Murine CD133+CD49flow/+ Cells Derived from ESCs Differentiate into Insulin Producing Cells in vivo

C l i n M e d International Library Citation: Ciriza J, Caneda C, McLelland B, Manilay JO (2014) Murine CD133+CD49flow/+ Cells Derived from ESCs Differentiate into Insulin Producing Cells in vivo. Int J Stem Cell Res Ther 1:004. doi.org/10.23937/2469-570X/1410004 Received: September 18, 2014: Accepted: December 28, 2014: Published: December 31, 2014 Copyright: © 2014 Ciriza J. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Ciriza et al. Int J Stem Cell Res Ther 2014, 1:004


Introduction
Worldwide, 347 million people suffer from diabetes [1] and the World Health Organization projects that diabetes will be the 7 th leading cause of death by 2030.In the United States, 25.8 million children and adults had diabetes in 2011, representing 8.3% of the total population [2], and costing $116B for treatment and diagnosis.Type 1 diabetes (TID) is an autoimmune disease that aggressively attacks the insulin producing cells (IPCs) in the pancreas, reducing the number of these cells to levels that force the patient to be dependent on daily insulin injections.Islet cell transplantation can be used for treatment, but sources of autologous and allogeneic donor insulin producing cells for transplantation are limited.Embryonic stem cells (ESCs) have been proposed as an alternative source of insulin producing cells [3].Major concerns about the use of ESC-derived cells for therapy include their teratoma formation ability and their possible immune rejection after transplantation.Teratoma formation prevented successful treatment of diabetic mice with ESC-derived cells [4] and immune rejection of insulin producing cells derived from ESCs was observed as early as 3 days post-transplantation into immunocompetent mice [5,6].To improve this outcome, the enrichment of definitive endodermal progenitors by flow cytometric sorting [7,8] could be used to deplete cultures of undifferentiated cells (which are the main sources of teratomas), while the prediction of the immunogenicity of insulin producing cells derived from ESCs (ESC-IPCs) could indicate possible preconditioning regimens for patients that received ESC-IPC transplants.
Several cell culture protocols have been described for the differentiation of murine nontransgenic ESCs into IPCs, [3] but three protocols appear to be the most highly utilized [9][10][11].The three protocols share the common initial step of embryoid body (EB) formation, but differ in the use of selection of nestin-expressing cells.When compared side by side, ESC-IPC differentiated using the Blyszczuk et al. [11] protocol were able to produce and release more insulin in vitro than those differentiated by the other two methods.However, they only rescued 33% of diabetic mice within 2 weeks and ultimately, these mice developed teratomas [12], likely due to the remaining undifferentiated ESCs in the cultures at the time of transplantation.Exhaustive microarray studies studying gene expression profiles of ESC-IPC at different stages of in vitro differentiation following the Blyszczuk et al. [11] protocol have shown that murine ESC-derived pancreatic cells appear to be blocked at the embryonic/fetal stage of development [13].
In contrast to fully mature IPCs, pancreatic progenitors generated from ESCs (ESC-PPs) might be more suitable for cellular replacement therapy in TID.A homogenous population of ESC-PPs identified by cell surface markers may not self-renew as stem cells do, but could differentiate into endocrine cells after transplantation [14].For example, Pdx1 + Cxcr4 + ESC-derived cells spontaneously differentiated

Abstract
Pancreatic progenitors have been identified and isolated during embryonic development showing their potential to differentiate into insulin producing cells after transplantation.Subsequently, they have been proposed as an alternative to the treatment of Type I diabetes.However, fetal pancreata represent a limited donor supply and we proposed that ESC-derived pancreatic progenitors are another possible option.We isolated committed pancreatic CD133 + CD49f low/+ progenitors, expressing neurogenin 3, from insulin producing cells derived from mouse embryonic stem cells in vitro.Here, we demonstrate their potential to differentiate into insulin producing cells in vivo while not forming teratomas.Our results can pave the way for future studies to ascertain the ability of in vitro ESC-derived pancreatic progenitors to be translated for clinical application for Type I diabetes treatments.
into IPCs in vivo, and led to a sustained correction of hyperglycemia in diabetic mice for a month with survival rates longer than three months [15].Late definitive endoderm cells can be identified by the combination of E-cadherin and Decay Accelerating Factor (DAF1/ CD55) cell surface markers on cells that co-express also Pdx1 [16] (Figure 1).Ngn3 + cells have been suggested to be another possible murine pancreatic progenitor population [17,18].CD133 + CD49f low/+ cells isolated from murine fetal pancreas have been described to express Ngn3 (Figure 1) as well as insulin-1, glucagon, somatostatin and pancreatic polypeptide after 2 days of co-culture with mitomycin C-treated mouse embryonic fibroblast (MEFs) [8].Moreover, transplanted CD133 + PDFGR β -cells isolated during murine pancreas development induced the formation of IPCs that expressed Cpeptide (a definitive feature of functional insulin production) within 7 days of culture and showed self-renewal properties after serial transplantation, suggesting they might be true pancreatic stem cells [19].Ngn3 + cells have been also detected in EBs from murine Ngn3-GFP ESC lines expressing insulin, glucagon and Pdx1, among other pancreatic markers, after sorting and differentiation in vitro [20].However, the identification with cell surface markers of non-transgenic pancreatic progenitor cells expressing Ngn3 from ESC-differentiation cultures and their application in vivo has not yet been reported.Here, we present the identification of CD133 + CD49f low/+ and E-cadherin + CD55 + pancreatic progenitor cells from a slightly modified ESC-IPC differentiation protocol.We show that CD133 + CD49f low/+ ESC-PPs can be transplanted into the kidney capsule of immunodeficient mice, produce insulin, and do not form teratomas.

Heterogeneous and immature ESC-IPCs are generated using current protocols
We initiated our studies with the ESC-IPC differentiation protocol described by Blyszczuk et al. [11] since it seemed to be the most efficient in terms of insulin production and release [12].This protocol includes the formation and growth of EBs for 5 days, their expansion for 9 days on gelatin-coated plates and their specific differentiation into IPCs for 18 days on collagen I coated plates with a defined differentiationmedium (Figure 2A).Healthy groups of cells could be observed in vitro after Day 5+16, with aslow as 5% FCS in the media (Figure 2A).Gene expression analysis showed similarities between ESC-IPCs and murine pancreatic β islet samples such as expression of insulin 2 and IAPP with very low expression of somatostatin and amylase (Figure 2B).Genes described to be expressedduring the development of insulin producing cells (Glut2, Nestin, Sox17, Isl1 and cytokeratin19 (CK19)) were also detected at different stages of the differentiation protocol including the final stage at Day 5+28 (Figure 2B).However, in contrast to β islet samples, insulin 1was notdetected at any time of differentiation (Figure 2B).
Immunohistochemistry analysis showed a high concentration of insulin accumulated inside the IPCs with low expression of c-peptide, in contrast to the β islet samples (Figure 2C).These data suggested that incomplete insulin processing at the protein level was occurring, and indicated that IPCs were not fully functional.Similarly, glucagon expression was much lower in IPCs than β islets (Figure 2C).The insulin release response in IPCs to glucose stimulation invitro was low and heterogeneous, with higher ratios of release at lower glucose concentrations in some experiments (Figure 2D).Finally, we detected that almost half of the IPCs at Day 5+28 of differentiation expressed SSEA-1 (Figure 2E), which suggested that they were teratomaforming cells [12].Based on these results, we next considered a different population of cells for further investigation.

Pancreatic progenitors can be detected in ESC-IPC cultures
We investigated if putative pancreatic progenitors could be identified in vitro during differentiation of ESCs into IPCs (Figure 1).First, using flow cytometry, we analyzed the presence of definitive endoderm progenitors with the cell surface marker phenotype.Ecadh + CD55 + , as previously described in vitro with other ES cell lines [16].Low percentages of Ecadh + CD55 + cells were detected at Day 5+16 of differentiation (mean ± SD: 0.93 ± 0.18 %, n=3) (Figure 3A).Almost all Ecadh + CD55 + cells were SSEA-1 + , suggesting a high teratoma formation ability potential if transplanted (Figure 3B).
We then searched for pancreatic progenitors that were more differentiated than definitive endoderm progenitors and with less potential of forming teratomas (Figure 1).In vivo, murine and human committed pancreatic progenitors have been identified by the CD133 + CD49f low/+ cell surface phenotype, and their potential to differentiate in vitro into insulin producing cells has been verified [8,19].We were able to quantify an average of 4.07 ± 1.29 % of CD133 + CD49f low/+ pancreatic progenitors at Day 5+16 in our IPC differentiation cultures (Figure 3C).This higher frequency of cells, compared to Ecadh + CD55 + cells, allowed for the collection of enough cells by flow cytometry to analyze their gene expression.CD133 + CD49f low/+ pancreatic progenitors did not express insulin-2, but expressed Ngn3, indicating that they belong to an earlier developmental stage, but were committed to differentiate into IPCs (Figure 3C).These progenitors did not express either pancreatic marker such as Pdx1or Pax4 by RT-PCR (data not shown).CD133 + CD49f low/+ pancreatic progenitors also expressed SSEA-1 (Figure 3D).The expression of SSEA-1 indicated their potential to form teratomas, which we then tested in vivo.

CD133 + CD49f low/+ pancreatic progenitors did not form teratomas and expressed insulin
First, we transplanted 10 6 SSEA1 + D3-ESCs under the kidney capsule of NSG mice and all mice developed teratomas by 28 days post-transplantation, as expected (Figure 4A, panel 1).Conversely, as a negative control for teratoma formation, we transplanted 300 islets isolated from adult pancreas and they did not form teratomas (Figure 4A, panel 5).We then tested the ability of our ESC-IPCs to form teratomas after transplantation.First, we confirmed that IPCs at Day 5+28 of differentiation expressed insulin 2, and then we transplanted 4-6x10 6 IPCs.All of the IPC-recipient mice developed teratomas by 28 days after transplantation (Figure 4A, panel 2).This result confirmed previous results [11] and justified the exploration of an alternative ESC-derived cell population for the replacement of insulin.
We hypothesized that more immature pancreatic progenitors could be generated, isolated and transplanted in vivo, and that after transplantation, they would not form teratomas and would produce insulin.As mentioned above, CD133 + CD49f low/+ cells have been demonstrated to be pancreatic progenitors (PPs) in mouse embryos [8] and we confirmed that CD133 + CD49f low/+ cells were generated in our in vitro cultures.We tested the function and potential of these CD133 + CD49f low/+ ESC-PPs to form teratomas in vivo.Since SSEA-1 expression is often equated with pluripotency and teratoma-forming ability in ESCs, we also compared the behavior of sorted SSEA-1-CD133 + CD49f low/+ ESC-PPs.We transplanted 5x10 4 pancreatic progenitors (SSEA1 + depleted or not) as previously described [8].No teratomas were observed in mice transplanted with either population (Figure 4A, panels 3,4).No significant differences were found amongst the sizes of the kidneys transplanted with ESC-PPs, islets or sham controls (Figure 4B), a result that correlated with a lack of teratoma forming ability in ESC-derived PPs.The absence of teratoma formation in recipients of CD133 + CD49f low/+ or SSEA-1-CD133 + CD49f low/+ ESC-PP cells was confirmed by histological analysis of kidney sections from the transplanted mice (Figure 4C, panels 3,4).In contrast, teratomas were clearly observed in kidneys transplanted with ESCs and ESC-IPCs (Figure 4C, panels 1,2).ESC-IPCs, CD133 + CD49f low/+ ESC-PPs, SSEA-1-CD133 + CD49f low/+ ESC-PPs and islets exhibited positive staining for insulin (Figure 5A, panels 2-5), whereas sections of kidney capsules transplanted with ESCs did not (Figure 5A, panel 1).Insulin production was confirmed via the observation of strong c-peptide expression in the kidneys thatwere transplanted with CD133 + CD49f low/+ ESC-PPs and control islets (Figure 5B, panels 1,3).SSEA-1 -CD133 + CD49f low/+ ESC-PPs showed positive, but lower expression of c-peptide (Figure 5B, panel 2).

Discussion
We generated ESC-IPCs that express endocrine genes such as insulin 2, glucagon and somatostatin, and our data showed that ESC-IPC cultures heterogeneously secrete insulin protein, corroborating the findings of other groups [12].Based on the lack of peptide signal by immunohistochemistry, we posit that the insulin is not completely processed intracellularly within ESC-IPCs and consequently, is not released properly outside the cell.One possible reason for this nonfunctional insulin release is incomplete differentiation of the ESCIPCs, resulting in a cell that is devoid of fully developed mechanisms for insulin processing.Other groups have suggested that the detection of insulin from ESC-IPCs is an artifact from apoptotic or necrotic cells [21] or a mixture of de novo insulin production mixed with insulin uptake from the medium [22] and this controversy is still unresolved.In our hands, insulin release from ESC-IPCs after glucose challenge varied widely from experiment to experiment, perhaps reflecting the presence of a heterogenous mixed of cells that are generated in each independent ESC-IPC culture [12].The immature nature of ESC-IPCs, their inconsistent release of insulin, and their propensity to In contrast, we confirmed that ESC-derived CD133 + CD49f low/+ pancreatic progenitor cells did not form teratomas after 28 days of in vivo transplantation.This was surprising, as 50% of CD133 + CD49f low/+ cells expressed SSEA1, which is often associated with pluripotency.It has been previously hypothesized that human pancreas cells that are positive for stage-specific embryonic antigen 4 (SSEA4) co-express ductal pancreatic progenitor markers and can be differentiated in vitro to pancreatic hormone-expressing cells [23].We hypothesize that SSEA1 could be a murine marker for pancreatic progenitors, sharing similarities to human SSEA4 as a murine ductal pancreatic progenitor marker.However, after transplantation in vivo, SSEA-1 -CD133 + CD49f low/+ ESC-PPs expressed lower c-peptide levels than CD133 + CD49f low/+ ESC-PPs, a result which suggests that some pancreatic progenitors may be depleted if the SSEA1 marker is used for isolation of PPs.Regardless of the expression of SSEA1, both unfractionated CD133 + CD49f low/+ and sorted CD133 + CD49f low/+ SSEA1 + populations did not form teratomas in vivo.
It is possible that the lack of teratoma-formation we observed is simply due to the transplantation of insufficient numbers of cells required to form teratomas in vivo.We do not feel this is the case, as we selected our cell dosage based on previous studies in which the same minimal cell dose resulted in 100% teratoma-forming ability after transplantation into immunodeficient mice [8,24,25].The lack of teratoma-forming ability and the clear expression ofinsulin and C-peptide in both unfractionated CD133 + CD49f low/+ and sortedCD133 + CD49f low/+ SSEA1 + populations lead us to conclude that PPs are a potentially feasible cellular therapy.
One challenge to the use of ESC-derived cells is that in vitro differentiation of ESCs to specific cell lineages may not necessarily produce cells that express the same cell surface markers as their in vivo-derived counterparts [26].To our knowledge, our study is the first to identify murine pancreatic progenitor populations derivedfrom ESCs in culture, and we have confirmed that ESCderived CD133 + CD49f low/+ cells arefunctionally similar to their in vivo counterparts in mouse embryos.Clearly, improvements in the generation of higher yields of pancreatic progenitor cells are required for practical application, and this could be achieved by discovery of methods to block unwanted lineage differentiation, as well as scaled up production and isolation by means of bioreactors and other stem cell separation technologies [27].
In conclusion, we have identified mouse ESC-derived pancreatic progenitor cells in vitro and discovered that they can survive and produce insulin in vivo.We envision that our phenotypic and functional data could be further utilized to study and improve the long-term survival of in vitro derived pancreatic progenitor cells and their ability to produce insulin in sufficient quantities to alleviate hyperglycemia in humans.

Mice
NOD.Cg-Prkdc scid Il2rg tm1Wjl /SzJ(NSG) and 129S2/SvPasCrl (129) mice were purchased from The Jackson Laboratory (Bar Harbor, ME, USA) and Charles River (Wilmington, MA, USA) respectively, or bred in house and housed in sterile microisolator cages with sterile feed and autoclaved water.Mice were euthanized by CO 2 asphyxiation or by cervical dislocation in the case of pancreatic islet isolation.All procedures were approved by the UC Merced Institutional Animal Care and Use Committee (IACUC).

Gene expression analysis
RNA from ESCs, EBs at Day 5+9 and differentiated cells at Days 5+16 and 5+28 was extracted using RNeasy® Protect Mini Kits (Qiagen, Valencia, CA) and from sorted pancreatic progenitors using Trizol (Invitrogen, NY), following the manufacturer's instructions.RNA samples were treated with DNAseI (Roche, San Francisco, CA) and the concentration of RNA was determined using a Nanodrop spectrophotometer (ThermoScientific, Waltham, MA).Next, cDNA was synthesized with SuperScriptIII First-Strand Synthesis System for RT-PCR Kit (Invitrogen, Grand Island, NY, USA).The gene expression of 17 genes was analyzed by PCR as previously described with β5-tubulin as the housekeeping gene [28].For Ngn3, Pdx1 and Pax4 gene expression analysis, clones containing the respective gene were purchased from Thermo Scientific (Waltham, MA, USA) and used as positive controls.PCR products were visualized after electrophoresis in ethidium bromide stained 2% agarose gels with a BioRadChemiDoc imaging system (BioRad, Hercules, CA, USA).Sizes of PCR products were compared with a 100 bp Low Scale DNA ladder (Fisher, Pittsburg, PA, USA).

Immunohistochemistry
Immunohistochemistry was performed on IPC-ESC grown on six well-plates at Days 5+16 or 5+28 of differentiation.Cells were collected by trypsinization and cytospun on slides (3x10 5 cells/ slide) with a Shandon Cytospin III cytocentrifuge at 1000 rpm for 8 minutes.INS-1 832/13 insulinoma cell line cultures (a gift from Dr. Chris Newgard, Duke University) were also prepared as controls for insulin expression.Samples were fixed with 4% paraformaldehyde (PFA) for 15 minutes at room temperature, rinsed 3 times with phosphate buffered saline (PBS) and blocked for 30 minutes at room temperature with blocking buffer consisting of 10% normal goat serum and 0.1 % Triton®-X-100 in PBS.After blocking, samples and controls were incubated at 4°C overnight with guinea pig antiinsulin, rabbit anti-c-peptide or rabbit antiglucagon.Samples stained with secondary antibodies alone were prepared as negative controls.The next day; samples were rinsed three times with PBS, incubated for 45 minutes at 37°C with anti-guinea pig IgG FITC or anti-rabbit IgG Atto 633, and stained with 0.5μg/ml DAPI.Finally, samples were sealed with Vectashield mounting medium (Vector, Burlingame, CA , USA) and visualized by fluorescence microscopy, using the Olympus BX51 Microscope, the 10x/0.30and 40x/0.75U Plan FLN lenses and the Image-Pro Plus 5.0 software.

Insulin release in vitro
At Day 5+28 of differentiation, IPC-ESC were cultured with insulin-free differentiation medium for 48 hours prior to challenge cells with different concentrations of glucose.Medium was changed again 3 hours before glucose challenge and cultures were washed five times with PBS.Cells were further incubated for 90 minutes with 2.5mM glucose in Krebs Ringer bicarbonate HEPES (KRBH) buffer consisting of 118mM NaCl, 4.7mM KCl, 1.1mM KH 2 PO 4 , 25mM NaHCO 3 , 3.4mM CaCl 2 , 2.5 MgSO 4 , 10mM HEPES, and 2mg/ml BSA at 7.4 pH.ESIPCs were challenged with KRBH containing 2.5 mM glucose, 5.5mM glucose and 50μM tolbutamide, or 27.7mM glucose for 1 hour.The supernatants from the different cultures were collected and cells were dissociated by trypsinization and resuspended in acid ethanol for protein extraction.Upon resuspension, cells were sonicated and incubated overnight at 4°C.Next day, acid ethanol suspension was neutralized with 1 M Tris Buffer, pH 7.5.Total protein content was determined by Bradford assay with the Coomassie Protein Assay Kit (ThermoScientific, Waltham, MA).Released and intracellular insulin levels were measured by means of ELISA with the Mercodia Ultrasensitive Mouse Insulin Enzyme-Linked Immunosorbent Assay (Mercodia Developing Diagnostics, Uppsala, Sweden) according to manufacturer instructions.

Flow cytometry analysis and sorting
Samples containing 10 6 single cells from ES or ES-IPCs cells at stages 5+16 or 5+28 were blocked with anti-CD16/32 for 5 minutes at 4 °C.Next, they were stained for 15 minutes at 4 °C with either an antibody cocktail containing anti CD133-PE, anti CD49f -APC and biotin anti SSEA-1, or the cocktail containing anti E-Cadherin-Alexa Fluor® 647, anti CD55-PE and biotin anti SSEA-1.For each antibody cocktail, "fluorescence minus-one" or fluorescently-labeled isotype antibody controls were also prepared.Finally, cells were resuspended in 0.1μg/ml DAPI before sorting to exclude dead cells.Cells were analyzed or isolated using a FACS Aria flow cytometer (Becton Dickinson, Franklin Lakes, NJ).Single color samples for dye compensation were prepared for each experiment.For each study, three independent experiments were performed.

Pancreatic islet isolation
After cervical dislocation, pancreata from 129 mice were infused with 3 ml/mouse of 0.8 mg/ml collagenase P solution (Roche Applied Science, Indianapolis, IN, USA), removed and digested for at least 16 minutes at 37 o C in the same collagenase P solution.Cell suspensions were filtered through a 70 micron cell strainer and washed several times with a solution consisting of 10mM HEPES (Fisher, Pittsburg, PA, USA), 10μg/μl Dnase I (Roche Applied Science, Indianapolis, IN, USA), 100 units/ml penicillin, 100μg/ml streptomycin, 1.7mM CaCl 2 (Fisher, Pittsburg, PA, USA) and 1.2mM MgCl (Fisher, Pittsburg, PA, USA) in HBSS (Invitrogen), until clumps were not visible in the supernatant.Cells were resuspended in 5 ml of Histopaque-1119 (Sigma-Aldrich, St. Louis, MO) and slowly added 5 ml of culture medium consisting of 25mM HEPES, 0.1mM non-essential amino acids (Invitrogen), 100 units/ml penicillin, 100μg/ml streptomycin, 10% FBS in RPMI 1640 (Invitrogen) on top of the Histopaque solution.Islets in the Histopaque-culture medium interface were collected and 2000 rpm for 20 minutes without brake spinning, washed three times with washing medium, and resupended in 5 ml of culture medium to further being collected by hand picking under a dissecting microscope [29].

In vivo engraftment
Eight to 10 week-old NSG mice were anesthetized with isofluorane and shaved.A vertical incision through dermal layers and peritoneum was made and the left kidney pulled out.After a small incision in the kidney capsule, 300 pancreatic islets, 4-5x10 6 ESC-IPCs, 5x10 4 pancreatic progenitors (SSEA1 + depleted or not) or 10 6 ESCs were inserted under the capsule.The kidney capsule was cauterized, the peritoneum sutured and skin stapled after transplantation.The post-operative analgesic buprenorphine was injected at doses of 0.05-0.1 mg/kg immediately after surgery and 18 hours later.Kidneys were collected 28 days post-transplantation and grafts were assessed by visualization and measurement of their diameter as well as histology.Sections between 5 to 6 microns from transplanted and nontransplanted kidneys were prepared at the UCSF Mouse Pathology Core for a fee.Immunohistochemistry was performed as described previously above.

Figure 1 :
Figure 1: Schematic representation of IPC differentiation: During differentiation from ESC to IPC, definitive endoderm and pancreatic progenitors can be isolated by means of different extracellular markers

Figure 3 :
Figure 3: Identification of definitive endoderm and committed pancreatic progenitors from in vitro IPC-ESC cultures by flow cytometry.A) Live cells at Day 5+16 were analyzed for the expression of E-cadherin and CD55.The gate shows the Ecadherin+CD55 + definitive endoderm progenitors.The data shown are representative of three independent experiments.B) Histogram analysis of SSEA1 expression on Ecadherin+CD55 + definitive endoderm cells.C) Live cells at Day 5+16 were analyzed for the expression of CD133 and CD55.The gate shows the CD133 + CD49f low/+ committed pancreatic progenitors.Data shown are representative of three independent experiments.D) Histogram analysis of SSEA1 expression on CD133 + CD49f low/+ committed pancreatic progenitors