Transplantation of autologous insulin-producing cells (IPCs) may be an ideal diabetic treatment for overcoming the limitations of current pancreas and islet transplantation, including the shortage of donors, immune reactions of the allograft, and low graft survival [29]. Various techniques and an alternative cell source have been actively reported for the development of new IPCs using differentiation from stem cells [30, 31], the proliferation of existing adult β-cells [32], and reprogramming of adult somatic cells [33, 34]. A recent report showed that cells isolated from the liver have high differentiation potential for the pancreatic lineage because the liver has the same origin as the pancreas. In this study, we isolated adult liver cells using a previously reported method [9, 12]. Adherent cells isolated from the liver acquired mesenchymal-like characteristics and considerable cellular plasticity, as has previously been reported [9]. These cells adhered to, and were cultured on, tissue culture plates, and their growth was maintained for more than 10 passages. Additionally, these human liver cells expressed mesenchymal stem-cell markers such as CD73, CD90, and CD105, but did not display hematopoietic markers (CD31 and CD45).
Numerous studies have been conducted in which differentiation of insulin-producing cells was effected by inducing the expression of transcription factors such as PDX1, NGN3, MAFA, and PAX4, which play important roles in pancreatic development [35, 36]. Recently, Ferber et al. showed that mature β-cell-like characteristics were induced when three pancreatic transcription factors, PDX1, PAX4, and MAFA, were sequentially supplemented in a direct hierarchical manner [16]. It is important that transcription factors are turned on and off at appropriate times for determining cell fate. PDX1 alone has little activity or it induces other pancreatic cells; however, it becomes a potent factor when it interacts with NEUROD1 [37, 38]. Another important transcription factor for β-cell-specific differentiation is MAFA [39-42]. Thus, we used PDX1, NEUROD1, and MAFA to induce IPC from liver cells. PDX1 at an MOI 500 was used as a key factor because PDX1 plays a crucial role in pancreatic organogenesis and beta cell function. The MOI of NEUROD1 and MAFA was optimized based on insulin promoter activity, insulin gene expression, and minimal adverse effects on infected cell viability. This combination increased the intensity of insulin promoter activity and mRNA level of insulin genes compared to that in cells treated with either PDX1 or NEUROD1, MAFA alone. In this study, PDX1 and NEUROD1 were added at the same time, and MAFA was sequentially induced 2 days later to maximize transdifferentiated cell maturation, as indicated by increased insulin gene expression. Insulin protein was confirmed to be expressed by immunohistochemistry, TEM image, and ELISA. Unlike fully mature beta cells, granule formation was not complete, but some aggregates of insulin protein were identified in TEM image. However, glucagon did not form an aggregate, and was distributed in the cytosol to a small degree.
It is known that cell-cell interaction of islets is important for insulin production and secretion [43, 44]. In our study, when cell sheets were formed, insulin gene expression and protein production were significantly increased. Also, several transcription factors were significantly increased in the cell sheet group. Interestingly, the expression of glucagon was dramatically reduced in the cell sheet group. In previous reports, Nkx6.1 reduces the expression of the glucagon gene, which explains why these transcription factors in the cell sheet group decreased glucagon expression and thus increased the purity of insulin-producing cell differentiation [45]. Insulin and C-peptide production and insulin secretion also significantly increased in the sheet group. Although the cell sheet’s ability to induce differentiation in IPC was confirmed in this study, further studies are required to determine the mechanisms underlying cell sheet formation and IPC differentiation and maturation.
A cell sheet-based tissue engineering approach has attracted attention in cell transplantation therapy because of its potential to deliver a large number of cells to the desired organ without surgical procedures [18-20]. Cell sheets maintain adhesive protein layers that facilitate attachment to other tissues. The cell sheet technique also makes the delivery of target molecule safe. In the successful clinical trial of intraportal islet transplantation by Shapiro et al., islet transplantation was exclusively conducted via the portal vein [22, 46-49]. When the islet is administered directly through the portal vein, many cells engraft in the liver through smaller venues, and the liver can supply sufficient blood in a near-physiological insulin delivery environment. However, several concerns remain regarding intraportal islet infusion, including procedure-related complications, bleeding, hepatic hypertension, thrombosis, and instant blood-mediated inflammatory reaction [21]. Various alternative sites for successful islet engraftment have been proposed for the islet or IPC, including the kidney capsule, omentum, skin, and cornea. Subcutaneous engraftment has also been proposed. Recently, Alejandro et al. reported that a patient who underwent islet transplantation on the omentum showed stable glycemic control without exogenous insulin and episodes of hypoglycemia at 12 months [50]. Although some locations may be advantageous in experimental models, their feasibility and translation into clinical settings is limited [51]. An ideal transplantation site should include a sufficient blood supply without any surgical risk. The liver is ideal for cell transplantation because of the high blood supply, near-physiologic insulin delivery environment, and long clinical application history. If the IPCs construct can be directly transplanted onto the liver surface, rather than via portal vein injection, this may be an ideal transplantation technique for ensuring safety and increasing survival and near-physiological insulin delivery. In this study, we confirmed that the IPC sheet improved the delivery efficacy of IPCs in the liver and decreased the risk of PV injection. Additionally, IPC sheet formation may enhance the differentiation function of IPCs in vitro by maintaining cell-cell interactions. To identify the potential of clinical applications for IPC sheet transplantation, we assessed not only glycemic control efficiency but also operative feasibility, bio-distribution, and liver toxicity, compared to the IPCs intraportal injection technique.
Mice transplanted with the IPC sheet showed a better control of blood glucose levels and weight gain than mice injected with IPCs. One and 2 weeks post-transplantation, immunohistochemical analyses showed that IPC sheet-transplanted livers stained positively for PDX1 and insulin. When it came to control of hyperglycemia, long term control of blood glucose levels was unfortunately not accomplished after IPC sheet transplantation into immunocompromised diabetic nude mice, even though transient decreased blood glucose levels and increased body weight were observed, with better outcomes than PV injection.
After injecting the IPCs suspension via the portal vein, and transplantation of the IPC sheet onto the liver of diabetic nude mice, cell biodistribution was assessed by in vivo/ex vivo fluorescence image tagging Qdot. The IPCs were detected in the liver in the cell sheet transplanted region of mice at days 2 and 7, while the intraportal injection group showed only a weak signal at day 2. This was confirmed by ex vivo fluorescence image and histology.
Increasing transplantation efficiency as well as reducing toxicity are important factors for clinical application. When we injected 106 differentiated IPCs through the portal vein, 80% of animals survived. In the preliminary study, mice transplanted with 1.5 × 106 and 2 × 106 IPCs died at rates of 40% and 90% during surgery, respectively (data not shown). However, no mice transplanted with 1, 3, or 5 layers of the IPC sheet died as a consequence of the operation. Mice injected with the IPCs through the portal vein showed liver toxicity and histologic damage to many parts of the liver margin. However, the IPC sheet showed no significant liver toxicity in our study. These results showed that directly transplanting the IPC sheet onto the liver surface is expected to significantly lower the risks related to current intra-portal vein injection, which include thrombosis, bleeding, and liver toxicity.
In conclusion, cell sheet formation enhanced the differentiation function and maturation of IPC in vitro. Additionally, parameters for clinical application such as distribution, therapeutic efficacy, and toxicity were favorable. The cell sheet technique may be used with IPCs derived from various cell sources in clinical applications.