved 120 d after implantation. The fibrotic action against A-P-A microcapsules was severe and the capsules retrieved by peritoneal lavage aggregated into clusters. In contrast, the surface-modified capsules were free-flowing and free of cell overgrowth. Secretion of insulin from rat islets within A-P-A-PEGA microcapsules responded well to changes in glucose concentration in a static glucose test. Intraperitoneal transplantation of the microencapsulated islets into streptozotocin-induced diabetic mice could maintain normal blood glucose levels in test animals for up to 200 d without immunosuppression.

The rescue effect of 15-deoxyspergualin on intraperitoneal microencapsulated xenoislets.

Because the development of surface neogrowth composed mainly of macrophages and fibroblasts precedes the recurrence of hyperglycemia in treated diabetic animals, the pericapsular macrophages may adversely affect the graft function of IP alginate-poly-L-lysine-alginate (A-P-A) microencapsulated islets. In order to clarify the role of pericapsular macrophages on late islet xenograft dysfunction, we investigated whether 15-deoxyspergualin (15-DSG), a macrophage inhibitor, has a rescue effect on the recurrent hyperglycemia in streptozotocin-induced diabetic mice that had been treated with IP transplantation of A-P-A microencapsulated rat islets. The mean duration of normoglycemia (whole blood glucose level below 8.3 mmol/l) in streptozotocin-induced diabetic mice treated with implantation of about 2200–2400 of A-P-A microencapsulated rat islets was 75 days. When the blood glucose levels were higher than 11.1 mmol/l for two consecutive determinations, 15-DSG at a dose of 0.625 mg/kg body weight or isotonic sodium chloride solution (control group) was given daily SC. The blood glucose levels decreased significantly from 13.9 ± 0.5 mmol/l to 11.0 ± 1.3 mmol/l (n = 18, p < 0.05) at the fourth day and to 7.6 ± 1.0 mmol/l (n = 18) at the 14th day of 15-DSG administration. That was not significantly different from the mean glycemic level during the normoglycemic period (7.6 ± 1.0 vs. 7.0 ±1.7 mmol/l, n = 18, p = NS). Isotonic sodium chloride solution injections did not reduce glycemic levels of mice in the control group. As another control, 10 streptozotocin-induced diabetic mice were given the same daily doses of 15-DSG for 14 days. 15-DSG did not decrease the blood glucose levels of diabetic mice in the control group. We further studied the effect of 15-DSG on the expression of interleukin-1β (IL-1β) in peritoneal exudate mononuclear cells (PEMCs) using reverse transcription-polymerase chain reaction. It was found that the mRNA of IL-1β was undetectable in PEMCs of 15-DSG-treated diabetic mice even after those cells were stimulated by lipopolysaccharides in vitro. Administration of 15-DSG at a daily dose of 0.625 mg/kg body weight from the 22nd to the 28th day after transplantation and 7 consecutive days every 3 weeks thereafter did not prolong graft survival of IP microencapsulated rat islets. Our data suggest that 15-DSG has a rescue effect when A-P-A microencapsulated islets have induced cellular overgrowth that threatens the survival of the graft. It is possible that the surface overgrowth composed of macrophages is involved in the pathophysiology of late failure of A-P-A microencapsulated xenogeneic islets.

Reduction in primary nonfunction of syngeneic islet transplants with nordihydroguaiaretic acid, a lipoxygenase inhibitor.

To study the effectiveness of a lipoxygenase inhibitor, nordihydroguaiaretic acid (NDGA), in the reduction of primary nonfunction, an insufficient number of syngeneic islets were transplanted underneath the renal capsule with NDGA administered daily for 4 weeks. After transplantation of the 150 islets, the decrement of blood glucose levels was significantly faster in the mice that had received NDGA than in the mice that had received no drug at all or dimethyl sulfoxide (DMSO) (p < 0.005, p < 0.05). The mean duration of temporary posttransplant hyperglycemia was 22.3 ± 3.2 (n = 10), 35.9 ± 2.3 (n = 14), and 33.7 ± 4.1 (n = 6) days for the respective groups. The diabetic mice that received 300 islets had their blood glucose levels decrease faster than those that received 150 islets (19.7 ± 1.6 vs. 35.9 ± 2.3 days, n = 14, p < 0.0001). There was no significant difference in the blood glucose reducing effect between the mice that received 150 islets with NDGA and the mice that received 300 islets [22.3 ± 3.2 (n = 10) vs. 19.7 ± 1.6 (n = 14) days, p > 0.05]. The insulin content of the graft from the mice treated with 150 islets and NDGA (3.02 ± 0.24 μg, n = 4) was higher than that from the mice that received 150 islets but no treatment (1.10 ± 0.26 μg, n = 15, p < 0.005) or that had been treated with DMSO (1.21 ± 0.30 μg, n = 4, p < 0.05). The insulin content of the pancreas remnant had no significant differences among the three groups. The net glucose-stimulated insulin secretion was 0.82 ± 0.14 vs. 0.20 ± 0.10 μIU/islet × 60 min (n = 8, p < 0.005) and 0.59 ± 0.08 vs. 0.04 ± 0.02 μIU/islet × 60 min (n = 8, p < 0.0001) for islets cultured without NDGA vs. with NDGA at 1 and 2 weeks, respectively. However, the insulin content of the cultured islets was similar between the two groups for up to 2 weeks of incubation (at 1 week: 0.71 ± 0.01 vs. 0.67 ± 0.04 ng/islet, n = 8, p > 0.05; at 2 weeks: 0.71 ± 0.02 vs. 0.80 ± 0.07 ng/islet, n = 8, p > 0.05). Serum leukotriene B4 (LTB4) concentrations before and between the fifth and seventh days after transplantation were determined. For diabetic mice that received 150 islets, serum LTB4 levels were 25,835 ± 3,335 and 27,631 ± 3,136 pg/ml (n = 4, p > 0.05). For diabetic mice that received 150 islets and NDGA, the corresponding figures were 22,401 ± 2,706 pg/ml and 27,530 ± 2,190 pg/ml (n = 8, p > 0.05). The graft histology revealed viable islet cells and networks of close vascular structures around the islets and did not reveal microscopic differences among the samples of all four groups. In conclusion, our data revealed that daily administration of NDGA for 4 weeks enhanced isoislet engraftment and preserved three times more mass of the islet beta cells in the isografts. This result indicates that NDGA reduces primary nonfunction of islet syngeneic grafts in diabetic mice.

The plasminogen-plasmin fibrinolytic system accelerates degradation of alginate-poly-l-lysine-alginate microcapsules in vitro

Abstract Inspired by the idea of “artificial cell” proposed by Dr T.M. Chang 30 years ago, Dr A.M. Suns group developed alginate-poly- l -lysine-alginate (A-P-A) microcapsules in 1980 to immunoisolate rat islet transplant which reversed hyperglycemia of streptozotocin-induced diabetic mice for a couple of weeks. Although reversal of hyperglycemia in various diabetic animal models and human subjects by A-P-A microencapsulated islets have been reported since 1980, the graft function was found to be variable in degree and duration. In 1994, Dr Soon-Shiong and colleagues presented a case of insulin-dependent diabetes (IDDM) who received a total of 960,000 A-P-A microencapsulated human islets intraperitoneally had a period, although for only 2 months, without the heed of exogenous insulin. We intend to identify and correct adverse factors that reduce graft survival of A-P-A microencapsulated islets to maintain long-term normoglycemia of treated diabetic patients. The pathogenesis of graft rejection of A-P-A microencapsulated xenoislet is not clear. It was noted that pericapsular cellular infiltration of A-P-A microencapsulated xenoislet grafts occurred as early as 28 days after intraperitoneal implantation. Many researchers found that the majority of pericapsular cellular components were macrophages and fibroblasts. Pericapsular macrophages may secrete proinflammatory cytokines, such as interleukin-1β which may suppress islet beta-cells and result in graft dysfunction. The inflammatory reaction induced by A-P-A microencapsulated islets may suffocate the islets by exhausting oxygen. Many unidentified factors are possibly involved in degradating A-P-A microcapsules in vivo. It is interesting to note that macrophages secreted urokinase plasminogen activator in response to inflammation. One of the principal players of the fibrinolytic system is plasminogen which binds to the exposed lysyl residues in fibrin. Polylysine was found to mimic the cofactor (enhancing) function of fibrin in tissue activator-induced and low molecular weight urokinase-induced plasminogen activation. Since A-P-A microcapsules were formed by polycondensation of poly- l -lysine and alginate, it contained terminal and interchain lysine residues on their surface. The exposed lysine residues of A-P-A microcapsule confer binding sites for plasminogen and its activators. If the membrane-bound tissue activator is activated for plasminogen activation, the catalytic domain of plasmin will cleave poly- l -lysine and disintegrate capsular membrane. Therefore, we hypothesize that the plasminogen-plasmin fibrinolytic system is involved in degradation of A-P-A microcapsules. To verify this hypothesis, we studied the lytic activity of the plasminogen-plasmin fibrinolytic system on the physical durability of A-P-A microcapsules in vitro.

Synergistic Effect of Hyperglycemia and p27(kip1) Suppression on Adult Mouse Islet Beta Cell Replication.

The complementary role of hyperglycemia and p27kip1 suppression on islet beta cell regeneration was investigated in a syngeneic mouse model. p27kip1 gene silencing was performed by infecting islets of C57BL/6 with shRNA lentiviral particles. At 54 hours after viral infection, p27kip1 protein content in cultured targeting islets was 22% of that in freshly isolated islets. Six days after transplantation to diabetic mice, targeting islet graft had considerably more cells with Ki67-staining nuclei than nontargeting islets. The mice in the targeting-islet group had a significantly shorter duration of temporary hyperglycaemia than mice in the non-targeting-islet group. The long-term ex vivo beneficial effect of p27kip1 silencing on graft function was also indicated by the significantly higher cumulative cure rate for diabetes in mice receiving 200 targeting islets than that in mice receiving 200 non-targeting islets. Our data suggest that hyperglycemia and persistent p27kip1 suppression have a synergistic effect on islet beta cell replication in adult mice.

Hyperglycemia In Vitro Up-Regulates Growth-Related Cell Cycle Proteins of Adult Mouse Pancreatic Islets

To study effects of glucose on growth-related proteins of adult islets, we cultured mice islets in medium containing either 5.5 mmol/L (LG) or 20 mmol/L (HG) glucose. Total islet proteins were processed for sodium dodecyl sulfate polyacrylamide gel and Western blotting using antibodies against beta-actin (housekeeping), p27kip1 (G1/G0 checkpoint), cyclin D1 (G1/S), cyclin B1 (G2/M), and FoxM1. At day 1, protein levels p27, B1, D1, and FoxM1 of islets on LG and HG were 0.48- and 0.63-fold; 7.09- and 11.58-fold; 1.25- and 1.38-fold; and 1.75- and 1.75-folds, the value of day 0, determinations respectively. At day 3, the proteins of p27, B1, D1, and FoxM1 of islets in LG and HG were 0.84- and 0.84-fold; 3.08- and 17.17-fold; 1.41- and 1.54-fold; and 0.83- and 1.17-fold of those on day 0, respectively. On day 7 the values were 1.19- and 1.09-fold; 3.15- and 14.81-fold; 0.86- and 1.44-fold; and 2.75- and 3.42-fold that of day 0, respectively. At day 1, the ratios of protein in islets after HG verse LG were 1.25, 2.38, 0.94, and 1.00 for p27, B1, D1, and FoxM1, respectively. At days 3 and 7, the protein ratios of HG/LG were 0.81 and 0.82, 5.47 and 2.64, 0.81 and 1.51, and 1.11 and 1.24 for p27, B1, D1, and FoxM1, respectively. In conclusion, adult mouse islets rapidly respond to cultivation by reducing p27 and increasing B1; HG attenuates p27 elevation but enhances B1 and D1 elevations, which favor islet entry into the cell cycle.

Prolonged Survival of Subcutaneous Allogeneic Islet Graft by Donor Chimerism without Immunosuppressive Treatment

The aim of this study was to investigate whether tolerance-induced protection of islets in the renal subcapsular space can also prevent subcutaneous allogeneic islets from being rejected. We used bone marrow stem cells from C57BL/6 (H2b) mice to construct donor chimerism in conditioned diabetic BALB/c (H2d) mice and investigated the effect of donor chimerism on engraftment and survival of subcutaneously transplanted allogeneic islets in streptozotocin-induced diabetic mice. We also studied the anti-inflammatory effect of mesenchymal stem cell on islet engraftment. Full but not low-grade or no donor chimerism was associated with successful engraftment of allogeneic islets and restoration of normoglycemia in the treated diabetic mice. The temporary hyperglycemia was 11 ± 1 versus 19 ± 5 days (p < 0.05) for the mice with full donor chimerism with transplanted islets in the renal subcapsular space versus the subcutaneous space, respectively. Cotransplantation of mesenchymal stem cell did not enhance alloislet engraftment. Full multilineage donor chimerism was associated with a higher transient expansion of CD11b+ and Gr-1+ myeloid progenitor cells and effector memory CD4 and CD8 T cells. In conclusion, full donor chimerism protected both renal subcapsular and subcutaneous allogeneic islets in this rodent transplantation model.

A removable tubing for implanting islet graft and studying immunosuppression

THE RENAL subcapsular space is the most common implantation site for experimental islet transplantation. However, the islet graft disperses underneath the renal capsule and is therefore not discerniable when the site is dissected for histologic examination. In this study, we examined the feasibility of using removable subcutaneous hybrid tubing for implantation of islet transplants in streptozotocin-induced diabetic mice.

In vitro evaluation of growth and anabolism for C3A/HepG2 hepatoma cells with logistic equation and linear regression expression.

SEVERAL bioartificial liver researchers adopt C3A/ HepG2 hepatoma cell line in their designs. Advantages of using this cell line in contrast to rat or porcine hepatocyte primary cells include stability, ease of proliferation, and possibility of long-term maintenance. Although some researchers have carried out comparative studies of C3A/HepG2 with other types of hepatocytes, there are no comprehensive surveys of the characteristics of C3A/ HepG2. Besides, to get a quantitative picture of threedimensional growth of hepatocyte, it is important to first realize the cell growth, toxin production and proteins synthesis kinetics in two-dimensional growth condition. In this study, we analyze the kinetics of concentration changes of ammonia (a key toxin of metabolism and a clinical index of liver detoxification function), albumin, and alpha-fetoprotein (a survival evidence of the hepatocyte) in the culture media, followed by finding out the relationship between cell density and those anabolites by using a linear type expression.