Sequestration and Synthesis: The Source of Insulin in Cell Clusters Differentiated from Murine Embryonic Stem Cells
http://www.100md.com
《干细胞学杂志》
Center for Biomedical Engineering, Brown University, Providence, Rhode Island, USA
Key Words. Diabetes ? Insulin ? Embryonic stem cell ? Glucose stimulation ? Immunostaining
Correspondence: Hyun Joon Paek, Ph.D., Center for Biomedical Engineering, Brown University, Providence, Rhode Island, USA. Telephone: 401-863-3262; Fax: 401-863-1753; e-mail: Hyun_Paek@brown.edu
ABSTRACT
The utility of pancreatic islet-like cells derived from embryonic stem cells (ESCs) in cell-replacement therapy for type 1 and type 2 diabetes remains controversial . Soria and colleagues originally derived insulin-producing cells from murine ESCs by using cell-trapping technology and demonstrated normalization of blood glucose in streptozotocin-induced diabetic mice. The following study by Lumelsky et al. , which manipulated culture conditions to derive insulin-releasing cell clusters (IRCCs) from murine ESCs without genetic modification, has been a subject of recent debate due to the presence of high-concentration exogenous insulin in the media used for the differentiation process and absence of pancreatic duodenal homeobox1 (Pdx-1) expression. In reproducing the protocol of Lumelsky et al. , Rajagopal et al. did not detect expression of messenger RNA (mRNA) for insulin 1 and detected only weak expression of insulin 2 mRNA. The absence of insulin 1 could be indicative of ectodermal origin of IRCCs rather than endodermal, for ectoderm expresses only insulin 2. IRCCs cultured with fluorescein isothiocyanate (FITC)–labeled insulin during differentiation showed positive immunostaining for insulin with a rhodamine-conjugated antibody. However, both FITC and rhodamine were colocalized, suggesting that the positive insulin staining observed was due to the presence of FITC-labeled insulin. To explain such results, Rajagopal et al. proposed that insulin secreted by IRCCs came from the media in which they had been cultured, not from de novo synthesis. In a separate series of experiments, Hori et al. maintained normoglycemia in streptozotocin-induced nonobese diabetic severe-combined immunodeficient mice over 3 weeks using implanted IRCCs. Glucose responsiveness of IRCCs was demonstrated through glucose tolerance testing of murine recipients. Hyperglycemia returned immediately after explantation of IRCCs . Insulin stored from sequestration of exogenous insulin is unlikely to lead to such observations, for secretion of exogenous insulin will diminish over time. More recently, Segev et al. demonstrated further differentiation and increased synthesis and secretion of insulin by IRCCs by adding a step of suspension culture at the end to the protocol of Lumelsky et al. . Furthermore, they detected expression of insulin mRNA, Pdx-1, and prohormone convertases by reverse transcription–polymerase chain reaction . Segev et al.’s results are thus consistent with de novo synthesis of insulin.
ESCs differentiation is a complex process that could be influenced in an unpredictable fashion by small changes in culture conditions . Nevertheless, the findings of different groups are difficult to reconcile. Resolution of these inconsistencies and varying perspectives over the source of insulin in IRCCs is all the more important given the current policy debates over the utility and ethics of ESC research.
MATERIALS AND METHODS
IRCCs were successfully derived using the protocol described above. The morphology of cells in each stage was similar to that reported by others.
Figure 1 shows the autoradiography print of the SDS-PAGE of insulin produced by islets and IRCCs cultured in medium containing 35S-cysteine. Tracks are shown for islet supernatant (lanes 1–3), IRCC supernatant (lanes 4–6), and IRCC lysates (lanes 7–9). Lanes 1, 4, and 7 represent total protein; lanes 2, 5, and 8 are from a precipitate of protein with rabbit anti-rat insulin antibody; and lanes 3, 6, and 9 show a control precipitate with a nonspecific primary antibody control, normal rabbit IgG. Autoradiography clearly showed bands for insulin and B-chain of insulin for rat islet-positive control (Fig. 1; lanes 1 and 2). Lane 3 shows bands in a high-molecular-weight range, which are a result of nonspecific binding with normal guinea pig IgG, although the identity of proteins is uncertain based only on their molecular weights. However, a combination of lanes 2 and 3 clearly demonstrates that insulin precipitation in lane 2 was not from nonspecific binding of normal rabbit IgG to insulin, because insulin bands from antibody-specific binding are not present in lane 3. In contrast, no detectable radiolabeled insulin was present in insulin secreted from IRCCs in either supernatant or the lysates (Fig. 1; lanes 4, 5, 7, and 8). Nonspecific binding of insulin to normal rabbit IgG is not observed in lanes 6 and 9. This is consistent with the position that IRCCs do not synthesize insulin de novo.
Figure 1. Autoradiography of immunoprecipitated insulin synthesized from 35S-radiolabeled cysteine. Islet supernatant (lanes 1–3), IRCCs supernatant (lanes 4–6), IRCCs lysates (lanes 7–9). Total protein secreted (lanes 1, 4, and 7) serves as a positive control. Immunoprecipitation using rabbit anti-rat insulin anti-body (lanes 2, 5, and 8) is the variable under study. Immunoprecipitation using normal rabbit IgG (lanes 3, 6, and 9) is a negative control. The absence of bands for insulin and B-chain of insulin in lanes 5 and 8, in contrast to lane 2, confirms the lack of de novo insulin synthesis. The bands and dark shades in lanes 1, 4 and 7 indicate the successful uptake of 35S-cysteine into the cells. Abbreviations: IgG, immunoglobulin G; IRCCs, insulin-releasing cell clusters.
Table 1 reports the quantities of insulin and C-peptide released along with predicted stoichiometric equivalence during glucose stimulation. C-peptide, released when proinsulin is cleaved to form insulin , is a convenient antigenic marker of insulin synthesis. In our hands, 10 IRCCs secreted approximately 9 ng of insulin during 1-hour stimulation in KRB buffer with 25 mM glucose, which was a rapid ninefold increase from the quantity of insulin secreted during 30-minute basal stimulation at 10 mM glucose. This is about 56% of the 16 ng released by 10 control islets and is comparable with quantities reported by Hori et al. . The relative increase in quantity of insulin secreted by IRCCs in 25 mM glucose buffer might have been even greater if the cells were incubated in a lower concentration of glucose than 10 mM for basal stimulation . Concentrations of C-peptide secreted along with insulin were measured by RIA with guinea pig anti-rat C-peptide antiserum. The quantity of C-peptide released along with the islets was only ~15% of the predicted equivalent, possibly due to the sensitivity of the RIA kit and the binding affinity of antibody used in this assay. The quantity of extracellular C-peptide secreted by IRCCs was <0.5% of the stoichiometric equivalent. These results are also consistent with release of stored insulin rather than de novo synthesis.
Table 1. Glucose-stimulated secretion of insulin (ELISA) and C-peptide (RIA)
As shown in Figure 2, immunostaining and immunogold labeling were positive for intracellular C-peptide in ~50% of IRCCs. For immunogold labeling, clusters of more than two gold dots suggested the presence of C-peptide within the cytoplasm of IRCCs. Positive immunostaining and immunogold labeling were observed in 50% of IRCCs (Figs. 2A, 2B). Controls did not show any nonspecific immunostaining and exhibited single dots from immunogold labeling in less than 5% of the cells (Figs. 2C–2F). The presence of intracellular C-peptide is consistent with de novo synthesis. Interestingly, transmission electron microscopy, used for immunogold labeling, also revealed that IRCCs lacked well-developed secretory granules compared with rat islet controls.
Figure 2. Immunostaining for intracellular C-peptide by (A) immunocytochemistry with guinea pig-derived anti-rat C-peptide primary antibody and (B) immunogold labeling using the same primary antibody (particles = 10 nm). Secondary antibody controls with either (C) rhodamine-conjugated goat anti-guinea pig IgG or (D) gold-labeled goat anti-guinea pig IgG. (E, F): Non-specific primary antibody control with normal guinea pig IgG and secondary antibodies. The orange color in (A) and dark gold dots (arrows) in (B) indicate the presence of at least some C-peptide in the IRCCs. Ultrastructure in (B) also reveals that IRCCs lack well-developed secretory granules. Abbreviations: C, cytosol; IgG, immunoglobulin G; IRCCs, insulin-releasing cell clusters; N, nucleus.
Table 2 contains the details of a mass-balance calculation for sequestered insulin. Assuming Fickian diffusion, the quantity of insulin in IRCCs fully equilibrated with media can readily be calculated as the product of the volume of a single IRCC and the maximum concentration of insulin in the media to which the cells are exposed. The volume of a single IRCC was 1.4 x 106 μm3 (1.4 x 10–6 ml) based on the assumption that an IRCC was hemiellipsoidal in shape with major and minor axes (r1 = 150 μm, r2 = 75 μm, r3 = 60 μm) obtained from optical microscopy. The total volume of 50 IRCCs that were used in glucose stimulation was 7 x 107 μm3 (7 x 10–5 ml). At 100% equilibration with an insulin concentration of 25 μg/ml, 50 IRCCs would contain 1.7 ng of insulin. However, the quantity of insulin released during glucose stimulation by 50 IRCCs was 48 ng (Table 1), 28 times greater than the amount they would contain if fully equilibrated with the highest concentration of insulin to which they had been exposed. If insulin was bound to intracellular macromolecules, and insulin is notoriously sticky, then concentration in the cell could exceed that in free solution. However, bound insulin is unlikely to be released upon glucose stimulation. This line of reasoning and our observation of the absence of well-developed secretory granules suggest that some mechanism other than sorption/release must be at play.
Table 2. Mass balance around IRCCs
DISCUSSION
In this study, radioisotopic labeling of insulin by IRCCs incubated in a medium containing 35S-cysteine was insignificant, as was extracellular secretion of C-peptide. However, IRCCs stained positive for intracellular C-peptide, and a mass balance suggested that sequestered insulin could not account for the total quantity of insulin secreted during glucose stimulation. We conclude that the sources of insulin secreted by IRCCs include both sequestration from media and de novo synthesis by IRCCs, and we speculate that the former phenomenon may actively suppress the latter.
ACKNOWLEDGMENTS
Lumelsky N, Blondel O, Laeng P et al. Differentiation of embryonic stem cells to insulin-secreting structures similar to pancreatic islets. Science 2001;292:1389–1394.
Rajagopal J, Anderson WJ, Kume S et al. Insulin staining of ES cell progeny from insulin uptake. Science 2003;299:363.
Hori Y, Rulifson IC, Tsai BC et al. Growth inhibitors promote differentiation of insulin-producing tissue from embryonic stem cells. Proc Natl Acad Sci U S A 2002;99:16105–16110.
Segev H, Fishman B, Ziskind A et al. Differentiation of human embryonic stem cells into insulin producing clusters. STEM CELLS 2004;22:265–274.
Sipione S, Eshpeter A, Lyon JG et al. Insulin expressing cells from differentiated embryonic stem cells are not beta cells. Diabetologia 2004;47:499–508.
Peck AB, Cornelius JG, Chaudhari M et al. Use of in vitro-generated stem cell-derived islets to cure type 1 diabetes: how close are we? Ann N Y Acad Sci 2002;958:59–68.
Kaczorowski D, Patterson ES, Jastromb WE et al. Glucose-responsive insulin-producing cells from stem cells. Diabetes Metab Res Rev 2002;18:442–450.
Hussain MA, Theise ND. Stem-cell therapy for diabetes mellitus. Lancet 2004;364:203–205.
Street CN, Sipione S, Helms L et al. Stem cell-based approaches to solving the problems of tissue supply for islet transplantation in type 1 diabetes. Int J Biochem Cell Biol 2004;36:667–683.
Weir GC, Bonner-Weir S. ?-cell precursor: a work in progress. Nat Biotechnol 2004;22:1095–1096.
Soria B, Roche E, Berna G et al. Insulin-secreting cells derived from embryonic stem cells normalize glycemia in streptozotocin-induced diabetic mice. Diabetes 2000;49:157–162.
Leon-Quinto T, Jones J, Skoudy A et al. In vitro directed differentiation of mouse embryonic stem cells into insulin-producing cells. Diabetologia 2004;47:1442–1451.
Halban P. Proinsulin processing in the regulated and the constitutive secretory pathway. Diabetologia 1994;37(suppl 2):S65–S72.(Hyun Joon Paek, Jeffrey R)
Key Words. Diabetes ? Insulin ? Embryonic stem cell ? Glucose stimulation ? Immunostaining
Correspondence: Hyun Joon Paek, Ph.D., Center for Biomedical Engineering, Brown University, Providence, Rhode Island, USA. Telephone: 401-863-3262; Fax: 401-863-1753; e-mail: Hyun_Paek@brown.edu
ABSTRACT
The utility of pancreatic islet-like cells derived from embryonic stem cells (ESCs) in cell-replacement therapy for type 1 and type 2 diabetes remains controversial . Soria and colleagues originally derived insulin-producing cells from murine ESCs by using cell-trapping technology and demonstrated normalization of blood glucose in streptozotocin-induced diabetic mice. The following study by Lumelsky et al. , which manipulated culture conditions to derive insulin-releasing cell clusters (IRCCs) from murine ESCs without genetic modification, has been a subject of recent debate due to the presence of high-concentration exogenous insulin in the media used for the differentiation process and absence of pancreatic duodenal homeobox1 (Pdx-1) expression. In reproducing the protocol of Lumelsky et al. , Rajagopal et al. did not detect expression of messenger RNA (mRNA) for insulin 1 and detected only weak expression of insulin 2 mRNA. The absence of insulin 1 could be indicative of ectodermal origin of IRCCs rather than endodermal, for ectoderm expresses only insulin 2. IRCCs cultured with fluorescein isothiocyanate (FITC)–labeled insulin during differentiation showed positive immunostaining for insulin with a rhodamine-conjugated antibody. However, both FITC and rhodamine were colocalized, suggesting that the positive insulin staining observed was due to the presence of FITC-labeled insulin. To explain such results, Rajagopal et al. proposed that insulin secreted by IRCCs came from the media in which they had been cultured, not from de novo synthesis. In a separate series of experiments, Hori et al. maintained normoglycemia in streptozotocin-induced nonobese diabetic severe-combined immunodeficient mice over 3 weeks using implanted IRCCs. Glucose responsiveness of IRCCs was demonstrated through glucose tolerance testing of murine recipients. Hyperglycemia returned immediately after explantation of IRCCs . Insulin stored from sequestration of exogenous insulin is unlikely to lead to such observations, for secretion of exogenous insulin will diminish over time. More recently, Segev et al. demonstrated further differentiation and increased synthesis and secretion of insulin by IRCCs by adding a step of suspension culture at the end to the protocol of Lumelsky et al. . Furthermore, they detected expression of insulin mRNA, Pdx-1, and prohormone convertases by reverse transcription–polymerase chain reaction . Segev et al.’s results are thus consistent with de novo synthesis of insulin.
ESCs differentiation is a complex process that could be influenced in an unpredictable fashion by small changes in culture conditions . Nevertheless, the findings of different groups are difficult to reconcile. Resolution of these inconsistencies and varying perspectives over the source of insulin in IRCCs is all the more important given the current policy debates over the utility and ethics of ESC research.
MATERIALS AND METHODS
IRCCs were successfully derived using the protocol described above. The morphology of cells in each stage was similar to that reported by others.
Figure 1 shows the autoradiography print of the SDS-PAGE of insulin produced by islets and IRCCs cultured in medium containing 35S-cysteine. Tracks are shown for islet supernatant (lanes 1–3), IRCC supernatant (lanes 4–6), and IRCC lysates (lanes 7–9). Lanes 1, 4, and 7 represent total protein; lanes 2, 5, and 8 are from a precipitate of protein with rabbit anti-rat insulin antibody; and lanes 3, 6, and 9 show a control precipitate with a nonspecific primary antibody control, normal rabbit IgG. Autoradiography clearly showed bands for insulin and B-chain of insulin for rat islet-positive control (Fig. 1; lanes 1 and 2). Lane 3 shows bands in a high-molecular-weight range, which are a result of nonspecific binding with normal guinea pig IgG, although the identity of proteins is uncertain based only on their molecular weights. However, a combination of lanes 2 and 3 clearly demonstrates that insulin precipitation in lane 2 was not from nonspecific binding of normal rabbit IgG to insulin, because insulin bands from antibody-specific binding are not present in lane 3. In contrast, no detectable radiolabeled insulin was present in insulin secreted from IRCCs in either supernatant or the lysates (Fig. 1; lanes 4, 5, 7, and 8). Nonspecific binding of insulin to normal rabbit IgG is not observed in lanes 6 and 9. This is consistent with the position that IRCCs do not synthesize insulin de novo.
Figure 1. Autoradiography of immunoprecipitated insulin synthesized from 35S-radiolabeled cysteine. Islet supernatant (lanes 1–3), IRCCs supernatant (lanes 4–6), IRCCs lysates (lanes 7–9). Total protein secreted (lanes 1, 4, and 7) serves as a positive control. Immunoprecipitation using rabbit anti-rat insulin anti-body (lanes 2, 5, and 8) is the variable under study. Immunoprecipitation using normal rabbit IgG (lanes 3, 6, and 9) is a negative control. The absence of bands for insulin and B-chain of insulin in lanes 5 and 8, in contrast to lane 2, confirms the lack of de novo insulin synthesis. The bands and dark shades in lanes 1, 4 and 7 indicate the successful uptake of 35S-cysteine into the cells. Abbreviations: IgG, immunoglobulin G; IRCCs, insulin-releasing cell clusters.
Table 1 reports the quantities of insulin and C-peptide released along with predicted stoichiometric equivalence during glucose stimulation. C-peptide, released when proinsulin is cleaved to form insulin , is a convenient antigenic marker of insulin synthesis. In our hands, 10 IRCCs secreted approximately 9 ng of insulin during 1-hour stimulation in KRB buffer with 25 mM glucose, which was a rapid ninefold increase from the quantity of insulin secreted during 30-minute basal stimulation at 10 mM glucose. This is about 56% of the 16 ng released by 10 control islets and is comparable with quantities reported by Hori et al. . The relative increase in quantity of insulin secreted by IRCCs in 25 mM glucose buffer might have been even greater if the cells were incubated in a lower concentration of glucose than 10 mM for basal stimulation . Concentrations of C-peptide secreted along with insulin were measured by RIA with guinea pig anti-rat C-peptide antiserum. The quantity of C-peptide released along with the islets was only ~15% of the predicted equivalent, possibly due to the sensitivity of the RIA kit and the binding affinity of antibody used in this assay. The quantity of extracellular C-peptide secreted by IRCCs was <0.5% of the stoichiometric equivalent. These results are also consistent with release of stored insulin rather than de novo synthesis.
Table 1. Glucose-stimulated secretion of insulin (ELISA) and C-peptide (RIA)
As shown in Figure 2, immunostaining and immunogold labeling were positive for intracellular C-peptide in ~50% of IRCCs. For immunogold labeling, clusters of more than two gold dots suggested the presence of C-peptide within the cytoplasm of IRCCs. Positive immunostaining and immunogold labeling were observed in 50% of IRCCs (Figs. 2A, 2B). Controls did not show any nonspecific immunostaining and exhibited single dots from immunogold labeling in less than 5% of the cells (Figs. 2C–2F). The presence of intracellular C-peptide is consistent with de novo synthesis. Interestingly, transmission electron microscopy, used for immunogold labeling, also revealed that IRCCs lacked well-developed secretory granules compared with rat islet controls.
Figure 2. Immunostaining for intracellular C-peptide by (A) immunocytochemistry with guinea pig-derived anti-rat C-peptide primary antibody and (B) immunogold labeling using the same primary antibody (particles = 10 nm). Secondary antibody controls with either (C) rhodamine-conjugated goat anti-guinea pig IgG or (D) gold-labeled goat anti-guinea pig IgG. (E, F): Non-specific primary antibody control with normal guinea pig IgG and secondary antibodies. The orange color in (A) and dark gold dots (arrows) in (B) indicate the presence of at least some C-peptide in the IRCCs. Ultrastructure in (B) also reveals that IRCCs lack well-developed secretory granules. Abbreviations: C, cytosol; IgG, immunoglobulin G; IRCCs, insulin-releasing cell clusters; N, nucleus.
Table 2 contains the details of a mass-balance calculation for sequestered insulin. Assuming Fickian diffusion, the quantity of insulin in IRCCs fully equilibrated with media can readily be calculated as the product of the volume of a single IRCC and the maximum concentration of insulin in the media to which the cells are exposed. The volume of a single IRCC was 1.4 x 106 μm3 (1.4 x 10–6 ml) based on the assumption that an IRCC was hemiellipsoidal in shape with major and minor axes (r1 = 150 μm, r2 = 75 μm, r3 = 60 μm) obtained from optical microscopy. The total volume of 50 IRCCs that were used in glucose stimulation was 7 x 107 μm3 (7 x 10–5 ml). At 100% equilibration with an insulin concentration of 25 μg/ml, 50 IRCCs would contain 1.7 ng of insulin. However, the quantity of insulin released during glucose stimulation by 50 IRCCs was 48 ng (Table 1), 28 times greater than the amount they would contain if fully equilibrated with the highest concentration of insulin to which they had been exposed. If insulin was bound to intracellular macromolecules, and insulin is notoriously sticky, then concentration in the cell could exceed that in free solution. However, bound insulin is unlikely to be released upon glucose stimulation. This line of reasoning and our observation of the absence of well-developed secretory granules suggest that some mechanism other than sorption/release must be at play.
Table 2. Mass balance around IRCCs
DISCUSSION
In this study, radioisotopic labeling of insulin by IRCCs incubated in a medium containing 35S-cysteine was insignificant, as was extracellular secretion of C-peptide. However, IRCCs stained positive for intracellular C-peptide, and a mass balance suggested that sequestered insulin could not account for the total quantity of insulin secreted during glucose stimulation. We conclude that the sources of insulin secreted by IRCCs include both sequestration from media and de novo synthesis by IRCCs, and we speculate that the former phenomenon may actively suppress the latter.
ACKNOWLEDGMENTS
Lumelsky N, Blondel O, Laeng P et al. Differentiation of embryonic stem cells to insulin-secreting structures similar to pancreatic islets. Science 2001;292:1389–1394.
Rajagopal J, Anderson WJ, Kume S et al. Insulin staining of ES cell progeny from insulin uptake. Science 2003;299:363.
Hori Y, Rulifson IC, Tsai BC et al. Growth inhibitors promote differentiation of insulin-producing tissue from embryonic stem cells. Proc Natl Acad Sci U S A 2002;99:16105–16110.
Segev H, Fishman B, Ziskind A et al. Differentiation of human embryonic stem cells into insulin producing clusters. STEM CELLS 2004;22:265–274.
Sipione S, Eshpeter A, Lyon JG et al. Insulin expressing cells from differentiated embryonic stem cells are not beta cells. Diabetologia 2004;47:499–508.
Peck AB, Cornelius JG, Chaudhari M et al. Use of in vitro-generated stem cell-derived islets to cure type 1 diabetes: how close are we? Ann N Y Acad Sci 2002;958:59–68.
Kaczorowski D, Patterson ES, Jastromb WE et al. Glucose-responsive insulin-producing cells from stem cells. Diabetes Metab Res Rev 2002;18:442–450.
Hussain MA, Theise ND. Stem-cell therapy for diabetes mellitus. Lancet 2004;364:203–205.
Street CN, Sipione S, Helms L et al. Stem cell-based approaches to solving the problems of tissue supply for islet transplantation in type 1 diabetes. Int J Biochem Cell Biol 2004;36:667–683.
Weir GC, Bonner-Weir S. ?-cell precursor: a work in progress. Nat Biotechnol 2004;22:1095–1096.
Soria B, Roche E, Berna G et al. Insulin-secreting cells derived from embryonic stem cells normalize glycemia in streptozotocin-induced diabetic mice. Diabetes 2000;49:157–162.
Leon-Quinto T, Jones J, Skoudy A et al. In vitro directed differentiation of mouse embryonic stem cells into insulin-producing cells. Diabetologia 2004;47:1442–1451.
Halban P. Proinsulin processing in the regulated and the constitutive secretory pathway. Diabetologia 1994;37(suppl 2):S65–S72.(Hyun Joon Paek, Jeffrey R)