Rebuilding a better home for transplanted islets

Abstract

Diabetes can be treated with β cell replacement therapy, where a patient is transplanted with cadaveric human islets to restore glycemic control. Despite this being an effective treatment, the process of isolating islets from the pancreas requires collagenase digestion which disrupts the islet extracellular matrix (ECM) and activates anoikis-mediated apoptosis. To improve islet survival in culture and after transplantation, the islet microenvironment may be enhanced with the addition of ECM components which are lost during isolation. Furthermore, novel β cell replacement strategies, such as stem cell-derived beta cell (SCβC) treatments or alternative transplant sites and devices, could benefit from a better understanding of how β cells interact with ECM. In this mini-review, we discuss the current understanding of the pancreas and islet ECM composition and review decellularization approaches to generate a native pancreatic ECM scaffold for use in both islet and SCβC culture and transplantation.

The extracellular matrix (ECM) is a network of proteins and polysaccharides which stimulates cells through structural and biochemical interactions. ECM molecules directly bind cell receptors, such as integrins, activating intracellular signaling cascades. The ECM can modulate signaling pathways by sequestering growth factors and affecting growth factor-receptor dynamics.¹ Through these mechanisms, ECM plays a significant role in cell health and identity. Because β cell replacement is used as a treatment for diabetic patients, understanding the role of ECM on islet and β cell survival and function could help optimize the treatment. In human islets, ECM has been found to affect β cell proliferation, differentiation, survival, and insulin secretion dynamics.² The culture of islets with purified matrix components has been shown to affect islet-specific gene expression.^3–5 Furthermore, the process of isolating islets from the pancreas involves collagenase digestion which destroys much of the native islet ECM, affecting islet health.^{6, 7}

Anoikis, meaning ‘without a home’, is an integrin-mediated form of apoptosis due to the absence of ECM in a tissue environment, and is a major contributor to cell death in isolated islets. High purity islet preps have high rates of apoptosis which can be reduced with anoikis inhibitors; impure preps, containing more intact ECM, have much better cell survival.^{8, 9} Islets cultured on purified ECM proteins: collagen (COL), laminins (LAM), and fibronectin (FN), have lower apoptosis rates and maintain better β cell function.^{10, 11} Soluble integrin-binding ligands, such as fibrinogen, RGD peptides and integrin antibodies, also reduce apoptosis in cultured islets.^{12, 13}

Islet ECM has been reported to contain COL I, III, IV, V and VI, as well as LAM and FN. Most studies established this compositional analysis through immunohistochemistry, and data originate from pancreata derived from a variety of different species.^{2, 14, 15} Because most native islet ECM is destroyed during isolation, it is difficult to use proteomic methods to further characterize islet-specific ECM.²

ECM can be isolated from tissues through a process called decellularization (decell), in which detergent-mediated cell lysis and subsequent washing removes cellular material, yielding native ECM. In this way, ECM has been extracted from many animal and human organs, and has been proposed for use in regenerative medicine for a variety of applications.^16–18 One approach is to use perfusion decell, by delivering detergents through the vasculature with the goal of preserving the macro- and microstructures of the organ.¹⁶ However, seeding cells or islets into a whole perfusion-decelled pancreas may be a challenge.^19–24 Another strategy is to solubilize the ECM with acidic pepsin digestion. The digested ECM can be neutralized and warmed to 37°C to form a hydrogel, or lyophilized to form a sponge.²⁵ These scaffolds require digestion and reformation of COL fibrils,²⁶ reorganizing the microstructure and potentially disrupting ECM protein functional domains. On the other hand, a hydrogel can be customized and easily incorporated into cell culture and transplantation models. The collagenous nature of native ECM hydrogel is advantageous for in vivo applications, where the neutralized digest can be injected as a liquid and form a gel at body temperature. Due to the potential challenges of seeding intact 3D scaffolds, hydrogel’s ease of use has practical advantages in tissue culture and transplantation platforms.

The pancreas of many species including mouse, rat, pig and human has been decellularized, generating pancreatic ECM (P-ECM). These studies provide clear evidence that pancreatic decell is achievable, resulting in a hypoimmunogenic matrix, free of DNA and cellular debris. The most common decell method utilized has been a perfusion approach, via vasculature or the pancreatic duct. Napierala et al. recently compared perfusion through the portal vein, aorta, or pancreatic duct of a porcine pancreas and found no significant difference in decell efficiency between the three routes.²¹ Detergents used have ranged from Triton X-100 to SDS to sodium deoxycholate, with no study specifically comparing the efficacy of these detergents on the pancreas, but with all studies having similar decell effectiveness. The length of detergent treatment appears to be dependent on species and organ size, with smaller organs (mouse, rat) requiring between 3–5 hours in detergent, and larger organs (pig, human) needing between 8 hours - 3 days.^{19–21, 23, 27–33} The detergent used and time exposed to detergent may affect the microstructure and function of retained matrix molecules,¹⁷ but has not been thoroughly studied for the pancreas.

Retention of ECM components has been assessed through a variety of methods, however few in-depth proteomic analyses of P-ECM composition have been published. Utilizing mass spectrometry-based proteomics with decelled mouse pancreas, Goh et al. identified 30 ECM-associated proteins in the P-ECM, including 17 COL and 7 LAM proteins.²⁹ Our group recently characterized the human P-ECM composition with quantitative mass spectrometry, identifying 120 ECM-associated proteins containing 33 COL proteins and 31 ECM glycoproteins, including 7 LAM proteins.³³ It may be important to expand this study to identify how decell treatments affect the P-ECM composition compared to native tissue, and how the P-ECM proteome varies among species. Utilizing human tissue with these techniques could benefit the understanding of the human pancreatic ECM composition, and create human-specific downstream applications of P-ECM.

The effect of P-ECM on cells and islets has been broadly investigated, but with diverse approaches and results. Cytocompatibility studies testing toxicity of the P-ECM on cell survival have indicated that the decelled matrix is not toxic to cells.^{19, 20, 29, 31}, ³³ When the MIN6 β cell line was perfused into the 3D mouse P-ECM, Goh et al. found cell survival and maintenance of insulin gene expression after 5 days.²⁹ Using stem cell-derived β cells (SCβCs), Wan et al. found that mouse SCβCs on rat P-ECM had a 2-fold increase in insulin gene expression compared to those without P-ECM.²³ Furthermore, Chaimov et al. found that hepatocytes transdifferentiated to β cells on P-ECM had an over 4-fold increase in insulin secretion.²⁷ When combined with islets, P-ECM has had variable effects. Two studies have found that islets cultured on or within P-ECM survived and functioned equally as well as islets in the absence of matrix, over a 4–5 day period.^{20, 31} Another group showed that islets within the 3D scaffolds could function, without a comparison to control islets.²¹ DeCarlo et al. performed a longer study in which rat islets cultured with rat P-ECM maintained insulin secretion up to 6 weeks, while naked islets began declining after 2 weeks.²⁸

Narayanan et al. recently published interesting results with decelled RIN5F (rat insulinoma) β cell cultures. Stem cells plated onto RIN5F-ECM and treated with RIN5F conditioned medium over 14 days differentiated into glucose-responsive insulin producing SCβCs, without a staged differentiation protocol or added growth factors.³⁴ This study mirrors other work demonstrating the instructive power ECM can have on cells,³⁵ and highlights the potential P-ECM may have on islets or SCβCs.

Few groups have explored the effects that P-ECM may have on transplanted islets. De Carlo et al. found that rat islets and rat P-ECM transplanted subcutaneously (SQ) within a PEG hydrogel reversed hyperglycemia up to 41 days in diabetic rats, while islets with P-ECM but no PEG only reversed hyperglycemia for a few days before returning to a diabetic state.²⁸ Chaimov et al. also transplanted islets and P-ECM hydrogel SQ within alginate capsules into diabetic mice. They observed a temporary drop in blood glucose, but not a long-term reversal of diabetes.²⁷ These are promising steps toward creating a competent SQ space for islet transplants, and P-ECM could play a role in enhancing the microenvironment of this space. More studies are required to assess the potential of P-ECM on transplanted islets or SCβCs, but based on the in vitro and limited in vivo studies, there may be real potential in improving long-term graft survival and function.

Following intrahepatic islet transplantation, a significant fraction of islet mass is lost after infusion due to instant blood-mediated inflammatory reaction (IBMIR) and microthrombosis, causing damage by a variety of postulated mechanisms.^{36, 37} Even with recent advances, only 44% of patients maintain insulin independence 3 years after infusion.³⁸ A better islet transplant site is needed to support islet survival and long-term reversal of hyperglycemia.³⁹ Many advances have been made in differentiating SCβCs,^40–44 however SCβC transplantation may encounter the same challenges that have beset cadaver islet transplant, unless a better strategy is identified.

Recent techniques being tested in clinical and experimental β cell replacement include exploration of novel transplant sites, such as the omentum,⁴⁵ vascularized SQ space,^46–49 and encapsulation strategies to protect islets from immune rejection.³⁹ Many of these approaches have been identified as having microenvironments that limit survival and engraftment of isolated cadaver islets. Thus, an improved microenvironment that supports robust islet survival and function is desirable^{39, 50} and could in part be enhanced by providing a matrix substrate such as P-ECM or a P-ECM hydrogel. As SCβCs undergo preclinical testing and enter clinical trials, the same challenges will need to be carefully considered about optimizing the environment of transplanted SCβCs to ensure long-term function and maintenance of β cell identity in vivo.

Abbreviations

COL

collagen

ECM

extracellular matrix

FN

fibronectin

LAM

laminin

P-ECM

pancreatic extracellular matrix

SCβC

stem cell-derived beta cell

SQ

subcutaneous