TISSUE-SPECIFIC STEM CELLS
Generation of Pancreatic Hormone-Expressing Islet-Like Cell
Aggregates from Murine Adipose Tissue-Derived Stem Cells
VIKASH CHANDRA,aSWETHA G,aSMRUTI PHADNIS,aPRABHA D. NAIR,bRAMESH R. BHONDEc
aNational Centre for Cell Science, Ganeshkhind, Pune, India;bBiomedical Technology Wing, Sree Chitra Tirunal
Institute for Medical Sciences and Technology, Thiruvananthapuram, Kerala, India;cTissue Engineering &
Banking Laboratory, National Centre for Cell Science, Ganeshkhind, Pune, Maharashtra, India
Key Words. Adipose tissue stem cells•Islet-like cell aggregates•Insulin•Pancreatic differentiation
The success of cell replacement therapy for diabetes
depends on the availability and generation of an adequate
number of islets, preferably from an autologous origin.
Stem cells are now being probed for the generation of
physiologically competent, insulin-producing cells. In this
investigation, we explored the potential of adipose tissue-
derived stem cells (ASCs) to differentiate into pancreatic
hormone-expressing islet-like cell aggregates (ICAs). We
initiated ASC culture from epididymal fat pads of Swiss
albino mice to obtain mesenchymal cells, murine epididy-
mal (mE)-ASCs. Subsequent single-cell cloning resulted in
a homogeneous cell population with a CD291CD441Sca-
11surface antigen expression profile. We formulated a 10-
day differentiation protocol to generate insulin-expressing
ICAs from mE-ASCs by progressively changing the differ-
entiation cocktail on day 1, day 3, and day 5. Our stage-
specific approach successfully differentiated mesodermic
mE-ASCs into definitive endoderm (cells expressing Sox17,
Foxa2, GATA-4, and cytokeratin [CK]-19), then into pan-
creatic endoderm (cells expressing pancreatic and duode-
nal homeobox [PDX]-1, Ngn3, NeuroD, Pax4, and glucose
transporter 2), and finally into cells expressing pancreatic
hormones (insulin, glucagon, somatostatin). Fluorescence-
activated cell sorting analysis showed that day 5 ICAs con-
tained 64.84% 6 7.03% PDX-11cells, and in day 10
mature ICAs, 48.17% 6 3% of cells expressed C-peptide.
Day 10 ICAs released C-peptide in a glucose-dependent
manner, exhibiting in vitro functionality. Electron micros-
copy of day 10 ICAs revealed the presence of numerous
secretory granules within the cell cytoplasm. Calcium algi-
nate-encapsulated day 10 ICAs (1,000–1,200), when trans-
planted i.p. into streptozotocin-induced diabetic mice,
restored normoglycemia within 2 weeks. The data pre-
sented here demonstrate the feasibility of using ASCs as a
source of autologous stem cells to differentiate into the
pancreatic lineage. STEM CELLS 2009;27:1941–1953
Disclosure of potential conflicts of interest is found at the end of this article.
Diabetes is a chronic degenerative disease resulting in lifelong
dependency on insulin treatments. Transplantation of cadav-
eric pancreatic islets to diabetic patients is the most preferred
cell replacement therapy available to treat type 1 diabetes .
However, the scarcity of transplantable human islets poses a
major obstacle in the widespread use of this therapy . This
has evoked a large-scale search for alternative sources of b
cells. Embryonic stem (ES) cells have been favorites in the
race because of their tremendous differentiation potential.
However, the ongoing ethical
involved in ES cell research limit its use in translational med-
icine. Hence, adult mesenchymal stem cells (MSCs) are now
being widely evaluated for their differentiation potential for
cell replacement therapy. Recent studies have shown that
insulin-secreting cells can be generated from ES cells [3–7],
and legal considerations
mouse/human bone marrow cells , rat hepatic oval cells
, pancreatic stem/progenitor cells , etc. However, scar-
city of the source and the invasive procedures required to iso-
late and culture these cells have limited their use. In this sce-
nario, adipose tissue (AT) is gaining much attention as a
prime source of MSCs for cell therapy by virtue of its ready
availability, enormous expandability, and ease of isolation
with minimum patient discomfort [11, 12]. AT is a remark-
able organ that regulates the fat mass and nutrient homeosta-
sis of the body. It releases a large number of bioactive media-
tors (adipokines) modulating homeostasis, blood pressure, and
lipid and glucose metabolism. Adipokines like leptin, adipo-
nectin, and visfatin are known insulin sensitizers and play
major roles in glucose homeostasis . Extrapancreatic insu-
lin-producing cells have been detected in the AT of diabetic
mice . Adipocytes from carp are also reported to express
insulin, which is termed carp adipocyte insulin . Studies
have also shown that the transcription factor Isl-1 and Pax-6,
Author contributions: R.R.B.: conception and design, financial support, administrative support, provision of study material and final
approval of manuscript; V.C.: conception and design, all bench work, collection and assembly of data, data analysis and interpretation,
and manuscript writing; S.G.: data analysis and interpretation and manuscript writing; S.P.: animal surgery; P.D.N.: provided
biocompatible capsules for transplantation studies.
Correspondence: Ramesh R. Bhonde, Ph.D., Tissue Engineering & Banking Laboratory, National Centre for Cell Science, NCCS
Complex, Ganeshkhind, Pune 411007, Maharashtra, India. Telephone: 91-20-25708079; Fax: 91-20-25692259; e-mail: rrbhonde@nccs.
res.in Received February 9, 2009; accepted for publication April 24, 2009; first published online in STEM CELLS EXPRESS May 7, 2009.
STEM CELLS 2009;27:1941–1953 www.StemCells.com
C AlphaMed Press 1066-5099/2009/$30.00/0 doi: 10.1002/stem.117
which play important roles in pancreatic endocrine develop-
ment, are also expressed in the proliferative population of
ASCs . All these features make ASCs the preeminent can-
didate to differentiate into the pancreatic endocrine lineage
for use in cell-based therapies for diabetes.
ASCs can easily be obtained from a patient’s own tissue,
isolated ex vivo, expanded, differentiated into insulin-produc-
ing cells, and transplanted back into the patient as an autolo-
gous transplant. Timper et al.  demonstrated the initial
prospects of differentiating human ASCs into insulin-, gluca-
gon-, and somatostatin-expressing cells. That study provided
preliminary data on the plasticity of ASCs to differentiate into
the pancreatic endocrine phenotype. However, the report
lacked in vitro/in vivo functionality assays of the differenti-
ated cells. Detailed studies need to be performed to fully
understand the pancreatic endocrine differentiation potential
of ASCs to develop novel cell-based therapies.
In the present study we demonstrate a three-step differen-
tiation protocol for murine ASCs to differentiate into func-
tional islet-like cell aggregates (ICAs) in defined serum-free
medium (SFM). Our 10-day protocol follows the conversion
of mesodermic ASCs into definitive endoderm (DE), then into
pancreatic endoderm, and finally into pancreatic hormone-
expressing ICAs. These ICAs when transplanted can restore
normoglycemia in streptozotocin
mice. Thus, the present findings present a step toward cell
replacement therapy for type 1 diabetes, by autologous trans-
plantation of ICAs differentiated from a patient’s own AT.
MATERIALS AND METHODS
ASC Isolation and Expansion
Swiss Albino mice (6–8 weeks old) were maintained at the Ani-
mal Facility of the National Centre for Cell Science on normal
chow and water. All experimental protocols were approved by
the Animal Ethical Committee of the institute. AT samples from
epididymal fat pads (n ¼ 10) were washed and digested in phos-
phate-buffered saline (PBS) supplemented with 2% bovine serum
albumin (BSA) and 0.2% collagenase type II (Sigma-Aldrich, St.
Louis, http://www.sigmaaldrich.com) prewarmed to 37?C for
45 minutes. The cells were centrifuged (400g for 4 minutes) at
room temperature (RT) in 1–2 ml sterile PBS containing 2% (v/
v) fetal bovine serum (FBS) (Sigma-Aldrich). Supernatant con-
taining mature adipocytes was discarded. The pellet containing
the stromal vascular fraction was resuspended in Dulbecco’s
modified Eagle’s medium (DMEM)/Ham’s F-12 (Gibco, Grand
Island, NY, http://www.invitrogen.com) with 10% FBS and anti-
biotics. The cells were plated at a density of 2 ? 105/cm2into
T25 culture flasks.
A serial dilution method was used to generate single-cell
clonogenic cultures. Briefly, 104isolated murine epididymal
(mE)-ASCs were suspended in 1 ml culture medium. Repeated
serial dilution was carried out to achieve a final dilution of 100
cells in 10 ml medium. For the single-cell culture, 100 ll diluted
cell suspension was transferred into each well of a 96-well plate
containing 200 ll culture medium.
In Vitro Differentiation of mE-ASCs Into ICAs
Differentiation was carried out in three stages as indicated in
Figure 2A (n ¼ 25). To induce differentiation, undifferentiated
mE-ASCs were trypsinized, diluted with SFM-A and centrifuged.
Cells were counted for initial seeding density and 1 ? 106cells/
cm2were resuspended in SFM-A and plated on ultralow attach-
ment tissue culture plates (Corning, Fisher Scientific Interna-
tional, Hampton, NH, http://www.fisherscientific.com) or small
glass petri plates (8 ? 106cells in a 200glass petri plate makes
?800–1,200 ICAs). SFM-A contained DMEM/F12 (1:1) (Gibco)
with 17.5 mM glucose, 1% BSA Cohn fraction V, fatty acid free
(#A8806; Sigma-Aldrich), 1? insulin-transferrin-selenium (ITS)
(5 mg/l insulin, 5 mg/l transferrin, 5 mg/l selenium), 4 nM activin
A, 1 mM sodium butyrate, and 50 lM 2-mercaptoethanol. The
cells were cultured in this medium for 2 days. On the third day,
the medium was changed to SFM-B, which contained DMEM/
F12 (1:1) with 17.5 mM glucose, 1% BSA, ITS, and 0.3 mM tau-
rine. The cell aggregates were cultured in this medium for
another 2 days and shifted to SFM-C on the fifth day. SFM-C
contained DMEM/F12 (1:1) with 17.5 mM glucose, 1.5% BSA,
ITS, 3 mM taurine, 100 nM glucagon-like peptide (GLP)-1 (am-
ide fragment 7–36; Sigma Aldrich), 1 mM nicotinamide, and 1?
nonessential amino acids (NEAAs). The cell aggregates were fed
with fresh SFM-C every 2 days for another 5 days. All chemicals
were purchased from Sigma Aldrich unless otherwise indicated.
Quantitative Real-Time Polymerase Chain Reaction
Tissue/cell samples were frozen in Trizol (Invitrogen, Carlsbad,
CA, http://www.invitrogen.com). Total RNA was isolated from
duplicate and triplicate samples, as per the manufacturer’s
instructions, measured on an ND-100 spectrophotometer (Nano-
Drop Technologies, Wilmington, DE, http://www.nanodrop.com),
and 2 lg RNA was used for cDNA synthesis per 20-ll reaction.
cDNA was amplified using a reverse transcription system kit
(ImProm-II Reverse Transcription System, #A3800; Promega,
Madison, WI, http://www.promega.com). The primer sequences
used for quantitative real-time polymerase chain reaction (qRT-
PCR) are summarized in supporting information Table 1. qRT-
PCR was performed in duplicate and triplicate with a 25-ll total
reaction mixture containing 1? Power SYBR Green PCR
Master Mix (Applied BioSystems, Foster City, CA, http://www.
appliedbiosystems.com), 600–750 nM for each forward and
reverse primer using 1/20th of the cDNA preparation. PCR ampli-
fication was carried out using the Applied BioSystems 7300
Real-Time PCR System (SDS v13.1; Applied BioSystems) and
stopped at 40 cycles (program: 2 minutes at 50?C, 10 minutes at
95?C, and 40 cycles of 15 seconds at 95?C and 1 minute at
60?C). For quantitative analysis of the fold change by qRT-PCR,
the initial Ct values were assigned to be 39 when transcript levels
were undetectable. All qRT-PCR results were normalized to glyc-
eraldehyde-3-phosphate dehydrogenase carried out in a duplex
reaction to correct differences in RNA input. For estimation of
the fold change by qRT-PCR when the initial transcript levels
were undetectable, the initial Ct value was assigned to be 39,
which would lead to a possible underestimation of the actual fold
change. The qRT-PCR values are shown as the mean ? standard
error of the mean (SEM) and represented as the fold difference
over detectable (Ct value of 39).
Immunophenotyping (Flow Cytometric Analysis)
Flow cytometry was performed on a BD FACSCanto flow cytom-
eter (BD Biosciences, San Diego, http://www.bdbiosciences.com),
and analysis was done with BD FACSDiva software (version
5.0). Cells were harvested in 0.25% trypsin-EDTA, washed in
flow cytometry buffer (1? PBS, 0.5% FBS, 0.5% BSA, 0.05%
sodium azide). Cells were stained with phycoerythrin (PE)- or flu-
orescein isothiocyanate-conjugated antibodies. ICAs were dissoci-
ated with 0.25% trypsin-EDTA (Sigma-Aldrich) at 37?C for 10–
12 minutes. Cells were washed with PBS containing 0.5% FBS,
fixed, and permeabilized with 90% chilled methanol for 15–20
minutes on ice. Cells were incubated in mouse antiproinsulin C-
peptide (1:100) at RT for 45 minutes followed by PE-conjugated
anti-mouse IgG for 30 minutes. Cells were incubated with PE-
conjugated anti-pancreatic and duodenal homeobox (PDX)-1 anti-
body (1:100) for 45 minutes at RT. A PE-conjugated isotype-
matched control was also performed for each experiment to avoid
nonspecific staining. The sources of antibodies and dilutions used
are summarized in supporting information Table 1.
Islet Neogenesis From ASCs
Immunocytochemistry (Confocal Microscopy)
Cultured cells or ICAs were fixed with freshly prepared 4% para-
formaldehyde (PFA) for 10 minutes at RT and permeabilized
with either 0.1% Triton X-100 for 3–5 minutes on ice or with
50% chilled methanol for 20 minutes on ice. Blocking was done
with blocking buffer (PBS, 0.5% FBS, and 0.5% BSA) for
30 minutes at RT. Primary antibodies were incubated overnight
at 4?C, washed with PBS, and then incubated with secondary
antibodies at RT for 1 hour. Slides were washed in PBS
and mounted with Vectashield (Vector Laboratories, Burlingame,
CA, http://www.vectorslab.com). 40,6-diamidoino-2-phenylindole
(Invitrogen) was used to visualize nuclei. The sources of antibod-
ies and dilutions used are summarized in supporting information
Table 1. Confocal images were captured using a Zeiss-LSM 510
laser scanning microscope (Carl Zeiss MicroImaging, Inc., Thorn-
wood, NY, http://www.zeiss.com) using a 63?/1.3 oil objective
with optical slices of ?1–2 lm. All results are representative
fields confirmed from at least five different experiments.
In Vitro Multilineage Differentiation Studies
Adipogenesis, chondrogenesis, and osteogenesis for mE-ASCs
was evaluated in the appropriate induction media according to the
manufacturer’s protocol (Cambrex, Walkersville, MD, http://
www.cambrex.com). The differentiation phenotype was docu-
mented using oil red O for adipocytes, saffranin-O staining for
chondrocytes, and alizarin staining for osteocytes. Dithizone
(DTZ) (Sigma-Aldrich) stain of 10 mg/ml dimethyl sulfoxide
concentration was used to stain ICAs.
Total Insulin Content and Release Assays
For the glucose-stimulated insulin release assay, about 200–300
day 10 ICAs were handpicked in an eppendorf tube. ICAs were
then washed three times with PBS and incubated with freshly
(120 mM NaCl, 5 mM Kcl, 2.5 mM CaCl2, 1.1 mM MgCl2, and
25 mM NaHCO3, with 10 mM HEPES buffer and 0.1% BSA)
without glucose for 3–6 hours. ICAs were incubated with 100 ll
KRBH buffer containing 5.5 mM glucose for 1 hour at 37?C. The
supernatant was collected and the same ICAs were further incu-
bated in 22 mM glucose for an additional hour at 37?C. C-peptide
concentrations were measured using a Mouse C-Peptide IþII EIA
Kit (#YK013; Yanaihra Institute Inc., Shizuoka, Japan, http://
www.yanaihara.co.jp) and the values were normalized to DNA con-
tent. Intracellular insulin was extracted by incubating ICAs/mouse
islets overnight in acid ethanol (18 ml 10M HCl/l, 70% ethanol) at
4?C. The total insulin content of ICAs was estimated after sonicat-
ing the ICA pellet in 200 ll acid/ethanol. For DNA estimation,
equal numbers of ICAs were taken and DNA was isolated using the
DNeasy tissue kit (Qiagen, Hilden, Germany, http://www1.qiagen.
com) as per the manufacturer’s instructions.
Hoechst 33342/Propidium Iodide Staining
for Cell Viability
To determine the viability of ICAs retrieved after 4 weeks of
transplantation, ICAs were washed with PBS and incubated with
100 lg/ml propidium iodide (PI) (Sigma-Aldrich) for 20 minutes.
ICAs were then fixed with freshly prepared 4% PFA for
10 minutes at RT, washed with PBS, and incubated with 0.1 mM
Hoechst 33342 (Sigma-Aldrich) . Confocal images were cap-
tured using a Zeiss-LSM510 laser scanning microscope (Carl
Zeiss MicroImaging, Inc.) using a 63?/1.3 oil objective with op-
tical slices of ?1–2 lm. Cell death was indicated by dual staining
of Hoechst 33342 (blue) and PI (red).
In Vivo Transplantation Studies
Male Swiss albino mice, aged 8–10 weeks, were used for trans-
plantation studies. STZ (Sigma-Aldrich) was injected i.p. at
160 mg/kg body weight, freshly dissolved in citrate buffer (pH
4.5). Blood glucose (BG) was measured from the snipped tail
using Accu-Chek (F. Hoffmann-La Roche Ltd, Basel, Switzerland,
http://www.roche.com). Only mice with BG levels stably >300
mg/dl after the STZ injection were used for transplantation stud-
ies. For transplantation, ?1,000–1200 day 10 ICAs were washed
with PBS, suspended in 100 ll of sodium alginate solution—1.2%
w/v alginic acid (Sigma) in 0.85% saline—packed into a biocom-
patible capsule—polyurethane-polyvinyl pyrrolidone interpenetrat-
ing network (PU-PVP-IPN)—developed by the Sree Chitra Tiru-
nal Institute for Medical Sciences & Technology (Kerala, India,
www.sctimst.ac.in)  (supporting information Fig. 2). The ICA-
alginate capsules were sealed and dipped into 0.1 N acetic acid
and CaCl2(0.15 M in distilled water) solution for gelling. These
encapsulated ICAs were then implanted into the peritoneal cavity
of diabetic mice. Fasting BG levels were measured regularly using
a glucometer after the mice had fasted for 6 hours.
A glucose tolerance test was performed by i.p. injection of
glucose (2 g/kg body weight) after a 6-hour fast, and BG levels
were measured at 15, 30, 60, 90, and 150 minutes postinjection
(n ¼ 3). After 4 weeks, the transplanted capsules were removed
from those mice whose hyperglycemia was rescued. ICAs were
collected after opening the capsule and total RNA was prepared
for qRT-PCR for the insulin gene transcript. Insulin and C-pep-
tide expression in the transplanted ICAs was detected by
Transmission Electron Microscopy
Mature ICAs at day 10 were fixed in 2.5% phosphate buffer (0.1 M,
pH 7.4) glutaraldehyde solution for 24 hours at 4?C. The ICAs were
then postfixed in 1% OsO4(osmium tetraoxide) for 2 hours at 4?C.
Samples were then dehydrated through a graded ethanol series and
propylene oxide and embedded with Quetol 651 embedding kit
(Polysciences, Inc., Warrington, PA, http://www.polysciences.
com). Ultrathin sections were cut with a diamond knife and stained
with 5% uranyl acetate and 2% lead citrate. Samples were observed
using a CM200 Philips Transmission electron microscope (Philips
Electronic Instruments, The Netherlands).
Values are expressed as the mean fold change ? SEM from three
different experiments unless otherwise indicated. The statistical
analysis was done using Student’s two-tailed t-test to determine
the significance between different conditions. Prism4 software
(GraphPad Software Inc., La Jolla, CA, http://www.graphpad.
com) was used for all analyses.
Isolation and Characterization of mE-ASCs From
Murine Epididymal Fat Pads
mE-ASCs were isolated from the fat pads of male Swiss
albino mice by previously described methods of ASC isola-
tion . The initial culture of the stromal vascular fraction
resulted in the growth of a plastic adherent cell population
with typical mesenchymal morphology. Although freshly iso-
lated mE-ASCs had a heterogeneous phenotype in culture,
single fibroblastoid cell populations that exhibited a homoge-
neous morphology were clonally expanded. Six clones of mE-
ASCs were evaluated in all the differentiation studies, and all
experiments were carried out using mE-ASCs of passages 4–
8. The homogeneity of the cloned population was confirmed
by the fluorescence-activated cell sorting (FACS) analysis of
mE-ASCs at passage 4, which showed uniform coexpression
of CD44, CD29, and Sca-1 surface markers (Fig. 1A). Unlike
human ASCs reported in earlier studies, mE-ASCs exhibited
CD73 (13.1% ? 8.4%), CD90 (4.3% ? 0.7%), and CD105
Chandra, G, Phadnis et al.
(10.4% ? 8.1%) (supporting information Fig. 1). mE-ASCs
were positive for the cytoskeletal proteins nestin, vimentin,
and a-smooth muscle actin, the extracellular matrix proteins
collagen I and fibronectin, and the proliferation marker Ki-67
(Fig. 1B). mE-ASCs exhibited in vitro competence to differ-
entiate into adipogenic, osteogenic, and chondrogenic lineages
upon specific induction as confirmed by Oil red O, alizarin
red, and saffranin O staining, respectively (Fig. 1C).
We postulated that mE-ASCs would be an ideal stem cell
candidate to differentiate into the pancreatic lineage because
expanded mE-ASCs at passage 4 with two-color flow cytometric analysis. (A): Clonally expanded mE-ASCs exhibited a uniform expression pro-
file for CD44, CD29, and Sca-1 surface antigens (n ¼ 6). Representative FACS profile of mE-ASCs showing dual positivity for CD44-CD29 and
CD44-Sca-1 as compared with the isotype control. (B): Immunofluorescence confocal images of mE-ASCs at passage 4, for the cytoskeletal
markers nestin (a), vimentin (b), a-SMA (c), the ECM proteins fibronectin (d) and collagen type I (e), and the proliferative marker Ki67 (f).
Cell nuclei were stained with DAPI. (C): mE-ASCs undergo in vitro differentiation into the adipogenic, chondrogenic, and osteogenic lineages.
Adipogenesis was demonstrated by the accumulation of neutral lipid vacuoles stained with oil red O (a). Calcium depositions in ECM stained
by alizarin red indicate osteogenic differentiation (b). Chondrogenesis was demonstrated by sulfated proteoglycans stained with saffranin O (c).
Scale bar, 20 lm. Abbreviations: a-SMA, a-smooth muscle actin; DAPI, 40,6-diamidoino-2-phenylindole; ECM, extracellular matrix; FACS,
fluorescence-activated cell sorting; FITC, fluorescein isothiocyanate; mE-ASC, murine epididymal adipose tissue-derived stem cell; PE,
In vitro characterization of mE-ASCs for mesenchymal stem cell markers. Expression of cell surface markers was assessed in clonally
Islet Neogenesis From ASCs
they exhibit transcript levels of the pancreatic genes (isl-1 and
pax-6) required for b-cell development . Using modified
protocols of  and , we formulated a 10-day differentia-
tion protocol for the conversion of mesodermic mE-ASCs to
endodermal pancreatic endocrine cells (Fig. 2).
Differentiation of mE-ASCs Into Definitive
mE-ASCs that proliferated as an adherent monolayer aggre-
gated into spherical cell clusters when the medium was
changed from serum-containing medium to day 1 SFM (SFM-A)
ferentiation protocol is divided into three stages. Serum-free DMEM/F12 medium supplemented with stage-specific growth factors was termed
SFM-A, SFM-B, and SFM-C. In the first stage, definitive endoderm differentiation was achieved with ITS, 4 nM activin A, 0.5 mM sodium butyr-
ate, and 50 lM 2-mercaptoethanol in low-adherent culture conditions. Pancreatic endoderm was induced in stage 2 with 0.3 mM taurine, and finally
pancreatic hormone-expressing ICAs were subsequently induced with GLP-1, niacin, and a supraphysiological level of taurine (3 mM) for 5 days.
(B): Phenotypic changes in mE-ASCs during the 10-day differentiation procedure (a-d) (see also supporting information data). Most of the day 10
ICAs stained with DTZ, a zinc-chelating agent known to selectively stain pancreatic b cells because of their high zinc content (e). Abbreviations:
BSA, bovine serum albumin; DMEM, Dulbecco’s modified Eagle’s medium; DTZ, dithizone; mE-ASC, murine epididymal adipose tissue-derived
stem cell; GLP-1, glucagon-like peptide 1; ICA, islet-like cell aggregate; ITS, insulin-transferrin-selenium; SFM, serum-free medium.
Generation of ICAs from mE-ASCs. (A): Schematic representation of stepwise protocol for generating ICAs from mE-ASCs. The dif-
Chandra, G, Phadnis et al.
(Fig. 2 and supporting information video 1). Cellular aggrega-
tion occurred as a gradual process and complete aggregates
were formed by 24–36 hours of incubation in SFM-A. The cells
formed tight clusters (100–200 lm) that resembled pancreatic
islets, and henceforth these clusters were termed ICAs. Immu-
nocytochemical and qRT-PCR analysis of ICAs after a 48-hour
exposure to SFM-A showed that ITS in combination with acti-
vin A and sodium butyrate induced mE-ASCs to differentiate
into DE. A majority of the differentiated cells (>70%) showed
greater expression of primitive endodermal markers, like a-
fetoprotein and nuclear b-catenin, at day 2 than undifferentiated
mE-ASCs (Fig. 3A). Endodermal differentiation was further
confirmed by immunostaining of day 2 ICAs with DE markers
like Sox17 (Fig. 3B.a), Foxa2 (Fig. 3B.b), and hepatocyte nu-
clear factor (HNF)-1b (Fig. 3B.c). The differentiation efficiency
for DE of day 2 ICAs was in the range of 20%–30%. We found
that 27.7% ? 3.8% (p ¼ .006) of cells expressed Sox17 and
22.3% ? 2.5% (p ¼ .004) of cells expressed Foxa2 within day
2 ICAs. We also observed considerable expression of the gut
tube marker HNF-1b in day 2 ICAs. qRT-PCR analysis of day
2 ICAs showed a greater transcript abundance of DE markers
like HNF-3b/FoxA2 (103-fold), GATA-4 (105-fold), and CK-
19 (103-fold) than in undifferentiated mE-ASCs (Fig. 3C), sub-
stantiating our immunostaining results.
Pancreatic Endoderm Differentiation of ICAs
To further investigate whether ICAs could differentiate into
pancreatic endoderm, we incubated day 2 ICAs in SFM sup-
plemented with taurine (SFM-B), an NEAA involved in the
development of pancreatic b cells [21, 22]. After 2 days of
incubation in SFM-B, we observed significantly greater
expression of pancreatic endoderm markers, like PDX-1
(102.5-fold), Ngn3 (102-fold), NeuroD (102.5-fold), Pax-4 (104-
fold), Pax-6 (102-fold), Nkx2.2 (105-fold), Nkx6.1 (103.5-
fold), and glucose transporter (Glut)-2 (107-fold), than in
undifferentiated mE-ASCs (Fig. 4A). FACS analysis to quan-
titate the percentage of PDX-1þcells in day 5 ICAs that had
undergone pancreatic endoderm differentiation revealed that
64.84% ? 7.038% (p ¼ .069) of cells in day 5 ICAs
expressed PDX-1, significantly greater than in undifferentiated
mE-ASCs (n ¼ 3) (Fig. 3D).
Maturation of ICAs Toward Pancreatic
ICAs matured to pancreatic hormone-expressing cells when
exposed to SFM supplemented with b-cell maturation factors
like GLP-1 and nicotinamide (SFM-C). After 5 days of incu-
bation in SFM-C, most of the day 10 ICAs stained positive
for DTZ, a zinc-chelating agent known to selectively stain
pancreatic b cells (Fig. 2B.e). Our data demonstrate higher
transcript levels of insulin, somatostatin, and panpolypeptide
genes in day 10 than in day 5 ICAs. We observed that gluca-
gon gene transcripts were only expressed in the ICAs by day
10 of differentiation. It was interesting to note that undifferen-
tiated mE-ASCs consistently exhibited low levels of somato-
statin and ghrelin transcripts (Fig. 5A). Further quantification
by flow cytometry showed that 48.17% ? 3% (p ¼ 0.001) of
cells constituted the C-peptide-positive fraction of day 10
ICAs (n ¼ 3). Undifferentiated mE-ASCs did not show any
expression of C-peptide (Fig. 5B). These results were further
confirmed by immunofluorescence staining of matured day 10
ICAs for islet-specific transcription factors and hormones.
Matured ICAs exhibited abundant expression of PDX-1, C-
peptide, insulin, glucagon, somatostatin, pancreatic polypep-
tide, and Glut-2 (Fig. 6A–E, G, H). A majority of the ICAs
generated showed coexpression of insulin, glucagon, and
somatostatin (Fig. 6F) similar to that observed during pancre-
atic development in mice/humans [31, 32]. In addition to en-
docrine hormones, fewer cells within ICAs were also found to
be positive for the exocrine hormone amylase (Fig. 6I). TEM
of day 10 ICAs revealed a structure typical of a secretory cell
with a few large vacuoles and electron-dense secretory
vesicles. These cells contained granules with a morphological
similarity to adult pancreatic islet b cells (Fig. 5C).
Further characterization of ICAs using qRT-PCR revealed
that the surface antigens expressed by undifferentiated mE-
ASCs (CD44, CD29, and Sca-1) were downregulated during
the course of differentiation. In addition, the downregulation of
the proliferative marker Ki-67 at the end of differentiation evi-
dently showed regressed cell proliferation and active differen-
tiation in mE-ASCs (Fig. 4B). The gene-profiling studies for
AT-specific markers during ICA differentiation showed inter-
esting results. Transcript levels of leptin and adiponectin in
mature ICAs were comparable with those of mE-ASCs. Strik-
ingly, however, the expression of the insulin mimetic adipokine
visfatin  showed significant upregulation in ICAs during
the course of differentiation (supporting information Fig. 2).
Static Stimulation and Total Insulin Content
of Mature ICAs
In this study, day 10 ICAs were first incubated with 5.5 mM
glucose for 1 hour to measure the basal level of C-peptide
release. The same ICAs were subsequently incubated for an
additional 1 hour in 22 mM glucose to measure glucose-
stimulated C-peptide release (n ¼ 3). The total C-peptide con-
tent of day 10 ICAs measured was up to 0.1852 ng/lg DNA
per 60 minutes when exposed to 5.5 mM glucose (p ¼
.0179). When stimulated with 22 mM glucose, the total C-
peptide content of ICAs increased to 0.275 ng/lg DNA per
60 minutes (p ¼ .05), confirming their in vitro functionality
to respond to glucose (Fig. 6J). The total intracellular insulin
content of day 10 ICAs was also analyzed and the value was
compared with that of mouse islets. It was observed that mE-
ASC-derived ICAs contained 0.4420 ? 0.2036 ng/lg DNA (p
¼ .162), compared with 3.488 ? 1.513 ng/lg DNA (p ¼
.2065) for insulin in adult mice islets (n ¼ 4) (Fig. 6K).
Mature ICAs Restore Normoglycemia
in Diabetic Mice
The physiological competence of mature day 10 ICAs to
maintain glucose homeostasis in vivo was evaluated by trans-
plantation of ICAs in STZ-induced diabetic mice. ICAs
(1,000–1,200) encapsulated in calcium alginate were packed
into PU-PVP-IPN biocompatible capsules  and trans-
planted into the peritoneal cavity of diabetic Swiss albino
mice (n ¼ 6) (supporting information Fig. 3). ICA-trans-
planted mice has significantly lower BG levels than sham-
operated and diabetic control mice over a time period of 2
weeks (Fig. 7A). Transplantation of undifferentiated mE-
ASCs (n ¼ 3) and day 2 ICAs (n ¼ 3) to STZ-induced dia-
betic mice showed a hypoglycemic effect. However, the mice
maintained a BG level of 200–250 mg/dl, even 4 weeks after
transplantation, and failed to restore normoglycemia within
the stipulated time, when compared with day 10 ICAs (Fig.
7A). An i.p. glucose tolerance test to measure glucose clear-
ance in the experimental mice group (n ¼ 3) demonstrated
that, in nondiabetic as well as ICA-transplanted mice, BG lev-
els shot up during the initial dose of glucose, which reached a
plateau level within 2 hours, compared with diabetic control
mice, which failed to show glucose clearance (Fig. 7B). ICAs
retrieved from the transplanted mice after 4 weeks showed
cluster integrity (Fig. 7C) and viability (supporting informa-
tion Fig. 4). The retrieved ICAs showed insulin and C-peptide
Islet Neogenesis From ASCs
ITS, activin A, and sodium butyrate, greater expression of AFP and nuclear b-catenin (markers of primitive endoderm) was observed in day 2
ICAs than in undifferentiated mE-ASCs. Cell nuclei were stained with DAPI. (B): Immunofluorescence analysis was performed on day 2 ICAs
for the expression of definitive endoderm markers like Sox17 (a), Foxa2 (HNF3B) (b), and HNF1B (c) (200? magnification). Nuclei were stained
with DAPI. Positive cells are pink (merged red of AlexaFluor 546 and blue of DAPI) in color at the nuclei, indicated by arrows. (C): Expression
of genes for definitive endoderm markers like Foxa2, GATA-4, and CK-19 was analyzed by qRT-PCR in day 2 and day 5 differentiated ICAs
and compared with undifferentiated mE-ASCs. All mRNA expression levels were normalized to the housekeeping gene GAPDH expression. Data
are presented as mean ? SEM fold increase over detectable (Ct, 39). (D): Flow cytometric analysis was performed to determine the expression
of PDX-1 in day 5 ICAs (lower panel) in comparison with undifferentiated mE-ASCs (upper panel). Isotype-matched antibody control was also
used to eliminate background staining (upper and lower left panels). Scale bar, 20 lm. Abbreviations: AFP, a-fetoprotein; CK-19, cytokeratin-19;
DAPI, 40,6-diamidoino-2-phenylindole; FSC, forward scatter; GAPDH, glyceraldehyde-3-phosphate-dehydrogenase; HNF, hepatocyte nuclear fac-
tor; ICA, islet-like cell aggregate; ITS, insulin-transferrin-selenium; mE-ASC, murine epididymal adipose tissue-derived stem cell; PDX-1, pan-
creatic and duodenal homeobox 1; PE, phycoerythrin; qRT-PCR, quantitative reverse transcription-polymerase chain reaction; SEM, standard
error of the mean; SFM, serum-free medium.
Characterization of mE-ASC-derived ICAs for endoderm markers. (A): After 2 days of treatment with SFM-A supplemented with
formed for pancreatic lineage genes like Pdx-1, Ngn3, NeuroD, Pax4, Nkx2.2, Nkx6.1, Pax6, Isl1, and Glut-2 at day 5 (D5) and day 10 (D10) of
differentiation. (A): Gene transcripts of ICAs were compared with those of undifferentiated mE-ASCs (passage 4, D0) and 8-week-old mouse
(Swiss albino) whole pancreas (m-Pan). Relative levels of gene expression were normalized to the GAPDH mRNA level. Transcript abundance is
shown here as the mean ? SEM fold increase for two preparations and represented as fold over detectable (Ct, 39). (B): Brief overview of tem-
poral sequence of pancreatic transcription factors known to be expressed during development. (C): qRT-PCR was performed to analyze the status
of mesenchymal markers like CD44, Cd29, Sca-1, vimentin, and nestin, and the proliferation marker Ki67 during the differentiation process.
Samples were collected at day 2 (D2), day 5 (D5), and day 10 (D10) and compared with undifferentiated mE-ASCs (UD). Relative levels of
gene expression were normalized to the GAPDH mRNA level and are presented here as the mean ? SEM fold increase over detectable (Ct, 39).
Abbreviations: qRT-PCR, quantitative reverse transcription polymerase chain reaction; cDNA, complementary DNA; GAPDH, glyceraldehyde-3-
phosphate-dehydrogenase; Glut-2, glucose transporter 2; ICA, islet-like cell aggregate; mE-ASC, murine epididymal adipose tissue-derived stem
cell; qRT-PCR, quantitative reverse transcription-polymerase chain reaction; SEM, standard error of the mean.
ICAs derived from mE-ASCs expressed various genes involved in pancreas development. SYBR Green-based qRT-PCR was per-
Islet Neogenesis From ASCs
matostatin, pan-pol, and ghrelin. Total RNA was isolated and cDNA was prepared for day 5 (D5) and day 10 (D10) samples which were com-
pared with undifferentiated mE-ASCs (passage 4, D0) and 8-week-old mouse (Swiss albino) whole pancreas (m-Pan). Relative levels of gene
expression were normalized to the GAPDH mRNA level and are presented here as the mean ? SEM fold increase over detectable (Ct, 39). (B):
Quantitative expression of C-peptide was evaluated by flow cytometry for undifferentiated mE-ASCs and day 10 ICAs along with isotype-
matched antibody control. (C): Transmission electron microscopy of day 10 ICAs showed cells that were polyhedral in shape and their spatial
relationship to adjacent cells. (D): Electron-dense secretory granules are seen throughout the cytoplasm of the cell. Scale bars, 2 lm (C) and
1 lm (D). Abbreviations: GAPDH, glyceraldehyde-3-phosphate-dehydrogenase; ICA, islet-like cell aggregate; pan pol, pancreatic polypeptide;
mE-ASC, murine epididymal adipose tissue-derived stem cell; PE, phycoerythrin; qRT-PCR, quantitative reverse transcription polymerase chain
reaction; SEM, standard error of the mean.
Expression of pancreatic endocrine hormone genes in ICAs. (A): SYBR Green qRT-PCR was performed for insulin, glucagon, so-
Chandra, G, Phadnis et al.
PDX-1 (A), C-peptide (B), insulin (C), glucagon (D), and somatostatin (E). (F): Immunofluorescence analysis for coexpression of insulin, gluca-
gon, and somatostatin in day 10 ICAs. Representative confocal optical slice of day 10 ICAs for insulin (red), glucagon (green), and somatostatin
(cyan). The pattern of coexpression was similar to that in the earliest stages of murine pancreatic development. Few cells within day 10 ICAs
were found to express pancreatic polypeptide protein (G), Glut-2 (H), and the exocrine hormone amylase (I). Cell nuclei were stained with
DAPI. Scale bar, 20 lm. (J): Static stimulation assay for day 10 ICAs. ICAs were sequentially treated with a low glucose (5.5 mM) and high
glucose (22 mM) concentration in triplicate for 1 hour. Cell supernatants were collected and analyzed for C-peptide release by EIA assay and
normalized to DNA content (p-values ? .05) (n ¼ 3). (K): Total intracellular insulin content of day 10 ICAs (p ¼ .162) was measured and com-
pared with that of normal adult mouse islets (p ¼ .2065) (n ¼ 3). Abbreviations: DAPI, 40,6-diamidoino-2-phenylindole; Glut-2, glucose trans-
porter 2; ICA, islet-like cell aggregate; PDX-1, pancreatic and duodenal homeobox 1.
Analysis of pancreatic hormone-expressing ICAs with immunofluorescence. Confocal optical section of day 10 ICAs stained for
Islet Neogenesis From ASCs
protein expression (Fig. 7E) as well as insulin gene transcripts
comparable with those of matured day 10 ICAs (Fig. 7D).
In the present study, we demonstrate the potential of ASCs to
differentiate into insulin-producing ICAs. Our experimental
design involves the growth and proliferation of multipotent
cells with a CD44þCD29þSca-1þsurface antigen expression
profile within the stromal vascular fraction of AT (murine epi-
didymal fat pads). These cells were then induced to differentiate
into the pancreatic endocrine lineage following stringent proto-
cols. We further demonstrated the in vitro and in vivo function-
ality of the differentiated ICAs, substantiating the potential of
using ICAs for possible cell replacement therapy for diabetes.
The first challenge of our study was to differentiate meso-
dermic mE-ASCs into the endodermal lineage, which is criti-
cal in generating pancreatic hormone-expressing cells. We
achieved this goal by inducing the cells to grow in SFM sup-
plemented with activin A, sodium butyrate, 2-mercaptoetha-
nol, and ITS (SFM-A). The high-glucose, SFM curbed cell
proliferation and enhanced differentiation. Previous reports
have demonstrated the potential of activin A to induce undiffer-
entiated human ES cells to undergo endoderm differentiation
restored normoglycemia within 2 weeks. Day 10 ICAs (1,000–1,200) were encapsulated in calcium alginate, packed into biocompatible capsules (made
of PU-PVP-IPN), and transplanted i.p. into diabetic mice. The control group of animals included a sham control of diabetic mice transplanted with an
empty capsule (n ¼ 4), diabetic mice transplanted with encapsulated undifferentiated mE-ASCs, and diabetic mice transplanted with encapsulated day
2 ICAs. (A): Overnight fasting blood glucose levels of mice in each experimental group for the 3-week time period are shown (B): A standard 2-hour
IP-GTT was performed on normal, STZ-treated diabetic, and ICA-transplanted diabetic mice. Glucose (2 g/kg body weight) was injected i.p. and glu-
cose readings were monitored at regular intervals. On retrieval from the transplanted mice, the ICAs exhibited cellular integrity and functionality. (C):
Phase contrast image of ICAs retrieved from transplanted mice after 4 weeks. (D): The retrieved ICAs show an equal level of insulin gene expression to
that of day 10 ICAs. (E): Immunostaining of the retrieved ICAs for the expression of C-peptide (green) and insulin (red). Cell nuclei were stained with
DAPI. Abbreviations: DAPI, 40,6-diamidoino-2-phenylindole; ICA, islet-like cell aggregate; IP-GTT, i.p. glucose tolerance test; mE-ASC, murine epi-
didymal adiposetissue-derived stem cell; PU-PVP-IPN, polyurethane-polyvinylpyrrolidone interpenetrating network; STZ, streptozotocin.
In vivo characterization of mE-ASC-derived ICAs. Transplantation of day 10 ICAs in STZ-induced diabetic Swiss albino mice (n ¼ 6)
Chandra, G, Phadnis et al.
[25, 6, 7]. We found that a 48-hour exposure to 10 ng/ml activin
A in combination with 1 mM sodium butyrate, a histone deace-
tylase inhibitor , could induce mE-ASCs to undergo endo-
dermal differentiation. This was confirmed by the expression of
definite endodermal markers like Sox17 , Foxa2 ,
HNF-1b, CK-19, and GATA-4 in the differentiated cells. We
observed that the endodermal differentiation efficiency in day 2
ICAs was in the range of 20%–25%, as revealed by Sox17 and
Foxa2 immunostaining (Fig. 3B). HNF-1b, required for visceral
endoderm differentiation , was expressed in most of the
cells of day 2 ICAs. ITS and 2-mercaptoethanol largely pro-
tected mE-ASCs from the stress induced by serum-free culture
conditions . The low cell attachment culture dish proved to
be highly efficient in helping cells to form three-dimensional
aggregates. By 24–36 hours after initial plating, mE-ASCs
formed clusters (ICAs) with a compact cellular aggregation.
We found that the initial plating density was a critical parame-
ter in forming good cell aggregates. An initial cells density of 2
? 106cells generated around 200–300 ICAs.
The second phase of our study was focused on detouring
ICAs to the pancreatic endodermal lineage. This was accom-
plished by adapting an earlier described protocol of supple-
menting taurine in the SFM cocktail (SFM-B) . Taurine is
an NEAA required for the development of functional b cells
[21, 22]. We found that a physiological dose of taurine
(0.3 mM) at day 3 of differentiation induced ICAs to acquire
traits of pancreatic endoderm, as confirmed by enhanced tran-
script levels of pancreatic endoderm markers like PDX-1,
Ngn3, NeuroD, Pax4, Nkx2.2, etc. in day 5 ICAs. Later
stages of differentiation largely focused on the maturation of
ICAs into pancreatic hormone-expressing cells. SFM cocktail
supplemented with GLP-1, nicotinamide, and supraphysiologi-
cal levels of taurine (3 mM) directed ICAs to mature into
pancreatic hormone-expressing cells by day 10. GLP-1 is an
incretin known to enhance the secretion of insulin from b
cells and also stimulates insulin synthesis, by increasing the
level of insulin mRNA and stabilizing it [24, 30]. Considering
the short plasma half-life of GLP-1, mainly because of pro-
teolytic degradation, we replenished GLP-1 in the designed
SFM (SFM-C) every 24 hours to continuously stimulate ICAs
to maturity. Immunofluorescence studies showed that a major-
ity of the cells constituting ICAs coexpressed insulin, gluca-
gon, and somatostatin. This observation is in accordance with
normal pancreatic development, in which immature islets are
known to coexpress pancreatic hormones [31, 32]. The differ-
entiation protocol standardized in our study is highly efficient
because it generates 57%–71% PDX-1þcells, whereas, to the
best of our knowledge, the maximum number of PDX-1þ
cells generated in any pancreatic differentiation study using
nonpancreatic stem cells is 15%–25% . In addition, our
protocol generates the maximum yield of C-peptide-positive
cells (47%–51%). Our results are encouraging because the
maximum efficiency of C-peptide-positive cells reported for
ICAs derived from nonpancreatic sources is only 2%–8% .
We further confirmed the in vitro functionality of ICAs
by static stimulation assays and quantified a stimulation index
of 1.67 for C-peptide. The total intracellular insulin content in
comparison with mice islets helped in determining the effi-
ciency of the ICAs generated. We found that, although mE-
ASCs showed appreciable potential to differentiate into the
pancreatic lineage, the efficiency of ICAs was only 12% with
respect to insulin content when compared with mature mice
islets. The ultrastructural features typical of an adult pancre-
atic b cell found within the day 10 ICAs strengthen our dif-
ferentiation approach .
For the in vivo functionality studies, we adopted the well-
established calcium alginate encapsulation method [34, 35],
and with a future translational approach in mind we made use
of biocompatible capsules made of PU-PVP-IPN  for pack-
ing the mature ICAs. Our initial studies have shown that a min-
imum of 1,000–1,200 ICAs per capsule per mouse is required
to achieve significant normoglycemia in STZ-induced diabetic
mice. The capsule exhibited biocompatibility and did not
evoke any immune rejection in implanted mice. The porosity
of the capsule is such that it easily allows nutrient exchange
between the packed ICAs and the peritoneal cavity. The cap-
sule significantly improved the ease of delivery of ICAs to dia-
betic mice with minimum surgical invasion. ICAs largely
improved the glucose homeostasis of diabetic mice, and nor-
moglycemia was restored within 2 weeks of implantation and
maintained for 1 month of tracking. The capsule also helped in
retaining the ICA architecture and viability. The transcript lev-
els of insulin were similar to that of day 10 ICAs even after re-
trieval from the mice after 4 weeks. Transplantation of undif-
ferentiated mE-ASCs showed moderate lowering of BG levels.
However, normoglycemia was not restored in these mice, sug-
gesting the need for pancreatic lineage-specific stimulation of
these cells. This is the first report adopting a truly translational
approach to cell replacement therapy adhering to pragmatic
ways of islet transplantation with minimal surgical invasion.
It was interesting to quantitate the level of adipokines dur-
ing the differentiation course of ICAs from mE-ASCs. Adipo-
kines like leptin, adiponectin, and visfatin have been reported to
play a significant regulatory role in glucose homeostasis. Visfa-
tin regulates b-cell function through extracellular biosynthesis
of nicotinamide mononucleotide . We found that the level
of adipokines remained stable throughout differentiation and
only the transcript level of visfatin, the major insulin mimetic
adipokine, was upregulated in ICAs. This observation calls for
detailed study to analyze whether adipokines have any regula-
tory role in glucose homeostasis in mE-ASC-derived ICAs.
Our study clearly demonstrates the potential of ASCs to
differentiate into the pancreatic lineage. The results mentioned
in our study are highly reproducible and of all the six clones
used in this study exhibited equal frequency of in vitro differ-
entiation and maturation into ICAs.
The inherent expression of the isl-1 gene in ASCs and
adipokine-mediated glucose regulation make ASCs the emi-
nent candidate for alternative cell therapy for diabetes. The
in vivo glucose regulation exhibited by ICAs holds great
promise for future translation research. Timper et al.  pro-
vided initial clues to the potential of ASCs to differentiate
into insulin-producing cells. Our study is a step forward in
evaluating the detailed differentiation and in vitro/in vivo
functional potential of ASCs into the pancreatic lineage,
incorporating translational approaches of cell replacement
therapy. However, a few questions still need to be answered.
What is the minimum threshold number of ICAs required to
achieve the same glucose regulation as in a normal pancreas?
How does coexpression of glucagon and somatostatin with in-
sulin influence glucose-dependent insulin release in ICAs? To
the best of our knowledge, this is the first detailed report
demonstrating the differentiation potential of ASCs into pan-
creatic hormone-expressing cells. We anticipate that our work
will add to the ongoing research to find alternative sources of
islets for cell replacement therapy in diabetes.
In conclusion, our study clearly demonstrates the potential of
ASCs to differentiate into functional ICAs that bring about nor-
moglycemia upon transplantation into experimental diabetic
Islet Neogenesis From ASCs
mice. Thus, our studies suggest ASCs as an alternative, autolo- Download full-text
gous source for cell replacement therapy in diabetes.
The authors wish to thank the Director of NCCS, Dr. G. C. Mis-
hra, for all the support. We wish to thank Dr. Anandwardhan
Hardikar for all the suggestions. We thank Swapnil Walke for
FACS analysis and Ashwini Atre for assistance with confocal
microscopy. We also thank Amit Jaiswal of IIT-Mumbai for
helping with the TEM studies. Vikash Chandra and Swetha G
are supported by a fellowship from the University Grants Com-
mission, Governmentof India.
DISCLOSURE OF POTENTIAL CONFLICTS
The authors indicate no potential conflicts of interest.
1 Shapiro AM, Lakey JR, Ryan EA et al. Islet transplantation in seven
patients with type 1 diabetes mellitus using a glucocorticoid free
immunosuppressive regimen. N Engl J Med 2000;343:230–238.
Gunnarsson R, Klintmalm G, Lundgren G et al. Deterioration in glu-
cose metabolism in pancreatic transplant recipients given cyclosporine.
Assady S, Maor G, Amit M et al. Insulin production by human em-
bryonic stem cells. Diabetes 2001;50:1691–1697.
Lumelsky N, Blondel O, Laeng P et al. Differentiation of embryonic
stem cells to insulin-secreting structures similar to pancreatic islets.
Soria B, Roche E, Berna ´ G et al. Insulin-secreting cells derived from
embryonic stem cells normalize glycemia in streptozotocin-induced di-
abetic mice. Diabetes 2000;49:157–162.
D’Amour KA, Bang AG, Eliazer S et al. Production of pancreatic hor-
mone-expressing endocrine cells from human embryonic stem cells.
Nature Biotechnol 2006;24:1392–1401.
Jiang J, Au M, Lu K et al. Generation of insulin-producing islet-like
clusters from human embryonic stem cells. Stem Cells 2007;25:
Tang DQ, Cao LZ, Burkhardt BR et al. In vivo and in vitro character-
ization of insulin-producing cells obtained from murine bone marrow.
Yang L, Li S, Hatch H et al. In vitro trans-differentiation of adult he-
patic stem cells into pancreatic endocrine hormone-producing cells.
Proc Natl Acad Sci U S A 2002;99:8078–8083.
10 Seaberg RM, Smukler SR, Kieffer TJ et al. Clonal identification of
multipotent precursors from adult mouse pancreas that generate neural
and pancreatic lineages. Nat Biotechnol 2004;22:1115–1124.
11 Zuk PA, Zhu M, Ashjian P et al. Human adipose tissue is a source of
multipotent stem cells. Mol Biol Cell 2002;13:4279–4295.
12 Scha ¨ffler A, Bu ¨chler C. Concise review: Adipose tissue-derived stro-
mal cells—basic and clinical implications for novel cell-based thera-
pies. Stem Cells 2007;25:818–827.
13 Rabe K, Lehrke M, Parhofer KG et al. Adipokines and insulin resist-
ance. Mol Med 2008;14:741–751.
14 Kojima H, Fujimiya M, Matsumura K et al. Extrapancreatic insulin-
producing cells in multiple organs in diabetes. Proc Natl Acad Sci U
S A 2004;101:2458–2463.
15 Roy SS, Mukherjee M, Bhattacharya S et al. A new cell secreting in-
sulin. Endocrinology 2003;144:1585–1593.
16 Timper K, Seboek D, Eberhardt M et al. Human adipose tissue-
derived mesenchymal stem cells differentiate into insulin, somatosta-
tin, and glucagon expressing cells. Biochem Biophys Res Commun
17 Herceg Z, Wang ZQ. Failure of poly(ADP-ribose) polymerase cleav-
age by caspases leads to induction of necrosis and enhanced apoptosis.
Mol Cell Biol 1999;19:5124–5133.
18 George S, Nair PD, Risbud MV et al. Nonporous polyurethane mem-
branes as islet immunoisolation matrices—biocompatibility studies. J
Biomater Appl 2002;16:327–340.
19 Di Rocco G, Iachininoto MG, Tritarelli A et al. Myogenic potential of
adipose-tissue-derived cells. J Cell Sci 2006;119:2945–2952.
20 Hardikar AA, Marcus-Samuels B, Geras-Raaka E et al. Human pan-
creatic precursor cells secrete FGF2 to stimulate clustering into hor-
mone-expressing islet-like cell aggregates. Proc Natl Acad Sci U S A
21 Boujendar S, Arany E, Hill D et al. Taurine supplementation of a low
protein diet fed to rat dams normalizes the vascularization of the fetal
endocrine pancreas. J Nutr 2003;133:2820–2825.
22 Cherif H, Reusens B, Ahn MT et al. Effects of taurine on the insulin
secretion of rat fetal islets from dams fed a low-protein diet. J Endo-
23 Janjic D, Wollheim CB. Effect of 2-mercaptoethanol on glutathione
levels, cystine uptake and insulin secretion in insulin-secreting cells.
Eur J Biochem 2005;210:297–304.
24 Wang Y, Egan JM, Raygada M et al. Glucagon-like peptide-1 affects
gene transcription and messenger ribonucleic acid stability of compo-
nents of the insulin secretory system in RIN 1046–38 cells. Endocri-
25 D’Amour KA, Agulnick AD, Eliazer S et al. Efficient differentiation
of human embryonic stem cells to definitive endoderm. Nat Biotech-
26 Candido EP, Reeves R, Davie JR. Sodium butyrate inhibits histone
deacetylasation in cultured cells. Cell 1978;14:105–113.
27 Kanai-Azuma M, Kanai Y, Gad JM et al. Depletion of definitive
gut endoderm in Sox17-null mutant mice. Development 2002;129:
28 Sasaki H, Hogan BL. Differential expression of multiple fork head
related genes during gastrulation and axial pattern formation in the
mouse embryo. Development 1993;118:47–59.
29 Barbacci E, Reber M, Ott MO et al. Variant hepatocyte nuclear factor
1 is required for visceral endoderm specification. Development 1999;
30 Fehmann HC, Habener JF. Insulinotropic hormone glucagon-like pep-
tide-I (7–37) stimulation of proinsulin gene expression and proinsulin
biosynthesis in insulinoma bTC-1 cells. Endocrinology 1992;130:
31 Polak M, Bouchareb-Banaei L, Scharfmann R et al. Early pattern
of differentiation in the human pancreas. Diabetes 2000;49:225–
32 Teitelman G, Alpert S, Polak JM et al. Precursor cells of mouse endo-
crine pancreas coexpress insulin, glucagon and the neuronal proteins
tyrosine hydroxylase and neuropeptide Y, but not pancreatic polypep-
tide. Development 1993;118:1031–1039.
33 Morini S, Braun M, Onori P et al. Morphological changes of isolated
rat pancreatic islets: A structural, ultrastructural and morphometric
study. J Anat 2006;209:381–392.
34 Hardikar AA, Risbud MV, Bhonde RR. A simple microcapsule gener-
ator design for islet encapsulation. J Biosci 1999;24:371–376.
35 Mallett AG, Korbutt GS. Alginate modification improves long-term
survival and function of transplanted encapsulated islets. Tissue Eng
Part A 2009;15:1301–1309.
36 Revollo JR, Ko ¨rner A, Mills KF et al. Nampt/PBEF/visfatin regulates
insulin secretion in b cells as a systemic NAD biosynthetic enzyme.
Cell Metab 2007;6:363–375.
Chandra, G, Phadnis et al.