Sie sind auf Seite 1von 12

Mol Neurobiol (2014) 50:673–684

DOI 10.1007/s12035-014-8680-2

Interaction Between Nonviral Reprogrammed Fibroblast Stem


Cells and Trophic Factors for Brain Repair
G. Liu & H. Anisman & J. Bobyn & S. Hayley

Received: 28 November 2013 / Accepted: 11 March 2014 / Published online: 28 March 2014
# Springer Science+Business Media New York 2014

Abstract There are currently no known treatment options play in virtually all neurodegenerative states. Hence, the pres-
that actually halt or permanently reverse the pathology evident ent results support the utility of using combined gene and cell-
in any neurodegenerative condition. Arguably, one of the most targeting approaches for neuronal pathology.
promising avenues for creating viable neuronal treatments
could involve the combined use of cell replacement and gene Keywords Stem cells . Fibroblast . Endotoxin .
therapy. Given the complexity of the neurodegenerative pro- Neurodegeneration . Induced pluripotent stem cells (iPSC)
cess, it stands to reason that adequate therapy should involve
not only the replacement of loss neurons/synapses but also the
interruption of multiple pro-death pathways. Thus, we pro- Introduction
pose the use of stem cells that are tailored to express specific
trophic factors, thereby potentially encouraging synergistic The self-renewing characteristic of neural stem cells and their
effects between the stem cell properties and those of the ability to generate neurons and glia have resulted in their
trophic factors. The trophic factors, brain-derived neurotropic consideration in cell replacement therapies in various neuro-
factor (BDNF), glial cell-derived neurotropic factor (GDNF), degenerative diseases and psychological diseases such as
fibroblast growth factor (FGF) 2, and insulin-like growth ischaemic and neoplastic lesions [1] and Parkinson's disease
factor (IGF) 1, in particular, have demonstrated neuroprotec- (PD) [2], Alzheimer’s disease (AD) [2], and multiple sclerosis
tive actions in a number of animal models. Importantly, we [2] as well as perhaps even neuropsychiatric illnesses [3, 4]. In
use a nonviral approach, thereby minimizing the potential risk particular, recent attention has been devoted to using the
for DNA integration and tumor formation. The present study newly developed induced pluripotent stem cell (iPSC) tech-
involved the development of a nonviral reprogramming sys- nology as a therapeutic option that avoids many of the caveats
tem to transform adult mature mouse fibroblasts into progres- of traditional stem cell methods [5]. Further, whether neural
sive stages of cell development. We also tailored these stem stem cells might be employed as the delivery vehicle for gene
cells to individually express each of the trophic factors, in- therapy in the central nervous system (CNS) has been pro-
cluding BDNF, GDNF, FGF2, and IGF1. Significantly, central posed [1], yet scant experimental evidence exists to support
infusion of BDNF-expressing stem cells prevented the in vivo this contention. To the best of our knowledge, no reports exist
loss of neurons associated with infusion of the endotoxin, concerning the use of iPSC-derived stem cells to deliver
lipopolysaccharide (LPS). This is particularly important in trophic factors. However, substantial data does indicate that
light of the role of inflammatory processes that are posited to the administration of brain-derived neurotropic factor (BDNF)
[6], glial cell-derived neurotropic factor (GDNF) [7], fibro-
blast growth factor (FGF) 2 [8], and insulin-like growth factor
Electronic supplementary material The online version of this article (IGF) [9] have a host of neuroprotective functions and may
(doi:10.1007/s12035-014-8680-2) contains supplementary material, promote neuroplasticity beneficial for neuropsychiatric ill-
which is available to authorized users.
nesses [3, 4]. Hence, the present investigation sought to assess
G. Liu : H. Anisman : J. Bobyn : S. Hayley (*) the effects of these trophic factors when delivered using a
Department of Neuroscience, Carleton University, 327 Life Sciences
nonviral-based stem cell method.
Research Building, 1125 Colonel By Drive, Ottawa, ON K1S 5B6,
Canada We chose to focus upon BDNF, GDNF, FGF2, and IGF1,
e-mail: sp.hayley@gmail.com given that each of these trophic factors has been demonstrated
674 Mol Neurobiol (2014) 50:673–684

to provide neuroprotective actions across a number of animal mammalian cells and adapted for higher solubility, brighter
models of neurodegeneration [3, 4, 6–9]. Most research has emission, and rapid chromophore maturation. Fusion of
been devoted to BDNF, showing not only its importance in the cDNA BDNF to the N terminus of ZsYellow1 preserved the
growth and differentiation of immature neurons [10, 11] but fluorescent properties of the native protein permitting the
also in adult neurogenesis and learning and memory [12, 13]. localization of the fusion protein in vivo. The target gene
As well, BDNF is influenced by chronic stress [14] and its cDNA BDNF was cloned into pZsYellow1-N1, so that it
reduction has been implicated in depression, along with AD was in the same frame with the ZsYellow1 coding sequence,
and PD [15]. The related neurotrophin, GDNF, has been with no intervening, in-frame stop codons. The inserted gene
particularly linked to PD, and its administration promoted cDNA BDNF also included the initiating ATG codon.
the survival of many types of neurons [16] and elicited struc- The nonviral vector comprising a reprogramming cassette
tural and functional recovery in rodent and nonhuman primate of four transcription factor genes, c-Myc, Klf4, Oct4, and
models of PD [17]. GDNF administration has even recently Sox2, termed pCAG2LMKOSimO (Addgene, 20866), was
shown some promise in clinical trials for PD [18]. Not sur- utilized for reprogramming somatic fibroblasts into iPSCs
prisingly, FGF2 has also been linked to neurodegenerative and [25]. This vector has the advantage of producing virus-free,
psychiatric conditions, with reductions evident in forebrain factor-removable iPSCs by using a single plasmid with a 2A-
ischemia [19] and its potential therapeutic use for major peptide-linked reprogramming cassette, c-Myc-Klf4-Oct4-
depressive disorder being studied [20]. Finally, IGF1 is known Sox2(MKOS)-IRES mOrange, flanked by loxP sites. These
to play a significant role in development, cell differentiation, reprogramming genes are transcribing from the universally
plasticity, and survival of immature and adult neurons [21, expressed synthetic CAG enhancer/promoter [25]. The bio-
22]. Interestingly, the IGF1 polymorphism rs972936 was as- marker, mOrange, was in the same reading frame as the
sociated with AD in a Han Chinese population [23], and a reprogramming cassette c-Myc, Klf4, Oct4, and Sox2.
single nucleotide polymorphism of the IGF1 gene was asso-
ciated with ischemic stroke in a Korean population [24]. Tissue Preparation for Primary Cell Culture
The aims of the present investigation were to assess (1)
whether nonviral reprogramming is an efficient tool to trans- Primary CD1 mouse pituitary and hippocampal cells as well as
form fibroblast cells into progressively staged stem cells and rat (Sprague Dawley) hippocampal cells were prepared from 2-
eventually dopaminergic neurons, (2) whether iPSC-derived month-old animals. Two-month-old adult mice fibroblasts
stem cells can be efficiency used to deliver trophic factors, (3) (AMFs) were isolated by microdissection of tail-tips. A matrix
whether there are synergistic or additive interactions between of 100×60-mm plates, six-well plates, or four-chambered slides
the reprogrammed fibroblast stem cells and the trophic gene were prepared by coating all culture vessels with poly-D-lysine
they are induced to express, and (4) if the reprogrammed (final coating concentrations from 1 μg/cm2 of the surface area,
fibroblast stem cells plus the trophic factor, BDNF, attenuates Millipore A-003-E) and laminin (final coating concentrations
the damage induced by a neuroinflammatory stimulus, name- from 1 μg/cm2 of the surface area, Invitrogen 23017-015). All
ly, lipopolysaccharide (LPS). Successfully addressing these tissues were collected with Dulbecco's modified Eagle's medium
aims would go far in promoting the view that enhanced gene (DMEM)/F-12 (Invitrogen 10565-018). Tissues were plated with
plus cell replacement therapies could be viable treatment 4 mL of the Hibernate®-E Medium without Ca2+ (BrainBits
options for neuronal conditions. LLC, HE-Ca) for brain tissues or DMEM/F-12 for fibroblasts.
Chipped small pieces of tissues were enzymatically digested by
papain (2 mg/mL of filter-sterilized, Worthington, cat. no.
Materials and Methods LS003119) for 45 min at 37 °C. Tissues were then subject to
centrifugation, and the cell-containing supernatant was resus-
Construction of Gene cDNAs pended in 5 mL of complete Hibernate®-E medium for brain
cells or knockout DMEM (Invitrogen 10829-018) for fibroblast
The full sequence of 750 bp complementary DNA (cDNA) for cells. After tubes were centrifuged for 4 min at 200×g and
BDNF from Mus musculus (GenBank BC034862), 723 bp supernatants removed, cell pellets were resuspended in
cDNA for GDNF from M. musculus (GenBank NM_010275), 41.5 mL Neurobasal® medium (Invitrogen 21103-049) for brain
465 bp cDNA for FGF2 from M. musculus (GenBank cells or 41.5 mL DMEM/F-12 (Invitrogen10565-018) for fibro-
NM_008006), and 480 bp cDNA for IGF1 from blast cells. A 50-mL stock of Neurobasal® medium or
M. musculus (GenBank NM_010512) was constructed using DMEM/F-12 was made up by mixing 50 μL of 10 μg/mL (final
PCR (Supplementary Table 1 for specific primers) and cloned 10 ng/mL) of basic fibroblast growth factors (bFGF; Invitrogen
into the pZsYellow1-N1 Vector (Clontech 632445) at the 13256-029), 7.5 mL of knockout™ serum replacement
ligation sites of XhoI and SalI. The gene for ZsYellow1 has (Invitrogen 10828-028), 0.5 mL of MEM nonessential amino
been codon-optimized to enhance its translation in acids solution (NEAA) (Invitrogen11140-050), 0.5 mL of L-
Mol Neurobiol (2014) 50:673–684 675

glutamine (Invitrogen 25030-081), and 91 μL (final 0.1 mM) of 21–28. This specific medium was made up of 50 mL
β-mercaptoethanol (Invitrogen 21985-023). All cells were incu- Neurobasal® medium with bFGF, heparin solution (Sigma
bated at 37 °C with 5 % CO2-humidified atmosphere, and media H3149, 50 μL of 2-mg/mL), N-2 supplement, glutaMAX™-
were freshly nourished every third day. I supplement, nonessential amino acids solution, knockout™
serum replacement, and β-mercaptoethanol. From days 29–
Nonviral Reprogramming for the Induction of Pluripotency 35, a dopamine neuronal progenitor medium was used to
direct the neural progenitor cells to further differentiate into
When the aforementioned cells were cultured to 85–90 % dopamine neural progenitors. This medium included FGF-8b
confluence, cells were transfected with the vector 20866 con- (100 ng/mL, Invitrogen PHG0271) and sonic hedgehog
taining four reprogramming genes with/without cDNA for (200 ng/mL, SHH, R&D systems 1314-SH-025), being dis-
BDNF, GDNF, IGF1, or FGF2. Transfection was carried out solved into the Neurobasal® medium, plus other nutrition
using lipofectamine® 2000 Transfection Reagent (Invitrogen contents—heparin solution, N-2 supplement without vitamin
11668-027). The cells were continuously cultured with the A (Invitrogen 12587-010), and NEAA. Finally, the dopamine
fresh appropriate media to support reprogramming into neural progenitor cells were further differentiated into mature
pluripotency. The cluster appearances of stem cell-like colo- dopamine neurons using a dopamine neuronal differentiation
nies expressing the mOrange-positive marker were noted at medium during days 36–50. This medium consisted of recom-
days 5–6 following transfection. The iPSCs were verified by binant human BDNF (50 μL of 25-μg/mL, Invitrogen
using positive alkaline phosphatase live stain (Invitrogen PHC7074), recombinant human GDNF (50 μL of 20-μg/
A14353) as well as an expression of stem cell stage-specific mL, Invitrogen PHC7045), ascorbic acid (50 μL of
antigens: SSEA-1, SSEA-3, Oct-4, and TRA-1-81 (Millipore 200 mM, Sigma A4403), dcAMP (dibutyryl cyclic-AMP;
SCR002). The iPSCs with the mOrange-positive marker were 50 μL of 1-mM, Sigma D0627) in 50 ml Neurobasal® medi-
further identified during days 17–20 by live staining with um, plus additions of heparin, N-2 supplement without vita-
TRA-1-81, and then these were transferred to cultures con- min A (Invitrogen 12587-010), and NEAA. Once again, all
taining a neural induction medium in freshly coated plates or media were changed every 3 days.
wells comprising a 0.1 % gelatin attachment factor solution
(Invitrogen S-006-100). Infusion of Stem Cells into the Mouse Brain
The transfection day for cDNA constructs was designated
as day 0. During days 1–14, the medium was freshly changed Male C57BL/6 mice (8–10 weeks of age) underwent stereo-
with complete N2B27 medium, which contained with taxic cannula implantation [27]. Briefly, mice were anesthe-
DMEM/F-12 (Invitrogen 11330-057) plus N-2 Supplement tized with oxygen-enriched isoflurane, placed in a stereotaxic
(Invitrogen 17502-048), B-27® Supplement (Invitrogen apparatus, and implanted with indwelling cannulae projecting
17504-044), knockout™ serum replacement, L-glutamine, directly into the lateral ventricle. After 1 week of recovery,
NEAA, and β-mercaptoethanol. In order to improve efficien- mice were then infused with lipopolysaccharide (LPS; 2 μg
cy of the reprogramming genes, the medium was supplement- dissolved in 4 μL saline) (In vivo Gen, San Diego, CA, USA)
ed with a CHALP molecule cocktail [26]. The CHALP cock- or saline. After a further week, mice received a second infu-
tail included CHIR99021 (GSK3β inhibitor, 3 μM, Stemgent sion, via the indwelling cannula. This infusion consisted of (1)
04-0004), PD0325901 (MEK inhibitor, 0.5 μM, Stemgent 04- vehicle (HBSS) or (2) reprogrammed dopamine neuron cells+
0006), hLIF (human leukemia inhibitory factor; 10 ng/mL, complete length cDNA BDNF. All infusions (4 μL) were
Millipore LIF1005), A-83-01 (TGF-β/activin/nodal receptor delivered at a rate of 0.4 μL/min over 10 min, so as to ensure
inhibitor, 0.5 μM, Stemgent, 04-0014), bFGF (100 ng/mL), adequate time for diffusion into the brain. Infusions were
and HA-100 (ROCK inhibitor, 10 μM, Santa Cruz, sc- delivered via polyethylene tubing connected to a microliter
203072). The medium was supplied with the Essential 8™ syringe (Hamilton; Reno, NV, USA) and a syringe pump
Medium (Prototype), which contained DMEM/F-12 (HAM) (Harvard Apparatus Pico Plus; Holliston, MA, USA). Mice
1:1, Essential 8™ Supplement (50X) (Invitrogen A14666SA), were placed in fresh cages on heated pads for recovery, after
N-2 Supplement, B-27® Supplement, knockout™ serum, L- the injector was removed, and drill holes were filled with bone
glutamine, NEAA, and β-mercaptoethanol from day 15 to 20. wax. All animals were sacrificed 2 weeks following the sec-
All media were changed every 3 days. ond intracerebroventricular (icv) infusion (reprogrammed
neurons vs vehicle). Thus, 3 weeks had elapsed since the
Induction of Neural Cells and Differentiation of Mature initial icv infusion of either LPS or vehicle, a time at which
Dopaminergic Neurons the neurodegenerative effects of LPS should be evident and
relatively stable [28], and 2 weeks had elapsed since infusion
Undifferentiated iPSCs were further transformed into neural of the iPSC-derived cells. All procedures of stereotaxic infu-
progenitor cells using a neural induction medium during days sion of cells were approved by the Carleton University
676 Mol Neurobiol (2014) 50:673–684

Committee for Animal Care and were conducted in adherence followed by peroxidase streptavidin (1:1,000, Jackson
to guidelines set out by the Canadian Council for the Use and ImmunoResearch Laboratories), also at room temperature
Care of Animals in Research. for 2 h. Sections were dehydrated, cover slipped using
Permount, and then visualized using diaminobenzidine
Quantitative Real-Time PCR (NeuN). For MAP2, the secondary antibodies (2 h incubation
at room temperature) were CY3 tagged anti-mouse (Abcam;
Initially, total RNA was isolated from cells or tissues using a 1:1,000) and visualized using a fluorescent microscope.
GeneJET RNA Purification Kit (Thermo Scientific K0732). In order to detect the fate of the infused reprogrammed
Thereafter, RNA was quantified using real-time PCR (qRT- stem cells, brain sections were directly imaged for the bio-
PCR), applying the TaqMan gene-specific primers and the marker, mOrange (exCitation and emission maxima are 549
LightCycler® 480 RNA Master Hydrolysis Probes kit and 565 nm), which is linked to the 20866 vector that contains
(Roche 04991885001). The primers (Supplementary the four reprogramming genes. As well, we assessed expres-
Table 1) were designed using the Integrated DNA sion of the yellow fluorescent protein (YFP)-linked BDNF
Technologies website software program. construct using the appropriate fluorescence filter (exCitation
and emission peaks at 514 and 527 nm, respectively).
In Vitro Immunocytochemistry

Stem cell markers were detected in cell culture samples by


Results
alkaline phosphatase live stain (Invitrogen A14353), in con-
junction with ES Marker Sample Kits (Millipore SCR002).
Verification of Complete Mouse cDNA Constructs
The latter kit comprises the monoclonal antibodies for the
analysis of the ES cell markers, including the SSEA-1,
The complete mouse cDNA constructs for BDNF, GDNF,
SSEA-3, Oct-4, and TRA-1-81 antigens. Immature neuronal
FGF2, and IGF1 were verified by Sanger Sequencing
cells were detected using anti-doublecortin (DCX) antisera
(Fig. 1a), using Applied Biosystem's 3730xl DNA Analyzer
(Invitrogen 48-1200), whereas mature neuronal cells were
at The McGill University Innovation Centre. The BDNF
identified by anti-MAP2 antibodies (Abcam ab32454), and
cDNA construct was previously characterized in our recent
finally, dopamine neuronal cells were assessed using anti-
paper by Liu et al. [27]. In the present study, the presence of
tyrosine hydroxylase (TH) antisera (ImmunoStar 22941). As
GDNF, FGF2, and IGF1 expression were confirmed by de-
well, parallel culture samples were subjected to anti-BDNF
tection of YFP after 1-week cell culture. Indeed, as depicted in
antibody (Abcam ab6201). The experiments were carried out
Fig. 1b, YFP staining was evident in adult mouse primary
according to the manufacturers’ standardization protocols and
hippocampal and pituitary cells as well as adult rat primary
choice of secondary antibodies.
hippocampal cells.
In Vivo Immunohistochemistry
Verification of In Vitro iPSCs from Reprogrammed Mature
Following perfusion with 4 % paraformaldehyde and Fibroblast
cryoprotection with a 20 % sucrose solution, serial coronal
sections (20 μm) were collected with a cryostat (Thermo Adult (2-month old) fibroblast cells of mice were collected
Fisher), slide mounted, and stored at −20° until immunohis- from tail-tips for cell culture. After transfection with vector
tochemical analysis. All tissues were immersed in heated 20866 comprising four reprogramming genes, Oct4/Sox2/
trisodium citrate buffer (pH 8.5) for antigen retrieval follow- Klf4/c-Myc, cells were transformed into iPSCs during days
ing preliminary washes with phosphate-buffered saline (PBS). 17–20 under optimal cell culture conditions (see “Materials
Subsequently, tissue was blocked for 1 h with 0.1 M tris- and Methods” for details). When the vector 20866 was intro-
buffered saline containing 0.1 M PBS containing 0.1 % sodi- duced into the fibroblast cells, mOrange-positive cells were
um azide, 0.3 % Triton X, and 2 % bovine serum albumin. converted to a stem cell-like morphology at days 5–6. The
NeuN and MAP2 are excellent markers of neurons for iPSCs were verified by using positive alkaline phosphatase
mature neurons [29]. Thus, brain sections were incubated live staining (panel d) as well as an expression of stem cell
overnight at 4 °C with mouse anti-NeuN or anti-MAP2 anti- stage-specific antigens: SSEA-1 (panel e), SSEA-3 (panel f),
bodies (Millipore and Abcam ab32454, respectively; 1:500) in Oct-4 (panel g), and TRA-1-81 (panel h) in Fig. 2.
0.01 M PBS with 0.3 % Triton X-100. In the case of NeuN, Furthermore, the colonies picked between days 17–20,
sections were rinsed in 0.01 M PBS and incubated with flourished on gelatin-coated plates (attachment factor is sterile
biotinylated goat anti-mouse IgG (1:500, Jackson 1X solution containing 0.1 % gelatin, Invitrogen S-006-100),
ImmunoResearch Laboratories) at room temperature for 2 h, further suggesting successful reprogramming.
Mol Neurobiol (2014) 50:673–684 677

Fig. 1 Verified complete cDNA a Complete cDNA structure confirmed by gene sequencing
construction of the multiple
genes. The complete mouse
cDNA constructions of GDNF,
FGF2, and IGF1 were cloned into
GDNF
a pZsYellow1-N1 vector and
were verified by Sanger
sequencing (a). These cDNA
constructions were transfected,
using lipofectamine 2000 reagent,
into adult mouse primary
hippocampal and pituitary cells as
well as adult rat primary
hippocampal cells. The cells were
cultured for a further week and
FGF2
then their gene expression was
verified by detection of yellow
fluorescent protein (YFP) (b)

IGF1

b Mouse hippocampus Mouse pituitary Rat hippocampus

GDNF

FGF2

IGF1

Additive Effects of Cellular Reprogramming Plus Trophic co-transfected with BDNF displayed increased BDNF mes-
Factor Transfection senger RNA (mRNA) relative to reprogrammed cells that
were not co-transfected with the trophic factor. The same
As depicted in Fig. 3a, expression of BDNF, GDNF, FGF2, holds true for panels B–D, with respect to GDNF, FGF2,
and IGF1 were each increased by a similar magnitude in iPSC and IGF1.
cells derived from adult fibroblasts that also co-expressed the In order to ascertain whether long-term expression of the
respective trophic factor. For instance, panel A shows that trophic factors was apparent, we determined trophic factor
cells that were reprogrammed by the 20866 vector and also mRNA in cell cultures after 10 days. Indeed, as shown in
678 Mol Neurobiol (2014) 50:673–684

Fig. 2 In vitro iPSCs yielded a fibroblast cells b vector 20866 mOrange marker c outgrowing iPSCs
from reprogrammed adult mouse
fibroblasts. Adult mouse tail-tip
fibroblast cells (a) were
transfected with vector 20866,
incorporating four
reprogramming genes Oct4/Sox2/
Klf4/c-Myc, and cultured for 17–
20 days. Clearly, the fibroblasts
were transformed into iPSC cells,
as indicated by their expression of
mOrange marker for vector 20866 d AP e SSEA-1 f SSEA-3
containing the four
reprogramming genes (b), their
prominent outgrowths (c), by
positive alkaline phosphatase live
staining (d), as well as an
expression of stem cell stage-
specific antigens: SSEA-1 (e),
SSEA-3 (f), Oct-4 (g), and TRA-
1-81 (h)
g Oct-4 h TRA-1-81

Fig. 3b, trophic factor expression was still elevated at Engineering iPSC-Derived Mature Neural Cells to Express
the 10-day time point in iPSC reprogrammed cells that BDNF for In Vivo Application
were also co-transfected with the particular trophic fac-
tor. Intermediate levels of mRNA elevation were appar- Our next step was to engineer the iPSC-derived mature
ent in cells that were only exposed to the reprogramming neurons to express the trophic factor, BDNF, so as to
process alone or were transfected with the trophic factor provide an efficient tool to use for in vivo cell replace-
(i.e., BDNF, GDNF, FGF2, IGF1) alone relative to vehicle- ment and delivery of a potential neuroprotective factor.
treated controls. Indeed, at 2 weeks following icv infusion (Fig. 5), in-
creased mRNA for hippocampal BDNF and cortical TH
was evident among mice that received infusion of the
Transformation of Reprogrammed iPSCs into Progressive reprogrammed neural cells along with the cDNA BDNF
Staged Neural Cells construct relative to those that received the reprogrammed
neural cells alone.
Given the aforementioned capability to utilize nonviral
reprogramming to transform adult mice fibroblasts into Central Infusion of Reprogrammed Stem Cells Co-Expressing
iPSCs, our next step was to produce specific neural cells. To BDNF Blunted the Neuronal Loss Induced by LPS
this end, iPSCs were first transformed into neural progenitor
cells during days 20–35 and subsequently (days 36–50) di- Given that neuroinflammatory processes are common to
rected to differentiate into mature neurons of an apparent most neurodegenerative diseases, we centrally infused the
dopaminergic phenotype. During this stage, these neurons endotoxin, LPS, and assessed the survival of NeuN+ and
appeared to form mature neural networks with distinct soma MAP2+ mature neurons in the surrounding brain paren-
and cellular morphology (Fig. 4). Moreover, these cells chyma. We were particularly interested in whether central
expressed the mature neuronal cell marker, microtubule- infusion of the reprogrammed stem cells that expressed
associated protein 2 (MAP2), as well as the dopamine rate BDNF would inhibit the damaging effects of the LPS
limiting enzyme, TH (Fig. 4). treatment. Indeed, administration of the iPSC-derived cells
Mol Neurobiol (2014) 50:673–684 679

Fig. 3 Interactions between a A BDNF B GDNF


in vitro iPSCs and multiple
neurotrophic factor genes. As
shown in a, the mRNA for the
a
neurotrophic factors, BDNF, a
GDNF, FGF2, and IGF2 were
each increased in a similar fashion

RFU
after 16-h co-transfected with b
their respective complete length b
cDNAs plus the 20866
reprogramming vector.
Specifically, in adult mouse
fibroblast cells, the specific
trophic factor +20866 (a, red
C FGF2 D IGF1
lines; duplicates) greatly elevated
the expression of the respective
trophic factor relative to treatment
with the 20866 vector alone (b, a
a
green lines; duplicates). Further
assessment over a 10-day interval
RFU

revealed that the augmented b


trophic factor expression was b
maintained in the reprogrammed
fibroblasts (b). Specifically, after
10 days in culture, the most robust
trophic factor expression was still
evident in cells subjected to co-
transfection with the respective
cDNA plus the 20866 b A BDNF B GDNF
reprogramming vector (a, red
lines) relative to cells treated with
the cDNA construct alone (b, a a
b
green lines), the vector 20866 b
alone (c, blue lines), or the vehicle
RFU

c
solution (HBSS) (d, black lines). c
All results above were carried on d
by RT-qPCR-TaqMan with
d
specific primers (see
Supplementary Table 1), and a
and b represented one of the three
experimental replications for ct
plots. All results of RNA levels C FGF2 D IGF1
were normalized to total levels of
a control housekeeping gene
a
GAPDH, whose primers are
presented in Supplementary a b
Table 1 b
RFU

c
c

d d

that expressed BDNF appeared to greatly reduce the loss ventricle extending into the adjacent striatum and lateral
of NeuN+ and MAP2+ neurons (Figs. 6 and 7, respective- septum, and this effect was not evident in mice that were
ly); these effects were apparent in the parenchyma (striatum infused with the reprogrammed neurons that expressed
and septal areas) around the site of icv LPS infusion. As BDNF (Fig. 6). As shown in Fig. 7, this finding was
would be expected, this effect was most apparent in the confirmed with high magnification MAP2 staining, that
hemisphere ipsilateral to the icv infusions. Indeed, in mice demonstrated reduced MAP2+ stained cells with LPS infu-
that received only the LPS infusion, there was an obvious sion and an increased signal in the mice that received the
loss of NeuN+ stained neurons around the ipsilateral lateral iPSC-derived cells with BDNF expression. Importantly,
680 Mol Neurobiol (2014) 50:673–684

a iPSC , Day 20 b Day 25 c Day 30 d neural progenitor, Day 35

e neurons, Day 37 f Day 40 g Day 43 h Day 47

i Day 50 j MAP2 k TH

Fig. 4 Transformation of reprogrammed iPSC cells into mature neurons. dopaminergic neurons, using a dopamine neuronal differentiation medi-
The iPSCs obtained from reprogrammed adult mouse fibroblasts were um, with clear neuronal networks beginning to form at this time (f–i).
transformed into neural progenitor cells following culture with a neural Expression of the mature neuronal cell marker, microtubule-associated
induction medium during days 20–35 (a–d). The neural progenitor cells protein 2 (MAP2) (i), and the dopaminergic neuronal marker, tyrosine
were further differentiated into dopaminergic neural progenitors using the hydroxylase (TH) (k), suggests that indeed these were mature neurons
dopamine neuronal progenitor medium from days 35–37 (e). Subsequent- with the potential to produce dopamine
ly, from days 38 to 50, the cells were differentiated into mature

A Hippocampus, BDNF B Cortical, TH

a a

b
RFU

Fig. 5 Enhanced hippocampal and cortical expression of BDNF and TH expression (b, green lines, duplicates). All results above were carried on
following in vivo infusion of reprogrammed neural cells. mRNA for by RT-qPCR-TaqMan with specific primers (see Supplementary Table 1),
BDNF within the hippocampus (a) and TH within the cortex (adjacent and Fig. 5 represented one of the three experimental replications for ct
to the site of infusion) (b) were enhanced by intracerebroventricular plots. All results of RNA levels were normalized to total levels of a
infusion of iPSC-derived mature neuronal cells (characterized in Figs. 3 control housekeeping gene GAPDH, whose primers are presented in
and 4) that co-expressed BDNF (a, red lines, duplicates) relative to Supplementary Table 1
infusion of the iPSC-derived neurons in the absence of BDNF co-
Mol Neurobiol (2014) 50:673–684 681

Fig. 6 Central infusion of iPSC-derived neuronal cells that co-express NeuN+ neurons in the ipsilateral hemisphere (ipsi; b) relative to the
BDNF blunted the loss of NeuN+ neurons induced by lipopolysaccharide contralateral hemisphere (contra; e) or the vehicle treated mice (a, d).
(LPS). Mice received icv infusion of LPS (2 μg) or vehicle (saline) and However, infusion of the BDNF co-expressing iPSC-derived neurons (c,
1 week later received a second icv infusion of either the iPSC-derived f) prevented this intra-hemispheric difference and reduced the degree of
cells engineered to co-express BDNF or vehicle (HBSS). All mice were neuronal loss relative to vehicle treatment
sacrificed after a further 2 weeks. Clearly, LPS infusion provoked a loss of

data was only used from animals in which cannula Identifying Centrally Infused iPSC-Derived Cells
placement was determined to be correctly located in
the lateral ventricle (this resulted in three animals being Following central infusion of the reprogrammed stem cells
excluded). plus cDNA BDNF 2 weeks earlier, positive signals for YFP

MAP2
x100

g
mOrange channel YFP channel Merge

a b c
h

d e f

Fig. 7 Centrally infused iPSC-derived neuronal cells that co-express Given that no such staining was observed in animals from the remaining
BDNF survived in vivo and prevented the loss of MAP2+ staining groups that did not receive the cellular infusions (i.e., vehicle and LPS
induced by LPS. The reprogrammed stem cells plus BDNF cDNA alone infused mice), it appears that at least a subset of the infused cells
infusions occurred 2 weeks earlier. As shown in the fluorescent channels survived for the 2-week post-infusion period. The right hand photomi-
with corresponding wavelengths for detection of mOrange (a, d) and YFP crographs (g–i) depict staining for the neuronal marker, MAP2. Corre-
(b, e), positive staining was observed for cells taken from two different sponding with the NeuN findings from Fig. 6, LPS infusion clearly
animals that received the iPSC-derived cellular infusions. Sections were reduced the presence of MAP2+ cells in the striatal region immediately
taken from around the site of infusion (in the striatal region immediately adjacent to the site of infusion (h) relative to vehicle infused mice (g).
adjacent to the lateral ventricle site of infusion) and the positive cells However, once again, infusion of the iPSC-derived neurons that
clearly co-localized both of the fluorescent markers (as shown in the contained the BDNF construct prevented/rescued this deficit, such that
merged photomicrographs of c and f), indicating the presence of the substantially increased MAP2+ staining was now evident (i)
YFP-linked BDNF and mOrange-linked 20866 reprogramming vector.
682 Mol Neurobiol (2014) 50:673–684

and mOrange were detected around the site of infusion In contrast to most iPSC-producing systems that use viral
(Fig. 7a, b and d, e), and these positive cells clearly co- approaches [37], we applied a novel single excisable nonviral
localized both of these markers (as shown in the merged vector containing a four-gene reprogramming cassette for
photomicrographs of panels c and f). Indeed, the top and transformation of iPSCs from adult mouse fibroblasts [25].
bottom photomicrographs of Fig. 7 (two representative brain Our results show that this system yields an efficient approach
sections from iPSC-injected animals) revealed that the to reprogram and transform adult mouse fibroblasts into
reprogrammed cells that expressed the 20866 vector and the iPSCs. Furthermore, the reprogrammed iPSCs were trans-
BDNF construct (evident in the respective mOrange and YFP formed into progressive stages of neural cells from immature
fluorescent channels) were indeed present in the striatal pa- to mature neurons, expressing MAP2 and NeuN. Of impor-
renchyma immediately in the vicinity of the infusion site. tance, these neuronal cells subsequently appeared to develop
Thus, at least a subpopulation of the infused reprogrammed mature neural networks, suggesting their potential ability to
cells did indeed appear to survive until the end of the study. make appropriate connections when infused into the brain.
Indeed, central infusion of these cells resulted in enhanced
gene expressions of BDNF within the hippocampus and TH
within the cortical region adjacent to the site of infusion.
Discussion However, most importantly, the central administration of
iPSC-derived neurons that expressed BDNF did appear to
Cell replacement and gene therapies hold tremendous promise prevent the loss of mature neurons that was provoked by
for a broad range of CNS disorders. Although the two repair LPS infusion. Moreover, at least some of the centrally infused
strategies have been considered separately, we currently pro- iPSC-derived cells appeared to survive for the 2-week period
pose their integration by engineering iPSC-derived neurons to following administration (as evidenced by cells that expressed
not only replace lost cells but also overexpress trophic factors the mOrange and YFP fluorescent signals). These data speak
(e.g., BDNF) that would be expected to facilitate the devel- to the possible importance of our cell+gene replacement
opment of new “brain circuits.” We were especially interested strategy in downregulating neuroinflammatory damage asso-
in producing iPSC-derived mature TH-expressing neurons ciated with an immune challenge. This is a particularly im-
and then engineering them to express one of several trophic portant finding, given that neuroinflammatory processes have
factors that have been linked to neuron survival and regener- been linked to virtually all neurodegenerative illnesses [38].
ation [18], namely, BDNF, GDNF, FGF2, and IGF1. Indeed, Our approach might potentially be applied to a variety of
the administration of these neurotropic factors has been ex- neurodegenerative diseases, given that deficiencies of BDNF
plored as potential therapy for neurodegenerative diseases [39], GDNF [40], FGF2 [41], or IGF1 [42], along with en-
such as PD, AD, and Huntington's disease [29–32]. Yet, most hanced pro-inflammatory cascades have been implicated in
existing methods do not allow for the long-term expression of numerous CNS disease states. Using our present strategy, one
these trophic factors and have been hampered by a myriad of could produce autologous transplants of iPSCs or mature
side effects [33]. In contrast, we hope to be able to induce neurons that overexpress various growth factors, which could
protracted expression of these trophic factors (through in- potentially neutralize pathology or provide alternate “brain
creased expression in transplanted neurons that integrate with circuits” to compensate for the endogenous pathological tis-
existing endogenous cells), and we expect this situation to sue. Indeed, it has already been reported that adult skeletal
provide a neural milieu more akin to “healthy conditions,” muscle-derived stem cells could provide a promising source
thereby minimizing the likelihood of adverse effects owing to for autologous transplantation for neurodegenerative diseases
widespread variations in trophic factor levels. [43]. A recent study also showed that the beneficial effects of
An important issue to consider regarding any kind of cell intravenous delivery of stem cells into rotenone-treated mice
replacement is the risk for the potential development of ab- (a model for PD) may result from the combined effects of
normally dividing cancer cells. Indeed, reprogrammed iPSC- neurotrophic and endogenous brain repair mechanisms [44].
derived cells do have the potential to form tumor cells [34], Moreover, human neural stem cells were found to be a suitable
and this risk is further magnified when viral delivery mecha- cell delivery vehicle for the introduction of GDNF into the
nisms are utilized, owing to the potential for DNA integration brain for treating PD; however, the caveat of using a lentivirus
[35]. To circumvent this potential problem, in the present to introduce GDNF into the cells was still apparent in this
investigation, we utilized a nonviral delivery mechanism, study [7]. Finally, recent studies have also found that stem
which we recently demonstrated to not alter host DNA (at cell-mediated gene delivery of BDNF, GDNF, or NT3 was
least at the gene sequence encoding the TH enzyme) [27]. safe and efficacious in stroke models [45]. We believe that
Thus, minimizing tumorigenic outcomes and using an iPSC using such stem cell-driven gene delivery approaches will
method that allows for autologous transplantation [36] should prove to be more effective (given that “optimized” new neu-
aid in the translation into the clinical realm. ronal cells can be delivered in addition to a gene/trophic factor
Mol Neurobiol (2014) 50:673–684 683

of interest) and safer than the existing viral systems that have administration in normotensive and hypertensive rats with an ische-
mic stroke. Neuroscience 250:253–262
been developed for gene delivery in animal models [46]. Yet,
10. Acheson A, Conover JC, Fandl JP, DeChiara TM, Russell M,
it remains to be determined which approach will ultimately be Thadani A, Squinto SP, Yancopoulos GD, Lindsay RM (1995) A
adopted in the clinical arena. BDNF autocrine loop in adult sensory neurons prevents cell death.
In summary, in the present study, we report that combining Nature 374:450–453
11. Huang EJ, Reichardt LF (2001) Neurotrophins: roles in neuronal
stem cells (as a cell replacement and gene delivery tool) with
development and function. Annu Rev Neurosci 24:677–736
BDNF, GDNF, FGF2, or IGF2 cDNA constructs resulted in 12. Yamada K, Nabeshima T (2003) Brain-derived neurotrophic factor/
additive increases in trophic factor expression. Importantly, TrkB signaling in memory processes. J Pharmacol Sci 91:267–270
when iPSC-derived neurons that expressed BDNF were cen- 13. Bekinschtein P, Cammarota M, Katche C, Slipczuk L, Rossato JI,
Goldin A, Izquierdo I, Medina JH (2008) BDNF is essential to
trally infused, there was an increase in BDNF and TH expres-
promote persistence of long-term memory storage. Proc Natl Acad
sion in vivo, coupled with a blunting of the neuronal loss Sci U S A 105:2711–2716
induced by bacterial endotoxin infusion. Once again, these 14. Greisen MH, Altar CA, Bolwig TG, Whitehead R, Wörtwein G
data raise the possibility of using combined gene+cell re- (2005) Increased adult hippocampal brain-derived neurotrophic fac-
tor and normal levels of neurogenesis in maternal separation rats. J
placement strategies involving iPSC cells to modulate endog-
Neurosci Res 79:772–778
enous neural circuitry, so as to overcome a variety of patho- 15. Siegel GJ, Chauhan NB (2000) Neurotrophic factors in Alzheimer's
logical states, particularly those with an inflammatory com- and Parkinson's disease brain. Brain Res Brain Res Rev 33:199–227
ponent. Future studies will be required to track the long-term 16. Rangasamy SB, Soderstrom K, Bakay RA, Kordower JH (2010)
Neurotrophic factor therapy for Parkinson's disease. Prog Brain Res
in vivo development of neural connections following trans-
184:237–264
plantation and ascertain the optimal endogenous milieu re- 17. Kordower JH, Bjorklund A (2013) Trophic factor gene therapy for
quired to facilitate proper integration of these modified cells. Parkinson's disease. Mov Disord 28:96–109
18. Allen SJ, Watson JJ, Shoemark DK, Barua NU, Patel NK (2013)
GDNF, NGF and BDNF as therapeutic options for neurodegenera-
Acknowledgments S.H. and H.A. are Canada Research Chairs and this tion. Pharmacol Ther 138:155–175
work was supported by grants from the Canadian Institutes of Health 19. Okada T, Kataoka Y, Takeshita A, Mino M, Morioka H, Kusakabe
Research. KT, Kondo T (2013) Effects of transient forebrain ischemia on the
hippocampus of the Mongolian gerbil (Meriones unguiculatus): an
Conflict of Interest The authors declare no conflicts of interest. immunohistochemical study. Zool Sci 30:484–489
20. Tumer CA, Gula EL, Taylor LP, Watson SJ, Akil H (2008)
Antidepressant-like effects of intracerebroventricular FGF2 in rats.
Brain Res 1224:63–68
References 21. Benarroch EE (2012) Insulin-like growth factors in the brain and their
potential clinical implications. Neurology 79:2148–2153
22. Ueno M, Fujita Y, Tanaka T, Nakamura Y, Kikuta J, Ishii M, Yamashita T
1. Müller FJ, Snyder EY, Loring JF (2006) Gene therapy: can neural (2013) Layer V cortical neurons require microglial support for survival
stem cells deliver? Nat Rev Neurosci 7:75–84 during postnatal development. Nat Neurosci 16:543–551
2. Martínez-Morales PL, Revilla A, Ocaña I, González C, Sainz P, 23. Wang W, Yu JT, Tan L, Liu QY, Wang HF, Ma XY (2012) Insulin-
McGuire D, Liste I (2013) Progress in stem cell therapy for major like growth factor 1 (IGF1) polymorphism is associated with
human neurological disorders. Stem Cell Rev 9:685–699 Alzheimer's disease in Han Chinese. Neurosci Lett 531:20–23
3. Faiz M, Nagy A (2013) Induced pluripotent stem cells and disorders 24. Kim HJ, Kim SK, Park HJ, Chung JH, Chun J, Yun DH, Kim YO
of the nervous system: progress, problems, and prospects. (2012) Polymorphisms of IGFI contribute to the development of
Neuroscientist. doi:10.1177/1073858413493148 ischemic stroke. Exp Ther Med 3:93–98
4. Vaccarino FM, Stevens HE, Kocabas A, Palejev D, Szekely A, 25. Kaji K, Norrby K, Paca A, Mileikovsky M, Mohseni P, Woltjen K
Grigorenko EL, Weissman S (2011) Induced pluripotent stem cells: (2009) Virus-free induction of pluripotency and subsequent excision
a new tool to confront the challenge of neuropsychiatric disorders. of reprogramming factors. Nature 458:771–775
Neuropharmacology 60:1355–1363 26. Yu J, Chau KF, Vodyanik MA, Jiang J, Jiang Y (2011) Efficient
5. Lieu PT, Fontes A, Vemuri MC, Macarthur CC (2013) Generation of feeder-free episomal reprogramming with small molecules. PLoS
induced pluripotent stem cells with CytoTune, a non-integrating One 6:e17557
Sendai virus. Methods Mol Biol 997:45–56 27. Liu G, Rustom N, Bobyn J, Litteljohn D, Rudyk C, Clarke M,
6. Wu CL, Chen SD, Yin JH, Hwang CS, Yang DI (2010) Osborn M, Anisman H, Hayley S (2013) Non-viral induction of
Erythropoietin and sonic hedgehog mediate the neuroprotective ef- neurons engineered to express brain derived neurotrophic factor:
fects of brain-derived neurotrophic factor against mitochondrial in- towards enhanced gene-cell therapy for stressor related pathology.
hibition. Neurobiol Dis 40:146–154 Stem Cells, Submitted
7. Wang S, Ren P, Guan Y, Zou C, Fu L, Zhang Y (2013) Inducible 28. Li G, Sun S, Cao X, Zhong J, Tong E (2004) LPS-induced degener-
regulation of GDNF expression in human neural stem cells. Sci ation of dopaminergic neurons of substantia nigra in rats. J Huazhong
China Life Sci 56:32–39 Univ Sci Technol Med Sci 24:83–86
8. Thau-Zuchman O, Shohami E, Alexandrovich AG, Leker RR (2012) 29. Giampà C, Montagna E, Dato C, Melone MA, Bernardi G, Fusco FR
Combination of vascular endothelial and fibroblast growth factor 2 (2013) Systemic delivery of recombinant brain derived neurotrophic
for induction of neurogenesis and angiogenesis after traumatic brain factor (BDNF) in the R6/2 mouse model of Huntington's disease.
injury. J Mol Neurosci 47:166–172 PLoS One 8:e64037
9. De Geyter D, Stoop W, Sarre S, De Keyser J, Kooijman R (2013) 30. Ebert AD, Barber AE, Heins BM, Svendsen CN (2010) Ex vivo
Neuroprotective efficacy of subcutaneous insulin-like growth factor-I delivery of GDNF maintains motor function and prevents neuronal
684 Mol Neurobiol (2014) 50:673–684

loss in a transgenic mouse model of Huntington's disease. Exp Neurol 39. Lühder F, Gold R, Flügel A, Linker RA (2013) Brain-derived neu-
224:155–162 rotrophic factor in neuroimmunology: lessons learned from multiple
31. Jin K, LaFevre-Bernt M, Sun Y, Chen S, Gafni J, Crippen D, sclerosis patients and experimental autoimmune encephalomyelitis
Logvinova A, Ross CA, Greenberg DA, Ellerby LM (2005) FGF-2 models. Arch Immunol Ther Exp (Warsz) 61:95–105
promotes neurogenesis and neuroprotection and prolongs survival in 40. Lundborg C, Westerlund A, Björklund U, Biber B, Hansson E (2011)
a transgenic mouse model of Huntington's disease. Proc Natl Acad Ifenprodil restores GDNF-evoked Ca(2+) signalling and Na(+)/K(+)-
Sci U S A 102:18189–18194 ATPase expression in inflammation-pretreated astrocytes. J
32. Sadagurski M, Cheng Z, Rozzo A, Palazzolo I, Kelley GR, Dong X, Neurochem 119:686–696
Krainc D, White MF (2011) IRS2 increases mitochondrial dysfunc- 41. Woodbury ME, Ikezu T (2013) Fibroblast growth factor-2 signaling
tion and oxidative stress in a mouse model of Huntington disease. J in neurogenesis and neurodegeneration. J Neuroimmune Pharmacol
Clin Invest 121:4070–4081 Sep 21
33. Ramaswamy S, Kordower JH (2012) Gene therapy for Huntington's 42. Mueller AM, Nassery A, Conlon H, Liu X, Jun E, Yoon BH,
disease. Neurobiol Dis 48:243–254 Cristofanilli M, Sadiq SA (2013) Effects of intraventricular metho-
34. Steinemann D, Göhring G, Schlegelberger B (2013) Genetic insta- trexate administration on cuprizone-induced demyelination in mice.
bility of modified stem cells—a first step towards malignant trans- Front Mol Neurosci 6:34
formation? Am J Stem Cells 2:39–51 43. Yang J, Wang X, Wang Y, Guo ZX, Luo DZ, Jia J, Wang XM (2012)
35. Araki R, Uda M, Hoki Y, Sunayama M, Nakamura M, Ando S, Sugiura Dopaminergic neuronal conversion from adult rat skeletal muscle-
M, Ideno H, Shimada A, Nifuji A, Abe M (2013) Negligible immuno- derived stem cells in vitro. Neurochem Res 37:1982–1992
genicity of terminally differentiated cells derived from induced pluripo- 44. Inden M, Takata K, Nishimura K, Kitamura Y, Ashihara E,
tent or embryonic stem cells. Nature 494:100–104 Yoshimoto K, Ariga H, Honmou O, Shimohama S (2013)
36. Lu X, Zhao T (2013) Clinical therapy using iPSCs: hopes and Therapeutic effects of human mesenchymal and hematopoietic stem
challenges. Genomics Proteomics Bioinforma 11:294–298 cells on rotenone-treated parkinsonian mice. J Neurosci Res 91:62–
37. Drews K, Jozefczuk J, Prigione A, Adjaye J (2012) Human induced 72
pluripotent stem cells—from mechanisms to clinical applications. J 45. Chen C, Wang Y, Yang GY (2013) Stem cell-mediated gene deliver-
Mol Med (Berl) 90:735–745 ing for the treatment of cerebral ischemia: progress and prospectives.
38. Choi DY, Liu M, Hunter RL, Cass WA, Pandya JD, Sullivan PG, Curr Drug Targets 14(1):81–89
Shin EJ, Kim HC, Gash DM, Bing G (2009) Striatal neuroinflam- 46. Coune PG, Schneider BL, Aebischer P (2012) Parkinson's disease:
mation promotes Parkinsonism in rats. PLoS One 4:e5482 gene therapies. Cold Spring Harb Perspect Med 2(4):a009431