- Email: [email protected]
Artificial Islets From Hybrid Spheroids of Three Pancreatic Cell Lines
Artificial Islets From Hybrid Spheroids of Three Pancreatic Cell Lines Y.H. Joa, I.J. Janga, J.G. Nemenoa, S. Leea, B.Y. Kima, B.M. Nama, W. Yanga, K.M. Leea, H. Kima, T. Takebeb,c,d, Y.S. Kime, and J.I. Leea,* a
Regenerative Medicine Laboratory, Center for Stem Cell Research, Department of Biomedical Science and Technology, Institute of Biomedical Science and Technology (IBST), Konkuk University, Seoul, Korea; bDepartment of Regenerative Medicine, Yokohama City University Graduate School of Medicine, Yokohama, Japan; cProject Leader of Advanced Medical Research Center, Yokohama City University, Yokohama, Japan; dPRESTO, Japan Science and Technology Agency, Kawaguchi, Japan; and eResearch Institute for Transplantation, Department of Transplantation Surgery, Yonsei University Health System, Seoul, Korea
ABSTRACT Pancreatic islets have been the focus of recent studies exploring the pathologic mechanisms of diabetes mellitus as well as more effective and radical treatments for this disease. Islet transplantation is a promising therapeutic strategy; however, isolation of pancreatic islets for this purpose has been challenging, because the technique is time consuming and technically difficult, and tissue handling can be variable. Pseudo-islets can be used as an alternative to naïve islets, but require cellular sources or artificial materials. In this study, pancreas-derived cells were used to generate pseudo-islets. Because the pancreas is composed of a variety of cell types, namely a cells, b cells, d cells, and other pancreatic cells that perform different functions, we used 3 different cell linesdNIT-1 (a b-cell line), a TC1 clone 6 (an a-cell line), and TGP52 (a pancreatic epithelial-like cell line)dwhich we cocultured in nonadhesive culture plates to produce hybrid cellular spheroids. These pseudo-islets had an oval shape and were morphologically similar to naïve islets; additionally, they expressed and secreted the pancreatic hormones insulin, glucagon, and somatostatin, as confirmed by reverse-transcription polymerase chain reaction and enzyme-linked immunosorbent assay. The results demonstrate that pseudo-islets that mimic naïve islets can be successfully generated by a coculture method. These artificial islets can potentially be used for in vitro tests related to diabetes mellitus, specifically, in drug discovery or for investigating pathology. Moreover, they can be useful for examining basic questions pertaining to cell-cell interactions and tissue development.
A
NIMAL testing involves using nonhuman animals for experiments that model processes occurring in humans. However, this approach can be controversial, owing in part to ethical issues regarding the use of large numbers of laboratory animals in experiments that often yield inconclusive results, thereby limiting their translational value. For these reasons, the so-called 3R principle (replacement, reduction, and refinement) was implemented in laboratory settings in 1959 [1]. This entails replacing live animals in experiments by adopting other strategies such as cell culture methods, reducing the number of animals used in an experiment, and refining methodology so that actual or potential pain and distress experienced by the animal can be minimized. Recently, the engineering of biomimetic tissues or organs has emerged as a promising alternative to animal testing [2,3]. For example, artificial islets have been biofabricated as a substitute for naïve islets to study pathologic processes and to establish 0041-1345/14/$esee front matter http://dx.doi.org/10.1016/j.transproceed.2013.11.074 1156
more effective and radical treatment for diabetes mellitus. Pancreatic islets consist of at least 5 different types of endocrine cells, including a, b, d, and ε cells, as well as pancreatic polypeptide (PP) cells, which secrete glucagon, insulin, somatostatin, ghrelin, and PP, respectively [4]. Type 1 diabetes mellitus results from the dysfunction of pancreatic b-cells [5]. Pancreatic islets can not be expanded using traditional cell culture techniques due to its three-dimensional (3D) Funding: Basic Science Research Program, through the National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science, and Technology (2010-0024188). Yong Hwa Jo and Ik Jin Jang contributed equally to this work and should be considered co-first authors. *Address correspondence to Jeong Ik Lee, DVM, PhD, DVetMedSci, DMed, Konkuk University, 120 Neungdong-ro (Hwayang-dong), Gwangjin-gu, Seoul 143-701, Korea. E-mail: [email protected]
Crown Copyright ª 2014 Published by Elsevier Inc. All rights reserved. 360 Park Avenue South, New York, NY 10010-1710 Transplantation Proceedings, 46, 1156e1160 (2014)
ARTIFICIAL ISLETS
1157
Fig 1. Pseudo-islet spheroids generated from pancreatic isletederived cell lines. Images of spheroids generated from cells seeded at densities of 3,000, 6,000, and 9,000 cells/well, obtained with the use of a light microscope with differential interference contrast optics. The pancreatic isletederived cell lines NIT-1, a-TC1 clone 6, and TGP52 were cocultured for 48 hours at ratios of 1:1:1 or 7:2:1, and each cell type cultured alone was used as a positive control group.
structure. A high-density suspension shaking (HDSS) culture method was developed that yields mixed cellular complexes containing 2 different cell types in nonadhesive plates. This method has enabled the production of 3-dimensional scaffoldfree cellular spheroids within a short culture period [6]. We have previously demonstrated disaggregation-expansionreaggregation technique in conjunction with HDSS to generate functional pseudo-islets that secrete insulin similar to those produced by naïve islets [7]. Given the physiologic and developmental link between the liver and pancreas, biomimetic tissues (eg, hybrid cellular spheroids) were produced from 2 commercial cell lines, ie, HepG2 cells (derived from hepatocytes) and RIN5F (derived from pancreatic islet cells) with the use of HDSS [8]. In addition, biomimetic pancreatic islets that are morphologically and physiologically similar to naïve islets were developed with the use of commercially available cell lines [9]. Current trends in diabetes mellitus prevalence indicate increasing morbidity and mortality rates worldwide. Therefore, there is an urgent need for a deeper understanding of the pathologic mechanisms underlying this disease, as well as for establishing more effective therapeutic strategies. In the present study, 3 different pancreatic isletederived cell linesda TC1 clone 6 (an a-cell line), NIT-1 (an insulinsecreting b-cell line), and TGP52 (a pancreatic epitheliallike cell line)dwere investigated as possible cell sources for the biofabrication of pseudo-islets with the use of a coculture method, as well as for their endocrine potential. MATERIALS AND METHODS Cell and Spheroid Culture Alpha-TC1 clone 6, NIT-1, and TGP52 cells originating from mouse pancreatic carcinomas (American Type Culture Collection, Manassas, Virginia, USA) were individually cultured or cocultured with the manufacturer’s recommended media incubated at 37 C with 5% CO2. These cell lines produce insulin, glucagon, and/or
somatostatin. The cells were seeded in 96-well nonadhesive Primesurface plates (Sumitomo Bakelite, Tokyo, Japan) at densities of 3,000, 6,000, and 9,000 cells per well.
Fluorescence Dye Staining NIT-1, a-TC1 clone 6, and TGP52 cells detached by trypsinization were labeled with Celltracker Green CMFDA (Green fluorescence; Invitrogen, Carlsbad, California, USA), Celltracker Red CMTPX (Red fluorescence; Invitrogen), and DAPI (4,6-diamidino2-phenylindole, blue fluorescence; Sigma-Aldrich, St Louis, Missouri, USA), respectively, according to the manufacturer’s instructions. Microphotographs of the biofabricated pseudo-islets were obtained with the use of Carl Zeiss Axio 200 invert fluorescence microscopy. Fluorescence images were gathered with emission fluorescence filtered by green, red, and blue emission.
Reverse-Transcription Polymerase Chain Reaction (RT-PCR) For each 20-mL PCR reaction, 1 mL cDNA template, equivalent to w100 pg total RNA, was mixed with Accupower Taq PCR Premix (Bioneer, Daejeon, Korea) and 5 pmol of the forward and reverse primers. The PCR conditions were as follows: 35 cycles of denaturation at 95 C for 30 seconds, annealing at 56 C for 30 seconds, and extension at 72 C for 30 seconds, followed by final extension at 72 C for 10 minutes. The primers were: insulin1 forward 50 -GTA ACC CCC AGC CCT TAG TG-30 and reverse 50 -GCA CTG ATC CAC AAT GCC AC30 ; insulin2 forward 50 -CAC CCA GGC TTT TGT CAA GC-30 and reverse 50 -GCT GGT GCA GCA CTG ATC TA-30 ; glucagon forward 50 -ACG CCC TTC AAG ACA CAG AG-30 and reverse 50 -GGC AAT GTT GTT CCG GTT CC-30 ; somatostatin forward 50 -TGA GGA CCT GCG ACT AGA CT-30 and reverse 50 -GGC TCC AGG GCA TCA TTC TC-30 ; and GAPDH forward 50 -CCG CAT CTT CTT GTG CAG TG-30 and reverse 50 -GAT GGG CTT CCC GTT GAT GA-30 .
Measurement of Insulin and Glucagon Levels To measure the levels of insulin and glucagon secreted by hybrid cellular pseudo-islets, the culture media were collected after a 2-day culture period. Insulin and glucagon levels in the supernate were
1158
JO, JANG, NEMENO ET AL
Fig 2. Distribution of constituent cell types within the pseudo-islet and spheroid size as a function of seeding density and cell type ratio. (A) Pseudo-islets were generated by coculturing NIT-1 (green), aTC1 clone 6 (red), and TGP52 (blue) cells for 48 hours at ratios of 1:1:1 and 7:2:1 and densities of 3,000, 6,000, and 9,000 cells/well. Spheroids were visualized after treatment with fluorescent dyes with the use of fluorescence microscopy. (B) Diameters of pseudo-islets were calculated as mean values of 12 samples. xP ¼ .0005, 1-way analysis of variance; *P ¼ .0009, 3,000 (1:1:1) vs 3,000 (7:2:1); zP ¼ .0001, 6,000 (1:1:1) vs 6,000 (7:2:1); KP ¼ .0001, 9,000 (1:1:1) vs 9,000 (7:2:1), Wilcoxon signed-rank test. measured with the use of an insulin enzyme-linked immunosorbent assay (ELISA) kit (Abcam, Cambridge, United Kingdom) and glucagon ELISA kit (R&D Systems, Minneapolis, Minnesota, USA), respectively, according to the manufacturers’ protocols.
correlations in insulin and glucagon levels among pseudo-islets. P values of <.05 were considered to be statistically significant.
RESULTS Statistical Analysis One-way analysis of variance (Graphpad Prism 5 for Windows; Graphpad Software, La Jolla, California, USA) was used to assess
Hybrid pseudo-islets were successfully produced from 3 different cell lines. NIT-1, a-TC1 clone 6, and TGP52 cells were cocultured at ratios of 1:1:1 and 7:2:1 on nonadhesive
ARTIFICIAL ISLETS
1159
amounts of glucagon than those seeded at 3,000 cells/well (Fig 4B), and this trend was observed regardless of cell type ratio.
DISCUSSION
Fig 3. mRNA expression levels for endocrine hormone genes in pseudo-islets. Insulin, glucagon, and somatostatin mRNA expression in pseudo-islets generated from NIT-1, a-TC1 clone 6, and TGP52 cell lines, cocultured at ratios of 1:1:1 and 7:2:1 and densities of 3,000, 6,000, and 9,000 cells/well, as analyzed by reverse-transcription polymerase chain reaction. GAPDH expression level was used as an internal control.
culture plates (Fig 1). Naïve pancreatic islets typically consist of 70% b-cells, 20% a-cells, and 10% other cell types [4]. Interestingly, in pseudo-islets formed from equal numbers of the 3 cell types, TGP52 and a-TC1 clone 6 cells localized to the center of the spheroids. In contrast, in pseudo-islets formed from a cell ratio of 7:2:1, TGP52 and a-TC1 clone 6 cells were uniformly distributed throughout the spheroid (Fig 2A). In addition, pseudo-islet size was proportional to the number of cells seeded, ie, larger spheroids were produced when a greater number of cells were seeded (Fig 2B). Pseudo-islet spheroids secreted insulin, glucagon, and somatostatin. RNA expression levels for these hormones in spheroids harvested after 2 days of culture, as well as protein levels in the culture media, were measured. Pseudoislets seeded at higher densities had higher levels of insulin1, glucagon, and somatostatin expression. However, there was no significant difference in insulin2 expression at different seeding densities (Fig 3). Interestingly, the expression of insulin and somatostatin, but not glucagon, varied as a function of cell type ratio (1:1:1 or 7:2:1). For example, higher levels of insulin were secreted by spheroids seeded at a higher density, but only when the constituent cell types were present at a 1:1:1 ratio (Fig 4A). Pseudo-islets seeded at 6,000 and 9,000 cells/well secreted correspondingly higher
Animal replacement strategies have significant benefits in tissue engineering, regenerative medicine, and most other areas of research. Aside from the ethical issues associated with the use of animals, research costs can be considerably reduced by using biomimetic tissues or organs instead of live animals. Products such as lung-, liver-, and heart-on-a-chip have recently been developed for drug screening [2,10,11]. However, these chips contain only 1 or 2 cell types, and therefore require further development to functionally mimic real organs. The human body contains w10 trillion cells, with 220 differentiated cell types [12]. The human endocrine system, encompassing many different organs and cell types, is among the most anatomically and physiologically complex of systems. The dysregulation of hormone secretion can lead to various diseases, including type 1 diabetes mellitus, which is characterized by the loss of functional b cells due to autoimmune responses, leading to the malfunction of pancreatic islets in the regulation of blood glucose levels [13]. The incidence of this disease is 8e17 per 100,000 individuals in Europe and the United States [14], making it a considerable public health challenge. Many animals, typically rodents, are sacrificed for the development of drugs and other treatment methods for diabetes [15]. The aim of the present study was to generate pseudo-islets that can mimic naïve islets. To achieve this end, 3 different pancreatic islet cell lines derived from a and b cells as well as pancreatic epithelial-like cells, were cocultured to yield pseudo-islet spheroids. Different cell type ratios were tested to determine the optimal combination that could recapitulate the morphologic and functional properties of naïve islets. The biofabrication of pancreatic islets has potential applications for diabetes treatment, because one currently employed therapeutic strategy involves the transplantation of pancreatic islets, which, once successfully incorporated by the recipient, can reestablish normal glucose regulation. Most research into treatment methods for diabetes has focused solely on insulin secretion while disregarding other glucose-regulating hormones. However, to understand the pathogenesis of diabetes, it is necessary to examine not only
Fig 4. Secretion of pancreatic hormones by pseudo-islets. Protein levels of (A) insulin and (B) glucagon secreted by pseudoislets generated from NIT-1, a-TC1 clone 6, and TGP52 cell lines cocultured at ratios of 1:1:1 and 7:2:1 and densities of 3,000, 6,000, and 9,000 cells/well, as analyzed by enzyme-linked immunosorbent assay.
1160
defective b cells, but also the functional interactions of other cells in the pancreas. The antagonistic actions of insulin and glucagon maintain homeostasis in the pancreas, and depend on the tight regulation of various pancreatic cell types, because glucagon deficiency can result in aberrant blood glucose levels even if insulin secretion is normal. Despite advances in the development of artificial organ chip systems, functional artificial endocrine tissues have yet to be developed. In the present study, pseudo-islets were evaluated in terms of endocrine secretion as a measure of their functional status. In spheroids formed from a 7:2:1 ratio of constituent cell types, a and b cells were randomly distributed (Fig 2), which was consistent with the morphology of naïve pancreatic islets [16]. At that ratio, insulin secretion was independent from seeding density (Fig 4), in contrast to spheroids formed from a 1:1:1 ratio, with which insulin and glucagon secretion was proportional to seeding density (P < .05). An appropriate cell composition is essential for making pseudo-islets that can functionally mimic naïve islets. The data from this study suggest that a 7:2:1 ratio is preferable to a 1:1:1 ratio, because hormone secretion did not require that a specific number of cells be present in the spheroids, at least in the case of insulin. It is not clear why this was not also true for glucagon, and future studies will determine whether other ratios besides 7:2:1 would be more suitable. In conclusion, mouse pseudo-islets that can mimic the secretory functions of naïve pancreatic islets were successfully generated; these can potentially be used as a substitute for live animals in studies investigating the pathogenesis of diabetes and potential treatment strategies for this disease. REFERENCES [1] Russell WMS, Burch RL. The principles of humane experimental technique. London: Methuen, 1959.
JO, JANG, NEMENO ET AL [2] Huh D, Matthews BD, Mammoto A, et al. Reconstituting organ-level lung functions on a chip. Science 2010;328:1662e8. [3] Huh D, Leslie DC, Matthews BD, et al. A human disease model of drug toxicityeinduced pulmonary edema in a lung-on-achip microdevice. Sci Transl Med 2012;4:159ra147. [4] Elayat AA, el-Naggar MM, Tahir M. An immunocytochemical and morphometric study of the rat pancreatic islets. J Anat 1995;186(Pt 3):629e37. [5] Acharjee S, Ghosh B, Al-Dhubiab BE, et al. Understanding type 1 diabetes: etiology and models. Can J Diabetes 2013;37: 269e76. [6] Lee JI, Sato M, Kim HW, et al. Transplantatation of scaffoldfree spheroids composed of synovium-derived cells and chondrocytes for the treatment of cartilage defects of the knee. Eur Cell Mater 2011;22:275e90. discussion 290. [7] Jeong JH, Lee JI, Ju MK, et al. Proliferation of pancreatic endocrine cells using disaggregation-expansion-reaggregation technology in isolated rat islets. Transplant Proc 2010;42:907e10. [8] Kim JY, Kim HW, Bae SJ, et al. Hybrid cellular spheroids from hepatocellular carcinoma and insulin-secreting cell lines. Transplant Proc 2012;44:1095e8. [9] Jo YH, Nam BM, Kim BY, et al. Pseudoislet of hybrid cellular spheroids from commercial cell lines. Transplant Proc 2013;45:3113e7. [10] Agarwal A, Goss JA, Cho A, et al. Microfluidic heart on a chip for higher throughput pharmacological studies. Lab Chip 2013;13:3599e608. [11] Prot JM, Bunescu A, Elena-Herrmann B, et al. Predictive toxicology using systemic biology and liver microfluidic “on chip” approaches: application to acetaminophen injury. Toxicol Appl Pharmacol 2012;259:270e80. [12] Nicholas W. In tiny cells, glimpses of body’s master plan. New York Times, December 18, 2001. [13] Cooke DW, Plotnick L. Type 1 diabetes mellitus in pediatrics. Pediatr Rev 2008;29:374e84. [14] Department of Health and Human Services. National diabetes statistics. Maryland, 2011. [15] National Statistics. Statistics of scientific procedures on living animals. London: 2009. [16] Pechhold K, Zhu X, Harrison VS, et al. Dynamic changes in pancreatic endocrine cell abundance, distribution, and function in antigen-induced and spontaneous autoimmune diabetes. Diabetes 2009;58:1175e84.
Regenerative Medicine Laboratory, Center for Stem Cell Research, Department of Biomedical Science and Technology, Institute of Biomedical Science and Technology (IBST), Konkuk University, Seoul, Korea; bDepartment of Regenerative Medicine, Yokohama City University Graduate School of Medicine, Yokohama, Japan; cProject Leader of Advanced Medical Research Center, Yokohama City University, Yokohama, Japan; dPRESTO, Japan Science and Technology Agency, Kawaguchi, Japan; and eResearch Institute for Transplantation, Department of Transplantation Surgery, Yonsei University Health System, Seoul, Korea
ABSTRACT Pancreatic islets have been the focus of recent studies exploring the pathologic mechanisms of diabetes mellitus as well as more effective and radical treatments for this disease. Islet transplantation is a promising therapeutic strategy; however, isolation of pancreatic islets for this purpose has been challenging, because the technique is time consuming and technically difficult, and tissue handling can be variable. Pseudo-islets can be used as an alternative to naïve islets, but require cellular sources or artificial materials. In this study, pancreas-derived cells were used to generate pseudo-islets. Because the pancreas is composed of a variety of cell types, namely a cells, b cells, d cells, and other pancreatic cells that perform different functions, we used 3 different cell linesdNIT-1 (a b-cell line), a TC1 clone 6 (an a-cell line), and TGP52 (a pancreatic epithelial-like cell line)dwhich we cocultured in nonadhesive culture plates to produce hybrid cellular spheroids. These pseudo-islets had an oval shape and were morphologically similar to naïve islets; additionally, they expressed and secreted the pancreatic hormones insulin, glucagon, and somatostatin, as confirmed by reverse-transcription polymerase chain reaction and enzyme-linked immunosorbent assay. The results demonstrate that pseudo-islets that mimic naïve islets can be successfully generated by a coculture method. These artificial islets can potentially be used for in vitro tests related to diabetes mellitus, specifically, in drug discovery or for investigating pathology. Moreover, they can be useful for examining basic questions pertaining to cell-cell interactions and tissue development.
A
NIMAL testing involves using nonhuman animals for experiments that model processes occurring in humans. However, this approach can be controversial, owing in part to ethical issues regarding the use of large numbers of laboratory animals in experiments that often yield inconclusive results, thereby limiting their translational value. For these reasons, the so-called 3R principle (replacement, reduction, and refinement) was implemented in laboratory settings in 1959 [1]. This entails replacing live animals in experiments by adopting other strategies such as cell culture methods, reducing the number of animals used in an experiment, and refining methodology so that actual or potential pain and distress experienced by the animal can be minimized. Recently, the engineering of biomimetic tissues or organs has emerged as a promising alternative to animal testing [2,3]. For example, artificial islets have been biofabricated as a substitute for naïve islets to study pathologic processes and to establish 0041-1345/14/$esee front matter http://dx.doi.org/10.1016/j.transproceed.2013.11.074 1156
more effective and radical treatment for diabetes mellitus. Pancreatic islets consist of at least 5 different types of endocrine cells, including a, b, d, and ε cells, as well as pancreatic polypeptide (PP) cells, which secrete glucagon, insulin, somatostatin, ghrelin, and PP, respectively [4]. Type 1 diabetes mellitus results from the dysfunction of pancreatic b-cells [5]. Pancreatic islets can not be expanded using traditional cell culture techniques due to its three-dimensional (3D) Funding: Basic Science Research Program, through the National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science, and Technology (2010-0024188). Yong Hwa Jo and Ik Jin Jang contributed equally to this work and should be considered co-first authors. *Address correspondence to Jeong Ik Lee, DVM, PhD, DVetMedSci, DMed, Konkuk University, 120 Neungdong-ro (Hwayang-dong), Gwangjin-gu, Seoul 143-701, Korea. E-mail: [email protected]
Crown Copyright ª 2014 Published by Elsevier Inc. All rights reserved. 360 Park Avenue South, New York, NY 10010-1710 Transplantation Proceedings, 46, 1156e1160 (2014)
ARTIFICIAL ISLETS
1157
Fig 1. Pseudo-islet spheroids generated from pancreatic isletederived cell lines. Images of spheroids generated from cells seeded at densities of 3,000, 6,000, and 9,000 cells/well, obtained with the use of a light microscope with differential interference contrast optics. The pancreatic isletederived cell lines NIT-1, a-TC1 clone 6, and TGP52 were cocultured for 48 hours at ratios of 1:1:1 or 7:2:1, and each cell type cultured alone was used as a positive control group.
structure. A high-density suspension shaking (HDSS) culture method was developed that yields mixed cellular complexes containing 2 different cell types in nonadhesive plates. This method has enabled the production of 3-dimensional scaffoldfree cellular spheroids within a short culture period [6]. We have previously demonstrated disaggregation-expansionreaggregation technique in conjunction with HDSS to generate functional pseudo-islets that secrete insulin similar to those produced by naïve islets [7]. Given the physiologic and developmental link between the liver and pancreas, biomimetic tissues (eg, hybrid cellular spheroids) were produced from 2 commercial cell lines, ie, HepG2 cells (derived from hepatocytes) and RIN5F (derived from pancreatic islet cells) with the use of HDSS [8]. In addition, biomimetic pancreatic islets that are morphologically and physiologically similar to naïve islets were developed with the use of commercially available cell lines [9]. Current trends in diabetes mellitus prevalence indicate increasing morbidity and mortality rates worldwide. Therefore, there is an urgent need for a deeper understanding of the pathologic mechanisms underlying this disease, as well as for establishing more effective therapeutic strategies. In the present study, 3 different pancreatic isletederived cell linesda TC1 clone 6 (an a-cell line), NIT-1 (an insulinsecreting b-cell line), and TGP52 (a pancreatic epitheliallike cell line)dwere investigated as possible cell sources for the biofabrication of pseudo-islets with the use of a coculture method, as well as for their endocrine potential. MATERIALS AND METHODS Cell and Spheroid Culture Alpha-TC1 clone 6, NIT-1, and TGP52 cells originating from mouse pancreatic carcinomas (American Type Culture Collection, Manassas, Virginia, USA) were individually cultured or cocultured with the manufacturer’s recommended media incubated at 37 C with 5% CO2. These cell lines produce insulin, glucagon, and/or
somatostatin. The cells were seeded in 96-well nonadhesive Primesurface plates (Sumitomo Bakelite, Tokyo, Japan) at densities of 3,000, 6,000, and 9,000 cells per well.
Fluorescence Dye Staining NIT-1, a-TC1 clone 6, and TGP52 cells detached by trypsinization were labeled with Celltracker Green CMFDA (Green fluorescence; Invitrogen, Carlsbad, California, USA), Celltracker Red CMTPX (Red fluorescence; Invitrogen), and DAPI (4,6-diamidino2-phenylindole, blue fluorescence; Sigma-Aldrich, St Louis, Missouri, USA), respectively, according to the manufacturer’s instructions. Microphotographs of the biofabricated pseudo-islets were obtained with the use of Carl Zeiss Axio 200 invert fluorescence microscopy. Fluorescence images were gathered with emission fluorescence filtered by green, red, and blue emission.
Reverse-Transcription Polymerase Chain Reaction (RT-PCR) For each 20-mL PCR reaction, 1 mL cDNA template, equivalent to w100 pg total RNA, was mixed with Accupower Taq PCR Premix (Bioneer, Daejeon, Korea) and 5 pmol of the forward and reverse primers. The PCR conditions were as follows: 35 cycles of denaturation at 95 C for 30 seconds, annealing at 56 C for 30 seconds, and extension at 72 C for 30 seconds, followed by final extension at 72 C for 10 minutes. The primers were: insulin1 forward 50 -GTA ACC CCC AGC CCT TAG TG-30 and reverse 50 -GCA CTG ATC CAC AAT GCC AC30 ; insulin2 forward 50 -CAC CCA GGC TTT TGT CAA GC-30 and reverse 50 -GCT GGT GCA GCA CTG ATC TA-30 ; glucagon forward 50 -ACG CCC TTC AAG ACA CAG AG-30 and reverse 50 -GGC AAT GTT GTT CCG GTT CC-30 ; somatostatin forward 50 -TGA GGA CCT GCG ACT AGA CT-30 and reverse 50 -GGC TCC AGG GCA TCA TTC TC-30 ; and GAPDH forward 50 -CCG CAT CTT CTT GTG CAG TG-30 and reverse 50 -GAT GGG CTT CCC GTT GAT GA-30 .
Measurement of Insulin and Glucagon Levels To measure the levels of insulin and glucagon secreted by hybrid cellular pseudo-islets, the culture media were collected after a 2-day culture period. Insulin and glucagon levels in the supernate were
1158
JO, JANG, NEMENO ET AL
Fig 2. Distribution of constituent cell types within the pseudo-islet and spheroid size as a function of seeding density and cell type ratio. (A) Pseudo-islets were generated by coculturing NIT-1 (green), aTC1 clone 6 (red), and TGP52 (blue) cells for 48 hours at ratios of 1:1:1 and 7:2:1 and densities of 3,000, 6,000, and 9,000 cells/well. Spheroids were visualized after treatment with fluorescent dyes with the use of fluorescence microscopy. (B) Diameters of pseudo-islets were calculated as mean values of 12 samples. xP ¼ .0005, 1-way analysis of variance; *P ¼ .0009, 3,000 (1:1:1) vs 3,000 (7:2:1); zP ¼ .0001, 6,000 (1:1:1) vs 6,000 (7:2:1); KP ¼ .0001, 9,000 (1:1:1) vs 9,000 (7:2:1), Wilcoxon signed-rank test. measured with the use of an insulin enzyme-linked immunosorbent assay (ELISA) kit (Abcam, Cambridge, United Kingdom) and glucagon ELISA kit (R&D Systems, Minneapolis, Minnesota, USA), respectively, according to the manufacturers’ protocols.
correlations in insulin and glucagon levels among pseudo-islets. P values of <.05 were considered to be statistically significant.
RESULTS Statistical Analysis One-way analysis of variance (Graphpad Prism 5 for Windows; Graphpad Software, La Jolla, California, USA) was used to assess
Hybrid pseudo-islets were successfully produced from 3 different cell lines. NIT-1, a-TC1 clone 6, and TGP52 cells were cocultured at ratios of 1:1:1 and 7:2:1 on nonadhesive
ARTIFICIAL ISLETS
1159
amounts of glucagon than those seeded at 3,000 cells/well (Fig 4B), and this trend was observed regardless of cell type ratio.
DISCUSSION
Fig 3. mRNA expression levels for endocrine hormone genes in pseudo-islets. Insulin, glucagon, and somatostatin mRNA expression in pseudo-islets generated from NIT-1, a-TC1 clone 6, and TGP52 cell lines, cocultured at ratios of 1:1:1 and 7:2:1 and densities of 3,000, 6,000, and 9,000 cells/well, as analyzed by reverse-transcription polymerase chain reaction. GAPDH expression level was used as an internal control.
culture plates (Fig 1). Naïve pancreatic islets typically consist of 70% b-cells, 20% a-cells, and 10% other cell types [4]. Interestingly, in pseudo-islets formed from equal numbers of the 3 cell types, TGP52 and a-TC1 clone 6 cells localized to the center of the spheroids. In contrast, in pseudo-islets formed from a cell ratio of 7:2:1, TGP52 and a-TC1 clone 6 cells were uniformly distributed throughout the spheroid (Fig 2A). In addition, pseudo-islet size was proportional to the number of cells seeded, ie, larger spheroids were produced when a greater number of cells were seeded (Fig 2B). Pseudo-islet spheroids secreted insulin, glucagon, and somatostatin. RNA expression levels for these hormones in spheroids harvested after 2 days of culture, as well as protein levels in the culture media, were measured. Pseudoislets seeded at higher densities had higher levels of insulin1, glucagon, and somatostatin expression. However, there was no significant difference in insulin2 expression at different seeding densities (Fig 3). Interestingly, the expression of insulin and somatostatin, but not glucagon, varied as a function of cell type ratio (1:1:1 or 7:2:1). For example, higher levels of insulin were secreted by spheroids seeded at a higher density, but only when the constituent cell types were present at a 1:1:1 ratio (Fig 4A). Pseudo-islets seeded at 6,000 and 9,000 cells/well secreted correspondingly higher
Animal replacement strategies have significant benefits in tissue engineering, regenerative medicine, and most other areas of research. Aside from the ethical issues associated with the use of animals, research costs can be considerably reduced by using biomimetic tissues or organs instead of live animals. Products such as lung-, liver-, and heart-on-a-chip have recently been developed for drug screening [2,10,11]. However, these chips contain only 1 or 2 cell types, and therefore require further development to functionally mimic real organs. The human body contains w10 trillion cells, with 220 differentiated cell types [12]. The human endocrine system, encompassing many different organs and cell types, is among the most anatomically and physiologically complex of systems. The dysregulation of hormone secretion can lead to various diseases, including type 1 diabetes mellitus, which is characterized by the loss of functional b cells due to autoimmune responses, leading to the malfunction of pancreatic islets in the regulation of blood glucose levels [13]. The incidence of this disease is 8e17 per 100,000 individuals in Europe and the United States [14], making it a considerable public health challenge. Many animals, typically rodents, are sacrificed for the development of drugs and other treatment methods for diabetes [15]. The aim of the present study was to generate pseudo-islets that can mimic naïve islets. To achieve this end, 3 different pancreatic islet cell lines derived from a and b cells as well as pancreatic epithelial-like cells, were cocultured to yield pseudo-islet spheroids. Different cell type ratios were tested to determine the optimal combination that could recapitulate the morphologic and functional properties of naïve islets. The biofabrication of pancreatic islets has potential applications for diabetes treatment, because one currently employed therapeutic strategy involves the transplantation of pancreatic islets, which, once successfully incorporated by the recipient, can reestablish normal glucose regulation. Most research into treatment methods for diabetes has focused solely on insulin secretion while disregarding other glucose-regulating hormones. However, to understand the pathogenesis of diabetes, it is necessary to examine not only
Fig 4. Secretion of pancreatic hormones by pseudo-islets. Protein levels of (A) insulin and (B) glucagon secreted by pseudoislets generated from NIT-1, a-TC1 clone 6, and TGP52 cell lines cocultured at ratios of 1:1:1 and 7:2:1 and densities of 3,000, 6,000, and 9,000 cells/well, as analyzed by enzyme-linked immunosorbent assay.
1160
defective b cells, but also the functional interactions of other cells in the pancreas. The antagonistic actions of insulin and glucagon maintain homeostasis in the pancreas, and depend on the tight regulation of various pancreatic cell types, because glucagon deficiency can result in aberrant blood glucose levels even if insulin secretion is normal. Despite advances in the development of artificial organ chip systems, functional artificial endocrine tissues have yet to be developed. In the present study, pseudo-islets were evaluated in terms of endocrine secretion as a measure of their functional status. In spheroids formed from a 7:2:1 ratio of constituent cell types, a and b cells were randomly distributed (Fig 2), which was consistent with the morphology of naïve pancreatic islets [16]. At that ratio, insulin secretion was independent from seeding density (Fig 4), in contrast to spheroids formed from a 1:1:1 ratio, with which insulin and glucagon secretion was proportional to seeding density (P < .05). An appropriate cell composition is essential for making pseudo-islets that can functionally mimic naïve islets. The data from this study suggest that a 7:2:1 ratio is preferable to a 1:1:1 ratio, because hormone secretion did not require that a specific number of cells be present in the spheroids, at least in the case of insulin. It is not clear why this was not also true for glucagon, and future studies will determine whether other ratios besides 7:2:1 would be more suitable. In conclusion, mouse pseudo-islets that can mimic the secretory functions of naïve pancreatic islets were successfully generated; these can potentially be used as a substitute for live animals in studies investigating the pathogenesis of diabetes and potential treatment strategies for this disease. REFERENCES [1] Russell WMS, Burch RL. The principles of humane experimental technique. London: Methuen, 1959.
JO, JANG, NEMENO ET AL [2] Huh D, Matthews BD, Mammoto A, et al. Reconstituting organ-level lung functions on a chip. Science 2010;328:1662e8. [3] Huh D, Leslie DC, Matthews BD, et al. A human disease model of drug toxicityeinduced pulmonary edema in a lung-on-achip microdevice. Sci Transl Med 2012;4:159ra147. [4] Elayat AA, el-Naggar MM, Tahir M. An immunocytochemical and morphometric study of the rat pancreatic islets. J Anat 1995;186(Pt 3):629e37. [5] Acharjee S, Ghosh B, Al-Dhubiab BE, et al. Understanding type 1 diabetes: etiology and models. Can J Diabetes 2013;37: 269e76. [6] Lee JI, Sato M, Kim HW, et al. Transplantatation of scaffoldfree spheroids composed of synovium-derived cells and chondrocytes for the treatment of cartilage defects of the knee. Eur Cell Mater 2011;22:275e90. discussion 290. [7] Jeong JH, Lee JI, Ju MK, et al. Proliferation of pancreatic endocrine cells using disaggregation-expansion-reaggregation technology in isolated rat islets. Transplant Proc 2010;42:907e10. [8] Kim JY, Kim HW, Bae SJ, et al. Hybrid cellular spheroids from hepatocellular carcinoma and insulin-secreting cell lines. Transplant Proc 2012;44:1095e8. [9] Jo YH, Nam BM, Kim BY, et al. Pseudoislet of hybrid cellular spheroids from commercial cell lines. Transplant Proc 2013;45:3113e7. [10] Agarwal A, Goss JA, Cho A, et al. Microfluidic heart on a chip for higher throughput pharmacological studies. Lab Chip 2013;13:3599e608. [11] Prot JM, Bunescu A, Elena-Herrmann B, et al. Predictive toxicology using systemic biology and liver microfluidic “on chip” approaches: application to acetaminophen injury. Toxicol Appl Pharmacol 2012;259:270e80. [12] Nicholas W. In tiny cells, glimpses of body’s master plan. New York Times, December 18, 2001. [13] Cooke DW, Plotnick L. Type 1 diabetes mellitus in pediatrics. Pediatr Rev 2008;29:374e84. [14] Department of Health and Human Services. National diabetes statistics. Maryland, 2011. [15] National Statistics. Statistics of scientific procedures on living animals. London: 2009. [16] Pechhold K, Zhu X, Harrison VS, et al. Dynamic changes in pancreatic endocrine cell abundance, distribution, and function in antigen-induced and spontaneous autoimmune diabetes. Diabetes 2009;58:1175e84.