Encapsulation and Immune Protection for Type 1 Diabetes Cell Therapy

Abstract

Type 1 Diabetes (T1D) involves the autoimmune destruction of insulin-producing β-cells in the pancreas. Exogenous insulin injections are the current therapy but are user-dependent and cannot fully recapitulate physiological insulin secretion dynamics. Since the emergence of allogeneic cell therapy for T1D, the Edmonton Protocol has been the most promising immunosuppression protocol for cadaveric islet transplantation, but the lack of donor islets, poor cell engraftment, and required chronic immunosuppression have limited its application as a therapy for T1D. Encapsulation in biomaterials on the nano-, micro-, and macro-scale offers the potential to integrate islets with the host and protect them from immune responses. This method can be applied to different cell types, including cadaveric, porcine, and stem cell-derived islets, mitigating the issue of a lack of donor cells. This review covers progress in the efforts to integrate insulin-producing cells from multiple sources to T1D patients as a form of cell therapy.

Graphical Abstract

1. Introduction

Type 1 Diabetes (T1D) is an autoimmune disease characterized by the destruction of insulin-producing β-cells in the islets of Langerhans located in the pancreas [1]. The onset of T1D begins during the thymic development of T cells, in which there is faulty negative selection by the medullary thymic epithelial cells (mTEC) and dendritic cells (DC), resulting in T cells that are autoreactive [2-8]. While the onset of islet inflammation is still not thoroughly defined, it is hypothesized that the gut microbiota and pancreas undergo compositional changes, resulting in β-cell apoptosis; this leads to the release of antigens, which are taken up by antigen-presenting cells (APCs) in the draining pancreatic lymph nodes (pLNs), resulting in the formation of T effector (Teff) cells that specifically target β-cells [9-15]. These Teff cells then migrate to the pancreas and begin the destruction of β-cells [13-15]. This creates a proinflammatory state resulting in neoautoantigen formation, creating more Teff cells that continue to target β-cells. Additionally, β-cells secrete CXCL1 and CXCL2, resulting in the recruitment of neutrophils and increasing the inflammatory response, and the original faulty T cell formation causes impaired T cell receptor (TCR) signaling, which in turn downregulates FoxP3 signaling and regulatory T cell (Treg) presence [1]. As a result, β-cells are continuously destroyed, and T1D clinical symptoms appear when 80-95% of β-cells have been killed [16].

Depletion of β-cells results in the reduction and eventual elimination of insulin production in the body, predominantly affecting adipose and muscle tissue. In a physiological state, glucose enters the β-cells and insulin gets released to the bloodstream. This insulin then binds to muscle and adipose cells, thus triggering glucose uptake through the insulin-dependent glucose transporter GLUT4 and the production of ATP [17]. Without insulin, glucose cannot be processed by muscle cells and adipocytes, resulting in proteolysis and lipolysis, respectively, to ensure that these tissues are maintained [18]. Additionally, glycogen stored in muscle and adipose tissue will be broken down through glycogenolysis, resulting in an excess amount of glucose that exceeds the renal threshold for glucose absorption, and excretion of glucose through the urine, otherwise known as glucosuria [19]. As a result of extremely high glucose in the bloodstream, patients will exhibit symptoms of polydipsia, polyuria, polyphagia, increased weight loss, and ketoacidosis [20]. Over time, the excess glucose in the bloodstream will cause high stress on endothelial cells, resulting in reactive oxygen species (ROS) that damage the cardiovascular system, leading to a wide variety of pathologies such as atherosclerosis, retinopathy, neuropathy, and nephropathy [21].

The current treatment for T1D is exogenous insulin injections, including insulin pumps [22]. Whereas these injections can be effective in controlling blood glucose levels, there are several issues with this management method. The efficacy of insulin injections is user-dependent, and T1D patients themselves must remain accountable to take multiple insulin shots every day and structure it in a way that works with their schedule [23]. Additionally, insulin injections do not fully recapitulate glucose dynamics, and there is a major risk for potentially deadly hypoglycemic episodes when blood glucose levels go to low [24]. The most promising solution to mitigate these drawbacks is a one-time transplant of insulin-producing cells that can respond to glucose stimuli on command. In this review, we will discuss the current methods of cell therapy, and particularly biomaterial-based encapsulation methods to improve the integration of multiple sources of insulin-producing cells to the host’s body.

2. Current Standard for Islet Transplantation

The first islet isolations and transplantations originated in the 1960s-1970s with rat islets, and shortly after in 1988 came a report for an automated method to isolate human islets [25, 26]. The landmark clinical islet transplantation study was in 2000, when James Shapiro and colleagues at the University of Alberta in Edmonton, Alberta, Canada reported that all patients in the trial, using the Edmonton Protocol, became insulin independent 1 year after transplant [27]. In the Edmonton Protocol, cadaveric islets are infused into the hepatic portal vein in under a heavy immunosuppression regimen of sirolimus (once a day), tacrolimus (twice a day), and daclizumab (five inductions across 8 weeks) [28], and since then over 1000 T1D patients have been transplanted with islets using this protocol [29]. 50% of patients remain insulin-independent for 1 year post-transplant and 25% remain insulin-independent for five years, and the majority of advancements in this protocol since its launch have been in improving the safety and localizing the immunosuppressive drug regimens [30]. Multiple follow-up studies since the first successful reporting have indicated success in the transplants long-term [31-33], and the success of this protocol led to a Phase 3 trial in the USA [34] and an international clinical trial [28]. This method has become monumental for cell therapy and the potential end of exogenous insulin injections, but it took many years for islet transplantation to be fully approved in the USA, while in multiple countries across Europe and Asia, as well as Canada and Australia, have provided islet transplantation as an approved treatment [35]. In June 2023, the first US FDA-approved cell therapy for T1D, Lantidra—developed by CellTrans— was released, which infused islets through the hepatic portal vein and reported improved islet survival and function with a modified Edmonton Protocol, particularly in patients who suffer from severe hypoglycemic episodes [36, 37].

Although highly promising, there are key limitations with the standards for islet cell transplantation. A major issue is that when infused into the intraportal vein, there is an instant blood-mediated immune response (IBMIR), in which platelets and complement factors will instantly activate, causing aggregation and coagulation, restricting access to necessary nutrients, and activating the immune system, leading to acute islet death, progressive islet loss, and an inability to reach insulin independence [38]. Due to the loss of islets during and after transplantation, an excess of islets—roughly enough from 2-4 donors per patient—are needed, creating a situation where there a not enough donors for the amounts of islets needed for transplantation [39]. Finally, the infusion of allogeneic islets will require patients to remain on chronic immunosuppression, which negatively impacts graft survival and the patient (e.g., nephrotoxicity) and makes the patient vulnerable to other ailments [40]. To combat these issues, methods to improve the engraftment of islets need to be employed. Particularly, other sources of insulin-producing cells and other transplant sites that pose less of a risk for IBMIR (Fig. 1) are being considered to mitigate the issue of a lack of cells.

Figure 1.

Schematic of three possible sources (cadaveric, porcine, and stem cell-derived islets) for insulin-producing cells that can be transplanted to a variety of sites in pre-clinical and clinical studies.

3. Recapitulating the Islet Microenvironment

Understanding the microenvironment in which insulin-producing islets reside may provide clues on how to improve their survival following transplantation. The islet microenvironment in the pancreas has been reviewed thoroughly by Aamodt et al. [41]. Briefly, the islets of Langerhans take up about 1-2% of the pancreas and comprise multiple endocrine, hormone-producing cells; in these cell clusters most notably exist glucagon-secreting α-cells and insulin-secreting β-cells. Each of the cell types in the islets of Langerhans communicates through cell-cell interactions to maintain glucose homeostasis in the body. The islets are integrated with the nervous system, which assists in regulating blood flow to the islets and necessary hormone secretion [42], and with an enriched vasculature, which provides constant transport of glucose, nutrients, and oxygen to them, as well as the ability for insulin and glucagon to enter the bloodstream [43]. Immediately around the islets, there is a peripheral extracellular matrix (ECM) comprising of multiple proteins including fibronectin, collagen, and laminin, all of which contribute to the development of β-cells, gene expression, and insulin secretion [44]. Finally, there are resident macrophages that communicate with endothelial cells on the vasculature surrounding islets, which work together to promote β-cell regeneration [45]. Each of these components help make up the islet microenvironment that promotes the development and maintenance of insulin-producing β-cells.

When islets are isolated for the purpose of transplantation, they are ripped away from the microenvironment, and this can be very damaging to the cells because without the vasculature and immediate transfer of nutrients, there is a significant risk for hypoxia [46]. In addition, the loss of cell adhesion signals from the extracellular matrix triggers cell apoptosis [47]. When islets are infused through the hepatic portal vein, they do not have this microenvironment readily available, and the activation of IBMIR contributes to the severe loss in islets [38]. Additionally, there are not enough cadaveric pancreata and islets that satisfy the necessary criteria for islet isolation and eventual transplantation, further putting a strain on the supply of insulin-producing cells [39, 48]. Altogether, there is a great need to investigate techniques that can preserve the transplanted islets post-isolation and increase their longevity after transplantation.

4. Efforts to Improve Cadaveric Islet Allogeneic Transplantation

Efforts to improve the engraftment and function of islet allografts for T1D therapy have been a key research area for years. Encapsulation in biomaterial platforms is a strategy that can support islets with a microenvironment that can both protect them from the recipient’s immune response and provide the necessary ECM cues and close association with vasculature. Specifically, this technology can promote the long-term survival and function of transplanted cadaveric islets, and efforts have been made to improve the vascularization, oxygenation, and immune protection of these islets when transplanted into diabetic patients. Alongside this, there has been exploration into several extrahepatic sites for islet transplantation, and many have shown promise for T1D correction, vascularization, and retrievability in the cases where there are severe side effects post-transplant, and the device needs to be removed.

4.1. Allogeneic Immune Response to Cadaveric Islets

When allogeneic islets are transplanted into a T1D patient, there will be a full immune response to the allograft, as reviewed by others [49]. In brief, the four stages of allogeneic rejection are IBMIR, inflammation, innate immune response, and full allograft rejection. In IBMIR, there is primarily platelet aggregation, complement activation, and coagulation in the blood vessels, greatly increasing the risk for thrombosis in the patient. IBMIR is evidently observed when cadaveric islets are infused into the intraportal vein, hence why there is extensive work looking into extrahepatic transplant sites. Within a few days of transplantation, proinflammatory signals are released, causing an increase in cytokine production, and the influx of neutrophils, which then send signals out to macrophages and dendritic cells to enter the site to phagocytose the islets and present antigens on their surfaces to recruit adaptive immune cells. The infiltration of helper and cytotoxic T cells further destroy the islets, recruit B cells that will produce antibodies against the allogeneic islets and differentiate T cells into memory T cells, and ultimately lead to overall allograft rejection. This entire process of allograft rejection will likely be amplified in T1D patients, who already possess Teff cells that are specifically targeting β-cells. A widely pursued strategy to mitigate this immune attack is encapsulating islets within a device that prevents immune interactions. Methods that have been tested over the years both for encapsulation and immune modulation will be discussed further below.

4.2. Encapsulation Strategies for Cadaveric Islet Transplantation

When islets are isolated, they are disconnected from the vasculature, and with their high metabolic rates, it is crucial for graft survival that they promptly have access to a blood supply in the patient. Proximity to vasculature in the patient allows oxygen and glucose to access the β-cells and stimulate the secretion of insulin into the bloodstream. However, what is also crucial for β-cell transplantation is protection from the host’s immune response. A promising strategy to allow for nutrient diffusion and protection from the host immune response is biomaterial encapsulation. The key features of any encapsulation device for islets are to be cytocompatible, protect the cells from the host’s immune system, and limit hypoxia [50]. There are a variety of encapsulation strategies across multiple size scales (Fig. 2) that have been investigated in cadaveric islet transplantation research. Key findings across these encapsulation methods will be discussed in this section.

Figure 2.

Nano-, micro-, and macroencapsulation devices designed for islet encapsulation and delivery.

4.2.1. Nanoencapsulation for Cadaveric Islets

One of the encapsulation methods that has been studied in islet transplantation is nanoencapsulation, in which islets are individually enclosed in biomaterial compositions at a thickness of <100 nanometers [51]. This allows for an overall smaller distance between the islets and the surrounding microenvironment, allowing for better nutrient transport to the islets and insulin excretion into the bloodstream. This nanoencapsulation should also protect islets from infiltrating immune cells [50]. Multiple materials have been used for these nanocapsules, and they have typically been deposited through a layer-by-layer (LbL) approach onto islets. Early nanoencapsulation studies go back to 2010, where islets were coated in nano-thin PEG layers, and it was found that compared to the control islets, insulin secretion improved with the nanocapsules [52]. The results from this study set a precedent for intraportal vein infusions, in which the nanoencapsulation could protect islets from IBMIR and retain their function. Zhi et al. performed a similar study, but used the natural polymers chitosan, chondroitin-4-sulfate, and alginate to coat islets, and it was found that this coating did not hamper insulin secretion and retained their function for 1 month post-transplant in rodent models [53]. Further studies sought to transplant nanoencapsulated islets in vivo and assess their function over time. Fukuda et al. used the LbL strategy in a study that established a technique of coating fibronectin and gelatin on mouse islets and transplanting them to the kidney capsule (k.c.), resulting in diabetes reversal and insulin response to glucose stimuli [54]. Similarly, Park et al. used the LbL technique to coat non-human primate (NHP) islets with poly(ethylene glycol) (PEG) and heparin, and transplant them to the hepatic portal vein of cynomolgus monkeys. When an intravenous glucose tolerance test (IVGTT) was performed at 1 month post-transplant, the monkeys with PEG + heparin-coated islets had the highest stimulated C-peptide concentration in the bloodstream compared to those with PEG-coated islets and unmodified control islets [55]. While these results are very promising and have also been studied in insulin-producing cells derived from porcine hosts or pluripotent stem cells [56, 57], there are concerns that nanoencapsulation can cause incomplete encapsulation of the islets, leaving them still permissible to immune cell destruction [58].

4.2.2. Microencapsulation for Cadaveric Islets

Microencapsulation allows for individual islets to be surrounded by a biomaterial but provides a larger coating of typically 500-1000 μm in diameter [50] for better immunoprotection. The most used material for islet microencapsulation has been alginate, a seaweed-based material that has a slow gelation time, low cytotoxicity, and low potential for fibrotic responses [59], and these encapsulated islets have typically been transplanted to the intraperitoneal (i.p.) space of rodents and NHPs, as due to their large volume this is the only place that can hold the microcapsules [60], and recent studies have reported that compared to the s.c. space, was less of a risk of a fibrotic capsule and a more reliable reversal hyperglycemia [61]. Several studies have optimized the geometry and composition of alginate-based microcapsules and tested across a variety of pre-clinical animal models. De Vos et al. tested alginate-polylysine (polylysine was tethered to increase the mechanical strength of the microcapsules [62]) microcapsule diameters of 800 and 500 μm, and results showed that when the microcapsules were smaller, more islets were not immunoprotected, and they could not successfully reverse diabetes when transplanted to the i.p. space in rats [63]. Other studies such as Omer et al. concluded that smaller alginate capsules cross-linked with BaCl₂ had less cellular overgrowth and were more stable, contradicting the results of the previous study [64]. In 2015, Veiseh et al. screened different capsule sizes, shapes, and materials, and results across both rodent and NHP models showed that larger, spherical microcapsules (~1.5 mm diameter) elicited less of a foreign body response, and specifically 1.5-mm alginate-Ba⁺ microcapsules restored normoglycemia for up to 180 days, 5-fold longer than the 500-μm alginate microcapsules [65].

Alginate microcapsules have typically been cross-linked with inorganic compounds or synthetic polymers, as briefly mentioned above. Optimizing the composition of the alginate microcapsules has been a priority of islet encapsulation research, primarily to minimize host immune responses and fibrotic capsule growth. Alginate is composed of mannuronic acid (M) and guluronic acid (G), and different batches of alginate will have different ratios of these components. Initial studies determined that high-M alginate elicited a smaller immune response than high-G alginate, and when high-M alginate was combined with BaCl₂, allogeneic islets could last for over 350 days in the i.p. space with no cellular overgrowth [66, 67]. Further studies in alginate formulations have been tested since then and were successful as allografts in the omental bursa of macaques [68].

In recent years, there has also been an effort to incorporate ECM components into alginate microcapsules. Although alginate is cytocompatible, it does not present pro-survival cues for the encapsulated islets. Multiple studies have reported the incorporation of collagen IV and laminin peptides in alginate microcapsules, and their ability to protect the encapsulated islets from inflammatory cytokines and to support the insulin secretion of islets [69-71]. These studies show that the incorporation of ECM ligands overall contributes to the immune acceptance of the islet grafts. The incorporation of ECM peptides into synthetic hydrogels is also a heavily researched topic, as the ligands provide bioactive cues for the encapsulated cells, and synthetic polymers are more tunable and have less lot-to-lot variability [72, 73]. Tomei et al. combined PEG, alginate, and the amino compound triethanolamine into a hydrogel that would uniformly coat islets of different sizes. This study reported that when optimized, the thickness of microcapsules would not harm islet viability, result in no delay in glucose-responsiveness, and maintain normoglycemia when transplanted to the k.c. in a syngeneic murine model for 100 days [74]. A follow-up study in 2022 also utilized PEG hydrogels for the coating of individual islets, but instead used an emulsion method, as the original method required a low pH that would harm NHP islet viability. Results indicated that the new emulsion coating method improved viability and insulin secretion compared to the previous coating method, however there was not any notable function when transplanted to the omentum of NHPs, showing a point for further optimization of PEG microencapsulation of islets [75]. Weaver et al. reported a study in which ECM peptides were screened as adhesive ligands in synthetic polymer microcapsules. Multiple fibronectin- and laminin-based peptides were examined, and it was found that when encapsulated with islets, in vitro insulin secretion was best when RGD was tethered to the PEG macromers. Further in vivo experiments showed that the synthetic hydrogels could maintain islet function longer than the conventional alginate microcapsules [73]. As there is early evidence that ECM peptide-tethered synthetic microgels might perform like alginate microcapsules, there is potential for this area of islet microencapsulation to be studied more, and improvements in the could be scaled up for future therapeutic purposes.

4.2.3. Macroencapsulation for Cadaveric Islets

Macroencapsulation is a strategy where larger amounts of islets can be encapsulated in a device that is at least 1 mm in diameter (roughly <10 mm in rodent models, <30 mm in non-human primate models [76], and >10 mm in human trials [50]) and is either intravascular or extravascular [76]. Studies investigating intravascular macrodevices, in which hollow fibers encapsulating islets are transplanted via vascular anastomosis to be in direct contact with host vasculature [50], have reported scaffolds with optimized porosity to reach optimal insulin secretion and the prevention of hypoxia [77, 78]. However, intravascular encapsulation and transplantation are complex, and there are high risks for infection, coagulation, and bleeding, so extravascular transplantation has been the primary focus for macroencapsulation of islets [50, 79]. However, because extravascular devices are large (mm-cm scale), the immunoisolating device prevents direct connection to vasculature and innervation, making the access to nutrients much harder and thus harming islet viability and function. One example of this challenge is illustrated in studies investigating the transplantation of the Beta-O₂ device; the overwhelming result from these studies was that islet functionality could not be maintained despite being viable, primarily due to low oxygenation, poor nutrient diffusion, immune cell infiltration, and collagen-dense fibrotic capsule formation [80-82]. To address the lack of close vascular apposition, several methods have been tested to improve the islets’ access to host vasculature in the transplant site, whether it be through co-delivery with vasculogenic growth factors, the co-delivery with cells or vessels, or pre-vascularization of the site.

Vascular endothelial growth factor-A (VEGF-A) has proven to be critical for islet vascularization in the microenvironment and providing isolated islets with the vascular environment to survive and function after transplantation [83-85]. In a 2013 study, 4-arm PEG-maleimide (PEG-4MAL) was conjugated to VEGF and ECM peptides to form a PEG + VEGF hydrogel. When encapsulated with islets and injected to the small bowel mesentery (s.b.m.) of mice, there was significantly improved revascularization of the graft compared to intraportal vein transplantation [86]. A later study showed that the PEG + VEGF system also worked with encapsulating rat islets and revascularizing in syngeneic recipients within 4 weeks [87]. Subsequent studies using the PEG + VEGF system examined the s.b.m., subcutaneous (s.c.) and gonadal fat pad (g.f.p.) of mice as clinically translatable extrahepatic transplant sites [88]. This study demonstrated that the g.f.p. facilitated the best revascularization and T1D correction, as well as a minimal immune response, showing the potential of the g.f.p. (and the equivalent omentum in larger mammals) as a transplant site. The omentum has continued to get studied as a transplant site for cadaveric islets, as it is already highly vascularized, and has shown potential for other vasculogenic or oxygen-generating hydrogels, making it highly versatile as well [89, 90].

Recent research in T1D cell therapy with cadaveric islets has focused on the s.c. space, primarily due to its minimal invasiveness and easy retrievability. Some work in the s.c. space has investigated device-less approaches; notably Pepper et al. reported a device-less method in which transplanted islets in mice after pre-vascularization with a medical-grade catheter resulted in >90% return to normoglycemia for at least 100 days [91]. While the allograft largely resulted in a sustained reversal of hyperglycemia, it is important to note that the lack of immunoprotection requires a form of systemic immunosuppression. Additionally, the s.c. space is poorly vascularized and elicits strong immune responses [88], so efforts to engineer macroencapsulation devices that can support vascularization of islets have been of high interest. The TheraCyte^™ device, a polytetrafluoroethane (PTFE)-based macroencapsulation device, has been used in several studies, in which islets are encapsulated within a 2-layer membrane that facilitates nutrient diffusion, neovascularization, and immune protection [92]. In a 2013 study, the macrocapsule was implanted in the s.c. space in a rat allogeneic model, and there was a 6-month survival in both immune deficient and immune competent mice with minimized CD8⁺ T cell infiltration [93]. Studies like this one set a precedent that devices targeting immunoisolation of the islets could protect the islets and ensure longer graft survival. PTFE has also been used in more complex encapsulation device designs, such as biphase cell delivery devices that had access to atmospheric oxygen and allowed for the refilling of cadaveric islets without an additional surgery [94]. There has also been investigation in other synthetic polymers for s.c. islet encapsulation; these include a triazole-zwitterionic hydrogel that promoted vascularization, inhibited fibrosis, and corrected diabetes up to 1 month in mice, a 3D-printed PLA device implanted with a growth factor enriched platelet gel, allowing for a pre-vascularized environment prior to transplanting islets and the maintenance of islet function, and PEG-based hydrogels involving methacrylic acid (MAA) or polymethacrylate (PMAA), both of which could greatly enhance vasculature in the s.c. space and restore normoglycemia in diabetic mice [95-98]. The use of natural polymers for s.c. islet transplantation in numerous pre-clinical models has also proven to be successful in allowing for the diffusion of nutrients and vascularization into the transplant site, and minimization of inflammatory immune responses. The most common natural polymers used in macrodevices reported in literature are alginate [99, 100], due to the reasons listed earlier, or collagen [101-105], as it is the most common protein in the ECM of the pancreas, making it a key component of the islet microenvironment [106]. There is also potential for alginate and collagen to be combined to a scaffold that improve vascularization in the s.c. space, as Mahou et al. demonstrated by embedding mesenchymal stromal cells and coating endothelial cells on alginate-collagen microspheres; a scaffold model like this one could incorporate islets, and studies can be performed analyzing their engraftment and function [107]. One notable novel macrodevice was developed by Wang et al., which comprised of a thermoplastic silicone-polycarbonate-urethane (TPSU) nanofibrous shell and an alginate core that encapsulated the islets. This device was tested across syngeneic, allogeneic, and xenogeneic rodent islets to mouse models and displayed islet function across 200 days, however there were issues in material and encapsulation inconsistencies [108]. The above studies indicate many directions for the s.c. transplantation of cadaveric islets, and more studies will optimize the transplantation protocols for future pre-clinical studies and clinical trials.

4.2.4. Incorporation of Immunomodulatory Proteins on Biomaterial Delivery Vehicles

All the devices discussed above focus on isolating the islets from the surrounding immune system. While they have shown progress in pre-clinical trials, there is also the potential for immunomodulation, in which the local immune environment is tuned to accept the graft at the transplant site without encapsulation in a device. An attractive aspect of this strategy is the possibility for direct revascularization/innervation of transplanted islets without the deposition of a collagen-dense fibrous capsule associated with encapsulating membranes and devices [82]. Fas ligand (FasL), is a member of the tumor necrosis factor (TNF) family that binds to the Fas receptor on cytotoxic T cells and induces their apoptosis [109], and promotes the homeostasis of regulatory T (Treg) cells [110]; due to its ability to establish a tolerogenic immune environment, it has been widely studied for islet immune acceptance. As early as 1996 researchers have been able to prevent rejection of islet allografts with myoblasts transfected with FasL (and thus were now FasL-expressing) in mice [111]. Subsequent studies further characterized FasL as a promising protein for islet protection after transplantation [112], and pancreatic islets were suspended in a biotin solution and incubated with streptavidin-FasL (SA-FasL) chimeric protein, allowing the SA-FasL to be immobilized onto the cell surface, to maintain Treg cells at the site long-term and establish a defense against effector T (Teff) cells [113]. Over the years, researchers also saw the value in tethering SA-FasL to biomaterial carriers encapsulating islets. Most notably, Headen et al. demonstrated that biotinylated PEG microgels functionalized with SA-FasL and co-implanted with allogeneic islets yielded allograft acceptance for 200 days only with a short transient rapamycin treatment and involved the generation of Treg cells [114]. Expanding upon this, Medina et al. reported a PEG-based hydrogel system that delivered FasL and IL-2D, an analog to IL-2, to further increase the Treg infiltration, which only induced short-term allograft survival, but set a precedent for the co-delivery of multiple proteins for modulating the immune microenvironment at the transplant site [115]. Results from the success of SA-FasL being tethered to PEG hydrogels in murine models led to a NHP study that was the first reported result of immunomodulatory biomaterials successfully transplanting islets and inducing immune acceptance in NHPs [116]. Due to the success of SA-FasL-tethered biomaterials in NHPs, there is great promise for this technology to be tested in humans in clinical trials in the years to come. The potential for immunomodulation also extends to other proteins such as PD-L1, which also serves in suppressing the adaptive immune response. Briefly, PD-L1 is the ligand involved in the programmed cell death-1 (PD-1) pathway, and it plays a key role in mitigating alloimmune responses, and conversely, blockades of the PD-1 pathway would greatly worsen the immune responses [117]. While recent, the publications that have come out in the past few years, in which PD-L1 is tethered to islet allografts or biomaterial carriers in a similar fashion as FasL, indicate promise in long-term graft survival, reversal of diabetes, and delayed immune rejection [117-119]. With their success in immune competent rodent models, it is possible these will translate into NHP studies.

4.3. Encapsulated Cadaveric Islets in Clinical Trials

Encapsulation devices across all three discussed strategies have made it to clinical trials. ViaCyte, formerly Novocell, launched a phase 1/2 clinical trial on allogeneic islets in PEG nanoshells, but this study had to be terminated and there are minimal results [120]. Clinical trials for islet encapsulation in alginate microcapsules have been reported since 1994, in which a patient with long-term insulin-dependent diabetes was transplanted twice with human islets encapsulated in alginate microcapsules in the i.p. space and was administered a low-dose cyclosporin and azathioprine immunosuppressive regimen. This patient displayed no adverse health effects, only 2 hypoglycemic blood glucose readings across 9 months, and overall health improved post-transplant [121]. A 2009 publication reported a similar study, but with four patients only administered anti-inflammatory and antioxidant medications rather than systemic immunosuppression. While blood glucose levels overall decreased, they were not consistently maintained, C-peptide levels in the bloodstream were inconsistent, multiple cytotoxic antibodies developed over time, and early islet loss required multiple infusions of islets [122]. Phase I clinical trials in Italy have also been examined with non-immunosuppressed patients, and the i.p. infusion of alginate-encapsulated islets was safe, and C-peptide was detected in the bloodstream 6 months post-transplant. However, there was still a continuous need for exogenous insulin injections, potentially because a smaller islet dose was administered for the safety and efficacy trial [123]. A more long-term study showed that four non-immunosuppressed patients transplanted with alginate-encapsulated islets had no immune responses, and while there was C-peptide content in the bloodstream, exogenous insulin injections remained prevalent and returned to the same regimens 7 years post-transplant [124]. While it was encouraging that there were no major immune responses without immunosuppression, none of these studies demonstrated long-term insulin independence. Finally, the β-Air device from Beta-O₂ has been used for clinical trials. The β-Air device incorporated a fillable oxygen tank, allowing for continuous oxygenation of the transplanted islets. Phase I of this clinical trial reported when the β-Air device was loaded with ~2000 IEQ/kg per patient, no immune system activation or rejection of the transplanted tissue. However, there were very low levels of C-Peptide detected in the bloodstream, likely due to 1) the fibrotic capsule formation around the device greatly hampered the diffusion of secretion, and 2) there was a non-clinically relevant dose of islets in the device, thus concluding that the β-Air device was safe but not functional [125]. ViaCyte has also developed a macrodevice that in recent years has been tested in clinical trials for stem cell-derived islets and will be discussed below in the advances of stem cell-derived islets.

5. Xenogeneic Islet Transplantation for T1D Cell Therapy

Despite efforts to improve the engraftment and immune acceptance of allogeneic islets, reliance on this cell source results in a shortage in donor supply. One route to circumvent this problem is the use of porcine islets for xenotransplantation. Porcine insulin is very similar to human insulin—only varying by one amino acid—and has been used to treat human T1D patients [126]. Pigs can also be bred easily, allowing for a much higher frequency of donors compared to humans. Protocols have been developed to reproducibly isolate porcine islets [127], and pre-clinical studies such as Potter et al. have shown that when neonatal and adult porcine islets were transplanted through the intraportal vein of NOD-SCID mice, there was no formation of amyloid—a common deposit associated with β-cell apoptosis—and a maintenance of normoglycemia for nearly 200 days [128]; this was different from the transplantation of human islets, which resulted in noticeable amyloid deposition and islet graft failure.

However, several key issues exist for the xenotransplantation of porcine islets to humans. Some of these include the inevitable human response to porcine islet xenoantigens such as galactose-α-(1,3)-galactose (α-Gal) and N-glycolylneuraminic acid (Neu5Gc), the concern for virus transmission, and similar concerns about IBMIR that exist with allogeneic islets. Extensive research over the past few decades has sought to tackle these issues.

5.1. Immune Response to Porcine Islet Xenografts

A primary concern of xenotransplantation is rejection of the graft. When a xenograft is first transplanted, there will typically be hyperacute reaction, characterized by the triggering of the complement pathway, platelet aggregation, coagulation, and infiltration of innate immune cells such as neutrophils [129]. Platt et al. demonstrated this in their study focusing on kidney and heart transplants from pigs to rhesus monkeys, which indicated that the deposition of rhesus IgM contributed to hyperacute rejection [130]. To evaluate the rejection of islet xenografts, studies have been performed across different species and transplant sites to map the timeline of xenograft rejection and mitigate this rejection through the delivery of inhibitory factors, genetic engineering of donor pigs, and most notably biomaterial encapsulation of porcine islets [131].

Infusion of porcine islets into the intraportal vein results in strong activation of the complement and coagulation pathways, causing platelet aggregation and thrombosis, and hyperacute rejection. Bennet et al. showed that when adult porcine islets were infused into the intraportal vein of cynomolgus monkeys, within 60 minutes, islets were embedded in clots, infiltrated with leukocytes, deposited with C3a and the membrane attack complex when the monkeys were not treated with the complement pathway inhibitor. However, when monkeys were treated with soluble complement receptor 1 (sCR1) at the time of transplant, there was much less insulin release, as high insulin release is indicative of islet death, and further concluding that anticoagulants are essential if intraportal infusion of porcine islets were to become clinical practice [132]. Bühler et al. transplanted adult porcine islets to the intraportal vein of baboons, but instead assessed novel immunosuppressive regimens of complement blockers (cobra venom factor), the removal of the anti-Gal antibody, and multiple T cell inhibitors (anti-CD154, anti-thymocyte globulin, and mycophenolate mofetil). This study showed that while there was inhibition of T cells and IBMIR commonly associated with intraportal vein infusion, the grafts still failed by day 28 primarily due to macrophage infiltration, further concluding that immunosuppression needs to also target macrophages [133]. Cardona et al. performed a similar study by infusing neonatal porcine islets through the intraportal vein of rhesus macaques and treating animals with CD28-CD154 blockade, and it was found that there was no Gal-specific hyperacute rejection and insulin independence was maintained for >140 days [134]. Notable milestones from this study were the fact that long-term diabetes correction was achieved in a xenogeneic transplant model, and the potential of neonatal porcine islets as opposed to adult porcine islets. Hering et al. corroborated these results by demonstrating that xenograft rejection of adult porcine islets to the intraportal vein of cynomolgus macaques was primarily mediated by T cell and macrophage infiltration as opposed to Gal-specific hyperacute rejection typically associated with intraportal vein infusion. In macaques that underwent rejection, it was found that utilizing CD154 antibodies for immunosuppression often caused thromboembolic events, indicating a need to investigate other immunosuppressive regimens [135]. Neonatal porcine islets continued to be used in multiple immune system blockade studies, including Thompson et al., in which the CD40/CD154 pathway was targeted by a therapy of chimeric anti-CD40 monoclonal antibody (Chi220) and anti-IL-2 receptor, resulting in protection of the xenograft for an average of 90 days and circumventing thromboembolism [136].

The findings from these NHP studies with porcine islets motivated further pre-clinical studies in rodents, such as Contreras et al. which became a precedent for genetic modification, and eventual encapsulation of islets. Porcine islets were genetically modified to overexpress Bcl-2, a gene repeatedly shown to decrease cytotoxicity caused by host antibodies and complement, and surface-modified with PEG and transplanted into NOD-SCID mice. Results showed that the surface- and gene-modified islets provided the best cytoprotection and corrected blood glucose levels compared to control islets [137]. Earlier studies also sought to investigate other transplant sites to avoid the IBMIR typically associated with infusion of cells into the portal vein. Transplants of fetal porcine islets underneath the k.c. of rats elicit infiltration of CD4⁺, ED1⁺, and ED2⁺ macrophages [138, 139]. Shortly after this discovery, Soderlund et al. evaluated the transplantation of fetal porcine islets under the k.c. of cynomolgus monkeys, some of which received cyclosporine and 15-deoxyspergualin immunosuppression and others did not. This study established a timeline of immune cell infiltration in response to the xenograft, namely that NHPs transplanted without immunosuppression had neutrophil infiltration at day 1, a clear necrotic area and infiltration of polymorphonuclear cells and CD8⁺ T cells at day 3, and larger numbers of macrophages infiltrating the graft at day 6. In contrast, immunosuppressed monkeys had significantly less CD8⁺ T cell infiltration by day 6. While the grafts in both sets of monkeys failed by day 12, it is also noteworthy that the immunosuppressed monkeys had much less CD8⁺ and CD68⁺ cell infiltration [140]. All these studies together show the potential of moving away from the intraportal vein as a transplant site for porcine islets and the immune responses associated with that transplant site.

5.2. Genetic Modification of Pigs

A promising strategy to circumvent the use of chronic immunosuppression is the genetic engineering of pig donors. A lingering concern in porcine islet transplantation has been the human and NHP antibody response to porcine antigens, so researchers have sought to inhibit the antigenicity in donor pigs and delivering islets from those pigs in pre-clinical studies. Komoda et al. used transgenic pigs overexpressing N-acetylglucosaminyltransferase-III (GnT-III), which has previously been proven to reduce porcine antigenicity to human natural antibodies [141], and when adult GnT-III porcine islets were transplanted under the k.c. of cynomolgus monkeys, there was prolonged survival of the xenograft compared to transplanted wild-type porcine islets, providing preliminary results that suppression of porcine antigens could play a role in xenograft survival [142].

Further genetic engineering studies have focused at knocking out α-Gal in pigs, and full porcine heart transplants to baboons have avoided hyperacute rejection, but transplants later failed primarily due to thrombotic microangiopathy [143]. Subsequent studies have centered on engineering transgenic pigs that either expressed or were knocked out for other key genes and enzymes that would aid in xenotransplantation. One of the most notable studies came in van der Windt et al., in which pigs were genetically engineered so that their islets would express a human complement-regulatory protein (hCD46), and those porcine islets were infused through the intraportal vein of diabetic cynomolgus monkeys. The authors reported that 80% of the monkeys were normoglycemic and insulin independent for 3 months, and one monkey was monitored for 12 months. Most notably, the expression of hCD46 reduced the need for chronic immunosuppression [144]. This became the first study to verify that transplanting porcine islets from transgenic pigs could result in the inhibition of the complement pathway, subsequent antibody-mediated rejection, and long-term normoglycemia with a reduced need for immunosuppression in primates. Subsequent studies investigated genetically engineering multiple genes in pigs, and with these studies came developments to rapidly accelerate the generation of transgenic pigs [145] and the modification of 3-5 genes [146]. Although these techniques have not resulted in long-term graft survival, it opened the possibility for overexpressing or knocking out multiple genes to find an optimal transgenic pig for xenotransplantation for multiple cell types and organs [147]. Additionally, studies have been conducted on examining human natural antibody titers for porcine antigens, which can further be used to narrow down which genes to focus on for generating transgenic pigs to avoid xenograft rejection [148].

5.3. Encapsulation of Porcine Islets

As stated earlier, transplantation of encapsulated islets is a cell therapy option with high potential, and the encapsulation of porcine islets is an attractive approach to shield the cells from xenorejection. A variety of natural and synthetic biomaterials have been used to encapsulate and deliver porcine islets. Cruise et al. encapsulated adult porcine islets in PEG-diacrylate (PEG-DA) and transplanted them in the i.p. space of athymic mice and Sprague-Dawley rats, leading to sustained normoglycemia for 110 days and 30 days, respectively [149]. While encouraging, it is important to note that PEG-DA and multiple other synthetic polymers cross-link through photopolymerization, which can reduce cell viability. Agarose has been frequently used as an encapsulation material for porcine islets across murine, canine, and NHP models, all of which have shown long-term normoglycemia [150-152]. While agarose has worked across several animal models, there are concerns that tissue will often react against agarose resulting in tissue adhesions and fibrous encapsulation, and it does not have optimal permeability that is essential for nutrient diffusion and blocking of complement factors [153]. Alginate has been the most used natural polymer for encapsulation of islets, and years of work has gone into the microencapsulation of both adult and neonatal porcine islets in murine, NHP, and clinical human studies.

5.3.1. Alginate Microencapsulation of Porcine Islets

Studies in the microencapsulation of porcine islets go back to the 1990s, and initially involved alginate tethered to poly-L-lysine (Alg-PLL)—which have tunable particle size, are more mechanically stable than traditional alginate microcapsules, and provide protection to the transplanted cells from immune rejection [154-156]—as the encapsulation material. Weber et al. transplanted encapsulated rat and dog islets in the i.p. space of NOD mice and showed the prevalence of CD4+ T cell infiltration [157]. While this was not done with porcine islets, it provides a foundation for the immune response of spontaneously diabetic, immune competent mice to xenografts. Alg-PLL continued to be used as a porcine islet microcapsule device in murine and NHP studies, the latter of which notably 7 out of the 9 cynomolgus monkeys achieved insulin independence for a range of 120-804 days without chronic immunosuppression [158, 159].

Over the years, studies have explored optimizing the alginate microcapsules by tethering different polymers or molecules to enhance stability or cell signaling. With the success of Alg-PLL microcapsule formulations, King et al. conducted a study screening different formulations of Alg-PLL microcapsules for their potential for fibrosis. These empty capsules were transplanted into C57BL/6 and Balb/c mice for one month with different timepoints along the way, and there was some indication that PLL could be contributing to cellular overgrowth and a stronger reaction on the capsule surface with cells [160]. This result was confirmed in a later study where alginate microcapsules were analyzed for the fibrotic reaction to them, and it was found that when barium was incorporated in the alginate microcapsules and PLL was excluded, the adult porcine islets were protected to the largest extent [161]. Other work contributed to the optimization of alginate microcapsule formulation and transplantation, namely that a double coat of alginate was more cytoprotective than a single coat [162], and the double-coated alginate microcapsules were used further in NHP study, which showed a reduction of hyperglycemia and no fibrotic responses before graft failure, likely due to hypoxia [163]. Additionally, a few studies have been done in which chitosan, another natural biomaterial, is tethered to the alginate microcapsules, and provided better cytoprotection and survival of the graft [164, 165]. However, chitosan is most soluble in highly acidic solutions which can harm cell viability, limiting its potential in therapeutic applications. Finally, recent studies have also sought to provide an extracellular matrix (ECM) that can help restore the microenvironment to islets after being isolated; Medina et al. functionalized alginate with ECM peptides derived from fibronectin, collagen, and laminin, and showed that RGD—a fibronectin-derived ligand—;supported the in vitro viability of porcine islets and significantly improved insulin secretion of the islets [166].

Success with the alginate microcapsules has led to clinical trials. As early as 2000, Elliott et al. reported a retrospective clinical trial that reports no transmission of porcine endogenous retroviruses (PERVs) after i.p. injection of alginate microcapsules with porcine islets [167]. Similarly, a 2007 study reported the first long-term function of neonatal porcine islets in alginate microcapsules in the i.p. space of one adult male for nearly 10 years [168]. Success in these initial studies launched the first clinical trials for alginate microencapsulated porcine islets by DIABECELL to the i.p. space, which also resulted in no PERV transmission, verifying safety in humans [169, 170]. Success in the Phase I clinical trial is a step forward for future clinical trials that will come with alginate-encapsulated porcine islets. Along with the encapsulation strategy, genetic engineering can also be extrapolated to engineer transgenic pigs with less PERV expression, making the xenotransplantation of porcine islets even safer.

5.3.2. Macroencapsulation of Porcine Islets

Porcine islet transplantation has been evaluated in clinical trials through the encapsulation in alginate microcapsules [167-170]. While promising, there are a couple of disadvantages of microcapsules that point research more towards macroencapsulation strategies. A significant limitation of microcapsules is that due to gravity, they may settle and clump together, causing hypoxia and cellular overgrowth, and their movement makes them difficult to recover after transplantation [171]. In contrast, the use of macrodevices could overcome these limitations. A key obstacle of macroencapsulation devices is subpar nutrient and insulin diffusion in and out of the device because of the fibrous capsule that forms around the device. To tackle this issue, Petersen et al. developed a hydroxymethylated polysulfone device that in in vitro studies improved insulin diffusion and did not release any PERVs from the porcine islets [172]. Researchers have also investigated bioartificial pancreas devices, primarily made of synthetic polymers. Teotia et al. investigated polysulfone-based devices and concluded that when polysulfone was combined with d-α-tocopheryl PEG 1000 succinate (TPGS) to make a hollow fiber membrane, normoglycemia could be maintained in diabetic mice for 30 days [173]. While this was a short-term study, it provided important material analysis for macroencapsulation devices that could be used in future research. Lee et al. investigated polydimethylsiloxane (PDMS) as the encapsulation device material and combined it with calcium peroxide (CaO₂) to formulate an oxygen-generating material. When encapsulated with neonatal porcine islets, there was higher cell viability and insulin secretion compared to standard PDMS scaffolds in vitro, and when transplanted to the s.c. space of Balb/c mice, neovascularization was observed, indicating potential for this site to successfully engraft the transplant [174]. Finally, Elliott et al. and Ludwig et al. used the PTFE TheraCyte device and Beta-O₂ bioartificial pancreas (PTFE + Alginate), respectively, and transplanted adult porcine islets in that device to the s.c. space of NHPs with no immunosuppression. Results from the TheraCyte device resulted in 8 weeks of normoglycemia in cynomolgus monkeys [175]. In contrast, results from the Beta-O₂ device showed up to 6 months of graft survival in rhesus macaques and release of C-peptide in response to glucose stimuli, but did not result in complete insulin independence, likely due to insufficient transport of nutrients and glucose [176].

Macroencapsulation devices made from natural polymers have also been investigated. Notably, Dufrane et al. implanted a collagen and alginate device encapsulating adult porcine islets in the s.c. space of diabetic NHPs, resulting in up to 6 months of normoglycemia, compared to non-encapsulated islets being rejected within 1 week. However, it was also noted that there was a strong humoral response to the encapsulated islets, namely that there was a significant increase in IgG anti-pig antibodies post-transplantation. While anti-pig antibodies decreased 2 months post-transplant in the NHPs with encapsulated islets, it is also important to note that both NHPs and humans naturally possess anti-pig antibodies, and this will always be a hurdle to get over for porcine islet transplantation [177]. Most recently, Ajima et al. reported an alginate macrodevice with adult porcine islets transplanted to the i.p. space of immune competent mice that restored normoglycemia for up to 200 days [178], and Duin et al. explored the combination of alginate and methylcellulose as a bioink for the encapsulation of neonatal porcine islets, and in in vitro studies showed success in insulin secretion for 4 weeks [179]. These studies show that macrodevices for porcine islets can have a wide variety of biomaterial compositions and can be transplanted to numerous sites, highlighting these devices as a promising translatable technology for T1D cell therapy.

5.4. Considerations regarding Porcine Islet Xenotransplantation

Pre-clinical studies in porcine islet encapsulation and transplantation forT1D have provided incredibly valuable advancements that will dictate the direction of this research, whether it be through attenuation of the host’s immune response, genetic engineering of cells and pig donors, or the encapsulation for immune protection in xenotransplantation. Multiple questions remain in the field. One involves whether the focus should be on adult or neonatal porcine islets. Whereas adult porcine islets are more mature and neonatal porcine islets need a few weeks to mature, studies have shown that when neonatal porcine islets are transplanted, the β-cell mass increases in transplant sites, likely due to a mixture of proliferation and differentiation [180-182]. Another relevant question is what the optimal transplant site is. Dufrane et al. transplanted porcine islets in alginate microcapsules in the i.p. space, s.c. space, underneath the k.c. of rats, and the latter two proved to be better sites than the i.p. space, because >13% of the capsules had broken in this transplant site, resulting in cellular overgrowth and CD68⁺ cell infiltration [183]. However, the k.c. is not a clinically translatable site for humans, further concluding that the s.c. space is likely the most promising transplant site to continue investigating. As stated earlier, there is a question of whether microcapsules or macrodevices are the best route for encapsulation methods. Cao et al. demonstrated through mathematical modeling that microcapsules provide the most oxygen to islets [184], but macrodevices are easier to implant and retrieve later if necessary. Finally, there is also the question if porcine islets are a better option than cadaveric islets, and studies have shown that while many porcine islets are needed to reach the necessary insulin secretion for humans [185], they perform better than human islets in pre-clinical studies [186]. With how readily available porcine islets are, and their promise in multiple xenotransplant settings and early clinical trials, they have significant potential to solve the issue of a lack of donor islets for T1D patients. However, there remains the risk of porcine islets being a xenograft and eliciting a strong innate and humoral immune response from the patient. Allogeneic cell transplantation remains a promising cell therapy resulting in a smaller immune response, and the following section will discuss human stem cell-derived islets as another attractive option for a large source of insulin-producing cells.

6. Stem Cell-Derived β [SC-β) Cells for T1D Therapy

A transformative solution for generating a replenishable source of β-cells is the differentiation from human pluripotent stem cells to insulin-producing β-cells. At first, it was postulated that pluripotent stem cells can be derived from the individual themselves, which can then be differentiated into β-cells, and when transplanted back to the body there is very low risk of immune rejection [131]. However, this would be an incredibly expensive method of transplantation with considerable regulatory challenges, and there is a risk of the patient’s autoimmunity to attack autologous cells. Due to this, allogeneic SC-β cells are currently being investigated in clinical trials, including those derived from human leukocyte antigen (HLA) knockout (KO) induced pluripotent stem cells (iPSCs), as a cell therapy for T1D [187, 188]. With heavily studied and optimized differentiation protocols, huge batches of insulin-producing cells that recapitulate many behaviors of adult islets can be manufactured using validated bioprocesses and transplanted when needed, eliminating the need for multiple donors.

6.1. Development of β-Cell Differentiation Protocols

Investigators have worked extensively on the differentiation of these cells and in the past decade have invented and optimized protocols that closely mimic the development of β-cells during the gestational period, as reviewed in Oliver-Krasinski et al. [189]. In the early 2000s, D’Amour et al. successfully differentiated endoderm cells and pancreatic hormone-expressing endocrine cells from human embryonic stem cells, thus establishing a foundation for the differentiation of β-cells from pluripotent stem cells [190, 191]. Nearly a decade later, Rezania et al. and Pagliuca et al. developed high-throughput protocols for differentiating β-cells from pluripotent stem cells. While not yet at the level of cadaveric primary human islets, the development of these protocols greatly improved the marker and gene expression, insulin response to glucose stimuli, and diabetes correction behavior in vivo of the stem cell-derived islets [192, 193]. Since then, numerous studies have been conducted to further investigate the signaling pathways involved in β-cell differentiation, what factors should be incorporated into the protocol to improve β-cell yield and maturation, and validation of the protocol across multiple cell lines. One of the crucial phases in differentiation is the pancreatic progenitor (PP) phase, in which pancreatic marker PDX1 and β-cell marker NKX6.1 are expressed in cells, and the PDX1⁺/NKX6.1⁺ population is a strong indicator of how the eventual β-cell yield will be. Nostro et al. reported that the induced expression of NKX6.1 is regulated by exposure to retinoic acid, fibroblast growth factor 10 (FGF10), and inhibitors of the bone morphogenetic protein (BMP) and hedgehog signaling pathways. With these modifications, the PPs were able to differentiate into endocrine cells when transplanted into NSG mice [194]. Groups have also sought to improve the endocrine cell and β-cell generation by targeting stages after PP generation. These efforts include the clustering of endocrine cells to form enriched aggregated β-cells, the development of an enriched serum-free media (ESFM) for the final β-cell phase, the disruption of the actin cytoskeleton to induce better endocrine cell differentiation, and the improvement in mature SC-β cell yield [195-198]. Finally, validating these protocols across multiple cell lines further ensures the reproducibility and potential for this procedure. Studies have used this protocol in multiple cell lines; most notably Millman et al. validated this study on stem cells from patients with T1D, indicating the potential for autologous transplantation and a low risk for immune rejection, and Hogrebe et al. reported a full differentiation protocol with specific criteria for different stem cell lines [199, 200].

Regarding transplant sites, the k.c. has consistently been used as the transplant site in murine models and has shown to be a site where transplanted SC-β cells could correct T1D. Maxwell et al. tested multiple cell doses (0.75M, 2M, 5M) and transplant sites (k.c., intramuscular, and s.c.), and showed that when 5M cells were transplanted underneath the k.c. of NSG mice, there was a reversal of diabetes and maintenance of normoglycemia across 22 weeks, and the SC-β cells matured in this site 14 weeks post-transplant denoted by the expression of C-peptide, MafA, MafB, and FAM159B. In contrast, lower cell doses and the other two transplant sites tested did not reverse hyperglycemia [201]. This study sets a precedent for future studies in the transplantation of SC-β cells, particularly in investigating alternative transplant sites, as the k.c. is not clinically translatable.

While all these studies show significant promise, there are a few key obstacles with SC-β cells. The marker expression and function of SC-β cells are not at the same level as cadaveric human islets. Insulin dynamics in response to glucose stimuli are often the same, but cadaveric islets secrete much more insulin than SC-β cells [200]. Additionally, with regards to in vivo studies, while the k.c. has been promising in murine models, this is not a clinically translatable site for the transplantation of SC-β cells to humans. There is a clear need to provide the SC-β cells a microenvironment where they can differentiate and mature, and examine clinically translatable, retrievable transplant sites. Multiple biomaterial platforms have been tested over the years and have shown promise in pre-clinical and clinical experiments, and an extensive list is reviewed in Neumann et al. [202]. Notable studies in the field will be discussed further in Section 6.3.

6.2. Genetic Engineering of SC-β Cells

To mitigate autoreactive and alloreactive immune responses, genetic engineering of SC-β cells has emerged as a promising strategy. As stated earlier, an alluring aspect of SC-β cell therapy is the potential for autologous transplantation, and researchers have investigated iPSCs from patients with monogenic diabetes, in which there are only mutations in a single gene. For example, Ma et al. investigated a mutation located in the INS locus from patients with neonatal diabetes, and discovered that when this mutation was corrected, the iPSCs from these patients demonstrated restored insulin production [203]. Maxwell et al. investigated the iPSCs of Wolfram syndrome (WS) patients, and specifically a diabetes-inducing mutation in Wolfram syndrome 1 (WFS1). When CRISPR/Cas9 was used to correct the pathogenic variant, the insulin production of the differentiated SC-β from WS-based iPSCs was much stronger than the uncorrected cells [204]. Finally, Lithovius et al. investigated correcting the SUR1 mutation in the K_ATP channel, and instead found that the mutant SC-β cells displayed higher insulin secretion than the gene-corrected SC-β cells [205]. While promising, these advances do not mitigate the fact that autologous transplants are incredibly expensive, and T1D is related to changes in multiple genes, so these single gene corrections could become overwhelming [206]. Due to this, experts in the field have chosen to predominantly focus on the potential of allogeneic SC-β cell transplants.

In allogeneic SC-β cell transplantation, the biggest obstacle to overcome is the alloimmune response. One of the most promising methods of preventing the alloimmune response is creating hypoimmune SC-β cells by knocking out the HLA Class I and II in iPSCs. Knocking out these HLA classes has been shown to work in iPSC-derived cardiomyocytes and could be a potential solution for multiple iPSC-derived cell therapies [207]. Briefly, HLA I and II are the major histocompatibility complexes (MHC) I and II, respectively, which are the cell surface markers that play key roles in adaptive immune cell activation. MHC I is present on all cell surfaces and is recognized by cytotoxic CD8⁺ T cells, while MHC II is only displayed on APCs and recruits helper CD4⁺ T cells that will recruit adaptive immune cells [207, 208]. To generate knockouts of HLA I and HLA II, CRISPR/Cas9 can be utilized to disrupt β-2 microglobulin (B2M) and Class II transactivator (CIITA), respectively, which will in turn reduce T cell activation [204]. Leite et al. demonstrated that by employing a B2M knockout, the activation of CD8⁺ T cells was greatly reduced [209]. Other studies went further by knocking out both B2M and CIITA, and delivering CD47—a cell surface marker that signals to macrophages to not be phagocytosed [210]— to create hypoimmune SC-β cells that could avoid the need for immunosuppression, survive and correct diabetes for 4 weeks in an immunocompetent, humanized mouse model, and the hypoimmune pluripotent stem cells could survive in the intramuscular region of rhesus macaques for 16 weeks, indicating significant potential for this technology [211, 212]. Furthermore, the addition of CD47 could potentially prevent IBMIR when the SC-β cells were infused through the intraportal vein, making this a key consideration if intraportal vein infusion remains a common transplantation method in this field [213]. The double knockout of HLA I and II continues to be one of the most promising solutions for evading alloimmune responses, as matching these two between patients is difficult, however more work will need to be done on optimizing the differentiation of knockout iPSC lines, as the differentiation parameters can vary a lot between different cell lines and each needs to be finetuned [200].

In addition to modifying HLA complexes in iPSCs, another direction of genetic engineering is to incorporate immune system markers into the iPSCs. Since 2012, it has been established that there are a few critical immune checkpoints in allorejection that need to be addressed, namely cytotoxic T lymphocyte antigen 4 (CTLA4) and PD-L1 [214]. Since the expression of these molecules could potentially avoid T cell alloreactivity, multiple studies have focused on engineering these markers onto transplanted cells as well as knocking out HLA classes. Castro-Gutierrez et al. demonstrated that when HLA I was knocked out and PD-L1 was overexpressed in SC-β cells, there was no demonstrable cytotoxic T cell activation [215]. Santini-González et al. corroborated these results by transplanting human SC-β cells in NOD mice that carried HLA I and demonstrated that while many T cells infiltrated to the transplant site, the SC-β cells overexpressing PD-L1 largely survived [216]. While these results were encouraging, it was also noted that the HLA I knockout and overexpression of PD-L1 would not avoid targeting from natural killer (NK) cells, so recent studies have also looked into genetically engineering iPSCs to express and secrete multiple cytokines and chemokines (i.e., IL-10, TGF-β, IL-2, and CXCL10), both to induce more tolerogenic environments and avoid the allorejection of the graft [217, 218].

6.3. Encapsulation of SC-β Cells

Encapsulation strategies for SC-β cells will need to provide immune protection and vascular support for the islets to remain viable and function properly after transplantation. Recent reports have indicated that when comparing encapsulated and non-encapsulated SC-β cells, the in vitro insulin secretion profiles and in vivo diabetes correction abilities were very similar [57]. From these results, we can understand that encapsulation of the SC-β cells will provide immunoprotection and not hamper the typical functions of islets. However, since the transplanted cells are usually still immature, the scaffolds in this scenario will also need to provide microenvironmental cues, and the grafts will need to be easily retrievable to prioritize the safety of the patient. In addition, multiple devices discussed below utilize cells that have not fully differentiated to SC-β cells, and the grafts will also need to be retrievable to assess the differentiating cell populations.

6.3.1. Microencapsulation of SC-β Cells

Several polymers have shown promising results in the maturation and function of SC-β cells for T1D therapy. Just as alginate has shown promise in the transplantation of porcine islets, it has also proven to be very successful in the transplantation of SC-β cells. Vegas et al. demonstrated that 1.5-millimeter alginate microcapsules loaded with SC-β cells and transplanted to the i.p. space of immune competent mice maintained normoglycemia for 100 days and elicited a reduced immune response compared to the other microcapsules tested[219]. Alagpulinsa et al. showed that when SC-β cells were encapsulated in alginate microcapsules along with CXCL12, an immunomodulatory cytokine, the graft could survive and correct diabetes in immune competent mice for over 150 days without immunosuppression[220]. While these studies show promise in the tunability and immune protection of alginate-based microcapsules, there have been concerns about the lack of retrievability of microcapsules in clinical settings to further examine the SC-β cells [76], and studies microencapsulating pancreatic endoderm cells resulted in a higher α-cell population rather than an insulin-producing beta cells population [202, 221]. Researchers often have wanted their grafts to be retrievable to assess maturation of the SC-β cells at some timepoint post-transplantation. As an alternative, macroencapsulation devices have been developed for the transplantation and retrieval of SC-β cells in pre-clinical and clinical studies.

6.3.2. Macroencapsulation ofSC-β Cells

Studies of SC-β cell transplantation have generally centered on macroencapsulation devices because evaluating the maturation into mature insulin-producing cells has been a critical marker of how effective the transplantation was as well as safety considerations in case the graft needs to be retrieved. Devices with fibrous membranes, with the aim of improving nutrient diffusion and reducing hypoxia, have been explored. For instance, Wang et al.—discussed earlier—reported that the nanofibrous TPSU shell + alginate core encapsulated with SC-β cells could maintain normoglycemia in immune deficient and immune competent mice for 120 and 60 days, respectively [108]. A separate study also used a fibrous structure but, in the reverse, in which nylon nanofibers were coated with a zwitterionically-modified alginate gel to prevent fibrotic reactions; this hydrogel allowed for straightforward mass transfer, low fibrotic responses, and the correction of hyperglycemia in SCID-beige mice for up to 238 days [222].

While these devices allow for sufficient nutrient transfer and retrievability in in vivo settings, electrospun nanofibers can have irregularities in pore size and fiber diameter, making them a more heterogeneous scaffold structure [223]. Formation of a more homogenous scaffold would greatly aid in ensuring the SC-β cells are receiving consistent microenvironmental cues and nutrients, and the cells all mature in similar ways. One way to obtain customized structures is through 3D printing to generate scaffolds for improved nutrient diffusion and immune cell blockage from SC-β cells [224]. Song et al. employed finite element modeling of cellular oxygen consumption to determine the SC-β cell cluster size that would not suffer from hypoxia and encapsulated those aggregates in a polylactic acid (PLA) + fibrin 3D-printed gel that remained stable in vivo for 12 weeks in the s.c. space [225]. More recently, Hwang et al. described a 3D printed polycaprolactone (PCL) macroporous capsule with decellularized porcine pancreas extracellular matrix (ECM) that was validated to maintain SC-β cells in culture and could be used in future in vivo experiments [226]. 3D printing could be a solution not only for incorporating different biomaterials into a single scaffold, but for engineering systems that can be specifically tuned for oxygen and insulin diffusion.

An active area of research in macroencapsulation has been investigating the potential of biomaterial scaffolds and the in vivo microenvironment to promote the differentiation of β-cells. The differentiation of hPSC-PPs into β-cells was observed when transplanting them under the k.c. of NSG mice [195], so efforts have centered on engineering biomaterial platforms that provide the microenvironmental cues for hPSC-PPs to differentiate into insulin-producing cells, thus avoiding the high variability that can come from differentiating in vitro. Studies in scaffolds that will support the differentiation of SC-β cells from PPs have spanned a variety of materials, including a collagen matrix with isolated microvessels [104], a microporous scaffold made of either poly(lactic-co-glycolide) (PLG) or PEG [227, 228], and a hybrid hydrogel platform combining amikacin hydrate and PEG diglycidyl ether (PEGDE) [229]. These scaffolds were able to support the differentiation of stem cell-derived PPs to insulin-producing islet-like cells, which could release insulin in response to glucose stimuli, deposit ECM on the surrounding scaffold, and could generate β-cells better than the traditional suspension cultures.

A heavily investigated material for macroencapsulation of SC-β cells is PTFE, and particularly the devices made by TheraCyte and ViaCyte. Bruin et al. assessed the differentiation of hESC-PPs to insulin-producing islet-like cells when transplanted either alone under the k.c. or in the TheraCyte device in the s.c. space. The results indicated that when transplanted to SCID-beige mice, both in vivo transplant sites were able to promote the differentiation of hESC-PPs, giving the first indication that using macroencapsulation devices with SC-β cells in the s.c. space is a feasible option for T1D therapy [230]. The TheraCyte device was also used in Kirk et al., which employed bioluminescent imaging to evaluate the maturation and containment of hESC-PPs in the device. Insulin was detected 7 weeks post-transplant, and it increased 17-fold over the course of 8 weeks, finally reversing diabetes at 20 weeks. Imaging of the device showed that the differentiated islet-like cells remained in the device for nearly 150 days, making it the longest tested period in murine models [231]. The Beta-O₂ device has also shown success in pre-clinical SC-β cell differentiation, in which hESC-PPs encapsulated in the β-Air were transplanted to the s.c. space of rats, and C-peptide could get detected as little as 1 week after transplant [232]. Finally, the Encaptra device by ViaCyte has also been tested in several pre-clinical models for the differentiation of pancreatic progenitor cells, and in a 50-week tracking period after s.c. implantation in mice, a functional β-cell mass was developed in 20 weeks, but it is worthy to note that there were more α-cells rather than β-cells in the final cell population [233]. While all these studies do show a lot of promise, there is a risk of teratoma formation when transplanting cells that are not fully differentiated, so this must be a consideration [234].

6.4. SC-β Cell Therapy in Clinical Trials

The clinical trials for SC-β cell therapy have shown significant potential for these cells as a T1D therapy, either with or without devices. ViaCyte’s development of the immune protective Encaptra (PEC-Encap, VC-01) device and the differentiation of the SC-derived pancreatic endoderm (PEC-01) population, led to the first clinical trials in 2014. The PEC-01 population did develop into endocrine cell populations, but there was excessive fibrosis around the device and no insulin secretion detected, leading to the end of that study [235, 236]. This device was modified to the PEC-Direct/VC-02 device, which allowed for more direct vascularization for the encapsulated cells, but with the condition that it would require a form of immunosuppression. Clinical trials for the PEC-Direct device demonstrated C-peptide production in response to glucose stimuli within 6-9 months post-transplant, and when explanted, most cells stained positive for endocrine markers, and the vessel allowed for vessel infiltration [237, 238]. While the VC-01 and VC-02 devices did not exhibit teratoma formation, the devices allowed the differentiation of PPs into different endocrine cell populations. Additionally, multiple regions around the device still had fibrous capsule deposition, affecting the diffusion of nutrients from the environment to the transplanted cells [239]. While the use of encapsulation devices for SC-β cell transplant is promising, there is room to improve particularly in ensuring the cells in the device become primarily endocrine and specifically insulin-producing cells and preventing fibrotic responses.

One direction where clinical trial strategy could go is by transplanting fully differentiated SC-β cells rather than SC-PPs, which has been successful in multiple transplant sites in NHPs [240, 241]. The most promising clinical trial thus far has been the Vertex VX-880 study, in which SC-β cells are infused through the intraportal vein under systemic immunosuppression [242]. Remarkably, 1 year after this trial began, two patients have reported better control of blood glucose levels, indicating further promise of SC-β cell transplantation [243]. However, being transplanted through the intraportal vein leaves the SC-β cells susceptible to similar issues as the Edmonton Protocol, namely coagulation cascades, the potential for cell mass loss, and the need for life-long immunosuppression. Considering Vertex’s recent acquisition of ViaCyte, there could be a development in the encapsulation of SC-β cells in immune protective devices for T1D therapy, and more clinical trials are expected as technologies develop over time [244].

7. Conclusion

Major strides have been made in the development of cell therapy for T1D. Cadaveric islets, porcine islets, and SC-β cells have been thoroughly investigated across several decades, and a variety of materials and techniques have been tested to encapsulate the cells to enhance their survival and function in the patient. While there has been immense progress in immunoisolation and immunomodulation of cadaveric islets, they are still plagued by the lack of sufficient donors and the inability to meet the demand for islet transplantation. Due to this, porcine islets or stem cell-derived islets could be cell-sourcing solutions. There needs to be further research done to mitigate allogeneic or xenogeneic immune responses, and to scale up the production of SC-β cells for further therapies. Additionally, a variety of encapsulation materials—both natural and synthetic biomaterials—have been tested and show potential across multiple pre-clinical and clinical trials but work still needs to be done on preventing fibrotic responses, improving vascular perfusion, recapitulating insulin dynamics with the transplanted cells, and avoiding the need for chronic immunosuppression. Therefore, biomaterial advances will be pivotal to the broad application of cell therapy in T1D.

Data Availability

No data was used for the literature discussed in this review.