Human iPSC‐derived neural precursor cells differentiate into multiple cell types to delay disease progression following transplantation into YAC128 Huntington's disease mouse model

Abstract Objectives To investigate whether human HLA‐homozygous induced pluripotent stem cell (iPSC)‐derived neural precursor cells (iPSC‐NPCs) can provide functional benefits in Huntington’s disease (HD), we transplanted them into the YAC128 transgenic HD mouse model. Materials and Methods CHAi001‐A, an HLA‐homozygous iPSC line (A*33:03‐B*44:03‐DRB1*13:02), was differentiated into neural precursor cells, and then, they were transplanted into 6 months‐old YAC128 mice. Various behavioural and histological analyses were performed for five months after transplantation. Results Motor and cognitive functions were significantly improved in transplanted animals. Cells transplanted in the striatum showed multipotential differentiation. Five months after transplantation, the donor cells had differentiated into neurons, oligodendrocytes and astrocytes. Transplantation restored DARPP‐32 expression, synaptophysin density, myelin basic protein expression in the corpus callosum and astrocyte function. Conclusion Altogether, these results strongly suggest that iPSC‐NPCs transplantation induces neuroprotection and functional recovery in a mouse model of HD and should be taken forward for clinical trials in HD patients.

with antisense oligonucleotides) showed promising results. 3,4 Conversely, human stem cell-based neuro-restorative or neurodegenerative strategies can offer an alternative therapeutic strategy.
To date, human embryonic stem cell-derived neural progenitor cells 5 or mouse induced pluripotent stem cells (iPSC)-derived neural stem cells 6 have shown beneficial effects on HD transgenic mice. These studies primarily focused on the replacement of neurons such as medium spiny neurons (MSNs).
iPSC-derived neural progenitor cells (iPSC-NPCs) self-renew and can differentiate into neurons as well as glia. 7 After spinal cord injury in mice, transplanted human iPSC-NPCs differentiated into the three major neural lineages (ie, neurons, astrocytes and oligodendrocytes), leading to functional recovery.
We previously reported that glial progenitor cell transplantation provided some benefits in HD mice. 8 Engrafting normal human glial progenitor cells into the striatum of newborn R6/2 HD mice substantially replaced the diseased striatal glia with both human glial progenitor cells and their derived astrocytes, slowing the disease progression and increasing the survival of R6/2 mice.
In this study, we investigated the therapeutic efficacy of healthy cord blood-derived human leukocyte antigen (HLA)-homozygous iPSC-NPCs on a YAC128 mouse HD model that exhibits striatal and cortical atrophy, as well as progressive deterioration of motor and cognitive functions. 9, 10 We found that five months after transplantation, the grafted cells predominantly differentiated into DARPP-32-, O4-or GFAP-positive cells, which had neuroprotective effects.
Surprisingly, most of the therapeutic benefits came from astrocytes.
Transplanted animals had higher striatum DARPP-32 expression and improved motor and cognitive functions. Our results suggest that iPSC-NPCs mainly differentiate into astrocytes, which delays neurodegeneration. Furthermore, given the benefits of using HLAhomozygous iPSCs to treat HLA-matched HD patients, our data strongly suggest that HLA-homozygous iPSC-derived NPCs have strong clinical merit and warrant future clinical trials.

| Lead Contact and materials availability
Further information and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact, Jihwan Song (jsong5873@gmail.com). This study did not generate new unique reagents.

| Cell line
We used the HLA-homozygous iPSC line, CHAi001-A, because it carries the most frequent HLA-homozygous haplotypes (A*33:03-B*44:03-DRB1*13:02). It was established from the frozen cord blood of a healthy donor using the episomal method. 11 These haplotypes cover about 9% of the Korean population and are also relevant to other Asian populations (Japanese, Chinese, etc). This iPSC line can also treat diverse Asian populations in the United States, etc This study was previously approved by the Institutional Review Board of CHA University (1044308-201511-BR-025-01).

| iPSC culture and differentiation into NPCs
CHAi001-A cells were maintained in StemFit Basic02 medium (Ajinomoto, Japan) supplemented with 100 ng/ml basic fibroblast growth factor (bFGF, Peprotech) and 10 umol/L Y27632 (ROCK inhibitor, Peprotech) for about seven days before they were treated with a TrypLE solution (GIBCO) for 5 minutes at 37°C in a CO 2 incubator.
The dissociated cells were cultured in the neural differentiation me-

| Flow cytometry analysis
When iPSCs reached ~80% confluency, we conducted fluorescenceactivated cell sorting (FACS). The cultures were dissociated into a single-cell suspension using 0.5× TrypLE solution comprised of TrypLE Select Enzymes (Thermo Fisher Scientific) and 0.5 mmol/L UltraPure EDTA (Thermo Fisher Scientific). Cells remained unfixed and were stained with PE-conjugated antigen-specific antibodies and corresponding isotypes using the manufacturer's recommended concentration. We used anti-SSEA4 (1:200) and anti-IgM isotype

| Quantification of differentiated cells in vitro
We measured the number of the differentiated cells NeuN, GFAPand O4-positive cells on plate using the IXMC (ImageXpress Micro Confocal) high-content imaging system (Molecular Devices). The differentiated cells on plate were delineated at 20X magnification (n = 3). The merged positive cells were counted by the scoring module of MetaXpress software (Molecular Devices).

| YAC128 transgenic mice and transplantation
All the experiments were performed on YAC128 transgenic and wildtype (FVB/N) mice maintained on the FVB/N strain background. 12 Mice were bred in the animal facility of CHA Bio Complex (Pangyo, Korea).
The Institutional Animal Care and Use Committee (IACUC 200019) of CHA University approved all experiments. YAC128 mice exhibit glutamate toxicity from three months of age 13,14 and motor function defects from six months of age. 9 In this study, we used six-month-old YAC128 mice to investigate whether HLA-iPSC-NPCs transplantation improved motor function deficits. In the transplanted group (TP, n = 10), we stereotaxically injected 4 μl of HLA-iPSC-NPCs (100 000 cells/μl) into the striatum of both hemispheres using the following coordinates: AP + 0.5 mm, ML ± 1.8 mm, and DV −3 and −4 mm from the bregma.
In the sham group (n = 10), we injected 4 μl of suspension medium (DMEM) parallelly into both hemispheres. In both groups, we intraperitoneally injected cyclosporine A (5 mg/kg, Sigma) three days before the transplantation and daily for months (until sacrifice).

| Behavioural tests
We performed motor function tests (rotarod and grip strength tests) and cognitive function tests (simple swim and novel object recognition tests) for five months to evaluate the effect of cell transplantation in six-month-old YAC128 transgenic mice. To investigate the cell survival and the changes of pathology development after transplantation, two or three mice from each group were sacrificed for histological analysis at different time points of the behavioural analysis (Wild type (WT): n = 8, Sham: n = 7, and TP: n = 8).

| Accelerating rotarod test
We used an accelerating rotarod protocol (San Diego Instruments) to test motor coordination and gait changes. Accelerations ranged from 0 to 45 rpm over a period of 2 minutes. After the training period (two trials per day for three days), mice were tested for three consecutive trials in a single day and allowed 1.5 hours of rest time between the trials. The rotarod was wiped clean with ethanol between each subject and trial.

| Grip strength test
To further assess motor function, we performed a grip strength test. 15 The apparatus (San Diego Instruments) comprised an adjustable grip (6 cm wide, 0-45) connected to a digital gauge. For this measure, the mice were lifted by the tail so that their forepaws grasped the grip. The mice were then gently pulled backward by the tail until they released the wire. The maximal force exerted before the mice lost their grip was recorded. Each mouse was tested nine times, and the average of the three highest scores was used for subsequent analyses.

| Simple swim test
We chose the simple swim test to measure cognitive impairments and procedural learning, as previously performed in HD animals. 9,16 Mice were placed in the centre of a linear swimming chamber (76 × 13 cm; water depth: 9 cm; platform: 6 × 13 cm) facing away from the escape platform. The amount of time required to reach the platform and the initial swimming direction were recorded for each trial. Swimming towards the platform was arbitrarily given a score of 0, whereas swimming away from the platform was given a score of 1. The mice were trained for three days in three pairs of two consecutive trials with an interval of 2 hours between each trial. For each mouse, the average of the three trials on the last day of testing was used for analysis. 16

| Novel object recognition test
The novel object recognition test measures rodent recognition memory. 17 The experimental apparatus consisted of a white rectangular open field (45 cm × 45 cm × 45 cm). Before training, mice explored the open field for 10 minutes and then rested for 5 minutes in the cage. Two identical objects were placed at 7 cm from both upper corners of the chamber, and the mice were placed in the lower-left corner of the chamber. They were then allowed to explore the field for 7 minutes. The mouse was then removed from the chamber and placed in the cage for 5 minutes. To evaluate the mouse's reaction to a new object, the upper-right corner object was replaced with an unfamiliar new object and the experimental animal was again placed in the lower-left corner for 5 minutes. The test was performed three and five months after transplantation. Exploratory activity in the experimental arena was measured using EthoVision XT 11.5 (Noldus).

| Immunocytochemistry and Immunohistochemistry
Cells were fixed using a 4% paraformaldehyde solution for iPSC characterization or 42 days after neural differentiation. YAC128 mice used in this experiment were sacrificed five months after cell transplantation when all the behavioural analyses were completed.
All mice were perfused with 5 μl/ml heparin and then with a 4% paraformaldehyde solution. Mouse brains were fixed in the 4% paraformaldehyde solution overnight and left to soak in 30% sucrose solution for about three days. The brain tissues were then frozen using an optimal cutting temperature compound and sectioned into 30 µm thick slices using a cryotome (CM3050, Leica, Germany). The cryosectioned brain tissues were stored in a cryoprotective solution   magnification, and then, the density of images was quantitated using Image J software (rsb.info.nih.gov, by W. Rasband).

| Morphology analysis
We measured the number of processes, the maximal process length, the outgrowth intensity, and the body area of GFAP-or Iba-1positive cells in striatal regions from three or six sections from three brains in each experimental group. We obtained images using a Nikon microscope or ImageXpress Micro Confocal high-content imaging system (Molecular Devices) and connected to the stage and the computer with the distance information in the z-axis. Each region was analysed using the outgrowth module of MetaXpress software (Molecular Devices).

| Cell counting
The unbiased stereological estimation of the total number of human Nuclei (hNu)-, DARPP-32-and Iba-1-positive cells, and the merged GFAP-and hNu-double-positive cells in the striatum was performed using the ImageXpress Micro Confocal high-content imaging system (Molecular Devices). The sections used for counting covered the entire striatum (2 mm 2 ). This generally yielded sections in a series. The positive cells were counted in all regions of the six to seven sections.
Sampling was performed using the ImageXpress Micro Confocal highcontent imaging system (Molecular Devices), which was connected to the stage and fed the computer with the distance information in the z-axis. The striatum was delineated at 20 × magnification. We

| Characterization and differentiation of HLAhomozygous iPSC
We assessed the effect of iPSC-NPCs transplantation on an HD animal model. We previously established ten most frequent HLAhomozygous iPSCs in the Korean population. They can be transplanted in 41.07% of the Korean population with minimal immune suppression. 11,19 First, we characterized iPSCs for transplantation.
Next, we applied the embryoid body-based neural differentiation method to differentiate these iPSCs into NPCs ( Figure 1E).

| HLA-iPSC-NPCs improved motor and cognitive function
To assess the efficacy of HLA-iPSC-NPCs against HD, we transplanted them into YAC128 transgenic mice. Figure 2A shows and GFAP + astrocytes (66.03%). Based on this analysis, it appears that astrocytes would play an important role in ameliorating the toxic environment in the HD brains, which were indeed supported by histological analyses (Figure 6). The remaining percentage of different cell types (~4.73%) has yet to be identified.
We examined whether the transplanted HLA-iPSC-NPCs rescued motor function. We performed the rotarod test and the grip strength test every month for five months after transplantation.
The TP group showed significantly better performance than the sham group in both the rotarod test ( Figure 2B, P < .001; see also Video S1) and the grip strength test ( Figure 2C, P < .01; see also Video S2). Interestingly, we observed a decline pattern after three months in the case of rotarod test, but further analysis revealed that no statistical differences were observed after three months (ie, 3 months vs. 4 months; 4 months vs. 5 month; and 3 months vs. 5 months) ( Figure S2).
We also assessed cognitive function by performing the simple swim test and the novel object recognition test. After transplantation, we performed the simple swim test every month for five months. Mice from the sham group reached the visible escape platform twice slower than WT mice. They also had aberrant swimming direction patterns, reminiscent of motor coordination and navigational memory impairments ( Figure 2D, Sham and TP: P < .05; see also Video S3). Interestingly, three months after transplantation, the motor coordination and navigational memory of the TG group continually improved compared to the sham group. Besides, four months after transplantation, the TP group reached the platform significantly faster than it previously did ( Figure 2D, P < .05; see also Video S3). We also performed the novel object recognition test three months after transplantation to measure short-term memory in a non-stressful experimental setting. As the heat map shows, the TP group explored the novel object significantly longer than the sham group. ANOVA analysis further demonstrated that transplantation significantly increased the novel object exploration time ( Figure 2E, P < .05; see also Video S4). Altogether, these results strongly suggest that HLA-iPSC-NPCs transplantation improved motor and cognitive functions in YAC128 TG mice.

| Transplanted HLA-iPSC-NPCs differentiated into neurons in YAC128 TG mice
We

| Transplanted HLA-iPSC-NPCs differentiated into oligodendrocytes in YAC128 TG mice
Next, we examined whether HLA-iPSC-NPCs differentiated into oligodendrocytes in the striatum. Five months after transplantation, we triple-stained hNu-positive cells with antibodies against O4 and MBP ( Figure 4A). As Figure 4A shows, we found hNu-, O4 and MBPpositive cells near the corpus callosum (CC). Besides, the immunohistochemical analysis using an antibody against MBP showed that transplantation significantly increased the density and thickness of the CC ( Figure 4B,C). Altogether, these results strongly suggest that HLA-iPSC-NPCs transplantation preserved myelin.

| Transplanted iPSC-NPCs differentiated into astrocytes in YAC128 TG mice
Finally, we examined whether HLA-iPSC-NPCs differentiated into astrocytes in the striatum.

| Transplanted HLA-iPSC-NPC-derived astrocytes showed environmental changes in the YAC128 TG mice
To investigate the cellular defects of YAC128 TG astrocytes, we examined the astrocytic Kir4.1 channel current which significantly decreases in the HD brain and HD transgenic animal models. 22 Astrocytic Kir4.1 channel deficit appears early in the disease process, and the elevated extracellular K+ levels impair spatial K+ buffering, which increases glutamate toxicity in the brain. 23,24 HLA-iPSC-NPCs transplants can potentially modulate several key components of glutamate toxicity ( Figure 6A). First, glutamate is transported into astrocytes along with Na + in exchange for K + . In HD, striatal astrocytes have a Kir4.1 channel deficit, and thus, the glutamate transporter dysfunction causes functional defects. Second, astrocytes play a crucial role in providing neurons with glutamine, an important precursor for glutamate and GABA synthesis (known as the glutamate-GABA-glutamine cycle).
Here, we examined the astrocytes' abnormal function in YAC128 TG animals and how HLA-iPSC-NPCs rescued it. In TP mice, GFAPor hGFAP-positive cells were Kir4.1-immunoreactive ( Figure 6B). By contrast, GFAP-positive astrocytes five months after transplantation had a significantly lower Kir4.1 level ( Figure 6B). Western blot analysis further confirmed that HLA-iPSC-NPCs transplantation restored Kir4.1 expression ( Figure 6B). Moreover, TP mice showed higher expression levels of the glutamate transporter EAAT2 in GFAP-or hGFAP-positive cells ( Figure 6C). Eleven-month-old TP mice had high expression levels of glutamine synthetase, as confirmed by immunostaining and Western blot ( Figure 6D).
These results suggest that astrocytes from the transplanted HLA-iPSC-NPCs restored the YAC128 TG astrocytes by two complementary mechanisms: rescuing Kir4.1 expression (thus increasing glutamate uptake) and increasing glutamine synthetase expression (thus increasing glutamine production).
Astrocytes play anti-inflammatory roles by releasing various factors to the microglia. 25 Therefore, we investigated whether astrocytes differentiated from the transplanted HLA-iPSC-NPCs could influence microglia polarization, thereby reducing inflammation. We found immunohistochemical evidence that TP mice had significantly lower IL1β levels (a pro-inflammation marker) and significantly higher Arginase1 (Arg1, an anti-inflammatory marker) levels than sham mice ( Figure 6E,F). Western blot analysis confirmed this result ( Figure 6G).
Results from the additional pixel analysis ( Figure S4) to quantify the intensity of each fluorescence images shown in Figure 6B-F were consistent with Western blot analysis.
In HD, expression of brain-derived neurotrophic factor (BDNF) is significantly reduced. In immunohistochemistry, we observed that hGFAP-positive cells were co-localized with BDNF-positive cells in TP group ( Figure S5A). Besides, TP mice had higher expression levels of tropomyosin receptor kinase B (TrkB, a BDNF receptor) than sham mice ( Figure S5B). Altogether, these results strongly suggest that HLA-iPSC-NPCs transplantation changed the environmental neurotoxicity of the mouse or human astrocytes.

| D ISCUSS I ON
This study demonstrates the multipotential differentiation of trans-  transgenic mice. Human astrocytes are larger and have more processes than mouse astrocytes. [37][38][39] The HLA-iPSC-NPC-derived astrocytes were larger and had more processes than mouse astrocytes.
Besides, mouse astrocytes in TP mice shared more morphological features with astrocytes from WT mice than with those from sham mice. In TP mice, we observed the respective functions of mouse astrocytes or human astrocytes. The present study focused on two aspects of astrocytic reduction in the environmental toxicity: (1) glutamate buffering and (2) anti-inflammatory actions.

| Glutamate toxicity control
Astrocytes remove the spillover of synaptic neurotransmitters such as glutamate. In HD, the expression of the two glia-specific transporters which accomplish this, EAAT1 and EAAT2 38-40, is low. Moreover, there is a loss of the astrocytic Kir4.1 channel in HD, which impairs K+ buffering and further increases glutamate toxicity. [22][23][24] TP mice displayed no loss of the astrocytic Kir4.1 channel and an increased expression of EAAT2.

| Rescue of the glutamate-GABA-glutamine cycle
In the brain, astrocytes are the most abundant cell type and are tightly associated with synapses, since they synaptic transmission and provide metabolic support to neurons. In HD, astrocyte metabolism is impaired.
In particular, the glutamate-GABA-glutamine cycle dysfunction reduces GABA production. 40,41 Thus, normalizing the impaired astrocyte metabolism could be beneficial to HD patients. We found that HLA-iPSC-NPCs transplantation increased glutamine synthase expression by astrocytes, implying a recovery of impaired astrocyte metabolism.

| Anti-inflammatory action
Astrocytes can also act as secretory cells by producing neurotransmitters, neuromodulators, neurohormones, cytokines, neurotrophic factors, etc Released cytokines and neurotrophic factors play anti-inflammatory roles in various neurodegenerative diseases. 25 For example, increasing the Quantification of DARPP-32 expression using Image J analysis (n = 3 in each group, *P < .05, ***P < .001). (H) Immunostaining and Western blot analyses for synaptophysin-positive area in the striatum. Quantification of synaptophysin-positive area using Image J analysis (Scale bar: 50 μm, n = 3 in each group, *P < .05, **P < .01). Data were analysed by two-way ANOVA followed by a Tukey post hoc test using the SPSS software production of BDNF by astrocytes reduced inflammation in transgenic HD mice. 42 Also, astrocytes release anti-inflammatory factors such as TGFβ, and HD astrocytes have low TGFβ mRNA and protein levels. 43 TGFβ release by astrocytes can decrease inflammation. Therefore, astrocyte differentiation by transplanted HLA-iPSC-NPCs could restore anti-inflammatory action. In line with this, we found that HLA-iPSC-NPC-derived astrocytes expressed BDNF and TrkB. Altogether, the normal function of differentiated astrocytes can improve the inflammatory environment of the HD brain.
In conclusion, transplanted HLA-iPSC-NPCs differentiate into the three major neural lineages (neurons, oligodendrocytes and astrocytes) and can exert neuroprotective effects in the HD YAC128 transgenic mouse model.
Using HLA-homozygous cells, we aim to develop an immune-compatible Quantification of MBP density and thickness of CC (n = 3 in each group, *P < .05). Data were analysed by two-way ANOVA followed by a Tukey post hoc test using the SPSS software allogeneic transplantation approach. Here, we showed that HLA-iPSC-NPC transplantation helped to recover several impaired functions. This approach could suppress immune rejection in allogeneic clinical settings, which would be promising for the future treatment of HD patients.

| DECL AR ATI ON OF INTERE S TS
JS is the founder and CEO of iPS Bio, Inc The other authors declare that they have no competing interests. Quantification of processes, maximal process length, and cell body area in hGFAP-positive cells and mGFAP-positive cells (for two regions in 6 brain sections of transplanted group (n = 1)). (D) Morphological differences between endogenous astrocytes in TP mice and WT mice. Scale bar: 20 μm. (E) Quantification of processes, max process length, outgrowth intensity and cell body area in sham, WT and TP mice (for two regions in eight brain sections of each group (n = 3 each), ***P < .001). Data were analysed by two-way ANOVA followed by a Tukey post hoc test or Student's t test using GraphPad's Prism software F I G U R E 6 HLA-iPSC-NPC-derived astrocytes modulate environmental toxicity in TP mice. (A) Diagram depicting the glutamate uptake and K+ channel (Kir4.1) buffering in differentiated astrocytes. Differentiated astrocytes of mice with transplanted HLA-iPSC-NPCs exhibited recovery of Kir4.1 channel (B), glutamate uptake (C) and glutamine synthetase (GS) (D) for astrocytic functional rescue. Double staining or Western blot analysis for Kir4.1 or EAAT2 or GS and GFAP in 11-month-old YAC128 transgenic mice with or without HLA-iPSC-NPCs transplants showed that differentiated grafted astrocytes were increased on all three markers compared to the endogenous astrocytes. Scale bar: 20 μm. (B, C and D. n = 3, each group, *P < .05, ***P < .001). (E) Double staining for the inflammatory marker IL-1β and Iba-1. Scale bar: 20 μm. (F) Double staining for an anti-inflammatory marker Arginase1 (Arg1) and Iba-1. Scale bar: 20μm. (G) Western blot analysis for Arg1 and IL-1β expression (n = 3 in each group, *P < .05, **P < .01). Data were analysed by two-way ANOVA followed by a Tukey post hoc test using the SPSS software