Diabetes mellitus has recently manifested a global trend in increased prevalence and is now a major public health problem around the world including in developing countries, such as China and India. It currently affects approximately 200 million people, and this number is likely to increase to 400 million by 2030 (Lock and Tzanakakis, 2007). Approximately 10% of these cases will have type 1 diabetes mellitus (T1D), caused by absolute deficiency of insulin-producing β cells. Therefore, a cellular therapy is the best prospect for a cure of T1D, provided autoimmunity to β cells can be controlled and there is a sufficient supply of insulin-secreting β cells for transplantation. However, the poor availability of donor islets has severely restricted the broad clinical use of islet transplantation. The lack of sufficient donor islets is why much attention has recently been paid to stem cells as a renewable source of β cells.
The term ‘stem cell’ was initially used in embryology in the late 19th century in the context of the origin of the blood system and gametes (Ramalho-Santos and Willenbring, 2007). Stem cells are undifferentiated cells that are capable of both self-renewal and giving rise to specialized functional cells. Depending on the developmental stages of their origin, stem cells can be divided into embryonic stem cells (derived from the inner cell mass of pre-implanted embryos) (Evans and Kaufman, 1981; Martin, 1981); epiblast stem cells (derived from post-implanted epiblast-stage embryos) (Brons et al, 2007; Tesar et al, 2007); germline-derived stem cells (derived from embryonic gonadial ridges or postnatal testes) (Shamblott et al, 1998; Kanatsu-Shinohara et al, 2004; Guan et al, 2006); induced pluripotent stem cells (from foetal or adult cells) (Takahashi and Yamanaka, 2006; Aoi et al, 2008; Hanna et al, 2008; Park et al, 2008) or adult stem cells (derived from postnatal tissues). Adult stem cells are a rare population in specific tissues but show powerful potential for regeneration. They can be further divided based on their tissue origin into a number of categories such as haematopoietic stem cells, neuronal stem cells, skin stem cells as well as mesenchymal stem cells. Unlike other tissue-specific stem cells, pancreatic stem cells (PSC) were proposed only relatively recently (Ramiya et al, 2000). However, despite intense research, the presence and origin of PSC are hotly debated. In order to understand the role and potential of PSC, better knowledge of pancreas development and function is required. In this review, we will particularly discuss several types of β-cell regeneration in physiological and pathophysiological conditions, and explore the mechanisms of regeneration of β cells. We hope that this will give readers a taste of this controversial but important area of research.
2. Pancreas development and physiology
The pancreas is an organ derived from endoderm. The endoderm is one of the three primitive germ layers formed during the early embryonic stage known as gastrulation. Taking the mouse as an example, the pancreas originates from the thickened endodermal epithelium along the dorsal and ventral surfaces of the posterior foregut. These thickenings can be identified histologically at embryonic day (E) 9.0-9.5 (Pictet et al, 1972). Subsequently, these epithelia evaginate into the surrounding mesoderm-derived mesenchymal tissue and form dorsal and ventral pancreatic buds. These buds continue to expand, branch and fuse as a result of gut rotation that brings the buds together. The fused developing pancreas continues to proliferate, differentiate and, ultimately, develop into the mature pancreas. The adult pancreas consists of digestive enzyme-secreting exocrine tissue, digestive enzyme-transporting ductal tissue and hormone-producing tissue located in the islets of Langerhans.
In humans, the dorsal bud can be detected as early as 26 days postcoitum (dpc, an equivalent stage to E9.5 mouse embryos), but insulin-positive cells are not visible until 52 dpc, approximately 2 weeks later than the equivalent stage seen in mice. The appearance of human insulin-positive cells precedes that of glucagon-positive cells at 8-10 weeks of development (Piper et al, 2004). All islet cells are detectable at the end of the first trimester in humans (Piper et al, 2004), but at later stages in mice (Herrera et al, 1991). These data indicate a human-mouse temporal difference in lineage development (Richardson et al, 1997), and this is supported by differences in gene expression patterns during developmental and disease processes in these two species (Fougerousse et al, 2000). More reviews of human pancreas development can be found elsewhere (De Krijger et al, 1992; Lukinius et al, 1992; Polak et al, 2000).
2.2. Pancreatic progenitors
In the thickened DE epithelium along the dorsal and ventral surfaces of the posterior foregut at E9.0-9.5, there are a group of cells express a parahox homeobox transcription factor (TF) gene termed Pdx1 (pancreas and duodenum TF 1). This is essential for pancreas expansion but not for pancreas initiation in early development and β-cell function in adults (Ohneda et al, 2000; Gu et al, 2002). The specification and differentiation of the various lineages in the pancreas along with transcriptional regulation has been extensively studied and reviewed (Wilson et al, 2003; Jensen, 2004; Servitja and Ferrer, 2004; Habener et al, 2005; Murtaugh, 2007; Best et al, 2008) and is briefly summarised in Figure 1.
Genetic lineage tracing experiment demonstrated that Pdx1-expressing (Pdx1+) cells are multipotent pancreatic progenitors because they give rise to exocrine, endocrine and duct tissues in the pancreas (Gu et al, 2002). These cells are located at the tip of the branching pancreatic tree marked by Pdx1+Ptf1a+(pancreas transcription factor 1a) Cpa1+(carboxypeptidase 1) (Zhou et al, 2007). To provide cues whether these cells are proliferative, transcriptome profiling analysis was performed showing that when DE cells commit to Pdx1+ pancreatic progenitors there are at least 28 out of 69 (40.6%) cell-cycle and cell-proliferation genes up-regulated (Figure 2) (Jiang et al, 2010).
Surprisingly, direct evidence of proliferation and self-renewal of these Pdx1+Ptf1a+Cpa1+ cells has not been produced. However, indirect evidence shows that these Pdx1+ cells may take up bromodeoxyuridine (BrdU), a thymidine analogue that may be incorporated into DNA during S-phase of the cell cycle, indicative of proliferation (Seymour et al, 2007). Unfortunately, due to the lack of a specific marker, this ability of purified Pdx1+ cells has not been examined in vitro either. Development of technologies that allow the demonstration of Pdx1+ cells to self-renew in vitro and to specify all pancreatic lineages will be not only important for pancreas developmental biology, but also crucial for a future ESC-derived cellular therapy for T1D. To serve this purpose, there are two ways to obtain purified Pdx1+ cells: from embryos in vivo or from directed ESC culture in vitro.
In humans, numerous PDX1+ cells can be easily detected in the pancreas between 8 and 21 weeks of age (Lyttle et al, 2008; Jeon et al, 2009). The number of PDX1+ cells colocalized with insulin or somatostain is progressively increasing during this period of development (Lyttle et al, 2008). Unfortunately again, these studies did not provide data to show whether the PDX1+ cells are generated by self-renewal themselves or commitment from their progenitors.
2.3. Islet progenitors
At around E9.5 in mice, a small group of cells in the thickened DE epithelium begin to express the basic helix-loop-helix TF neurogenin 3 (Ngn3) (Gradwohl et al, 2000; Gu et al, 2002; Xu et al, 2008). Accumulating evidence indicates that Ngn3-expressing (Ngn3+) cells in the pancreas are islet progenitors that give rise to all islet lineage cells because: (1) in Ngn3 knockout mice, islet cells do not develop (Gradwohl et al, 2000); (2) gene lineage tracing shows that Ngn3+ cells give rise to all pancreatic endocrine cells (Gu et al, 2002); (3) in adult pancreas, Ngn3+ cells can be activated by partial duct ligation and (4) after injection into a foetal pancreas in vitro, purified Ngn3+ cells differentiate into all islet cell types (Xu et al, 2008). Whereas mouse Ngn3 mRNA expression peaks around E15.5 (equivalent to weeks 7-8 in humans), human NGN3 expression was low prior to 9 weeks, but from 9 weeks onward, its expression increased sharply and remained high untill 17 weeks (Jeon et al, 2009). Although a few studies showed that Ngn3+ cells were proliferative (Jensen et al, 2000; Oliver-Krasinski et al, 2009), the recent genetic clonal analyses of mosaic analysis with double marker (MADM) demonstrated that Ngn3+ cells are quiescent and these cells give rise to a single islet cell type (Desgraz and Herrera, 2009). Consistent with this, a recent seminal study shows that the expression of Ngn3 inhibits proliferation by inducing cyclin-dependent kinase inhibitor 1a (Cdkn1a) (Miyatsuka et al, 2011). Our recent study demonstrate that in addition to induction of Cdkn1a, numerous genes involved in cell-cycle and cell-proliferation are down-regulated (95/127 or 74.8%) (Figure 3) (Jiang et al, 2010). Even though there is no literature reporting whether Ngn3+ cells proliferate in vitro, it is not likely that Ngn3+ cells are the so-called PSC.
The islets are composed mainly of α, β, δ, ε and PP cells (Figure 1) that secrete glucagon, insulin, somatostatin, ghrelin and pancreatic polypeptide respectively (Jorgensen et al, 2007). These hormones are generally responsible for the regulation of glucose homeostasis. For this function, there is a set of fine tuned paracrine interactions among these endocrine cells which is summarized in Figure 4. In adult humans, there are 2,000–3,000 β cells/islet of Langerhans, with approximately 1 million islets scattered throughout the pancreas (Stefan et al, 1982). The β cells sense the fluctuation of blood glucose levels and secrete insulin in a manner dependent on the glucose concentration. Insulin regulates circulating blood glucose concentrations through its actions on peripheral tissues, such as to inhibit hepatic glucose release and stimulate glucose uptake and storage by skeletal muscle and adipocyte tissue.
3. Regeneration of β cells occurs physiologically and pathophysiologically
β cells are indeed observed to be regenerated at least during pregnancy, partial pancreatectomy and obesity. These observations lead to the birth of the PSC concept (Bonner-Weir and Sharma, 2002). The existence of PSC is also inferred from the continued function of islets after transplantation (Ryan et al, 2002; Ryan et al, 2005). Since there is no convincing evidence of contribution of haematopoietic stem cells to islet cells (Wagers et al, 2002) nor that β cells are long-lived (Bonner-Weir, 2000), the continued function of transplanted islets suggests that PSC reside inside the islets and/or the functional β cells are capable of self-renewal.
3.1. Regeneration of β cells during pregnancy
To cope with physiological demand, pancreatic β cells do regenerate during pregnancy in humans and experimental animals. For example, the uptake of BrdU increases 3-fold at E10 and 10-fold at E14 in islets of pregnant rats (Parsons et al, 1992), providing indirect evidence that proliferation of islet cells contributes significantly to the increase of islet volume (Hellman, 1960; Van Assche, 1974). However, there only is a 2-fold increase in islet volume and a 3-fold increase in BrdU labelling at E15.5 maternal mouse islets (Karnik et al, 2007). This discrepancy may reflect a species difference in regeneration capacity or a difference in sensitivity of detection methods. During human pregnancy, both an increase in volume of maternal islets and hyperplasia of ‘β’ cells have also been observed (Van Assche et al, 1978), but direct evidence of proliferation in these islets is still lacking. In rodents and humans, the proliferation of β cells during pregnancy may be stimulated by prolactin and placental lactogens (Nielsen et al, 1999).
Recently, genetic studies provided molecular insights into how β-cell proliferation occurs during pregnancy. The pregnancy hormone prolactin suppresses the transcriptional co-activator menin, encoded by the gene multiple endocrine neoplasia type 1 (MEN1), resulting in β-cell proliferation. This is demonstrated by experiments in which a short infusion of prolactin is sufficient to reduce menin expression and stimulate proliferation of mouse islet cells (Karnik et al, 2007). In addition, the pregnancy hormones prolactin and placental lactogen also induce expression of tryptophan hydroxylase 1 (Tph1), encoding the enzyme Tph1 essential for serotonin production, which induces BrdU incorporation into isolated islet cells (Kim et al, 2010). However, it is still unknown whether proliferation comes from functional β cells and/or from PSC within islets. Future research should be directed to recapitulate this effect in vitro with purified subpopulations of islet cells, from which novel stimuli and molecular pathways may be identified. Defining these pathways may establish a platform on which novel strategies can be developed for a cure of T1D.
3.2. Regeneration of β cells during obesity
β cells regenerate in response to pathological processes such as obesity. For example, up to 10-fold increase in β-cell mass has been observed in obese rodents, responding to their insulin resistance (Butler et al, 2003a). Double staining of pancreas sections from obese mice and humans can detect insulin-producing cells that express Ki-67, a marker strictly associated with cell proliferation (Butler et al, 2003a; Butler et al, 2003b), indicating that regeneration may occur in the islets. Again, the regeneration capacity seems to be significantly greater in mice than in humans, although the underlying mechanism is unclear yet. Studies of one obese mutant mouse line, termed Ay, showed that reduction of menin contributes to adaptive β-cell proliferation (Karnik et al, 2007). Taken together, these data suggest that a similar mechanism for β-cell regeneration may operate in physiological pregnancy and pathophysiological obesity in mice. It would be very interesting to determine whether this mechanism is also at work during human pregnancy or obesity.
3.3. Regeneration of β cells after partial pancreatectomy
Like many other organs in the body, islets do regenerate in response to injury, in this case, pancreatectomy. In rats 4 weeks after 90% pancreatectomy, for example, there is a regeneration to 27% and 45% of sham-operated pancreas and islet mass, respectively (Bonner-Weir and Sharma, 2002). However, there are species differences in regeneration capacity. Even a 50% pancreatectomy in adult dogs would cause impaired fasting glucose in the short term (Matveyenko et al, 2006) and diabetes mellitus in the longer term (Stagner and Samols, 1991). Likewise, a 50% pancreatectomy in adult humans also leads to subsequent obesity and diabetes mellitus (Robertson et al, 2002). These studies again indicate that there is a difference between species in their capacity to mount islet regeneration: this is much more powerful in rodents than in larger mammals. Additional studies are needed to confirm this capacity difference and understand its underlying mechanism. Furthermore, it is less clear to what extent islet regeneration contributes to maintain β-cell mass in adult humans from existing islet cells and how much is from cells of any other origins. This knowledge is critical for a viable strategy to promote β-cell regeneration both in vivo and in vitro. Additionally, it is unknown whether the PSC also contribute to this type of regeneration. In the following sections, we will examine several possible mechanisms that may lead to β-cell generation and regeneration.
4. Evidence that β cells are generated from PSC
In addition to the observations of long-term survival of transplanted islets mentioned above, substantial in vitro evidence has indicated that “pluripotent PSC” may be present in all three major pancreas compartments, i.e. the ductal epithelium (Cornelius et al, 1997; Ramiya et al, 2000; Suzuki et al, 2002), islets, and acinar tissue (Zulewski et al, 2001; Seaberg et al, 2004). This evidence comes from studies of both rodent and human pancreas. For example, a potential PSC candidate has been purified by flow cytometry in the developing and adult mouse pancreas. These cells are identified by expression of the receptor for hepatocyte growth factor, c-Met, and absence of blood cell surface markers such as CD45, TER119, c-Kit, and Flk-1. These cells can differentiate into multiple pancreatic lineage cells from individual cells in vitro and give rise to pancreatic endocrine and acinar cells in vivo following transplantation (Suzuki et al, 2004). However, the in vivo localization and the molecular characteristics of these c-met expressing cells are largely unknown and clonogenesis at the single cell level has not been established.
4.1. PSC may house in ductal epithelium
Over the last several years, we investigated the differentiation and proliferation of foetal mouse pancreatic cells, believed to be a rich source for potential PSC. We first demonstrated in vitro that bone morphogenetic proteins-2, -4, -5 and -6, members of the transforming growth factor β superfamily, can promote the proliferation of pancreatic precursors and the development of pancreatic cystic epithelial colonies containing β cells (Jiang et al, 2002; Jiang and Harrison, 2005a), a process partially recapitulating an in vivo developmental stage. In addition, we also found that various isoforms of epidermal growth factors can stimulate colony formation (Jiang and Harrison, 2005b). These data indicate that extracellular signalling molecules, including various families of growth factors, modulate fate changes of pancreas precursor/stem cells. However, currently these colony-forming cells may at most be considered precursors because their self-renewal has not been demonstrated in vitro (F.X.J. unpublished observations).
Recently, in vitro and in vivo experiments have indicated that PSC are localized to the ductal epithelium. Bonner-Weir and colleagues were the first to report that adult human pancreatic ductal epithelial cells can form islet-like clusters and differentiate into insulin-secreting β cells (Bonner-Weir et al, 2000). Ramiya and colleagues reported that transplantation of in vitro generated islet-like structures from mouse PSC in the ductal epithelium can reverse diabetes mellitus (Ramiya et al, 2000). Another set of experiments used cultures of “pancreatic ductal cell aggregates” that were left over from pancreas digests after purification of human islets for transplantation. From these cultures, fibroblast-like cells grew out; these have been termed “pancreatic mesenchymal stem cells (pMSC)”. These cells can undergo at least 12 passages and express a range of bone marrow-derived MSC markers including CD13, CD29, CD44, CD54, CD105, α6 integrin subunit (also known as CD49f) and Thy1 (also known as CD90). These pMSC can give rise to cells of at least two germ layer origins including endoderm-derived cells, but not convincingly pancreatic lineage cells (Seeberger et al, 2006). In vivo, the large, small, and centrolobular ducts of the rat pancreas contain foci of cells that express the neural stem cell–specific marker, nestin, but do not express the ductal marker cytokeratin 19 (CK19) (Zulweski et al. 2001), suggestive of “islet progenitor” cells.
However, all the above studies have used mixed cell populations and have failed to demonstrate clonogenesis. Using culture conditions suitable for generating neurospheres ex vivo, mouse pancreatic ductal cells gave rise to neurosphere-like structures that can subsequently be differentiated into several types of islet cells including β cells (Seaberg et al, 2004). The molecular phenotype of progenitor cells for these islet cells remains unknown. On the other hand, after pancreatic duct ligation, numerous CK19+ ductal cells are regenerated and then Ngn3+ cells are observed, transplantation of the latter has further resulted in their differentiation into functional β cells (Xu et al, 2008), suggesting that the regeneration process may resemble that of embryonic pancreas development.
These studies did not explore the immediate origin of Ngn3+ cells in adult pancreas and whether the cells that give rise to these Ngn3+ cells possess PSC features. Recently, the use of the in vivo genetic tracing Cre-loxP system has generated further knowledge in this regard. In these experiments, Cre expression was directed by the promoter of carbonic anhydrase II, a marker of mature ductal cells, resulting in the excision of the stop cassette (Rosa-loxP-stop-loxP-lacZ) in the transgenic Rosa26 (R26R) mice. Therefore β-galactosidase activity is expressed in the cells that express Cre (ductal cells). At 4 weeks, β-galactosidase is detected in many ducts, patched acinar cells and 35-40% of islet cells (Bonner-Weir et al, 2008). These data provide direct evidence that adult ductal epithelial cells can give rise to islet cells, at least in mice. It is still unknown whether the carbonic anhydrase II-expressing cell population is homogenous or heterogenous, and whether this is a differentiation process from PSC or a transdifferentiation process from mature ductal cells. Anyway, these results should be repeated by independent laboratories and even in other experimental animals.
4.2. PSC may reside in the islets
Accumulated in vitro evidence suggests that a subpopulation of islet cells has PSC potential. A well-planned in vitro study found that a subset of unspecified cells in mouse islets and ducts named as pancreas-derived multipotent precursor cells give rise to neurosphere-like structures and subsequently could differentiate into islet β-like cells. This was a minor property, as most of these cells developed into neural lineage cells (Seaberg et al, 2004). The phenotype of these potential PSC cells was not defined.
In rat and human islets, a distinct population of nestin+ cells that do not express the hormones insulin, glucagon, somatostatin or pancreatic polypeptide has been identified. When cultured in vitro, these cells proliferate extensively (~8 months), give rise to cells that express liver and exocrine pancreas markers, such as α-fetoprotein and pancreatic amylase, and display a ductal/endocrine phenotype with expression of CK19, neural-specific cell adhesion molecule, insulin, glucagon, and PDX1. These nestin+ putative progenitor cells may therefore participate in the neogenesis of islet endocrine cells (Zulewski et al, 2001), mediated at least partially by glucagon-like peptide-1, an incretin hormone derived from processing proglucagon (Abraham et al, 2002). However, recent studies in vivo indicate that Nestin+ cells are mostly restricted in non-endodermal-derived cells (Lardon et al, 2002; Selander and Edlund, 2002). In addition, from donated adult human islets, outgrown fibroblast-like cells do not express hormones, but proliferate readily and give rise in vitro to hormone-expressing cell aggregates. However, these aggregates cannot be induced to become typical functional islet cells (Gershengorn et al, 2004). Furthermore, a rigorous genetic-based lineage tracing in mice under the control of Pdx1 or rat insulin promoter (RIP) demonstrated that neither PDX1- nor RIP-expressing cells contribute significantly to these fibroblast-like cells in vitro (Chase et al, 2007). In contrast, human islet RIP-expressing cells could be dedifferentiated to fibroblast-like cells in which insulin expression was not detectable and proliferate in vitro up to 16 population doublings (Russ et al, 2008). Whether this discrepancy is due to species differences remains to be clarified. Because of its critical importance, such tracing experiments should be reproduced by independent research groups. Further investigations are required to resolve the inconsistent results from the current studies.
4.3. PSC may locate in the exocrine tissue
In the clinic, a large population of nonendocrine pancreatic acinar cells would be discarded after purification of islets from donated pancreas for transplantation. The possibility of making use of these cells has attracted significant interest in recent years. After co-transplantation with foetal pancreatic cells under the kidney capsule of immunodeficient mice, these nonendocrine pancreatic epithelial cells have been shown to be capable of endocrine differentiation though without evidence of β-cell replication or cell fusion. These experiments suggest the existence of PSC or progenitor cells within the acinar compartment of the adult human pancreas (Hao et al, 2006). More recently, analysis using the Cre/loxP-based tracing system demonstrated that amylase/elastase-expressing acinar cells can give rise to insulin-positive cells in a suspension culture (Minami et al, 2005). However, because clonal assay of these amylase/elastase-expressing cells and their intermediate steps have not been investigated, this study may simply reveal that mouse and rat pancreatic acinar cells are able to transdifferentiate into surrogate insulin-expressing cells (Baeyens et al, 2005; Minami et al, 2008). This possibility was further supported by a recent study which shows that mouse acinar cells can be directly re-programmed in vivo to β-like cells with just three transcription factor genes, namely, Pdx1, Ngn3 and MafA (Zhou et al, 2008). On the other hand, similar lineage tracing experiments in vivo demonstrate that after 70-80% pancreatectomy, pre-existing mouse pancreatic acinar cells do not contribute to regeneration of islet β cells (Desai et al, 2007). The contradiction of in vitro and in vivo findings requires further reconciliation.
5. Evidence that islet β cells are capable of self-replication
There are several pieces of strong evidence demonstrating that islet β cells act as functional “stem” cells to reproduce themselves. Using RIP-driven reporter genes to genetically trace the fate of functional insulin-secreting cells, Dor and colleagues (Dor et al, 2004) first revealed that adult mouse pancreatic β cells are duplicated by RIP-expressing cells within the islets, either physiologically or after partial pancreatectomy. This study assumed that all RIP-expressing cells in adult islets are functional β cells and did not exclude the presence of PSC. Similarly, by using a transgenic model, in which the expression of diphtheria toxin was directed by RIP to β cells, diphtheria expression results in apoptosis of 70%-80% of β cells, destruction of islet architecture and, finally, diabetes mellitus. Withdrawal of diphtheria expression led to a significant regeneration of β-cell mass and a spontaneous normalization of blood glucose levels and islet architecture. Simultaneously, RIP-based lineage tracing analysis indicated that the proliferation of 20-30% surviving ‘β’ cells played a major role in this regeneration and in recovery of euglycemia (Nir et al, 2007).
Using the more sophisticated MADM system in mice known as RIP-CreER; Rosa26GR/ Rosa26RG, each RIP-expressing clone has been demonstrated to consist of 5.1±5.4 or 8.2±6.9 cells after one or two months of chase (Brennand et al, 2007). These RIP-expressing clones have been interpreted as further evidence of regeneration of functional β cells. An additional loss-of-function study following knockout of the Hnf4α (hepatocyte nuclear factor 4α) gene suggested that the β-cell regeneration may involve the Ras/Erk signalling cascade (Gupta et al, 2007) and ultimately be regulated by cycling modulators including cyclin D2 (Georgia and Bhushan, 2004). Taken together, further identification and characterization of the so-called self-replicative or dedifferentiative RIP-expressing cells both in vivo and in vitro will be urgently needed because they may hold the key for a regenerative therapy for type 1 diabetes mellitus.
Again using thymidine-based lineage tracing, β cells were demonstrated to be produced within an islet by rare self-renewing cells with a slow replication-refractory period (Teta et al, 2007), although the identity of these unique cells and the length of their replication-refractory period remain to be determined. The frequency of these self-renewal cells can be significantly increased after partial pancreatectomy or during pregnancy. Further studies should determine the molecular signature and biological potential of these replicating self-renewal cells. Because of ethical issues, similar studies cannot be performed in human islet tissues, but such investigation should at least be repeated in larger mammals.
Nevertheless, the β-cell population in the adult islets is in fact functionally heterogeneous (Heimberg et al, 1993; Pipeleers et al, 1994; Szabat et al, 2009). By using a dual fluorescence reporter mouse line, a few Ngn3+ cells in the developing pancreas have been observed to coexpress insulin (Hara et al, 2006). In humans a few NGN3+ cells in the foetal pancreas are also observed to coexpress insulin from 10 to 21 weeks of age (Lyttle et al, 2008). Consistent with these studies, the insulin gene expression has been detected from Pdx1+ progenitors through Ngn3+ cells during development to mature islet cells (Jiang et al, 2010) (Figure 5).
An early study indicated that the RIP-expressing cells in the developing pancreas gave rise to other islet cell types in addition to β cells (Alpert et al, 1988). The CD105+CD37+CD90+ mesenchymal stromal cells present in adult human islets have been shown to express a low level of insulin mRNA (Davani et al, 2007). After considering all these studies, therefore, lineage tracing studies under the control of other mature β-cell specific transcription factor gene promoters such as Pdx1 or MafA should be performed. Just like the β-cell line Min6 cells (Miyazaki et al, 1990), the glucose-responsive β cells in the islets may indeed duplicate themselves if the Pdx1- or MafA-expressing cells can be shown to be proliferative similar to the RIP-expressing cells.
6. The identity of PSC is inconclusive
Whereas investigation of β-cell duplication as a mechanism of islet regeneration has attracted great attention in recent years, much progress has been made to identify PSC. However, their identity is still not known. This is due, in addition to the knowledge gap that the signals required for late stage differentiation of functional β cells are largely unknown, at least partially to lack of the following factors: specific cell surface markers to characterize and purify cells that may have PSC potential; a simple, effective and reproducible in vitro assay to examine self-renewal and differentiation potential of a purified cell population; in vivo functional assays to determine biological function for both experimental animals and humans. Finally, well-agreed general criteria for defining a PSC are required.
7. Future directions for PSC research
We believe arguably that the successful establishment of in vitro culture and in vivo functional assay systems (Daniel et al, 1989; Dexter, 1989) facilitates the investigation of haematopoietic stem cells and subsequently leads to the application in clinics as a stem cell therapy. Therefore our future work should reach a consensus that PSC at least satisfy the following criteria: (1) clonogenesis should be demonstrated at the level of single cells sorted by flow cytometry; (2) sorted single cells should self-renew in vitro; (3) clonogenic cells should give rise in vitro to more than one specialized cell lineage and (4) these cells should differentiate in vivo into several functional cell types after transplantation. In the future, the investigation of PSC should be particularly encouraged based on three considerations. First, differentiation of embryonic stem cells into insulin-expressing cells was initially exciting and promising, but safety consideration aside, the generation of functional β cells proves extremely difficult and progress has been far slower than originally expected. PSC and dedifferentiation or transdifferentiation of other cell types in the pancreas may therefore provide an alternative renewable source of surrogate β cells. Second, β-cell duplication theory was mostly established from rodent models. It cannot completely exclude the existence of PSC, and it assumed that all insulin-expressing cells in the adult islets are uniformly functional β cells, but in fact these cells are heterogeneous. Last, as there is a significant difference in regeneration capacity between rodent and human islets, it may be wise not to directly extrapolate regeneration data from rodents to humans. Unlike previous studies that were only performed on in vivo or in vitro experiments without targeting specific cell types, future PSC work may need to employ integrated approaches, for example, applying cell surface markers to target particular cell populations; examining in vitro their potential for self-renewal and clonogenesis, using genetic approaches to dissect molecular mechanisms of their phenotypic changes in vivo and finally to examine their lineage contribution and biological function. Performing such integrated research is likely beyond the capacity of most individual laboratories, and therefore requires multi-discipline, multi-laboratory and even multi-national collaborations including the participation of the pharmaceutical sector to generate a therapeutic grade of surrogate β cells for a replacement therapy to diabetes mellitus.
8. Note to add
On the eve of submission, an important study from Dr Derek van der Kooy’s laboratory sheds a shiny ray of new light on this inconclusive topic. By using state-of-art genetic lineage tracing techniques, Smukler and colleagues now provide conclusive evidence that the PSC cells (originally called as pancreas-derived multipotent precursor cells) in the adult pancreas were derived from the embryonic pancreatic lineage, but not from the neural crest, as previously supposed. The PSC cells express insulin, along with an array of markers typical of islet progenitors; are distinct from mature functional ells; and give rise to non-beta cells in vivo. In addition to mouse PSC cells, human PSC cells have also been identified. After transplantation to diabetic mice, both mouse and human PSC cells could ameliorate their diabetes mellitus (Smukler et al, 2011).
The authors are supported by grants from Juvenile Diabetes Research Foundational International (4-2006-1025), the Diabetes Research Foundation of Western Australia, the University of Western Australia, and the Medical Research Foundation of Royal Perth Hospital.