Developmental alterations in Huntington’s disease neural cells
and pharmacological rescue in cells and mice
The HD iPSC Consortium
*
Reprints and permissions information is available online at http://www.nature.com/reprints/index.html.
Correspondence should be addressed to L.M.T. (lmthomps@uci.edu).
*
The complete list of authors is as follows: Ryan G Lim
1
, Lisa L Salazar
2
, Daniel K Wilton
3
, Alvin R King
4
, Jennifer T Stocksdale
5
,
Delaram Sharifabad
4
, Alice L Lau
5
, Beth Stevens
3
, Jack C Reidling
5
, Sara T Winokur
4
, Malcolm S Casale
5
, Leslie M
Thompson
1,,2,,4–,6
, Mónica Pardo
7
, A Gerardo García Díaz-Barriga
7
, Marco Straccia
7
, Phil Sanders
7
, Jordi Alberch
7
, Josep M
Canals
7
, Julia A Kaye
8
, Mariah Dunlap
8
, Lisa Jo
8
, Hanna May
8
, Elliot Mount
8
, Cliff Anderson-Bergman
9
, Kelly Haston
8
, Steven
Finkbeiner
8,,10
, Amanda J Kedaigle
11
, Theresa A Gipson
12,,13
, Ferah Yildirim
11
, Christopher W Ng
12,,13
, Pamela Milani
11
,
David E Housman
12,,13
, Ernest Fraenkel
11
, Nicholas D Allen
14
, Paul J Kemp
14
, Ranjit Singh Atwal
15
, Marta Biagioli
15
, James F
Gusella
15
, Marcy E MacDonald
15
, Sergey S Akimov
16
, Nicolas Arbez
16
, Jacqueline Stewart
16
, Christopher A Ross
16,,17
, Virginia
B Mattis
18
, Colton M Tom
18
, Loren Ornelas
18
, Anais Sahabian
18
, Lindsay Lenaeus
18
, Berhan Mandefro
18
, Dhruv Sareen
18
&
Clive N Svendsen
18
1
Department of Biological Chemistry, University of California, Irvine, Irvine, California, USA.
2
Department of Psychiatry and Human Behavior, University of California, Irvine, Irvine, California, USA.
3
F.M. Kirby Neurobiology Center, Boston Children’s Hospital, Harvard Medical School, Boston, Massachusetts, USA.
4
Department of Neurobiology and Behavior, University of California, Irvine, Irvine, California, USA.
5
UCI MIND, University of California, Irvine, Irvine, California, USA.
6
Sue and Bill Gross Stem Cell Center, University of California, Irvine, Irvine, California, USA.
7
Department of Biomedical Sciences, Faculty of Medicine and Health Sciences, University of Barcelona, Barcelona Spain; August Pi
i Sunyer Biomedical Research Institute (IDIBAPS), Barcelona, Spain; and Networked Biomedical Research Centre for
NeuroDegenerative Disorders (CIBERNED), Barcelona, Spain.
8
Gladstone Institutes and the Taube/Koret Center of Neurodegenerative Disease Research, Roddenberry Stem Cell Research Program,
San Francisco, California, USA.
9
Sandia National Laboratories, Livermore, California, USA.
10
Departments of Neurology and Physiology, University of California, San Francisco, San Francisco, California, USA.
11
Department of Biological Engineering, Massachusetts Institute of Technology, Cambridge, Massachusetts, USA.
12
Center for Cancer Research, Massachusetts Institute of Technology, Cambridge, Massachusetts, USA.
13
Department of Biology, Massachusetts Institute of Technology, Cambridge, Massachusetts, USA.
14
School of Biomedical Sciences, Cardiff University, Cardiff, UK.
15
Center for Human Genetic Research, Massachusetts General Hospital, Boston, Massachusetts, USA.
16
Division of Neurobiology, Department of Psychiatry, Johns Hopkins University School of Medicine, Baltimore, Maryland, USA.
17
Departments of Neurology, Neuroscience and Pharmacology, Johns Hopkins University School of Medicine, Baltimore, Maryland,
USA.
18
The Board of Governors Regenerative Medicine Institute and Department of Biomedical Sciences, Cedars-Sinai Medical Center,
Los Angeles, California, USA.
AUTHOR CONTRIBUTIONS
Designed the experiments: R.G.L., L.L.S., D.K.W., J.C.R., S.T.W., L.M.T., J.A., J.M.C., N.D.A., P.J.K., J.A.K., S.F., F.Y., D.E.H.,
E.F., J.F.G., M.E.M., S.S.A., N.A., C.A.R., V.B.M. and C.N.S. Generated iPSC lines in study: L.O., A.S., L.L., B.M. and D. Sareen.
iPSC culture and neuronal differentiation: V.B.M., L.L.S., A.R.K., J.T.S., C.M.T., S.S.A., J.A.K., H.M. and M.D. Carried out
experiments: R.G.L. and T.A.G., RNA-seq; L.L.S., Isx-9 qPCR and NEUROD1 overexpression; A.L.L., M.P., A.G.G.D.-B., M.S. and
P.S., mouse neurodevelopment studies; A.G.G.D.-B., comparison between mouse and human data; V.B.M. and C.M.T., cell counts;
V.B.M., immunocytochemistry; F.Y., R.S.A. and M.B., ChIP; S.S.A., Cell Titer-Glo cell survival assay; S.S.A., N.A. and L.L.S.,
NEUROD1 knockdown; N.A. and J.S., cell culture and transfection of mouse primary neurons and nuclear condensation assay; S.S.A.,
Western analysis; E.M., J.A.K., M.D. and H.M., Isx-9 neuron assays; D.K.W., Isx-9 synaptic assays; J.C.R. and D. Sharifabad, mouse
Isx-9 studies. Analyzed the data: R.G.L., L.L.S., D.K.W., B.S., J.C.R., M.S.C., S.T.W., L.M.T., J.A.K., M.D., H.M., L.J., D.K.W.,
C.A.-B., S.F., A.J.K., T.A.G., F.Y., C.W.N., P.M., D.E.H., E.F., J.F.G., M.E.M., S.S.A., N.A., C.A.R., V.B.M. and C.N.S. Wrote the
manuscript: R.G.L., L.L.S., D.K.W., J.C.R., S.T.W., L.M.T., J.M.C., N.D.A., P.J.K., J.A.K., K.H., S.F., A.J.K., T.A.G., P.M., D.E.H.,
E.F., M.E.M., J.F.G., S.S.A., C.A.R., V.B.M. and C.N.S. A list of authors by individual consortium group appears in the
Supplementary Note.
COMPETING FINANCIAL INTERESTS
The authors declare no competing financial interests.
Note: Any Supplementary Information and Source Data files are available in the online version of the paper.
HHS Public Access
Author manuscript
Nat Neurosci
. Author manuscript; available in PMC 2017 September 22.
Published in final edited form as:
Nat Neurosci
. 2017 May ; 20(5): 648–660. doi:10.1038/nn.4532.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
Abstract
Neural cultures derived from Huntington’s disease (HD) patient-derived induced pluripotent stem
cells were used for ‘omics’ analyses to identify mechanisms underlying neurodegeneration. RNA-
seq analysis identified genes in glutamate and GABA signaling, axonal guidance and calcium
influx whose expression was decreased in HD cultures. One-third of gene changes were in
pathways regulating neuronal development and maturation. When mapped to stages of mouse
striatal development, the profiles aligned with earlier embryonic stages of neuronal differentiation.
We observed a strong correlation between HD-related histone marks, gene expression and unique
peak profiles associated with dysregulated genes, suggesting a coordinated epigenetic program.
Treatment with isoxazole-9, which targets key dysregulated pathways, led to amelioration of
expanded polyglutamine repeat-associated phenotypes in neural cells and of cognitive impairment
and synaptic pathology in HD model R6/2 mice. These data suggest that mutant huntingtin impairs
neurodevelopmental pathways that could disrupt synaptic homeostasis and increase vulnerability
to the pathologic consequence of expanded polyglutamine repeats over time.
HD is characterized by motor abnormalities, psychiatric symptoms and cognitive deficits. It
is caused by a CAG repeat expansion in the huntingtin gene (
HTT
) encoding a
polyglutamine tract expansion in the huntingtin protein (HTT)
1
. CAG repeats of 40 or more
cause symptoms and those above ~60 cause juvenile-onset HD. Mutant HTT (mHTT) is
implicated in a spectrum of cellular aberrations
2
. Neuropathology includes cortical atrophy
and loss of striatal medium spiny neurons
1
. Age of clinical onset is primarily predicted by
the extent of CAG repeat expansion and considered to result from cumulative toxic insults,
together with environmental and other genetic factors
3
.
There may be deficiencies of neurodevelopment and differentiation of neural stem and
progenitor cells in the adult striatum throughout life
4,5
. Adult neurogenesis appears to be
impaired in the striata of HD patients, with increased cell proliferation
6
and an absence of
adult-born neurons
7
, suggesting that neurogenesis may be initiated but maturation is
impaired. Neuroimaging scans of premanifest HD-affected brains detect changes in striatal,
cortical and whole-brain volume before symptoms
8–10
. Head circumference is less in
premanifest HD subjects than unaffected subjects
11
.
Patient-derived induced pluripotent stem cells (iPSCs), somatic cells reprogrammed to a
pluripotent state, provide an important resource for deciphering mechanisms underlying
neurological disease (for example, ref. 12) and aid in the design of new therapeutics. iPSCs
can be differentiated into multipotent neural stem progenitor cells that produce mature
neural subpopulations (for example, ref. 13). We previously reported that differentiated HD-
derived iPSCs display expanded CAG-associated phenotypes
14
.
We used neural cells differentiated from HD patient-derived iPSC lines with juvenile-onset
CAG repeat expansions (60 and 109 repeats) to explore the effects of pathological
HTT
alleles by using parallel cultures for unbiased omics analyses. This discovery-based
approach revealed consistent deficits related to neurodevelopment and adult neurogenesis,
suggesting that specific gene networks represent potential therapeutic interventions. We
tested a small molecule, isoxazole-9 (Isx-9), that targets several of the dysregulated gene
Page 2
Nat Neurosci
. Author manuscript; available in PMC 2017 September 22.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
networks. Isx-9 normalized CAG repeat-associated phenotypes in both juvenile-repeat and
adult-onset repeat HD iPSC-derived neural cultures, as well as cognition and synaptic
pathology in R6/2 mice, demonstrating that these deficits might be reversed and synaptic
homeostasis improved.
RESULTS
Differentiation of iPSCs with expanded HD and non-disease CAG repeats
iPSC lines from non-disease and HD fibroblasts with CAG repeats from 21 to 33 (non-
disease) and 60 and 109 (juvenile-onset HD range) were reprogrammed using non-
integrating episomal factors (Supplementary Table 1)
15
and pluripotency and normal
karyotypes confirmed. iPSC lines were differentiated for 56 d into mixed neural cultures
containing neurons, glia and neural progenitors as described
14
. Mixed cultures broadly
represent cell types and progenitors in human brain, as opposed to pure cultures of a single
cell type. Lines similarly expressed germ-lineage and cell type–specific markers
(Supplementary Figs. 1 and 2). Cell quantification in HD and non-disease lines was
performed (Supplementary Fig. 1a–e), with total counted cells expressing glial fibrillary
acidic protein (GFAP, 16.1%), neuronal markers TUJ1 (26.5%) and MAP2ab (14.8%), and
striatal marker DARPP-32 (14.3%) (Supplementary Fig. 1b–e), as an average among lines.
Oligodendrocyte (Supplementary Fig. 1f), endoderm (SOX17, FOXA2), mesoderm
(MYO1), and microglia (IBA1) (Supplementary Fig. 2a, b) markers were absent. iPSC lines
had a similar overall composition from staining data, including nestin (Supplementary Fig.
1a). At earlier times, nestin-expressing neural progenitor cells may persist longer in the
expanded 109Q repeat line, and nestin-positive cells were more susceptible to cell stress
after acute brain-derived neurotrophic factor (BDNF) withdrawal
15
, suggesting early and
subtle effects exerted by expanded CAG repeats.
RNA-seq analysis of differentiated iPSCs
We used unbiased whole-genome and multi-platform approaches in parallel iPSC cultures.
Whole-transcriptome analysis (RNA-seq) was performed, and principal component analysis
showed separation of HD 109Q and 60Q lines (two clones each) from non-disease 33Q, 28Q
and 21Q lines, indicating minimal variability within groups and a maximal variance
explained by disease and non-disease differences (Fig. 1a). Statistical analysis using the
Bioconductor package DESeq (RNA-seq differential expression) revealed 1,869
differentially expressed genes (DEGs) between HD and non-disease lines (Supplementary
Table 2). Hierarchical clustering (Fig. 1b) of log
2
-transformed gene expression values
showed that samples clustered by repeat length, with distinct separation and expression
patterning among the samples. Independent clonal lines derived from individual subjects
clustered tightly together, with low variability within groups (Fig. 1b).
RNA-seq analysis suggests altered neurodevelopment in HD lines
Ingenuity pathway analysis (IPA; Supplementary Fig. 2c) was used to investigate biological
changes by examining genes from DESeq analysis. Of the 1,869 DEGs, 543 (29%)
(Supplementary Table 2) centered on development. The top three were
cellular development
,
Page 3
Nat Neurosci
. Author manuscript; available in PMC 2017 September 22.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
nervous system development and function
and
tissue development
(Fig. 2a). Over half of the
DEGs (59%) were associated with nervous system development and function.
Some of these neural developmental genes and their physical and regulatory interactions are
depicted in Figure 2b (Supplementary Table 3), with
NEUROD1
as a highly enriched hub.
HTT interacts physically with the products of two genes that are among the most
downregulated in the RNA-seq data set:
NEUROD1
and
GAD1
(refs. 16,17). Expression of
the proneuronal basic helix-loop-helix gene
NEUROD1
was decreased across all HD lines
(Supplementary Table 2).
NEUROD1
regulates neurodevelopment
18
and adult
neurogenesis
19
. Of genes in this network (Fig. 2b and Supplementary Table 3), several
encode proteins that regulate
NEUROD1
expression
(POU4F
,
NEUROG2
,
ASCL
and
REST
) and are dysregulated or predicted to have aberrant activities at the protein level. RNA
expression levels for several, including
NEUROD1
, were validated by quantitative PCR
(qPCR) (Supplementary Fig. 3a–f).
IPA also predicted the activation or inhibition of upstream regulators (Supplementary Table
4), including several implicated in HD, such as increased REST activation and decreased
BDNF signaling
20
. REST-mediated epigenetic remodeling regulates the developmental
switch of synaptic NMDA receptors from GluN2B to GluN2A subunits, thus decreasing
GRIN2B levels
21
;
GRIN2B
was downregulated in the RNA-seq data set. Several genes
essential for neurogenesis, including
NEUROD1
,
NEUROG2
and
ASCL1
and the
microRNA miR-124, are predicted to have inhibited activity. Genes encoding proteins
involved in the transforming growth factor (TGF)-β pathway (
TGFB2
,
TGFB3
,
TGFB3R
)
were upregulated in the differentiated HD iPSCs, with predicted activation of upstream
regulators TGFβ1 and TGFβ1R. TGFβ signaling is implicated as a ‘switching factor’ whose
levels and temporal activity require tight regulation during development to determine
neuronal cell fate
22
. This signaling network was disrupted in neural stem cells from HD
patient iPSCs
23
. Highlighted in this network were genes overlapping the IPA-predicted
regulators and genes that are major hubs with direct connections to other genes in the larger
network, suggesting an association with HD.
Cellular pathways related to neuronal function are altered in HD iPSC-neural cells
IPA also identified dysregulated pathways relevant to neuronal development and maturation,
including axonal guidance, Wnt signaling, Ca
2+
signaling, neuronal CREB signaling, and
glutamate and GABA receptor signaling (Fig. 3a), that are altered in HD models or human
HD
2
. Axonal guidance pathways regulate connectivity, integrating actions of guidance cues
and receptors. Four families of molecules and receptors provide axon guidance cues
24
,
including netrins, slits, ephrins and semaphorins, and genes within each family are
dysregulated in the RNA-seq data set (Supplementary Table 2), with most downregulated.
Intracellular calcium signaling promotes axonal and dendritic outgrowth, connecting
alterations in axonal guidance and calcium-signaling pathways, and is critical for neuronal
synaptic activity–regulated transcription and maturation
25
. Widespread dysregulation of
Ca
2+
signaling pathways in the HD samples (Fig. 3b) included genes for members of the
NMDA- and AMPA-type glutamate receptors, nicotinic acetylcholine receptor subunits,
several subunits of the voltage-gated Ca
2+
channel CACNA1, and the plasma membrane
Page 4
Nat Neurosci
. Author manuscript; available in PMC 2017 September 22.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
Ca
2+
-ATPase, in addition to the calcium sensors and downstream effectors CAMKII, CALM
and CREB (Fig. 3b). Among downregulated DEGs encoding proteins in the glutamate and
GABA signaling pathways are glutamate decarboxylases and
GAD1
and
GAD2
, which
encode GABAergic neuron markers. Several other DEGs in the HD iPSC-neural cultures are
involved with GABA synthesis, release, reuptake or degradation. As glutamate is transported
into neurons via SLC1A3 and SLC1A6, downregulating the corresponding genes would be
predicted to yield reduced glutamate substrate for GABA synthesis. This is significant:
GABAergic neurons are the most vulnerable to mHTT toxicity
1
. The dysregulated genes
display integrated features and potential effects on nervous system development and
function, including clustering of DEGs encoding proteins involved in CREB, Wnt, axonal
guidance signaling and synaptic function (Supplementary Fig. 3g, h).
Comparing HD iPSC lines to mouse striatal gene expression indicates altered maturation
We next compared the 1,869 DEGs from the RNA-seq analysis to orthologs (1,647 mouse
genes) involved in mouse striatal development derived from microarray mouse
developmental data using identifiers from Xspecies, a cross-species ortholog identifier and
retrieval tool. This comparison yielded 679 overlapping genes. Hierarchical clustering
showed that the HD60Q and HD109Q RNA-seq samples clustered with the proliferative
germinal zone samples whereas non-disease samples clustered with the postmitotic mantle
zone samples (Fig. 4a). Thus, differentiated HD lines may mature slower than non-disease
lines or improperly shut off gene expression in earlier stages of development.
Enriched GO terms for the set of 679 genes (Fig. 4b) included
synaptic transmission
(GO:
0007268),
nervous system development
(GO:0007399),
CNS development
(GO:0007417),
and a series of more specific processes, such as
axonal guidance
(GO:0007411),
cell
adhesion
and
locomotor behavior
(GO:0007626); these were similar to the categories
enriched in the RNA-seq analysis. Using the hierarchical cluster algorithm with average
linkage for gene expression values, we obtained 13 clusters from the 679 genes
(Supplementary Fig. 4). Activation
z
-scores predicted that most of the processes have
decreased or inhibited function in the HD lines (Supplementary Fig. 5a). Analysis with
Metacore to visualize functional connections between genes in the sets yielded 296 nodes
and also pointed to developmentally described transcription factors (DLX2, c-MYC,
DACH1, MEF2A, hASH1) and relevant pathway regulators (TGFβ receptor) (Fig. 4c). Gene
cluster I (Supplementary Fig. 5b) contains
NEUROD1
and related genes that are expressed
more highly in non-disease repeat samples.
Finally, we compared HD DEGs to genes that change in expression during human striatal
maturation
26
. Genes implicated in development of human striatum are regulators of genes
differentially expressed in our HD neural cells (Supplementary Fig. 5c), indicating a core
network of genes that both contribute to the maturation of the human striatum and are
altered in HD iPSC-derived neural cells.
Page 5
Nat Neurosci
. Author manuscript; available in PMC 2017 September 22.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript