From Beta Cell Replacement to Beta Cell Regeneration

Abstract

Diabetes is affecting more than 25.8 million people in the United States, causing huge burden on the health care system and economy. Insulin injection, which is the predominant treatment for diabetes, is incapable of replenishing the lost insulin-producing beta cell in patients. Restoring beta cell mass through replacement therapy such as islet transplantation or beta cell regeneration through in vitro and in vivo strategies has attracted particular attentions in the field due to its potential to cure diabetes. In the aspect of islet transplantation, gene therapy, stem cell therapy, and more biocompatible immunosuppressive drugs have been tested in various preclinical animal models to improve the longevity and function of human islets against the posttransplantation challenges. In the islet regeneration aspect, insulin-producing cells have been generated through in vitro transdifferentiation of stem cells and other types of cells and demonstrated to be capable of glycemic control. Moreover, several biomarkers including cell-surface receptors, soluble factors, and transcriptional factors have been identified or rediscovered in mediating the process of beta cell proliferation in rodents. This review summarizes the current progress and hurdles in the preclinical efforts in resurrecting beta cells. It may provide some useful insights into the future drug discovery for antidiabetic purposes.

Diabetes, type 1 and type 2 altogether, is affecting 25.8 million people in the United States and causing huge burden on health care and economy.¹ Type 1 diabetes (T1D) is caused by the autoimmune destruction of insulin-producing beta cells and is usually treated by multiple daily insulin injections. Type 2 diabetes (T2D) is caused genetics and lifestyle factors. Patients with T2D initially do not need insulin. But as beta cell function declines over time, many T2D patients eventually take insulin. However, insulin injection is still seen by many health care providers as a last resort to treat their diabetic patients,² probably due to the fact that insulin cannot replenish the lost islets. The need of improvement in diabetes care calls for more efficacious strategies to replenish the insulin-producing cells. The Edmonton protocol of human islet transplantation, which is the most successful islet replacement therapy so far, has helped more than 1000 T1D patients since its debut in 1999.³ However, islet transplantation has made limited progress recently due to the shortage of islet donors and the poor immunosuppression in clinics. Xenograft islet transplantation, which is an alternative strategy exploring new sources of islets, caused safety concerns due to the intense xenospecific rejection and the risk of xenopathogens.⁴ Recently, more efforts have been put onto the regeneration of functional beta cells from both in vitro and in vivo perspectives and several remarkable discoveries in preclinical studies have been reported. Hereby, we reviewed the recent advances in β cell replacement and regeneration in combating diabetes. Hopeful this review may provide some useful insights in the future drug discovery for antidiabetic purposes.

Islet Transplantation

Islet transplantation is the most successful islet replacement therapy so far, achieving far better glycemic control than daily insulin injections. The first successful trial of human islet transplantation was reported in 1990, restoring normoglycemia in 5 out of 9 diabetic patients for more than 100 days.⁵ However, most of the early trials of human islet transplantation failed to sustain normoglycemia in the islet recipients for more than 1 year. In 2000, Shapiro and colleagues reported a steroid-free protocol of human islet transplantation, which was latterly referred as the Edmonton protocol, and remarkably achieved insulin-independence in 7 patients for a median duration of 11.9 months.⁶ The Edmonton protocol greatly improved the islet transplantation and was latterly adapted as the golden standard by islet transplant centers around the world. Up to 2012, more than 1000 patients received pancreatic islet transplantation. In 6 selected transplantation centers, more than 50% of patients remained insulin-independent for more than 5 years following islet transplantation.³

Though meeting great success, human islet transplantation has been constantly criticized in the past 2 decades for the use of multiple pancreases in the surgery, despite the current shortage of pancreas donors. The situation can be changed if a better strategy is developed to prevent the graft loss during and after the islet transplantation. Islet loss is usually caused by 2 reasons, the immune rejection from the islet-recipient and the primary nonfunction (PNF) of the islet grafts.^7,8 The immune rejection is the main cause of the islet loss and is characterized by the immune recognition and a subsequent destruction of islet allografts by the islet recipient. The PNF summarizes all the nonimmune reasons including the cytotoxic drugs, the hypoxia, the inflammatory cytokines, the poor revascularization, and so on, causing the loss of function of the islets. Immunosuppressive drugs or a combination of these drugs have been used clinically to prevent the rejection of islet grafts. However, this strategy has 2 major weaknesses: (1) the immunosuppressive drugs do not discriminate the islet-reactive T cells from the host immunity and consequently compromise the whole-body immunity and the quality of life of islet recipients, and (2) most immunosuppressive drugs are detrimental to the function of transplanted islets and cause the PNF of the transplanted islets. Neutralizing antibodies against CD3, CD20, and thymocyte globulin has been developed and tested in clinical trials. Other antibodies against CD40, CD48, interleukin-1β (IL-1β), lymphocyte function-associated antigen 1 (LFA1) were explored preclinical studies.^9-11 These highly specific antibodies do not cause the PNF of the transplanted islets but still fail to specifically suppress the islet-reactive immune rejection while leaving the whole-body immunity alone.

A long-standing obstacle of studying human islet transplantation is the lack of proper in vitro and in vivo models closely fitting the human immune rejection. Luckily, the humanized mouse model developed by Shultz and colleagues in the Jackson Laboratory could potentially resolve this issue.¹² The first generation of humanized mouse involved the infusion of mature human immunocytes in the immunodeficient mouse but suffered from high risk of graft versus host diseases (GVHD).¹³ The second generation of humanized mouse involved the infusion and engraftment of immune progenitor cells such as hematopoietic stem cells.^14,15 These cells subsequently developed into mature immunocytes in mouse without causing GVHD.¹² The humanized mouse model provided a convenient tool to explore a potential solution in preventing immune rejection of human islet transplantation and increased the translatability of current preclinical studies. New information was soon gathered and new strategies were developed to address the islet-reactive immune rejections. For example, mesenchymal stem cells (MSCs) isolated from various sources have been extensively studied in T1D animals before with a focus on the prevention of the onset of T1D.^16,17 New studies demonstrated that in humanized mice reconstituted with mature human peripheral blood mononuclear cells (PBMCs), human bone marrow derived MSCs cotransplanted with human islets prevented the T-cell infiltration and activation in the transplantation site and protected islet grafts by secreting immunosuppressive factors and recruiting regulatory T cells (Tregs), suggesting a potential application in future clinics.^18,19 For another example, Tregs were known to cause immune tolerance in organ transplantations. Recently studies demonstrated that human Tregs could be expanded in vitro and infused into humanized mice receiving human islet transplantation by suppressing the proliferation and differentiation of alloreactive T cells.²⁰ Also, through extensive screening of drug conjugations and drug derivatives, new immunosuppressive compounds with less detrimental impacts to human islets were tested in a humanized model.²¹ Last but not the least, gene therapy with a focus on increasing the resistance of human islets to the posttransplantation challenges is still a plausible strategy for the target evaluation in the research of human islet transplantation in preclinical studies.^22-25

Beta Cell Regeneration

Human beta cells do not increase throughout lifetime unless during a few exceptions including embryonic development, pregnancy, T2D, and so on. All these scenarios have been under close examination to identify a potential beta cell-specific mitogenic factor. Many factors, including glucagon like peptide-1 (GLP-1), gastric inhibitory peptide (GIP), insulin-like growth factor 1 (IGF1), epidermal growth factor (EGF), hepatocyte growth factor (HGF), serotonin, prolactin, placental lactogen, and so on, have been screened but limited success has met so far.²⁶ Meanwhile, a hot debate about the sources of beta cell proliferation, whether they are from the replication/expansion of preexisting beta cells or regeneration from progenitor cells or other types of cells (differentiation or transdifferentiation), is ongoing in academia. Knowing the sources of beta cell proliferation may be of critical importance in finding the right strategy to replenish beta cell loss to treat diabetes.

With the aid of the gene delivery vectors and a moderately controlled culturing condition, researchers successfully induced insulin-producing beta cells from various cell types including embryonic stem cells, MSCs, induced pluripotent stem cells and pancreatic stem cells,^27-29 even though some of these cells were not favorably seen as the sources of beta cells as suggested by the up-to-date lineage tracing studies. Several groups independently demonstrated that these in vitro generated beta cells were capable of responding to the glucose stimulation and restoring normoglycemia in diabetic animals.³⁰ Several factors, including pancreatic and duodenal homeobox 1 (Pdx-1), neurogenin 3 (Ngn3), paired box gene 4 (Pax4), and aristaless related homeobox (Arx), were pivotal in the in vitro beta cell regeneration, and not surprisingly most of these factors overlapped with the signals during the early development of pancreas.³¹ However, the results from in vitro regeneration should be explained with caution because most of these findings were hardly reproducible in vivo, suggesting different sources of beta cell regeneration. Lineage tracing studies also showed no convincing evidence that beta cells could regenerate from stem cells rather than self-expansion in adult animals.^32,33 Meanwhile, most of these studies utilized viral vectors to express the beta cell regeneration/trophic factors,^34,35 which is hard to translate into human trials. The risks of contamination, immunogenicity, and tumorigenicity of using genetically modified cells remain unsolved to advance this technique into clinical applications.

The in vivo attempts to promote beta cell regeneration focused on identifying a potential circulating factor. Incretins especially GLP-1 secreted by intestine were frequently linked to the replication and regeneration of beta cells. GLP-1 has a half-life of less than 2 minutes, and its mimetics such as exenatide and liraglutide have been marketed for T2D for years. In diabetic rats, a 2-day infusion of GLP-1 increased the cell proliferation of pancreatic islets and the level of serum insulin, suggesting a functional beta cell proliferation.³⁶ Increasing intracellular insulin was also observed in the human islets after a 5-day coincubation with GLP-1, suggesting potential human translatability.³⁷ The GLP-1 mimetics, exenatide, and liraglutide demonstrated similar effect in increasing beta cell mass in various animal studies.^38,39

Beside intestine, liver may be another key organ in the beta cell regeneration. Previous studies suggested that the liver-specific knockout of insulin receptor promoted compensatory beta cell replication in rodents.⁴⁰ A follow-up study revealed the central role of the hepatic activation of extracellular regulated kinase (ERK) in the beta cell proliferation of the T1D mice but failed to identify the circulating factor.⁴¹ In 2013, on a mouse model of insulin resistance and dramatic beta cell increase, Melton and colleagues identified the first circulating factor causing beta cell proliferation in mouse named it betatrophin.⁴² Liver specific expression of betatrophin led to a functional beta cell proliferation in normal and diabetic mice. However, the betatrophin receptor, presumably in pancreas, is yet to be discovered. A recent study also found that overexpression of mouse betatrophin failed to increase the beta cell mass of transplanted human islet.⁴³ Whether such discrepancy is due to the differences between human betatrophin receptor and mouse betatrophin receptor or that betatrophin is simply just another rodent-only discovery awaits further exploration.

Though beta cell proliferation has not been thoroughly understood, alpha cell proliferation has been reported in many studies with a well-characterized mechanism. Because alpha cell and beta cell share a common ancestor during the embryonic development, several groups have been working on the alpha-to-beta conversion in vivo.⁴⁴ Glucagon receptor (GCGR) antagonist and antibodies were known to substantially increase the alpha cell mass in the pancreas of mice and monkeys,^45,46 providing a convenient platform for the exploration of in vivo alpha-to-beta conversion. Among these studies, common features included an improvement in glycemic control, increased pancreatic α cell mass, and increased serum levels of GLP-1 and fibroblast growth factor 21 (FGF21), the latter of which was known to reduce blood glucose in an insulin-independent way.⁴⁷ Notably, similar results were observed in GCGR antisense oligonucleotides (ASO) treated Zucker diabetic fatty (ZDF) rats.⁴⁸ The latter study demonstrated that the GCGR ASO reverse the diabetes phenotype by the dual action of decreasing blood glucose and improving pancreatic β cell function. The α-cell hyperplasia may cause certain safety concerns but is transient after withdrawing the GCGR ASO.⁴⁸ Moreover, ISIS Pharmaceuticals recently reported the positive data from a phase 2 study of GCGR ASO based therapy as shown by the remarkable reductions in hemoglobin A1c (HbA1c) and the improvement in pancreatic function in T2D patients in, suggesting the potential of anti-GCGR therapy in treating diabetes.⁴⁹ Meanwhile in the preclinical research area, Collombat and colleagues did excellent work in converting alpha cells to functional beta cells in mouse pancreas. They demonstrated the plasticity of pancreatic duct cells and showed that pancreatic duct cells might be a source of both alpha and beta cells in mice by reexpressing Ngn3, adopting a glucagon-positive phenotype and going through a Pax4-mediated alpha-to-beta cell conversion.^44,50,51 A brief summary of the potential targets and the current progress in beta cell regeneration is listed in Table 1.

Table 1.

Potential Targets and Current Progress in Beta Cell Regeneration.

Targets	Properties/therapeutics	Progress
Betatrophin	Soluble factor/replacement therapy	Betatrophin increased beta cell mass in mouse studies.42 Failed to increase beta cell mass in human islets after renal transplantation into mouse.43
Insulin receptor	Receptor	Liver specific knockout of insulin receptor increased beta cell mass in a T1D mouse model through ERK pathway.40,41 Direct intervention may be risky. A circulating factor needs to be identified.
GCGR	Receptor/antagonist, antibody, ASO	GCGR antagonists and antibodies increased GLP-1 and improved beta cell function in mouse, rat, and monkey.45,46,48 Positive data of GCGR ASO in a phase 2 human trial.49 The risk of α cell hyperplasia might cause safety concerns. A circulating factor needs to be identified.
Pdx-1, Ngn3, Pax4, Arx, etc	Transcription factors	May be involved in the beta cell regeneration from stem cells or other types of cells.44,51,53 The druggability is relatively weak for transcription factors compared with other targets.

Conclusions

The cure of diabetes relies on the replacement of insulin for glycemic control, the relief of autoimmunity, and the replenishment of functional beta cells.⁵² Beta cell replacement therapy by human islet transplantation is the only way of replenishing beta cells so far, though facing extensive challenges of the limited donors and the intense immune rejection. Beta cell regeneration, holding great promise in future, is under preclinical evaluation. Specifically, the in vitro generated insulin-producing cells need to resolve the safety concerns of contamination, immunogenicity, and tumorigenicity to be translated into a clinical advance. The recent advance in regenerative medicine and the development of a standard operational procedure for cell therapy especially the T-cell-based immune therapy may provide some useful inspirations for this aspect. For the in vivo efforts to regenerate beta cells, the major hurdle is the lack of a druggable target directly increasing beta cell mass, through either self-replication or regeneration, in a convincing and safe way. Betatrophin needs more evidence and probably a thorough validation in a nonhuman primate model to demonstrate it is not a rodent-only discovery. Meanwhile, the identification of the betatrophin receptor and its human homolog may substantially benefit the drug discovery process regarding this interesting target. Pancreatic duct-alpha-beta conversion is another interesting aspect to work on and it is likely that the pancreatic duct-alpha-beta conversion is mediated by other factors rather than betatrophin because betatrophin did not affect the alpha cell mass in mouse pancreas.^42,43 However, the druggability of this pathway is still weak due to the lack of a circulating factor. Meanwhile, antiglucagon strategy caused a well-characterized alpha cell hyperplasia but did not affect beta cell mass.⁴⁵ Whether this pathway can be combined with the pancreatic duct-alpha-beta conversion to regenerate beta cells and whether the alpha cell hyperplasia can be managed in harmless way, both require further exploration. Moreover, researchers in the field may recognize that our final goal is to regenerate human beta cells and our research does not end with mouse islet regeneration. Currently, human islets are readily available from the Integrated Islet Distribution Program overseen by the Juvenile Diabetes Research Foundation. A standard protocol for the in vitro culture and the functional characterization of the human islet is available. The humanized animal models for the human islet transplantation and the studies using nonhuman primates to test the efficacy of antidiabetic drugs have been reported and optimized.^18,45 All of this progress may help us to find a better standard to study beta cell regeneration in academia and industry, which is the actual demonstration of increasing human β-cell mass using the in vitro or in vivo models. Such improvement may substantially benefit the early-stage drug discovery for the beta cell regeneration.