As for the non-MSC compartment, HSCs are presently being used for the treatment of diabetes in clinical trials for diabetes, but from a different angle: the reeducation of the immune system. In clinical trials conducted on 15 newly diagnosed type I diabetic patients, HSCs were mobilized, collected from peripheral blood, and cryopreserved. Patients were then subjected to a very aggressive immunosuppressive regime prior to the reinfusion of their own stem cells, which were assumed to be at a stage prior to the onset of the disease. As a result of this treatment, 14 out of 15 patients became insulin-free for extended periods of time. Others are already transplanting hematopoietic cells directly into the pancreas of diabetic patients, but the claimed beneficial effects of this treatment on the overall glycemic control still need to be independently validated– and a mechanistic explanation investigated. As was the case with MSCs, since the hematopoietic compartment is derived from the mesoderm, there are doubts that HSCs may become true endodermal or ectodermal derivatives. To this date, there is no conclusive evidence in one way or another that transplanted HSCs that migrate to target tissues help in their regeneration by differentiation/ transdifferentiation, cell fusion, or simply by supporting endogenous regeneration through revascularization and/or as feeders.

The latest cell type to join our arsenal of adult stem cells is the amniotic fluid stem (AFS) cell. These cells are naturally shed by the embryo to the surrounding amniotic fluid, and can be easily cultured in vitro. In a recent breakthrough report, De Coppi et al. described a novel subpopulation of amniotic fluid cells with a multilineage differentiation potential spanning the three embryonal layers. These cells can be expanded almost indefinitely without loss of pluripotency, and they coexpress markers of both human embryonic stem cells (Oct3/4, SSEA-4, telomerase, and others) and MSCs (CD105, CD73, and CD90). Unlike huES cells, however, AFSs did not form teratomas in immunocompromised recipients. Their proven ability to generate endodermal derivatives bodes well for the field of pancreatic regeneration, although current efforts at differentiating these cells into pancreatic endocrine cell lineages have not been reported yet.

Transiently Immortalized Beta Cells

A potential alternative to the use of adult stem cells for directed differentiation into beta cells would be to define the conditions to expand fully differentiated beta cells.  we have already reviewed the literature on beta cell regeneration in vivo. However, in vitro “re-differentiation” of beta cells after de-differentiation and expansion phases was a less than convincing strategy, due to the fact that the end product could not stand comparison with the real beta cell. A different approach to the same end was presented in 2005 by Narushima et al., who transfected primary human beta cells with loxP-flanked sequences for the simian virus 40 large T antigen (SV40T) and the human telomerase reverse transcriptase (hTERT). These cells could be expanded without senescence for more than 50 passages in vitro. During the expansion phase, expression of insulin and other critical beta cell genes was downregulated. However, upon Cre-mediated “reversion,” the two immortalizing genes were removed and the cell adopted a quiescent, functional beta cell phenotype again. Reverted cells were able to permanently correct hyperglycemia in streptozotocin-treated SCID mice without any further replication in vivo . A complementary technology that could help the clinical translation of these findings is that of transducible TAT-Cre whose use in this context would preclude the use of viral vehicles to deliver the recombinase.

Transient immortalization of beta cells, as described by Narushima et al. A double round of retroviral transfection allows for the selection of cells that express both the hTERT and the SV40 large T antigen genes, which confer immortality. An EGFP marker can be used to sort these cells. This can be reverted by introducing the Cre recombinase, which will excise out the immortalizing genes in a process that will bring together a constitutive promoter and a neomycinselectable marker. These “reverted” cells can be selected using G418 and by sorting for EGFPnegative signal.

Mesenchymal stem cells • Cord blood stem cells • Amniotic fluid stem cells • Hematopoietic stem cells • Transient immortalization

The general notion of “adult stem cells” describes populations of cells with poorly defined characteristics that can be found in specific niches of most adult tissues. They are thought to be involved in the native turnover of their tissue, and occasionally show a great mobilization potential in response to insults, leading to regeneration. Unfortunately, our ability to study these cells in their native environment is very limited: we are biased by their apparent behavior after we isolate and expand them in vitro. It can be argued that the majority of the expandable “adult” stem cells described thus far are just the result of an adaptation to in vitro culture, which is probably a misleading indication of their actual nature and potential in vivo. Indeed, stem cell niches tend to be complex and maintain a delicate equilibrium between many different compartments. However, when we extract these cells from their native environment, we disrupt this complexity and favor the expansion of those that will better adapt to whichever culture conditions are set. Since researchers tend to favor attachment to plastic as a desirable characteristic (due to ease of cultivation and passage), perhaps it is not surprising that the end result, regardless of the tissue of origin, is usually quite similar: fibroblastic-like mesenchymal stem cells (MSCs) with a relatively limited differentiation potential and the ability to be propagated extensively without senescence. Only recently have we started to look into the possible in vitro replication of the natural complexity seen in stem cell niches. This would include not only the various participating cell types, but also the right combination of extracellular components. If successful, we might be able to accomplish the ultimate goal of expanding true, “ready-to-differentiate” tissue-specific stem cells – and not their surrogate, artifactual derivatives – for therapeutic purposes. This chapter is not intended as an exhaustive review of adult stem cells and their niches, particularly in view of the fact that very little, if anything, is known about the elusive pancreatic stem cell. Instead, we will describe the attempts at differentiating pancreatic cells from stem cells of “adult” (as opposed to embryonic) origin. In general, these cells will be characterized by some degree of in vitro expansion and differentiation potential, even if typically narrower than that of embryonic stem cells.

Mesenchymal Stem Cells

Most studies on beta cell differentiation from adult stem cells are based on MSCs. Originally identified in the bone marrow, these cells are multipotent, adhere to plastic, self-replicate for many passages, and can be easily derived from virtually any organ or tissue in the body, perhaps with the exception of peripheral blood. They have an elongated morphology resembling that of fibroblasts and have the potential to differentiate both in vitro and in vivo into a variety of connective fates, including adipose tissue, bone, and cartilage. At one point, it was thought that mesenchymal cells could be as pluripotent as their embryonic counterparts. The general view at this point is that, being mesodermal in origin, these cells might require more than simple soluble factor-driven protocols in order to force their differentiation into endodermal and ectodermal lineages. Earlier this decade, Verfaillie and collaborators isolated and characterized a subset of bone marrow-derived MSCs (multipotent adult progenitor cells [MAPCs]) that could have been an exception to the above rule. When injected into mouse blastocysts, these cells contributed extensively to all three germ layers in a manner consistent with that of a truly totipotent stem cell type. Other groups, however, have had difficulties replicating the methods originally described to establish these cell lines, and the technology has failed to become mainstream.

A colony of human MSCs derived from the umbilical cord blood . After passaging, the cells grow in a monolayer and adopt a mesenchymal-like morphology . Kindly provided by Prabakar Kamalaveni and Luca Inverardi

Due to an explosion on reports on MSC-like cells isolated from almost every possible tissue source, the International Society for Cellular Therapy established recently a set of minimal criteria for MSC standardization. These criteria are the following :

1. MSCs must adhere to plastic when maintained in standard culture conditions.
2. They must express the surface markers CD73, CD90, and CD105.
3. They must not express the surface markers CD34, CD45, CD14, CD11b, CD79 a , and CD19.
4. They must be negative for HLA-DR (major histocompatibility complex class II).
5. They must readily differentiate into osteoblasts, adipocytes, and chondroblasts following standard in vitro methods.

Interestingly, expression of the pluripotent Oct3/4 transcription factor has been occasionally observed in MSCs . However, the prevailing view is that Oct3/4 expression in adult tissues is not representative of pluripotency beyond the inner cell mass (ICM) stage, but rather an indication that this factor is not as embryonic-specific as originally thought.

Given their ease of differentiation toward connective phenotypes, MSCs are already in the pipeline for potential regenerative therapies involving the replacement of lost/damaged tissues such as the skin, the heart, and the gastrointestinal (GI) tract, among others. While we will now review the efforts made thus far at using these cells for beta cell differentiation, there is an increasingly widespread credence that the best use of MSCs in this context is not as building blocks, but rather as immunomodulatory and pro-angiogenic agents. Indeed, MSCs are well known for their ability to secrete many cytokines and growth factors, including vascular endothelial growth factor, brain-derived neurotrophic factor, nerve growth factor, basic fibroblast growth factor, insulin-like growth factor-1, hepatocyte growth factor (HGF), and others. In different contexts, these factors have been found to interact with local microenvironments and have anti-apoptotic, morphogenetic, mitogenic, and angiogenic effects that may favor engraftment and/or endogenous regeneration. A growing area of interest in the field of islet transplantation is that of cocultivation and/or cotransplantation of adult islet tissues with MSCs.

Signal-Driven Approaches

Many groups have attempted the directed differentiation of MSCs of various origins into beta cells by means of adding specific combinations of soluble factors to the culture medium. The general assumption has been that, since these cells are already committed along one of three major differentiation pathways (mesoderm), standard protocols for the differentiation of ES cells may not work as effectively. This, however, remains to be tested.

In 2004, Chen et al. treated rat MSCs with nicotinamide and beta-mercaptoethanol, obtaining “islet-like cluster cells” (ICCs), a term used by many people in the field to describe cellular aggregates (perhaps with the bias that this is how islets look in culture after isolation). When transplanted into syngeneic diabetic recipients, the authors observed a reduction in overall glycemic levels. One year later, Choi et al. reported robust expression of all pancreatic endocrine hormones, as well as glucoseregulated insulin secretion in vitro, after the exposure of bone marrow-derived rat MSCs to pancreatic extracts from rats subjected to partial (60%) pancreatectomy. The rationale behind these experiments, as described in the chapter “Stem Cell Differentiation: General Approaches,” is that pancreatic insults induce a regenerative response that might be accompanied by the release of islet-specific growth factors. Other protocols for the in vitro differentiation of MSCs include cultivation with HGF and activin A (which resulted in the expression of small levels of insulin from human MSCs, as well as glucagon, glucokinase, and a number of markers of early pancreatic development); the use of fibronectin and pellet suspension techniques; the coculture with pancreatic tissues in an artificial extracellular matrix; and a combination of conophylline and betacellulin (BTC)-delta4, which substantially accelerated the pancreatic differentiation of murine bone marrow-derived MSCs when compared with basal differentiation methods making use of standard activin A plus BTC. Cells differentiated in this way were able to reduce hyperglycemia in transplanted animals for up to 4 weeks, an outcome comparable to that resulting from transplanting rat MSCs treated with exendin-4 and nicotinamide.

Genetic Manipulation

As is the case for embryonic stem cells (see the chapter “Embryonic Stem Cells and Pancreatic Differentiation”) and transdifferentiation (see the chapter “Transdifferentiation”), the master pancreatic regulator Pdx1 remains a key player in all approaches involving genetic manipulation of MSCs. Thus, upon transfection of bone marrow-derived rat MSCs with a constitutively active Pdx1 cassette, Sun et al. observed expression of insulin, glucagon, and somatostatin. Transplantation of these cells into diabetic animals resulted in longer survival and maintenance of body weight, even if normoglycemia was not achieved. A similar approach was also used by Li et al. with human MSCs, where the Pdx1 gene was introduced via an adenoviral vector. Several relevant beta cell genes, including Ngn3, insulin, glucagon, glucokinase, and GLUT-2 were significantly up-regulated, and transduced cells exhibited a marginal ability to respond to glucose stimulation in vitro. Upon transplantation, these cells were able to reverse hyperglycemia in diabetic mice for up to 6 weeks. A triple adenovirus-mediated transfection with Pdx1, Hlxb9, and Foxa2 resulted in the detection of insulin in vitro, but with just a small proportion of cells expressing the fundamental component of the pro-insulinprocessing gene PC 1/3. Another group used retroviral vectors to ensure longterm expression of the Pdx1 gene in human bone marrow-derived MSCs, which led to the expression of all four pancreatic hormones but not NeuroD 619 (see the chapter “Pancreatic Development”). Despite the absence of this critical marker, these cells were able to regulate insulin secretion in vitro. In addition, when transplanted into diabetic animals, NeuroD was activated and a significant reduction in hyperglycemic levels was observed. More recently, Masaka et al. developed a chemical protocol based on the administration of HGF and FGF-4, which resulted in the generation of a bipotential, self-renewable “hepato-pancreatic” progenitor cell population. When Pdx1 was overexpressed in these cells, insulin was produced and an in vitro glucose-regulated response observed. In line with the above observations, the intrahepatic transplantation of mouse MSCs expressing a human ectopic insulin cassette reduced glucose levels for up to 6 weeks in diabetic recipients .