Stem Cell Reports

Article

Human iPSC-derived neural progenitor cells secreting GDNF provide

protection in rodent models of ALS and retinal degeneration
Alexander H. Laperle,1,2 V. Alexandra Moser,1,2 Pablo Avalos,1 Bin Lu,1 Amanda Wu,1 Aaron Fulton,1
Stephany Ramirez,1 Veronica J. Garcia,1 Shaughn Bell,1 Ritchie Ho,1 George Lawless,1 Kristina Roxas,1
Saba Shahin,1 Oksana Shelest,1 Soshana Svendsen,1 Shaomei Wang,1,* and Clive N. Svendsen1,*
1Cedars-Sinai Board of Governors Regenerative Medicine Institute, Cedars-Sinai Medical Center, Los Angeles, CA, USA
2These authors contributed equally
*Correspondence: [email protected] (C.N.S.), [email protected] (S.W.)
https://doi.org/10.1016/j.stemcr.2023.03.016

SUMMARY

Human induced pluripotent stem cells (iPSCs) are a renewable cell source that can be differentiated into neural progenitor cells (iNPCs)
and transduced with glial cell line-derived neurotrophic factor (iNPC-GDNFs). The goal of the current study is to characterize iNPC-
GDNFs and test their therapeutic potential and safety. Single-nuclei RNA-seq show iNPC-GDNFs express NPC markers. iNPC-GDNFs
delivered into the subretinal space of the Royal College of Surgeons rodent model of retinal degeneration preserve photoreceptors and
visual function. Additionally, iNPC-GDNF transplants in the spinal cord of SOD1G93A amyotrophic lateral sclerosis (ALS) rats preserve
motor neurons. Finally, iNPC-GDNF transplants in the spinal cord of athymic nude rats survive and produce GDNF for 9 months,
with no signs of tumor formation or continual cell proliferation. iNPC-GDNFs survive long-term, are safe, and provide neuroprotection
in models of both retinal degeneration and ALS, indicating their potential as a combined cell and gene therapy for various neurodegen-
erative diseases.

INTRODUCTION nal cord of the well-characterized SOD1G93A (hereto,

SOD1) transgenic rodent model of ALS, neither benefit mo-
Several neurodegenerative diseases, including Parkinson’s tor neuron survival nor functional measures. This suggests
and Alzheimer’s, and amyotrophic lateral sclerosis (ALS), that an additional strategy may be required for neuronal
involve unique genetic and environmental risk factors, protection. As glial cell line-derived neurotrophic factor
while others such as retinitis pigmentosa (RP) involve (GDNF) has protective effects on dopamine and motor neu-
various genetic mutations. Replacing the damaged cell rons (Henderson et al., 1994; Lin et al., 1993; Zurn et al.,
population is one treatment approach; however, trans- 1994), we have genetically engineered fNPCs to stably
planted cells do not readily form new long-distance synap- secrete GDNF (fNPC-GDNF) (Capowski et al., 2007). Unlike
tic connections with their targets or integrate with second- fNPCs alone, this combined cell and growth factor ex vivo
ary retinal neurons in the adult environment. Therefore, therapy could protect spinal motor neurons in the SOD1
protection of remaining host cells is a more practical ALS rat, as well as dopamine neurons in a Parkinsonian
approach and suggests the need for a protective interven- rat model (Behrstock et al., 2006; Klein et al., 2005; Suzuki
tion that is beneficial across various neurodegenerative et al., 2007). In addition, delivery of fNPCs releasing GDNF
diseases. to the ALS rat motor cortex protected motor neurons,
ALS involves progressive motor neuron death in the spi- slowed disease progression, and extended lifespan (Thom-
nal cord and motor cortex, leading to paralysis and death, sen et al., 2018). RP involves progressive loss of photorecep-
typically within 2–5 years of diagnosis (Hulisz, 2018). tors, ultimately leading to blindness. Using the well-estab-
There are currently no effective treatments. As glial cells lished Royal College of Surgeons (RCS) rat model of retinal
are compromised in ALS, one neuroprotective strategy is degeneration, we have shown that fNPCs can protect
to provide new glial support cells (Bruijn et al., 1997; Lep- vision and photoreceptors and that GDNF-secreting fNPCs
ore et al., 2008). These can be derived from the human fetal provide enhanced protection (Gamm et al., 2007).
forebrain and then sorted to isolate glial-restricted progen- We previously expanded a single fetal cortical sample
itors (GRPs) (Lepore et al., 2011; Sandrock et al., 2010). under current Good Manufacturing Practice (cGMP) to
Alternatively, we have shown that human cortical fetal- derive a neural progenitor cell line, termed CNS10-NPC,
derived neural progenitor cells (fNPCs) can be expanded which was lentivirally transduced to stably produce
in culture as free-floating spheres (Svendsen et al., 1998) GDNF and banked as a clinical product (termed CNS10-
and can differentiate into astrocytes in vitro and in vivo NPC-GDNF) (Shelley et al., 2014). This product was used
(Das et al., 2016; Gowing et al., 2014; Svendsen et al., for a first-in-man cell and gene therapy for ALS with deliv-
1997). While GRP and fNPC transplants survive in the spi- ery to the lumbar spinal cord in a recently completed

Stem Cell Reports j Vol. 18 j 1629–1642 j August 8, 2023 j ª 2023 The Authors. 1629
This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
phase I/IIa safety trial (clinical trials.gov: NCT02943850). (Shelley et al., 2014; Svendsen et al., 1998), transduced
The primary endpoint of safety at 1 year was met, with with lentivirus to stably express GDNF (Capowski et al.,
no negative effect of the transplant on motor function, 2007; Shelley et al., 2014), then propagated and banked un-
and grafts survived and produced GDNF for up to der cGMP as the CNS10-NPC-GDNF clinical cell lot, and
42 months post-transplantation (Baloh et al., 2022). fNPC-GDNFs were banked as a research-grade cell lot
CNS10-NPC-GDNF is now being delivered to the motor used in this study. To produce iPSC-derived NPCs similar
cortex of ALS patients in a phase I/IIa clinical trial to these fNPCs, a new and substantially shorter protocol
(NCT05306457). Finally, a current phase I/IIa clinical trial was developed. An iPSC line was generated by nucleo-
is transplanting CNS10-NPC into the subretinal space of fecting healthy PBMCs with nonintegrating oriP/EBNA1
RP patients (NCT04284293). plasmids, which allowed for episomal expression of reprog-
Fetal-derived NPCs and their derivatives are a promising ramming factors (Barrett et al., 2014). Using dual SMAD
treatment for Parkinson’s disease, ALS, and RP, as well as inhibition (Chambers et al., 2009), a monolayer of neuroe-
other neurological conditions such as Huntington’s dis- pithelial progenitors was efficiently generated, which were
ease, stroke, and dementia (Andres et al., 2011; Goldberg then collected and transitioned into neurospheres in the
et al., 2017; McBride et al., 2004). However, this product presence of epidermal growth factor (EGF), fibroblast
has some limitations, including low fetal tissue availabil- growth factor 2 (FGF2), and leukemia inhibitory factor
ity that can hinder large scale-up, thus restricting phase (LIF) (Figure 1A). iNPC neurospheres were expanded by me-
III trials and commercialization. Our lab and others have chanical passage using a tissue chopper or mesh chopper.
shown that induced pluripotent stem cells (iPSCs) provide After passage 16, iNPCs were dissociated to single cells
a renewable, scalable, and safe cell source for deriving po- and lentivirally infected to stably express GDNF. Trans-
tential cell products (Rosati et al., 2018; Sareen et al., duced cells reformed as iNPC-GDNF neurospheres that
2014; Tsai et al., 2015). Peripheral blood mononuclear were used in the current studies.
cells (PBMCs) from an individual’s blood sample can be To demonstrate consistency in the production process,
reprogrammed into iPSCs, which proliferate in culture, two independent batches of iNPC-GDNFs were produced.
differentiate into multiple tissue types, and can be cryo- G-band karyotyping revealed that one batch remained
preserved and thawed (Barrett et al., 2014). We have karyotypically normal, while the other batch acquired a
demonstrated that human iPSCs cultured in specified me- karyotypic anomaly not present in the originating iPSC
dia could generate neural-specific cultures that were prop- line with nearly 100% trisomy of chromosome 12, termed
agated as spheres (termed EZ spheres), which could iNPC-GDNF-T12 (Figure S1A). To avoid complications that
engraft and differentiate into astrocytes in the rodent spi- could arise from karayotypic abnormality, subsequent
nal cord and retina (Ebert et al., 2013; Sareen et al., 2014; studies were performed with the karyotypically normal
Tsai et al., 2015). iNPC-GDNF batch, which remained normal up to 19 pas-
Here, we developed and tested an iPSC-based therapy as sages. To identify any potential remaining pluripotent cells
an alternative to fetal-derived products. We generated in the differentiated batches, pluripotency genes POU5F1
iPSC-derived NPCs (termed iNPCs), which were transduced (OCT4), NANOG, SOX2, and KLF4 were plotted from sin-
to express GDNF (iNPC-GDNF), characterized, and gle-nuclei RNA sequencing (snRNA-seq) data (Figure S1B).
compared with fNPC-GDNFs. iNPC-GDNFs were trans- While individual cells expressed some of these markers at
planted into the subretinal space of RCS rats where they low levels, no cell expressed all four or any combination
protected photoreceptors and vision, the spinal cord of of OCT4, NANOG, and KLF4 (Figure S1C). As expected,
SOD1 rats where they protected motor neurons, and the SOX2 was expressed throughout most cells (Figure S1B),
spinal cord of nude rats where they showed long-term sur- as it is also a NPC marker (Zhang and Cui, 2014).
vival and safety. Based on safety and efficacy, iNPC-GDNFs snRNA-seq was performed to assess the similarity be-
can be pursued as a promising combined cell and gene ther- tween iNPC-GDNF and fNPC-GDNF neurosphere cultures.
apy for multiple neurodegenerative diseases. Following batch normalization, UMAP (uniform manifold
approximation and projection) was applied to identify
clusters of cell types in an unbiased manner, which identi-
RESULTS fied 11 separate clusters (Figure 1B, Table S1). When split by
sample source, fNPC-GDNFs were found in all clusters
iNPC-GDNFs and derived astrocytes resemble fNPC- except for 2 and 4, which were almost exclusively iNPC-
GDNFs and derived astrocytes GDNFs, while clusters 8 and 10 were entirely fNPC-GDNFs
NPCs derived from a single human fetal cortex were main- (Figure 1C, Table S1). Importantly, the overlap of a popula-
tained as free-floating spheres (termed neurospheres), tion of iNPC-GDNF clusters in most fNPC-GDNF clusters
expanded by mechanical passage using a tissue chopper demonstrates the remarkable similarity in some cell

1630 Stem Cell Reports j Vol. 18 j 1629–1642 j August 8, 2023

 A

B C

D F

Figure 1. iNPC-GDNF and fNPC-GDNF profiles

(A) Protocol to generate and expand iNPC-GDNFs.
(B) Unbiased UMAP clustering of cells from fNPC-GDNF and iNPC-GDNF neurospheres split into 11 defined clusters.
(C) Clustering split by cell type shows that iNPC-GDNFs clustered similarly to fNPC-GDNFs in many but not all cases.
(D and E) Feature plots show expression of (D) cortical neural progenitor markers and (E) astrocyte markers.
(F) Immunocytochemistry on fNPC-GDNFs and iNPC-GDNFs differentiated for 7 days in culture shows production of S100b and GFAP, with
DAPI (blue). Scale bar represents 75 mm.
See also Table S1 and Figure S1.

Stem Cell Reports j Vol. 18 j 1629–1642 j August 8, 2023 1631

 A B C

D E F G

Figure 2. iNPC-GDNFs are neuroprotective in the RCS rat model of retinal degeneration
(A) Schematic of experiment RCS rats (n = 13) that received subretinal injection of iNPC-GDNFs at postnatal day (P) 21–23. The fellow eye
received either balanced salt solution injection (sham, n = 4) or no treatment (n = 9).
(B and C) Visual function was evaluated by (B) OKR and (C) ERG, which showed significant preservation in cell-treated compared with
control animals (****p < 0.0001, **p < 0.01, *p < 0.05, one-way ANOVA with Tukey’s HSD, ns: not significant).
(D and E) Cresyl violet-stained retinal images show photoreceptor preservation in (D) iNPC-GDNF-treated retina with potential grafted cells
(arrowheads), compared with (E) sham. d1, d2, e1, and e2 are high-power images from areas in boxed outlines.
(F) Percent of retina with at least two layers of photoreceptors in iNPC-GDNF-treated vs. sham (***p < 0.001 via unpaired t test with
Welch’s correction).

(legend continued on next page)

1632 Stem Cell Reports j Vol. 18 j 1629–1642 j August 8, 2023

populations. Indeed, markers of cortical NPCs were highly compared with controls at P60 and P90 (Figure 2C). ERG
expressed throughout all cells in this analysis (Figure 1D). does show visual impairment compared with a wild-type
Exclusively iNPC-GDNF clusters tended to express markers rat, as ERG was compromised already when treatment
for the immature progenitors (VIM, SOX2, NESTIN). Inter- started. But, critically, iNPC-GDNF treatment provided sus-
estingly, clustering was often driven by cell cycle phase tained protection of both ERG and OKR over the study
(Figure S1D), again indicating strong similarities between duration.
cell types, as this distinction does not commonly appear To assess protection of retinal morphology at P90, the
in more diverse datasets. Cluster 0 cells showed high length of preserved photoreceptors was measured against
expression of astrocyte-related genes (Figure 1E). Given the whole length of cresyl violet-stained retinal sections,
that this cluster had a low contribution in iNPC-GDNF which showed that iNPC-GDNF treatment preserved three
neurospheres (Table S1), we wanted to assess their differen- to six photoreceptor layers, compared with most controls
tiation potential. iNPC-GDNFs and fNPC-GDNFs were with only one layer remaining (Figures 2D and 2E). Quan-
differentiated as monolayer cultures for 7 days, with subse- tifying the outer nuclear layer (ONL) showed that iNPC-
quent immunocytochemistry showing that both produced GDNFs protected an average of 30.59% ± 3.6% of the
the astrocyte markers glial fibrillary acidic protein (GFAP) retina, indicating roughly one-third of the retinal sections
and S100b (Figure 1F). Flow cytometry further demon- retained at least two layers of photoreceptors (Figure 2F).
strated that 80% of differentiated iNPC-GDNFs acquired Further, staining with the human-specific nuclear antibody
GFAP production (Figure S1E). Collectively, results demon- MAB1281 confirmed iNPC-GDNFs had robust survival and
strate that while some differences exist in fetal- and iPSC- extensive migration in the subretinal space (Figure 2G)
derived NPCs, there also are substantial similarities. and, notably, demonstrated multiple layers of photorecep-
tors associated with donor cells at the injection site (Fig-
iNPC-GDNFs preserve vision and photoreceptors in the ure 2G1) and with migrating donor cells (Figure 2G2).
RCS rat model iNPC-GDNF treatment also preserved cone photoreceptors,
Fetal- and iPSC-derived NPCs can protect vision following a revealed by immunofluorescence with a cone arrestin anti-
single subretinal injection into the RCS rat (Gamm et al., body that showed inner and outer segments and dense
2007; Tsai et al., 2015; Wang et al., 2008). Importantly, cone pedicles (Figure 2H). In contrast, controls had only
GDNF-secreting fNPCs provide enhanced vision protec- weak staining for cones, with no visible segments and
tion, and GDNF delivery can preserve photoreceptors cone pedicles. Retinal connections, based on wide and
(Gamm et al., 2007; Garcı́a-Caballero et al., 2018). To eval- disperse synaptophysin staining, were evident in cell--
uate the efficacy of iNPC-GDNFs, RCS rats at postnatal day treated retinas, compared with narrow staining in controls
(P) 21–23 received a single unilateral subretinal injection of (Figure 2H). Strong GFAP staining in control retinas indi-
cells, and the fellow eye served as the control, with cates a reactive Müller glia response (Figure 2I). This was
balanced salt solution (sham) injection or no treatment substantially reduced in the treated retina and mainly
(Figure 2A). located in the optic nerve fiber layer, suggesting that cell
Visual acuity measured by the optokinetic response treatment reduced Müller glia activation. The pan-GFAP
(OKR) showed that sham-treated and untreated eyes had antibody also stained grafted iNPC-GDNFs, indicating dif-
a significant decrease in visual acuity from P60 to P90 ferentiation into astrocytes.
(Figure 2B). However, there was almost complete rescue Immunofluorescence with human-specific antibodies for
with iNPC-GDNF treatment, with visual acuity over time Nestin and astrocytes (SC123) was used to characterize
remaining around 0.55 cycle/degree (Figure 2B), compara- grafted iNPC-GDNFs, which showed that cells remained
ble to wild-type rats (McGill et al., 2007). Electroretinog- as neural progenitors or differentiated into astrocytes
raphy (ERG), measuring the average response of the whole (Figures S2A and S2B). Consistent with our prior study
retina to light stimulation, revealed that iNPC-GDNF- (Gamm et al., 2007), a few grafted cells migrated into the
treated eyes had significantly higher b-wave amplitude inner retina, but most formed a layer of cells or lump

(G) Staining with human nuclear marker MAB1281 shows extensive distribution of grafted iNPC-GDNFs; g1 and g2 are high-power images
from areas in boxed outlines.
(H) Synaptophysin and cone arrestin staining show cones with segments and pedicles (arrowheads) were preserved in cell-treated retina
compared with controls.
(I) GFAP staining shows reactive Müller glia in controls and cell-treated retina, GFAP labels iNPC-GDNFs, and host Müller glia (stars, arrows
point to Müller glial end feet). Scale bars represent (D, E, and G) 500 mm and (H and I) 75 mm. INL, inner nuclear layer; IPL, inner plexiform
layer; ONL, outer nuclear layer. d1, e1, and g1 indicate regions near injection site, d2, e2, and g2 indicate regions distal from injection site.
See also Figure S2.

Stem Cell Reports j Vol. 18 j 1629–1642 j August 8, 2023 1633

 A B C

E F

G H

(legend on next page)

1634 Stem Cell Reports j Vol. 18 j 1629–1642 j August 8, 2023

within the subretinal space. While Ki67 staining indicates ImageJ analysis revealed a significantly elevated ratio of to-
proliferating host cells in the degenerative environment, tal number of motor neurons (transplanted/non-trans-
only a few Ki67-positive human cells demonstrated planted) in treated compared with untreated animals
the limited proliferation of transplanted iNPC-GDNFs (Figure 3F). We have previously shown that fNPC-GDNF
(Figure S2C). specifically protect large motor neurons (<600 mM) in the
ALS rat spinal cord and motor cortex (Suzuki et al., 2007;
iNPC-GDNFs are neuroprotective in SOD1 ALS rat Thomsen et al., 2018). Here, the number of ChAT+ motor
To evaluate whether iNPC-GDNFs are protective in SOD1 neuron across all size bins was significantly increased be-
ALS rats, cells were injected unilaterally into the lumbar tween transplanted and non-transplanted sides in treated
spinal cord, with untreated SOD1 rats serving as controls animals, with no difference across size bins between
(Figure 3A). Treated and untreated groups had similar the left and right spinal cord in untreated animals
body weights over the study course (Figure S3A). Basso, (Figures 3G and 3H). This indicates that iNPC-GDNFs pro-
Beattie, and Bresnahan (BBB) locomotor scores were vide neuroprotection of motor neurons in the ALS rat spi-
collected weekly, starting 1 week before transplantation, nal cord.
to assess hindlimb function and disease onset. There was
no significant difference in BBB scores for transplanted/ iNPC-GDNFs show long-term survival and safety in the
non-transplanted hindlimbs in treated and untreated nude rat spinal cord
rats, indicating no delay in disease onset (Figures 3B To examine iNPC-GDNF long-term survival and safety,
and 3C). cells were transplanted unilaterally into the lumbar spinal
To assess iNPC-GDNF survival and host motor neurons, cord of athymic nude rats (Figure 4A). These rats have a
rats were euthanized at disease onset, and tissue was spontaneous mutation in the Foxn1 gene and thus lack
analyzed by immunohistochemistry. Grafted cells stained T cells, rendering them immunodeficient to enable xeno-
positive for the human-specific cytoplasmic marker graft survival without immunosuppression. iNPC-GDNF
(SC121) with some cells co-staining for the neural progen- transplants survived well in both gray matter and
itor marker Nestin (Figure S3B). Minimal co-staining of white matter spinal cord regions (Figures S4A and S4B)
SC121 with the neuronal marker Neun (Figure S3C) and produced GDNF (Figure 4B) for up to 9 months. Histo-
coupled with extensive co-staining for GFAP (Figure S3D) pathological examination of hematoxylin and eosin stain-
indicate a primarily glial identity of the grafted cells. ing showed no structural abnormalities indicative of
SC121 staining confirmed iNPC-GDNFs in the treated spi- cancerous growth in any section across all animals. Ki67-
nal cord, but not the non-treated side, with engraftment postive cells were observed at 3 months but not 9 months
around choline-o-acetyl transferase (ChAT)-positive host post-transplantation, indicating reduced proliferation over
motor neurons (Figure 3D). Robust GDNF production was time (Figures S4C and S4D). At 9 months post-transplanta-
seen in the iNPC-GDNF-treated spinal cord but not in the tion, some iNPC-GDNFs retained Nestin expression (Fig-
non-transplanted side (Figure 3E). While all transplanted ure S4E) with relatively little overlap of Neun (Figure S4F).
animals showed surviving grafts, the location and extent Strong staining from an antibody that recognizes human
varied (Figures S3E and S3F), and it was observed that better and rat GFAP (pan-GFAP) and from a human-specific
graft targets correlated with greater motor neuron counts. GFAP antibody (SC123) demonstrated that cells differenti-
To assess whether iNPC-GDNFs protected host motor neu- ated primarily into astrocytes (Figure S4G), and higher
rons, spinal cord sections containing SC121+ grafts were magnification shows clear astrocyte morphology for
quantified for ChAT+ motor neuron numbers and size. iNPC-GDNFs co-labeled with SC121 and GFAP (Figure 4C).

Figure 3. iNPC-GDNFs are neuroprotective in the SOD1 ALS rat

(A) Schematic of experiment. 10 male SOD1 rats at 70–95 days were unilaterally transplanted (left, L) with 10,000 iNPC-GDNFs in three
sites 2 mm apart in the lumbar spinal cord and compared with 9 untreated animals.
(B) Kaplan-Meier shows probability of onset times of treated vs. untreated animals are not significantly different.
(C) Hindlimb BBB scores at disease onset are not significantly different.
(D) Immunohistochemistry shows SC121+ human grafts surrounding host ChAT+ motor neurons, with DAPI (blue).
(E) GDNF staining of transplanted compared with non-transplanted spinal cord.
(F) Ratio of ChAT+ motor neurons averaged per animal (n = 7 untreated and n = 10 treated animals).
(G and H) ChAT+ neuron size in (G) untreated animals (n = 6, average of four sections per animal) and (H) treated animals (n = 8, average of
at least eight sections per animal); *p < 0.05 via unpaired t test with Welch’s correction (F) or multiple paired t tests corrected with the
Holm-Sı́dák method (G and H); ns: not significant. Scale bars represent (D) 75 mm and (E) 250 mm.
See also Figure S3.

Stem Cell Reports j Vol. 18 j 1629–1642 j August 8, 2023 1635

 A B

D E

Figure 4. iNPC-GDNF spinal cord long-term grafts

(A) Schematic of experiment. Rats (n = 10) were unilaterally transplanted with 100,000 iNPC-GDNFs in three sites 2 mm apart in the lumbar
spinal cord, with euthanasia at 3 months (n = 1) and 9 months (n = 9) post-transplantation.
(B and C) Grafts at 9 months were stained for (B) GDNF and (C) pan-GFAP and SC121 with DAPI (blue).
(D and E) SC121 and IBA1 staining (D) and quantification of IBA1+ cells with a paired t test (E) (average per animal, minimum of four
sections analyzed per animal). ns: not significant. Scale bars represent (B) 75 mm, (C) 20 mm, and (D) 50 mm.
See also Figure S4.

To examine effects of long-term iNPC-GDNF transplants (Figure 4D). While microglia number varied across animals,
on spinal cord inflammation, immunohistochemistry was there was no significant increase in microglia surrounding
performed for human cells and the microglial marker IBA-1 the SC121+ graft compared with the contralateral spinal

1636 Stem Cell Reports j Vol. 18 j 1629–1642 j August 8, 2023

cord (Figure 4E), indicating that iNPC-GDNFs did not (Liu et al., 2013). While fewer cells in cluster 0 suggests
enhance long-term inflammation. Collectively, results that iNPC-GDNF neurospheres are initially not as specified
demonstrate that long-term iNPC-GDNF transplants can toward a glial fate as fNPCs, iNPC-GDNFs do develop into
survive, differentiate into glia, and are safe with no evi- an astrocytic fate after differentiation in culture or post-
dence of tumor growth. transplantation, indicating cells likely had an early glial
progenitor fate. After extended passaging, the iNPC-
GDNF line used throughout this study remained karyotyp-
DISCUSSION ically normal; however, one batch of cultured iNPC-GDNFs
developed trisomy 12, as reported with cultured pluripo-
Neurodegeneration encompasses many different diseases tent human embryonic stem cells (Draper et al., 2004).
and cell types with diverse genetic and environmental Interestingly, we have also shown that cultured fNPCs
causes that elude pathway-targeted therapies. Instead, can develop a trisomy, on chromosomes 7 and 19, but still
NPCs that engraft and differentiate into supportive glia, do not develop tumors after transplantation (Sareen
and which can be genetically engineered to produce a et al., 2009).
growth factor, are broadly applicable to many diseases. A Human iPSC-derived neural progenitors have been pre-
focal delivery approach is critical as GDNF cannot cross viously examined by our lab and others. Our EZ sphere
the blood-brain barrier, and systemic delivery has been protocol required expansion to at least passage 10 and a
associated with adverse effects (Thomsen et al., 2017). A complex 14-week differentiation process to generate speci-
combined cell and gene therapy approach based on growth fied neural progenitors (Ebert et al., 2013; Sareen et al.,
factor release is particularly promising for sporadic ALS, 2014; Tsai et al., 2015). The new process developed here in-
where there is no known gene mutation for targeted gene volves a simple 10-day monolayer treatment with the dual
therapy approaches. This approach can also be used to treat SMAD inhibitors LDN193189 and SB431542 to efficiently
retinal degeneration, regardless of specific mutations. We generate neuroepithelial progenitor cells, which are then
have shown that fNPC-GDNFs are a powerful cell and adapted to suspension culture and expanded. This simple
gene therapy product for several neurological disorders. and fully defined process for iNPC generation is now
But unlike fetal-derived cells, iPSC derivation from an indi- primed for transition to cGMP production.
vidual’s blood or skin avoids low tissue availability and Similar to iPSC-derived neural progenitor transplants in
permits for personalized medicine (Svendsen, 2013). the RCS rat (Tsai et al., 2015), this study shows that iNPC-
Furthermore, while fetal-derived transplants in the brain GDNFs survive and migrate extensively in the subretinal
may not require long-term immunosuppression (Mendez space. We have previously shown that neural progenitor
et al., 2005), patient-specific iPSCs could avoid side effects transplants do not provide retinal cell replacement (Jones
that can arise from even short-term immunosuppression. et al., 2016; Tsai et al., 2015). Rather, cells can differentiate
We now have developed human iPSCs that stably produce into astrocytes that could improve the diseased environ-
GDNF as a promising future cell and gene therapy. ment. A single injection of cells protected 30% of the retina
Though senesce in vitro can be disadvantageous as fNPCs and preserved visual function. fNPCs can regulate the
cannot be banked indefinitely (Wright et al., 2006), immune response by inhibiting microglial activation, pro-
reduced proliferation over time in vivo is advantageous to mote antioxidant effects, phagocytose retinal pigment
reduce tumorigenicity potential, as we have shown in the epithelium outer segments, and release trophic factors
rodent central nervous system (CNS) (Gowing et al., FGF2 and IGF-1 (Jones et al., 2016; Tsai et al., 2015). In addi-
2014; Ostenfeld et al., 2000) and critically in the human tion to these mechanisms of action, GDNF provides direct
spinal cord up to 42 months post-transplantation (Baloh protection of photoreceptors via GDNF receptors, as for
et al., 2022). In contrast, while continued proliferation motor neurons (Jomary et al., 2004; Trupp et al., 1996).
with iPSCs facilitates large scale-up (Ausubel et al., 2011), GDNF also indirectly protects photoreceptors via GDNF re-
a risk of tumor formation exists if pluripotent cells remain ceptors on retinal Müller glia (Hauck et al., 2006), which
within the iNPC cultures (Yamanaka, 2020). However, this then increase production of bFGF, BDNF, and GDNF (Har-
study shows that iNPC-GDNFs have no combined expres- ada et al., 2003).
sion of pluripotency markers OCT4, NANOG and While we demonstrated EZ sphere-derived NPC engraft-
KLF4 in vitro. Furthermore, iNPCs are non-tumorigenic ment in the rat spinal cord (Sareen et al., 2014), we had
post-transplantation in this study and our previous studies not demonstrated efficacy in a disease model. Here we
(Sareen et al., 2014; Tsai et al., 2015), and we have now show that iNPC-GDNFs engraft in the SOD1 rat spinal
extended this finding to 9 months. Reduced tumorigenic cord and protect motor neurons. Interestingly, there was a
potential upon lineage differentiation has also been shown significant increase in ChAT+ neuron numbers across all
with iPSC-derived cardiomyocytes compared with iPSCs size bins between transplanted and non-transplanted sides

Stem Cell Reports j Vol. 18 j 1629–1642 j August 8, 2023 1637

in treated animals, in contrast to fNPC-GDNF that specif- iNPC-GDNF efficacy following delivery to both the spinal
ically protect large motor neurons (>600 mM) in the spinal cord and motor cortex in ALS rats. An ex vivo combined cell
cord and motor cortex (Suzuki et al., 2007; Thomsen and gene therapy approach permits delivery of both support-
et al., 2018). Disparities in cell effectiveness on motor ive astrocytes and a protective neurotrophic factor without
neuron survival warrant further investigation and may in vivo genetic manipulations (Okano, 2022a). This approach
relate to differences in fetal and iPSC-derived NPC snRNA- has broad utility across multiple neurodegenerative condi-
seq profiles. While motor neurons were protected, neither tions and CNS injuries, highlighting the need for scalable
iNPC-GDNFs in this study or fNPCs in prior studies signifi- development of cell-based therapies (Okano, 2022b).
cantly protected hindlimb motor function (Suzuki et al., iNPC-GDNFs can protect diseased cells and function.
2007), which may be due to the severity of this transgenic These cells show long-term engraftment and GDNF produc-
rat model. tion, as well as differentiation into astrocytes. Grafted spinal
Motor neuron protection correlated with graft location, cords showed no evidence of cancerous growths or uncon-
with the greatest protective effect achieved when the graft trolled cell proliferation. Long-term safety data, coupled
was most closely associated with the ventral horn. Off- with the very low levels of pluripotency genes expressed
target grafts did not provide improvement in motor neuron in vitro, suggests that iNPC-GDNFs are a safe, scalable, and
number or size, highlighting that transplant delivery and renewable source for cell transplantation therapies. Our cur-
targeting are critical to achieve neuroprotection. This was rent efficacy and safety studies provide a strong rationale
evident in our phase I/IIa clinical trial delivering CNS10- to pursue the use of iNPC-GDNFs. Generating a GMP
NPC-GDNF to the ALS patient spinal cord, in which dorsal bank of iNPC-GDNFs followed by investigational new
grafts may have contributed to cell reflux, neuroma forma- drug-enabling safety studies with the clinical product are
tion at the dorsal root ganglia, pain in some participants, the next steps required to move this promising combined
and lack of overall effects on motor neurons and function cell and gene therapy into clinical trials for various neurode-
(Baloh et al., 2022). generative diseases.
iNPC-GDNFs have clear potential, though some limita-
tions exist. Off-site or excessive GDNF levels can cause side
effects in patients as well as rare cerebellar Purkinje cell loss EXPERIMENTAL PROCEDURES
and aberrant neuronal sprouting in animal models (Geor-
Resource availability
gievska et al., 2002; Hovland et al., 2007; Nutt et al., 2003).
Corresponding authors
While we confirmed GDNF is safe up to 42 months in the hu-
Clive Svendsen: [email protected], Shaomei Wang:
man spinal cord (Baloh et al., 2022), future studies should
[email protected].
develop controllable GDNF release (Akhtar et al., 2018) to Materials availability
tailor GDNF transgene expression for an individualized pa- Requests for raw and analyzed data and materials are promptly re-
tient dose and to permit gene shut-off in the case of severe viewed by the Cedars-Sinai Board of Governor’s Regenerative Med-
side effects. We previously demonstrated that lenti-GDNF icine Institute to verify if the request is subject to any intellectual
transduction with a viral titer similar to that used here property or confidentiality obligations. Patient-related data not
yielded fNPCs with about two to four inserted gene copies included in the paper may be subject to patient confidentiality.
per cell (Capowski et al., 2007; Shelley et al., 2014). While Any data and materials that can be shared will be released via a ma-
the location of GDNF transgene integration was not deter- terial transfer agreement.
mined, we have extensively confirmed that transduced cells Data and code availability
are safe following transplantation. A caveat to our studies is All transcriptomic data in this study are available in the GEO repos-
itory: GSE214210. R code is available on GitHub: https://github.
that RP and ALS rat models were treated before overt disease
com/shaughnmb/2022_laperle_et_al.
onset, so treatment still needs to be assessed at a clinically
relevant stage after disease onset. Further, while iNPC-
GDNFs protected motor neurons in the SOD1 rat, disease Ethics statement
onset was not significantly delayed. Given that both spinal All cell lines and protocols in the present study were used in accor-
and cortical motor neurons die in ALS and that upper motor dance with guidelines approved by the Stem Cell Research Over-
sight committee and institutional review board under the auspice
neurons play a critical role in disease progression (Thomsen
institutional review board and Stem Cell Research Oversight proto-
et al., 2014), treating more diseased sites is likely required for
cols Pro00032834 (iPSC Core Repository and Stem Cell Program)
a greater therapeutic benefit. fNPC-GDNF transplants in the
and Pro00021505 (Svendsen Stem Cell Program). All animal
ALS rat motor cortex delayed disease onset and extended life- work was approved and performed under the guidelines of the
span (Thomsen et al., 2018), providing the basis for our Cedars-Sinai Medical Center Institutional Animal Care and Use
current clinical trial delivering CNS10-NPC-GDNF to the Committee under protocols 8517 for spinal cord and 7611 for
motor cortex of ALS patients. Future studies should evaluate retina. All animals in retinal studies were treated in accordance

1638 Stem Cell Reports j Vol. 18 j 1629–1642 j August 8, 2023

with the ARVO Statement for the Use of Animals in Ophthalmic mals received daily intraperitoneal injection of dexamethasone
and Vision Research. for 2 weeks (1.6 mg/kg per day) post-surgery and were adminis-
tered cyclosporine A in the drinking water (210 mg/L). Animals
Cell culture were euthanized at P90 by CO2 inhalation. RCS rats were tested
The iPSC line (named CS02iCTR-Tn11) was generated by the by OKR and ERG at approximately P60 and P90, per our published
Cedars-Sinai iPSC Core using previous protocols (Barrett et al., protocols (Gamm et al., 2007; Wang et al., 2008). See supplemental
2014). iPSCs were maintained in E8 medium on Matrigel and experimental procedures for full details.
passaged every 5 days using Versene. iPSCs at passage 17–35 were SOD1G93A rats
used. For differentiation, iPSCs at 80% confluency were singular- At day 70–95, male rats (n = 10) were injected with 5,000 iNPC-
ized with Accutase and plated onto Matrigel-coated six-well plates GDNF cells/ml at 2 mL per site, into three sites spaced 2 mm apart
at 200,000 cells per cm2 in E8 medium with 5 mM Y27632. Cultures in the left lumbar spinal cord (X = 0.75 mm, Z = 1.65 mm), along
were induced toward a neural progenitor cell using dual SMAD in- with untreated rats (n = 9) as controls. Rats were immunosup-
hibition with differentiation media (supplemental experimental pressed with daily intraperitoneal injections of cyclosporine
procedures for full details) for 10 days with daily media changes. (10 mg/kg). Rats were euthanized at disease onset, defined by
Cultures were then treated with Versene and gently lifted with a two consecutive BBB scores %15, by transcardial perfusion with
cell scraper into SEFL media (Stemline with EGF, FGF2, LIF and 4% PFA. Body weight was monitored weekly. A modified BBB loco-
heparin) (supplemental experimental procedures for full details). motor test (Basso et al., 1995) assessed hindlimb function. Kaplan-
Aggregates were transferred to a poly-HEMA-coated T25-flask to Meier was performed to assess onset.
establish neurospheres. Differentiated iNPCs were expanded in Nude rats
SEFL media as neurospheres in poly-HEMA-coated tissue culture Male immunodeficient athymic nude rats (n = 10) (Charles River)
flasks for up to 30 passages. Using a McIlwain tissue chopper (Shel- at 100 days were injected with 50,000 iNPC-GDNF cells/ml as
ley et al., 2014; Svendsen et al., 1998) or in-house mesh chopping above. Animals were euthanized at 3 months post-transplantation
device, spheres were passaged weekly. iNPCs were transduced (n = 1) to confirm engraftment and 9 months (n = 9), by transcar-
with lentivirus GDNF (0.125 pg of p24 per cell) to create the dial perfusion.
iNPC-GDNFs as reported (Capowski et al., 2007; Shelley et al., Retinal and spinal cord analysis
2014). iNPC-GDNF neurospheres were collected at passage 29, Retinal sections were processed as published (Tsai et al., 2015). See
dissociated with TrypLE, resuspended in cell freezing medium, supplemental experimental procedures for full details. The length
and cryopreserved. of ONL protection was measured on cresyl violet-stained retinal
montage sections (more than two layers of ONL, six retinas, six sec-
Immunocytochemistry tions/retina) against the whole retinal length using Java-based im-
Cryopreserved vials containing single-cell suspensions of fNPC- age processing software (ImageJ; National Institutes of Health, Be-
GDNFs or iNPC-GDNFs were plated on poly-l-ornithine-treated thesda, MD). Spinal cords were processed as previously described
Matrigel-coated glass coverslips. Cells were differentiated in Stem- (Gowing et al., 2014). See supplemental experimental procedures
line medium with 2% B27 supplement for 7 days, with medium ex- for full details.
change every third day. Cells were fixed in 4% paraformaldehyde ChAT+ cell quantification
(PFA), washed in phosphate buffered saline (PBS), permeabilized Four sections adjacent to identified graft sites and co-labeled for
in 1% Triton X-100 in PBS, and stained overnight at 4 C with SC121 and ChAT+ host motor neurons were imaged at 10x using
GFAP and S100b antibodies in 5% normal donkey serum, the Leica DFC365 FX camera, Leica DM6000 B microscope, and Le-
0.125% Triton X-100 and PBS. Samples were washed in PBS and ica Application Suite Advanced Fluorescence 3.2.0.9652 program.
stained with secondary antibodies for 2 h at room temperature, fol- Images from treated animals (n = 8, average of at least eight sec-
lowed by a DAPI nuclear counterstain. Images were acquired using tions per animal) and untreated animals (n = 6, average of four sec-
a Leica DM6000B microscope with a 20x objective. tions per animal) were used for motor neuron counts using the
Freehand Selections tool and Region of Interest Manager in ImageJ.
Single-nuclei RNA sequencing Images also underwent automated size analysis of motor neuron
Nuclei were isolated from fNPC-GDNF and iNPC-GDNF neuro- areas using IMARIS software with manual thresholding. Untreated
spheres. For single-nuclei library preparation and sequencing, (n = 2) and treated (n = 2) animals were removed from the IMARIS
the standard 10x protocol was used per the "Chromium analysis due to insufficient contrast in the ChAT staining for the
NextGEM Single Cell 3’ Reagent Kits v3.1 User Guide, Rev D." software to accurately distinguish positive staining from
See supplemental experimental procedures for full details. background.
Regressive H&E (nude rats)
Animals For hematoxylin and eosin (H&E) staining, mounted sections were
RCS rats dried and then defatted and rehydrated in dH2O. Sections were
Pigmented dystrophic RCS rats (rdy+, p+) (n = 13) received subreti- stained with hematoxylin for 10 min and washed in running tap
nal injection of 2 mL of iNPC-GDNFs at 15,000 cells/mL in balanced water. Slides were dipped 5–10 times in 1% HCl in 70% EtOH,
salt solution (BSS) at postnatal day (P) 21–23. The fellow eye served then washed in running tap water and dipped in 0.5% lithium car-
as the control, with either BSS injection (sham, n = 4) or no treat- bonate in ammonia water (0.1%) for 30 s. Sections were washed in
ment (n = 9) with our published protocol (Tsai et al., 2015). Ani- tap water and stained with eosin (diluted 1:5 in dH2O) for 5–10 s.

Stem Cell Reports j Vol. 18 j 1629–1642 j August 8, 2023 1639

Slides were then dehydrated in 95% EtOH, three changes of 100% Andres, R.H., Horie, N., Slikker, W., Keren-Gill, H., Zhan, K., Sun,
EtOH, three changes of Xylene, and coverslipped with mounting G., Manley, N.C., Pereira, M.P., Sheikh, L.A., McMillan, E.L., et al.
media. (2011). Human neural stem cells enhance structural plasticity
and axonal transport in the ischaemic brain. Brain 134,
1777–1789.
Statistics
Use of statistical tests is described in legends. GraphPad Prism 9 Ausubel, L.J., Lopez, P.M., and Couture, L.A. (2011). GMP scale-up
software was used for all calculations. Error bars represent stan- and banking of pluripotent stem cells for cellular therapy applica-
dard deviations. Statistical tests include one-way ANOVA with Tu- tions. Methods Mol. Biol. 767, 147–159.
key’s HSD, unpaired t test with Welch’s correction, and multiple Baloh, R.H., Johnson, J.P., Avalos, P., Allred, P., Svendsen, S., Gow-
paired t tests with Holm-Sı́dák correction. Survival curves were ing, G., Roxas, K., Wu, A., Donahue, B., Osborne, S., et al. (2022).
compared by Log rank Mantel-Cox test (chi square 0.6011, df 1, Transplantation of human neural progenitor cells secreting
p value 0.4381) and Gehan-Breslow-Wilcoxon test (chi square GDNF into the spinal cord of patients with ALS: a phase 1/2a trial.
0.7974, df 1, p value 0.3719). Significance is considered at Nat. Med. 28, 1813–1822.
p < 0.05. Barrett, R., Ornelas, L., Yeager, N., Mandefro, B., Sahabian, A., Le-
naeus, L., Targan, S.R., Svendsen, C.N., and Sareen, D. (2014). Reli-
SUPPLEMENTAL INFORMATION able generation of induced pluripotent stem cells from human
lymphoblastoid cell lines. Stem Cells Transl. Med. 3, 1429–1434.
Supplemental information can be found online at https://doi.org/
Basso, D.M., Beattie, M.S., and Bresnahan, J.C. (1995). A sensitive
10.1016/j.stemcr.2023.03.016.
and reliable locomotor rating scale for open field testing in rats.
J. Neurotrauma 12, 1–21.
AUTHOR CONTRIBUTIONS
Behrstock, S., Ebert, A., McHugh, J., Vosberg, S., Moore, J.,
Conceptualization: A.H.L., P.A., S.W., and C.N.S. Methodology: Schneider, B., Capowski, E., Hei, D., Kordower, J., Aebischer, P.,
A.H.L., A.M., P.A., V.J.G., S.B., R.H., G.L., K.R., O.S., B.L., S.R., and Svendsen, C.N. (2006). Human neural progenitors deliver glial
S.W., and C.N.S. Investigation: A.H.L., A.M., S.W., and C.N.S. Visu- cell line-derived neurotrophic factor to parkinsonian rodents and
alization: A.H.L., A.M., A.W., A.F., S.R., K.R., S.B., V.J.G., S.S, S.W., aged primates. Gene Ther. 13, 379–388.
and C.N.S. Funding acquisition: C.N.S. Project administration: Bruijn, L.I., Becher, M.W., Lee, M.K., Anderson, K.L., Jenkins, N.A.,
S.W. and C.N.S. Supervision: A.H.L., A.M., S.W., and C.N.S. Copeland, N.G., Sisodia, S.S., Rothstein, J.D., Borchelt, D.R., Price,
Writing: A.H.L., A.M., S.S., S.W., and C.N.S. D.L., and Cleveland, D.W. (1997). ALS-linked SOD1 mutant G85R
mediates damage to astrocytes and promotes rapidly progressive
ACKNOWLEDGMENTS disease with SOD1-containing inclusions. Neuron 18, 327–338.

This work was supported in part by grants from the California Capowski, E.E., Schneider, B.L., Ebert, A.D., Seehus, C.R., Szulc, J.,
Institute for Regenerative Medicine (CIRM) DR2A-05320 to CNS Zufferey, R., Aebischer, P., and Svendsen, C.N. (2007). Lentiviral
and CIRM-EDUC2-12638 to SR and the Board of Governors vector-mediated genetic modification of human neural progenitor
Regenerative Medicine Institute. We thank the Cedars-Sinai Ge- cells for ex vivo gene therapy. J. Neurosci. Methods 163, 338–349.
nomics Core for assistance with snRNA-seq, the Cedars-Sinai Cy- Chambers, S.M., Fasano, C.A., Papapetrou, E.P., Tomishima, M., Sa-
togenetics Core for karyotype analysis, and Jovita Dimas-Harms delain, M., and Studer, L. (2009). Highly efficient neural conver-
and Jake Inzalaco for immunohistochemistry assistance. Sche- sion of human ES and iPS cells by dual inhibition of SMAD
matics were created using biorender.com signaling. Nat. Biotechnol. 27, 275–280.
Das, M.M., Avalos, P., Suezaki, P., Godoy, M., Garcia, L., Chang,
CONFLICT OF INTERESTS C.D., Vit, J.-P., Shelley, B., Gowing, G., and Svendsen, C.N.
(2016). Human neural progenitors differentiate into astrocytes
The authors declare no competing financial interests. Intellectual
and protect motor neurons in aging rats. Exp. Neurol. 280, 41–49.
protection-patent ‘‘Cortical Neural progenitor cells from iPSCs’’
No. 62/924,523 was filed October 22, 2019. Draper, J.S., Smith, K., Gokhale, P., Moore, H.D., Maltby, E., John-
son, J., Meisner, L., Zwaka, T.P., Thomson, J.A., and Andrews, P.W.
Received: November 7, 2022 (2004). Recurrent gain of chromosomes 17q and 12 in cultured hu-
Revised: March 24, 2023 man embryonic stem cells. Nat. Biotechnol. 22, 53–54.
Accepted: March 27, 2023 Ebert, A.D., Shelley, B.C., Hurley, A.M., Onorati, M., Castiglioni, V.,
Published: April 20, 2023 Patitucci, T.N., Svendsen, S.P., Mattis, V.B., McGivern, J.V., Schwab,
A.J., et al. (2013). EZ spheres: a stable and expandable culture sys-
tem for the generation of pre-rosette multipotent stem cells from
REFERENCES human ESCs and iPSCs. Stem Cell Res. 10, 417–427.
Akhtar, A.A., Gowing, G., Kobritz, N., Savinoff, S.E., Garcia, L., Gamm, D.M., Wang, S., Lu, B., Girman, S., Holmes, T., Bischoff, N.,
Saxon, D., Cho, N., Kim, G., Tom, C.M., Park, H., et al. (2018). Shearer, R.L., Sauvé, Y., Capowski, E., Svendsen, C.N., and Lund,
Inducible expression of GDNF in transplanted iPSC-derived neural R.D. (2007). Protection of visual functions by human neural pro-
progenitor cells. Stem Cell Rep. 10, 1696–1704. genitors in a rat model of retinal disease. PLoS One 2, e338.

1640 Stem Cell Reports j Vol. 18 j 1629–1642 j August 8, 2023

Garcı́a-Caballero, C., Lieppman, B., Arranz-Romera, A., Molina- Lepore, A.C., Rauck, B., Dejea, C., Pardo, A.C., Rao, M.S., Rothstein,
Martı́nez, I.T., Bravo-Osuna, I., Young, M., Baranov, P., and Her- J.D., and Maragakis, N.J. (2008). Focal transplantation-based astro-
rero-Vanrell, R. (2018). Photoreceptor preservation induced by in- cyte replacement is neuroprotective in a model of motor neuron
travitreal controlled delivery of GDNF and GDNF/melatonin in disease. Nat. Neurosci. 11, 1294–1301.
rhodopsin knockout mice. Mol. Vis. 24, 733–745. Lepore, A.C., O’Donnell, J., Kim, A.S., Williams, T., Tuteja, A., Rao,
Georgievska, B., Kirik, D., and Björklund, A. (2002). Aberrant M.S., Kelley, L.L., Campanelli, J.T., and Maragakis, N.J. (2011). Hu-
sprouting and downregulation of tyrosine hydroxylase in lesioned man glial-restricted progenitor transplantation into cervical spinal
nigrostriatal dopamine neurons induced by long-lasting overex- cord of the SOD1 mouse model of ALS. PLoS One 6, e25968.
pression of glial cell line derived neurotrophic factor in the stria- Lin, L.F., Doherty, D.H., Lile, J.D., Bektesh, S., and Collins, F.
tum by lentiviral gene transfer. Exp. Neurol. 177, 461–474. (1993). GDNF: a glial cell line-derived neurotrophic factor for
Goldberg, N.R.S., Marsh, S.E., Ochaba, J., Shelley, B.C., Davtyan, midbrain dopaminergic neurons. Science 260, 1130–1132.
H., Thompson, L.M., Steffan, J.S., Svendsen, C.N., and Blurton-
Liu, Z., Tang, Y., Lü, S., Zhou, J., Du, Z., Duan, C., Li, Z., and Wang,
Jones, M. (2017). Human neural progenitor transplantation res-
C. (2013). The tumourigenicity of iPS cells and their differentiated
cues behavior and reduces a-synuclein in a transgenic model of de-
derivates. J. Cell Mol. Med. 17, 782–791.
mentia with lewy bodies. Stem Cells Transl. Med. 6, 1477–1490.
McBride, J.L., Behrstock, S.P., Chen, E.-Y., Jakel, R.J., Siegel, I.,
Gowing, G., Shelley, B., Staggenborg, K., Hurley, A., Avalos, P., Vic-
Svendsen, C.N., and Kordower, J.H. (2004). Human neural stem
toroff, J., Latter, J., Garcia, L., and Svendsen, C.N. (2014). Glial cell
cell transplants improve motor function in a rat model of Hunting-
line-derived neurotrophic factor-secreting human neural progeni-
ton’s disease. J. Comp. Neurol. 475, 211–219.
tors show long-term survival, maturation into astrocytes, and no
tumor formation following transplantation into the spinal cord McGill, T.J., Lund, R.D., Douglas, R.M., Wang, S., Lu, B., Silver,
of immunocompromised rats. Neuroreport 25, 367–372. B.D., Secretan, M.R., Arthur, J.N., and Prusky, G.T. (2007). Synge-
neic Schwann cell transplantation preserves vision in RCS rat
Harada, C., Harada, T., Quah, H.M.A., Maekawa, F., Yoshida, K.,
without immunosuppression. Invest. Ophthalmol. Vis. Sci. 48,
Ohno, S., Wada, K., Parada, L.F., and Tanaka, K. (2003). Potential
1906–1912.
role of glial cell line-derived neurotrophic factor receptors in
Müller glial cells during light-induced retinal degeneration. Neuro- Mendez, I., Sanchez-Pernaute, R., Cooper, O., Viñuela, A., Ferrari,
science 122, 229–235. D., Björklund, L., Dagher, A., and Isacson, O. (2005). Cell type anal-
ysis of functional fetal dopamine cell suspension transplants in the
Hauck, S.M., Kinkl, N., Deeg, C.A., Swiatek-de Lange, M., Schöff-
striatum and substantia nigra of patients with Parkinson’s disease.
mann, S., and Ueffing, M. (2006). GDNF family ligands trigger in-
Brain 128, 1498–1510.
direct neuroprotective signaling in retinal glial cells. Mol. Cell Biol.
26, 2746–2757. Nutt, J.G., Burchiel, K.J., Comella, C.L., Jankovic, J., Lang, A.E.,
Laws, E.R., Lozano, A.M., Penn, R.D., Simpson, R.K., Stacy, M.,
Henderson, C.E., Phillips, H.S., Pollock, R.A., Davies, A.M., Le-
et al. (2003). Randomized, double-blind trial of glial cell line-
meulle, C., Armanini, M., Simmons, L., Moffet, B., Vandlen, R.A.,
derived neurotrophic factor (GDNF) in PD. Neurology 60, 69–73.
Simpson LC corrected to Simmons L., et al. (1994). GDNF: a potent
survival factor for motoneurons present in peripheral nerve and Okano, H. (2022a). A combined stem-cell-gene therapy strategy for
muscle. Science 266, 1062–1064. ALS. Nat. Med. 28, 1751–1752.
Hovland, D.N., Boyd, R.B., Butt, M.T., Engelhardt, J.A., Moxness, Okano, H. (2022b). Transplantation of neural progenitor cells into
M.S., Ma, M.H., Emery, M.G., Ernst, N.B., Reed, R.P., Zeller, J.R., the human CNS. Trends Mol. Med. 28, 897–899.
et al. (2007). Six-month continuous intraputamenal infusion Ostenfeld, T., Caldwell, M.A., Prowse, K.R., Linskens, M.H., Jau-
toxicity study of recombinant methionyl human glial cell line- niaux, E., and Svendsen, C.N. (2000). Human neural precursor cells
derived neurotrophic factor (r-metHuGDNF) in rhesus monkeys. express low levels of telomerase in vitro and show diminishing cell
Toxicol. Pathol. 35, 1013–1029. proliferation with extensive axonal outgrowth following trans-
Hulisz, D. (2018). Amyotrophic lateral sclerosis: disease state over- plantation. Exp. Neurol. 164, 215–226.
view. Am. J. Manag. Care 24, S320–S326. Rosati, J., Ferrari, D., Altieri, F., Tardivo, S., Ricciolini, C., Fusilli, C.,
Jomary, C., Darrow, R.M., Wong, P., Organisciak, D.T., and Jones, Zalfa, C., Profico, D.C., Pinos, F., Bernardini, L., et al. (2018). Estab-
S.E. (2004). Expression of neurturin, glial cell line-derived neuro- lishment of stable iPS-derived human neural stem cell lines suit-
trophic factor, and their receptor components in light-induced able for cell therapies. Cell Death Dis. 9, 937.
retinal degeneration. Invest. Ophthalmol. Vis. Sci. 45, 1240–1246. Sandrock, R.W., Wheatley, W., Levinthal, C., Lawson, J., Hashi-
Jones, M.K., Lu, B., Saghizadeh, M., and Wang, S. (2016). Gene moto, B., Rao, M., and Campanelli, J.T. (2010). Isolation, character-
expression changes in the retina following subretinal injection of ization and preclinical development of human glial-restricted pro-
human neural progenitor cells into a rodent model for retinal genitor cells for treatment of neurological disorders. Regen. Med. 5,
degeneration. Mol. Vis. 22, 472–490. 381–394.
Klein, S.M., Behrstock, S., McHugh, J., Hoffmann, K., Wallace, K., Sareen, D., McMillan, E., Ebert, A.D., Shelley, B.C., Johnson, J.A.,
Suzuki, M., Aebischer, P., and Svendsen, C.N. (2005). GDNF deliv- Meisner, L.F., and Svendsen, C.N. (2009). Chromosome 7 and 19
ery using human neural progenitor cells in a rat model of ALS. trisomy in cultured human neural progenitor cells. PLoS One 4,
Hum. Gene Ther. 16, 509–521. e7630.

Stem Cell Reports j Vol. 18 j 1629–1642 j August 8, 2023 1641

Sareen, D., Gowing, G., Sahabian, A., Staggenborg, K., Paradis, R., in the SOD1G93A ALS rat but has adverse side effects. Gene Ther.
Avalos, P., Latter, J., Ornelas, L., Garcia, L., and Svendsen, C.N. 24, 245–252.
(2014). Human induced pluripotent stem cells are a novel source Thomsen, G.M., Avalos, P., Ma, A.A., Alkaslasi, M., Cho, N., Wyss,
of neural progenitor cells (iNPCs) that migrate and integrate in L., Vit, J.-P., Godoy, M., Suezaki, P., Shelest, O., et al. (2018). Trans-
the rodent spinal cord. J. Comp. Neurol. 522, 2707–2728. plantation of neural progenitor cells expressing glial cell line-
Shelley, B.C., Gowing, G., and Svendsen, C.N. (2014). A cGMP- derived neurotrophic factor into the motor cortex as a strategy to
applicable expansion method for aggregates of human neural treat amyotrophic lateral sclerosis. Stem Cell. 36, 1122–1131.
stem and progenitor cells derived from pluripotent stem cells or
Trupp, M., Arenas, E., Fainzilber, M., Nilsson, A.S., Sieber, B.A., Gri-
fetal brain tissue. J. Vis. Exp., 51219.
goriou, M., Kilkenny, C., Salazar-Grueso, E., Pachnis, V., and Aru-
Suzuki, M., McHugh, J., Tork, C., Shelley, B., Klein, S.M., Aebischer, mäe, U. (1996). Functional receptor for GDNF encoded by the
P., and Svendsen, C.N. (2007). GDNF secreting human neural pro- c-ret proto-oncogene. Nature 381, 785–789.
genitor cells protect dying motor neurons, but not their projection
Tsai, Y., Lu, B., Bakondi, B., Girman, S., Sahabian, A., Sareen, D.,
to muscle, in a rat model of familial ALS. PLoS One 2, e689.
Svendsen, C.N., and Wang, S. (2015). Human iPSC-derived neural
Svendsen, C.N. (2013). Back to the future: how human induced progenitors preserve vision in an AMD-like model. Stem Cell. 33,
pluripotent stem cells will transform regenerative medicine. 2537–2549.
Hum. Mol. Genet. 22, R32–R38.
Wang, S., Girman, S., Lu, B., Bischoff, N., Holmes, T., Shearer, R.,
Svendsen, C.N., Caldwell, M.A., Shen, J., ter Borg, M.G., Rosser,
Wright, L.S., Svendsen, C.N., Gamm, D.M., and Lund, R.D.
A.E., Tyers, P., Karmiol, S., and Dunnett, S.B. (1997). Long-term sur-
(2008). Long-term vision rescue by human neural progenitors in
vival of human central nervous system progenitor cells trans-
a rat model of photoreceptor degeneration. Invest. Ophthalmol.
planted into a rat model of Parkinson’s disease. Exp. Neurol. 148,
Vis. Sci. 49, 3201–3206.
135–146.
Wright, L.S., Prowse, K.R., Wallace, K., Linskens, M.H.K., and
Svendsen, C.N., ter Borg, M.G., Armstrong, R.J., Rosser, A.E., Chan-
Svendsen, C.N. (2006). Human progenitor cells isolated from the
dran, S., Ostenfeld, T., and Caldwell, M.A. (1998). A new method
developing cortex undergo decreased neurogenesis and eventual
for the rapid and long term growth of human neural precursor
senescence following expansion in vitro. Exp. Cell Res. 312,
cells. J. Neurosci. Methods 85, 141–152.
2107–2120.
Thomsen, G.M., Gowing, G., Latter, J., Chen, M., Vit, J.-P., Staggen-
Yamanaka, S. (2020). Pluripotent stem cell-based cell therapy-
borg, K., Avalos, P., Alkaslasi, M., Ferraiuolo, L., Likhite, S., et al.
promise and challenges. Cell Stem Cell 27, 523–531.
(2014). Delayed disease onset and extended survival in the
SOD1G93A rat model of amyotrophic lateral sclerosis after sup- Zhang, S., and Cui, W. (2014). Sox2, a key factor in the regulation
pression of mutant SOD1 in the motor cortex. J. Neurosci. 34, of pluripotency and neural differentiation. World J. Stem Cells 6,
15587–15600. 305–311.
Thomsen, G.M., Alkaslasi, M., Vit, J.P., Lawless, G., Godoy, M., Zurn, A.D., Baetge, E.E., Hammang, J.P., Tan, S.A., and Aebischer, P.
Gowing, G., Shelest, O., and Svendsen, C.N. (2017). Systemic in- (1994). Glial cell line-derived neurotrophic factor (GDNF), a new
jection of AAV9-GDNF provides modest functional improvements neurotrophic factor for motoneurones. Neuroreport 6, 113–118.

1642 Stem Cell Reports j Vol. 18 j 1629–1642 j August 8, 2023