A stepwise approach to deriving functional β-cells from human embryonic or induced pluripotent stem cells

Introduction

The pancreas is a vital organ comprising both endocrine and exocrine roles. Among its various cellular constituents, the endocrine clusters known as islets of Langerhans have a pivotal role in controlling blood glucose concentrations. Beta (β) cells are essential pancreatic cells responsible for glucose homeostasis by matching fluctuating insulin demands and adjusting rates of insulin secretion [1]. Insulin is the most potent anabolic hormone in the human body, counteracting the hyperglycemic effects of glucagon and adrenaline. Following food intake, glucose levels increase in the bloodstream and induce a rise in the circulating insulin levels. The latter stimulates insulin receptors that are mostly abundant in the liver, muscle, and adipose tissue. This activation enhances anabolic glucose-consuming pathways like glycolysis, glycogen production, and lipid synthesis. Therefore, insulin release is essential for proper glucose uptake in various tissues.

Consequently, insulin release is acutely regulated as β cells adapt the rate of insulin secretion to metabolic needs. Glucose-stimulated insulin secretion (GSIS) occurs in a biphasic manner: it is composed of an immediate primary phase and a slower secondary phase. First, a sharp increase of circulating insulin is observed as the hormone is discharged from readily releasable granules. Glucose diffusion into the β-cell facilitated by GLUT2 induces a rapid rise in intracellular ATP levels (through glycolysis and the Krebs cycle). Following the increased cytoplasmic ATP/ADP ratio, the K-ATP channels close, causing membrane depolarization and intracellular calcium release. Ca²⁺ triggers decreased electrostatic repulsion between the secretory granules and the plasma membrane, activating exocytosis of the pre-existing insulin granules [2]. In addition to the triggering role of Ca²⁺, elevated cAMP levels are also required for a normal insulin secretory response to glucose [3]. The second phase of insulin release is slower: it requires continuous glucose stimulation, which induces transcriptional modifications leading to proinsulin peptide synthesis, granule build-up, and vesicle priming (Figure 1). In the Golgi complex of β cells, proinsulin is cleaved to form mature/active insulin. The cleavage product known as connecting peptide (C-peptide) is a short peptide of 31 amino acid residues that gets stored in secretory vesicles alongside insulin. Its short, 30-min half-life makes C-peptide a reliable marker for β cell function [4].

Figure 1:

Islet cell architecture and signaling pathways mediating glucose-stimulated insulin secretion (GSIS). (1) Glucose uptake into β cells occurs via diffusion through the GLUT2 transporter. It undergoes glycolysis and the tricarboxylic acid (TCA) cycle, resulting in ATP production. Increased ATP to ADP ratio leads to the closure of the K-ATP channels, membrane depolarization, and influx of Ca²⁺ through the voltage-gated calcium channels. Ca²⁺ triggers exocytosis of the insulin granules. (2) Prolonged stimulation by glucose induces transcription of insulin gene in the β cell nucleus. Insulin is processed and packaged into vesicles that become primed for immediate release – figure generated on BioRender by C.F.

Thus, functional β-cells should acutely respond to an elevation of circulating glucose and stop insulin release once blood glucose levels return to baseline. These responsive β cells constitute the functional β cell mass. Healthy β cells shall adapt their number, size, and differentiation state to meet metabolic demands [5].

Functional β cell mass is depleted in cases of chronic metabolic distress like diabetes. With a rapidly rising prevalence, diabetes has been extensively studied and classified. Type 1 diabetes (T1D) is caused by autoimmune destruction of β-cells, leading to absolute insulin deficiency [6]. Type 2 diabetes (T2D) is characterized by dysfunctional β cells and insulin signaling, which is often associated with obesity and lifestyle factors [7], [8], [9]. We also note rare monogenic cases of diabetes (MODY) and neonatal diabetes, etc. [10]. While oral medications and lifestyle changes are the most prescribed remedy for T2D, poor diabetes control remains as most patients require a combination of multiple drugs [11]. Likewise, T1D management is not optimal as it requires exogenous insulin delivery. Such a procedure is coupled with challenging management of insulin dosing and leads to poor overall glycemic control [12], 13]. Traditional treatments for diabetes, including insulin therapy and oral hypoglycemic agents, primarily manage symptoms but do not address the underlying β-cell dysfunction or loss [14]. Therefore, developing regenerative therapies that can restore β-cell function or replace lost β-cells represents a promising approach to achieving long-term diabetes management and potential cure. This enhanced glucose homeostasis can be achieved by islet cell transplantation [15]. Deriving pancreatic β cells in vitro is essential for successful islet transplantation, for it addresses the following challenges and avenues for growth:

1. Limited Donor Availability. Cadaveric donors are the primary source of islets for transplantation. However, donor pancreases are limited, and the demand far exceeds the supply. In vitro-derived β cells can address this shortage.

2. Immunosuppression. Transplanted islets face immune rejection unless patients receive immunosuppressive drugs. In vitro-derived β cells can be genetically modified to reduce immunogenicity, thus improving transplantation outcomes.

3. Drug Testing and Research. In vitro-derived β cells serve as valuable models for studying diabetes pathogenesis, drug testing, and understanding β cell biology. Researchers can explore new therapies and optimize transplantation protocols.

4. Personalized Medicine. Patient-specific β cells can be generated from induced pluripotent stem cells (iPSCs). These β cells can then be used for personalized transplantation, minimizing immune rejection [12].

5. Quality Control. In vitro-derived β cells allow rigorous quality control, ensuring that transplanted islets are functional, viable, and safe.

6. Functional Restoration. Successful islet transplantation restores insulin secretion and glucose regulation, as well as reduces the need for exogenous insulin injections.

In short, deriving pancreatic β cells in vitro is essential for advancing islet transplantation as a viable treatment option for diabetes. It addresses donor shortages, improves transplantation outcomes, and accelerates research and drug development.

The following review focuses on the potential of embryonic stem cells (ESCs) and iPSCs to differentiate into functional β-cells. ESCs are pluripotent cells derived from the inner cell mass of blastocysts. They can differentiate into various cell types of the three germ layers (endoderm, mesoderm, and ectoderm) [16]. Their pluripotency makes them a valuable tool in regenerative medicine and holds significant promise for generating β cells. However, several problems remain with the use of ESCs, such as moral sensitivity regarding the use of human embryos and the immune rejection of the cells by recipients after transplantation. Human induced pluripotent stem cells (hiPSCs), on the other hand, are generated by reprogramming somatic cells to a pluripotent state. This reprogramming is achieved through the introduction of specific transcription factors such as OCT4, SOX2, KLF4, and c-MYC. hiPSCs offer several advantages over hESCs, including the ability to generate patient-specific cells, which reduces the risk of immune rejection and bypasses the ethical issues associated with hESCs. Recent advancements in hiPSC technology have significantly improved the efficiency and safety of the reprogramming process, making hiPSCs a promising alternative for β-cell generation [17], [18], [19].

This review synthesizes the latest advancements and persistent challenges in deriving functional β-cells from hESCs and hiPSCs. It explores differentiation pathways, key transcription factors, and novel experimental techniques, while also discussing the functional maturation of these cells and their translational and clinical applications. By integrating recent research findings, this review aims to highlight both the potential and limitations of stem cell-derived β-cell therapies for diabetes. The insights provided will contribute to a deeper understanding of β-cell differentiation and identify future research directions necessary for optimizing these regenerative therapies.

Differentiation stages (from ESC to insulin-producing cell)

In vitro engineered β-like cells possess limited functionality compared to primary human β-cells, from which we note deficient GSIS. However, Rezania et al. revolutionized the field with their seven-stage protocol that efficiently converts human embryonic stem cells (hESCs) into insulin-producing cells (IPCs) [17]. In the following section, we investigate the master genes controlling the differentiation process, and we expand upon them throughout the review. From stage 1 (definitive endoderm) to stage 7 (mature β cell), the researchers used several vitamins, growth factors, and special media to allow for cellular differentiation (Figure 2). While the endoderm-derived marker FOXA2 is expressed as early as in stage 1, the differentiating cells engage in the pancreatic lineage around stage 3 following the expression of PDX1. This pancreatic homeodomain transcription factor (PDX1) is a key marker of pancreatic progenitor. It is also known to precede the expression of NKX6.1 (homeobox transcription factor). However, their co-expression is restricted to β cell lineage [20]. The endocrine marker (NGN3) gives rise to hormone-expressing cells. While transiently expressed in the developing pancreas, its downstream target NEUROD1 persists in β cells as of stage 4 [17]. Mature β cells are finally characterized by high MAFA expression (a basic leucine zipper transcription factor). MAFA is absent in developing β cells and has an essential role in maintaining the homeostasis of mature β cells [21]. Such mature cells are expected to be insulin⁺/glucagon⁻/somatostatin⁻.

Figure 2:

A 7-stage differentiation protocol from ES to β cell. Row 1: Differentiation stages, number and title. Row 2: Growth media used in the protocol. Row 3: Color-coded genes referencing their stages of expression. The most important β cell markers are PDX1 (pancreatic), NKX6.1 (pancreatic), NEUROD1 (endocrine), and MAFA (mature β cell). Adopted with permission from [17].

Characteristics of IPCs (from ESCs) [17]

Insulin secretory mechanism in S7 cells

Having INS expressions similar to those in human islets, S7 cells successfully secrete amounts of insulin comparable to those of human β cells (Figure 3A). Thus, S7 cells exhibit functional insulin secretory mechanisms. However, they still have higher proinsulin content and lower C-peptide amounts compared with human islets. This data suggests inefficient proinsulin processing in S7 cells.

Figure 3:

Stage 7 (S7) cells possess functional similarities to human β cells but are considered immature. (a) Insulin secretory mechanism in S7 and human islet cells. Hormone content relative to DNA content was measured; decreased C-peptide and increased proinsulin levels in S7 cells were noted. (b) Calcium signaling traces performed after 20 mmol/L glucose stimulation followed by 30 mmol/L KCl. Baseline glucose level was always 3 mmol/L. Percentages of responsive cells are indicated in grey. (c) S7 cells reverse diabetes in vivo. S7 cells were transplanted under the kidney capsule of SCID-beige mice with streptozotocin (STZ)-induced diabetes. Fasting blood glucose levels showed a reversal of diabetes 40 days post-transplantation and a fast return to hyperglycemia after the removal of the engrafted kidney (red arrow). Adopted with permission from [17].

Temporal progression of gene expression in β cell differentiation [22]

The aforementioned protocol has been thoroughly implemented across different laboratories. Therefore, the need to understand the specifics of each differentiation stage increased. Researchers investigated the temporal progression of gene expression throughout the protocol. Do all cells enter the different stages at the same time? Do they all follow the same orderly path? A single-cell qPCR study conducted by Peterson et al. aims to understand the linearity of differentiation by researching the following: do cells always gain the pancreatic β cell-specific marker NKX6.1 before the endocrine progenitor marker NEUROD1 [22] ? More broadly, they are investigating possible lineage bifurcations and the impact of heterogeneity in hESC differentiation.

Genetic regulations by mafa in mature β cells

After establishing that the path towards β cell differentiation is non-linear, researchers aimed to answer the questions: what causes the final lineage commitment to β cells and what inhibits the cells’ dedifferentiation back to early stages? As GSIS is key in mature pancreatic β cells, and as this function remains flawed in S7 cells from Rezania et al. researchers investigated the voltage-gated Ca²⁺ channels crucial for insulin release [23]. Expectedly, these channels’ expression is associated with the transcription of MAFA, the master regulator of gene expression of mature β cells [24].

Effects of Cavγ4 downregulation

Downregulation of Cavγ4 leads to β-cell dedifferentiation. New β-cell dedifferentiation marker aldehyde dehydrogenase 1a3 (Aldh1a3) [29] was found to be highly enriched in Cavγ4-silenced INS-1 cells. Aldh1a3 was initially recognized as a cancer precursor marker and later suggested as a β cell dedifferentiation marker in T2D [29]. Indeed, Cavγ4 deletion causes β cell dedifferentiation without affecting cell viability or proliferation. Thus, upon exposure of β cells to chronic oxidative stress, such as in glucose toxicity, MAFA levels become markedly reduced, inducing lower Cavγ4 levels and causing β cell dedifferentiation. The connection between glucotoxicity, differentiation, gene expression, and the functional consequences (Ca²⁺ signaling and insulin secretion) was elucidated. The evidence strongly suggests that Cavγ4 is a direct target of MAFA, which could be implicated in long-term regulation of β cell differentiation and functional status (Figure 4).

Figure 4:

Schematic of MAFA’s regulation of Cavγ4 and the downstream cascade on insulin secretion in healthy and dedifferentiated β cells. Glucotoxicity in dedifferentiated β cell leads to reduced MAFA signaling, downregulation of its direct target Cavγ4, diminished L-type Ca²⁺ channels and intracellular Ca²⁺ signaling and finally impairment of glucose-stimulated insulin secretion (GSIS). Adopted with permission from [23].

Endocrine cell clustering in β cell differentiation [30]

While in vitro cell culture conditions have been optimized to generate mature, functional β cells, the outcomes achieved thus far fall short of attaining perfection. Nair et al. investigated the importance of endocrine cell clustering during pancreatic islet organogenesis [30]. The researchers isolated and reaggregated immature β-like cells to form islets of enriched β-clusters (eBCs). In the following, we scrutinize the role of cellular reorganization involving endocrine clustering.

First, endocrine clustering was proven to be attained during human and rodent postnatal development, especially concerning β cell maturation. Indeed, in vivo studies have shown that EP first appear scattered throughout the pancreatic epithelium and only aggregate into islets once becoming mature [31]. Opposingly, in vitro hESC differentiation protocols result in mixed populations of PDX1+/NKX6.1+ combined with mature and immature endocrine cells subject to several lineage bifurcations (as mentioned in Section 4). Thus, the study conducted by Nair et al. proves the imminent role of endocrine cell clustering for hESC-derived β cell maturation.

To obtain C-peptide eBCs, researchers isolated INS⁻GFP⁺ immature β-like cells using fluorescence-activated cell sorting (FACS) and aggregated them into clusters. Islet-like cluster formation was further enhanced by extending the culture time for the reaggregated cells. To control for the extended culture, eBCs were compared to stage 7 cells from Rezania et al.’s protocol with prolonged cell cultivation. This negative control for endocrine clustering was referred to as non-enriched clusters (NECs).

Application of pluripotent stem cell-derived secretomes in β-cell differentiation

The use of secretomes derived from pluripotent stem cells (PSCs), which include a variety of bioactive molecules such as growth factors, cytokines, and extracellular vesicles, has emerged as a promising method to boost the efficiency and functionality of IPC differentiation.

Novel protocols using IPSCs

Recent advancements in stem cell research have significantly enhanced protocols for differentiating iPSCs into functional insulin-producing β-cells. These advancements are pivotal as they strive to emulate the complex signaling environment necessary for β-cell development.

Future avenues in β cell differentiation

As the field rapidly advances, the differentiation of β cells from PSCs remains complex. One of the long-term challenges in transplanting SC-derived β cells for therapeutic applications remains the selective differentiation into a β-cell only population. To this day, most protocols result in intermingled β-like, α-like, and δ-like cells [21]. Such unwanted mixtures have tumorigenic potentials, affecting the safety of cell therapies. Therefore, Nair et al.’s proposed eBCs [30] lead the way to a more secure alternative for islet cell therapy.

While researchers implement new protocols and study the genetic variations during β cell differentiation, it is important to note that the optimized route of cell transplantation is also extensively studied. Some argue that fully functional β cells will be sensible to the stresses associated with transplantation and recurrent autoimmunity in T1D [45]. However, immune isolation strategies against autoimmune destruction include macro- and micro-encapsulation devices. Yet, all these approaches also increase the distance between the grafted cells and the blood circulation, creating possibly damaging hypoxic environments for the graft [45]. As a cure for T1D, cell replacement therapy may still require conjunction with autoimmunity-suppressing therapies. As for T2D, a combination of antidiabetic drugs and stem cell-based therapy was revealed to be a promising treatment, at least in mice [46].

In recent decades, deriving pancreatic β cells from human stem cells has significantly progressed. Here are the major advancements not limited to our current discussion:

1. PSCs, including ESCs and iPSCs, serve as an ideal alternative source for generating pancreatic β cells. These cells can be differentiated into functional β cells in vitro.

2. Recent advances enable the differentiation of stem cell-derived islets (SC-islets) with high endocrine purity. These SC-islets include hormone-secreting β, α, and δ cells that undergo GSIS. In animal studies, these SC-islets have successfully treated or prevented diabetes [47].

3. Researchers can now study patient-derived SC-islets with diabetes. This approach allows the exploration of gene-editing strategies to reverse diabetes-causing gene variants.

4. Techniques like CRISPR-Cas9 have been employed to modify stem cells and enhance their functionality as β cells. These gene-edited cells hold promise for transplantation and diabetes treatment [48].

5. Researchers have developed microcapsules that protect stem cell-derived β cells from immune system attack. These encapsulated cells show improved glucose metabolism and reduce inflammatory fibrotic tissue buildup [49], 50].

6. Efforts are ongoing to improve the maturation and functionality of stem cell-derived β cells. Researchers aim to achieve glucose responsiveness and insulin secretion comparable to primary human islets.

Our current knowledge of β cell differentiation from PSCs is rapidly evolving, hESCs and iPSCs have proven to be immensely valuable. A fixed protocol was established selecting for the main master regulatory genes of β cell differentiation. The protocol initiates the embryonic stem cell into pancreatic lineage commitment, fosters the acquisition of endocrine precursors and ultimately allows the expression of mature β cell markers. Yet, this protocol requires refinement for better yield and functional qualities of the generated β cells.

Furthermore, the β cell differentiation path was unveiled as non-linear and exhibited several bifurcations. The pancreatic β cell progenitor NKX6.1 and the endocrine precursor can be gained independently from one or the other, without any sequestered order required as previously believed. Yet, their co-expression persists to be essential for β cell differentiation. However, even at its last stage, β cell differentiation is susceptible to variations. The key mature β cell marker MAFA was proven imperative to maintain the differentiation state of the β cells. MAFA has been identified to act on Cavγ4 gene expression, preventing the cell’s dedifferentiation.

Though studying individual genetic markers of β cells is extremely helpful to understanding β cell morphogenesis, it was confirmed that recapitulating pancreatic islet clustering led to improved physiological responses of differentiating β cells (Figure 5). This approach mimics in vivo clustering of mature β cells and allows for enhanced glucose-stimulated insulin secretion.

Figure 5:

Conclusive summary of β cell differentiation pathway. Middle row: Major steps of β cell differentiation from human embryonic stem cells (hESCs). Main transcription factors included based on Rezania et al.’s study [17]. Upper path: Possible bifurcation in the acquisition of NKX6.1 and NEUROG3 found by Peterson et al. [22]. Brown dash arrows: Possible dedifferentiation of end-stage β cells in case of MAFA downregulation [23]. Lower path: enriched β cell clusters improved β cell function [30]. Images generated on BioRender by C.F.

Finally, several protocols were designed to improve the yield and functionality of β-cells derived from iPSCs. However, achieving GSIS comparable to primary human β-cells remains a critical challenge. These findings collectively underscore the intricacies of β cell differentiation, pointing toward avenues for further advancement in regenerative medicine for diabetes treatment.