Engineering synucleinopathy‐resistant human dopaminergic neurons by CRISPR‐mediated deletion of the SNCA gene

Abstract An emerging treatment for Parkinson's disease (PD) is cell replacement therapy. Authentic midbrain dopaminergic (mDA) neuronal precursors can be differentiated from human embryonic stem cells (hESCs) and human induced pluripotent stem cells (iPSCs). These laboratory‐generated mDA cells have been demonstrated to mature into functional dopaminergic neurons upon transplantation into preclinical models of PD. However, clinical trials with human fetal mesenchephalic cells have shown that cell replacement grafts in PD are susceptible to Lewy body formation suggesting host‐to‐graft transfer of α‐synuclein pathology. Here, we have used CRISPR/Cas9n technology to delete the endogenous SNCA gene, encoding for α‐synuclein, in a clinical‐grade hESC line to generate SNCA +/− and SNCA −/− cell lines. These hESC lines were first differentiated into mDA neurons, and then challenged with recombinant α‐synuclein preformed fibrils (PFFs) to seed the formation for Lewy‐like pathology as measured by phosphorylation of serine‐129 of α‐synuclein (pS129‐αSyn). Wild‐type neurons were fully susceptible to the formation of protein aggregates positive for pS129‐αSyn, while SNCA +/− and SNCA −/− neurons exhibited significant resistance to the formation of this pathological mark. This work demonstrates that reducing or completely removing SNCA alleles by CRISPR/Cas9n‐mediated gene editing confers a measure of resistance to Lewy pathology.


| INTRODUCTION
Parkinson's disease (PD) is a common neurological disorder that is caused by the loss or dysfunction of specific neuronal cell types. The dopaminergic neurons of the substantia nigra pars compacta that innervate the dorsal striatum are significantly affected in this condition. The degeneration of the nigrostriatal dopaminergic pathway is largely responsible for the motor symptoms of PD.
An experimental therapy for PD is transplantation of fetal ventral mesencephalic cells into the striatum (Lindvall et al., 1988(Lindvall et al., , 1990(Lindvall et al., , 1994. Although some patients experienced long-term alleviation of motor symptoms (Li et al., 2016;Ma et al., 2010), a significant proportion suffered from graft-induced dyskinesias (Hagell et al., 2002;Piccini et al., 2005), which was attributed to serotonergic hyperinnervation from the graft (Politis et al., 2010). Furthermore, most grafts older than 10 years acquired Lewy body pathology suggesting that host-to-graft spread of disease may be occurring (Kordower, Chu, Hauser, Freeman, & Olanow, 2008;Li et al., 2008;Mendez et al., 2008). The burden of Lewy bodies in the graft correlated with a decrease in the symptomatic benefit to the patient (Chu & Kordower, 2010). These clinical observations highlight the need for cell therapies that are resistant to the formation of Lewy bodies, which are predominantly composed of the protein α-synuclein (Spillantini et al., 1997). Such diseaseresistant cells will be particularly important for patients with young-onset Parkinson's or genetic forms of the condition with substantial α-synuclein burden, such as SNCA multiplications (Farrer et al., 2004) and GBA mutation carriers (Neumann et al., 2009).
In cell culture and animal models of PD, synucleinopathy can be initiated by inoculation with preformed fibrils (PFFs) of recombinant α-synuclein protein. In primary rodent neurons, molecular pathology is observed by monitoring the phosphorylation of α-synuclein at position serine-129. This is a prominent posttranslational event that is found in Lewy bodies, and is an early event during the aggregation of α-synuclein (Fujiwara et al., 2002). If neurons lack endogenous expression of α-synuclein they are not susceptible to α-synuclein PFF seeding (Volpicelli-Daley et al., 2011), and it has been described that neurons with low endogenous expression of α-synuclein are spared in PD (Braak, Ghebremedhin, Rüb, Bratzke, & Del Tredici, 2004). Furthermore, when Snca −/− mice were stereotactically injected with α-synuclein PFFs they exhibited no signs of synucleinopathy or neurodegeneration unlike their wild-type counterparts (Luk et al., 2012).
The differentiation of human embryonic stem cells (hESCs) and induced pluripotent stem cells (iPSCs) into midbrain dopaminergic (mDA) neurons is now well understood and described Kriks et al., 2011), and the route to clinical trials for hESC/iPSC-derived mDA cell therapy for Parkinson's has been mapped out (Barker, Parmar, Studer, & Takahashi, 2017). However, these grafts will express wild-type levels of α-synuclein protein making them susceptible to synucleinopathy with a similar kinetics to fetal mesencephalic grafts. Here, we show that deleting alleles of the SNCA gene in hESCs and iPSCs reduces or eliminates α-synuclein expression from mDA neurons and confers resistance to α-synuclein PFF induced formation of Lewy-like pathology. This work sets out a strategy to produce disease-resistant mDA grafts for Parkinson's that will last the life time of the individual, and will be particularly important for patients with an aggressive synucleinopathy.

| hESC and iPSC culture
Approval for the use of hESCs used in this study was granted by the MRC Steering Committee for the UK Stem Cell Bank and for the Use of Stem Cell Lines (ref, . RC17 hESCs (~passages 25-30) and AST18 iPSCs were maintained in self-renewing conditions on Laminin-521 (L521, 5 μg/ml, Biolamina) coated 6-well plates (Corning), and fed daily with StemMACS iPS-Brew XF (iPS-B, Miltenyi Biotec). Every 2 or 3 days, once the hESCs reached 70%-90% confluency, the cells were passaged as clumps with EDTA (0.5 mM, Thermo Fisher Scientific) at a split ratio of 1:3 to 1:8. See Supporting Information Table S1 for all catalogue numbers.

| Cloning of gRNA plasmids
For each gRNA, a 100 bp double-stranded DNA fragment was made by PCR of a pair of partially complementary single-stranded DNA, of which the complementary sequence was the gRNA (Mali et al., 2013). The fragment was then cloned into BbsI-HF (NEB) digested pSpCas9n-2A-Puro plasmid (PX462, Addgene) (Ran et al., 2013)

| TOPO-cloning and DNA sequencing
PCR products of putative SNCA knockout clones were purified with Wizard ® SV Gel and PCR Clean-Up System (Promega), incubated with Taq DNA polymerase at 72°C for 10 min, and inserted into pCR ™ 4-TOPO ® vector (TOPO ® TA Cloning ® Kit for Sequencing, Thermo Fisher Scientific) according to manufacturer's instructions. The plasmids were respectively transfected into DH5α competent E. coli (NEB) according to manufacturer's instructions. Carbenicillin (Fisher) was used at 100 μg/ml in LB/agar plates for selection of transformed cells. The next day, clones were picked into liquid LB-containing 100 μg/ml carbenicillin. After overnight incubation, plasmids were isolated with QIAprep Spin Miniprep Kit (QIAGEN). The purified plasmids were digested with EcoRI (NEB) and electrophoresed on 1% agarose gel. The plasmids of potential SNCA knockout clones (using T7 primer) and the purified PCR product of the potential α-synuclein homozygous knockout clones were sent for Sanger sequencing at the MRC PPU DNA Sequencing and Services (Dundee).

| Gene expression analysis by RT-qPCR
Total RNA was isolated and DNaseI-treated from self-renewing hESCs and day 44 mDA neurons with the MasterPure ™ Complete DNA and RNA Purification Kit (Epicentre) following the manufacturer's instructions. cDNA was synthesized from 500 ng of total RNA using the SuperScript ™ IV Reverse Transcriptase (Thermo Fisher Scientific). qPCR was performed using the Roche LightCycler ® 480 System with the Universal Probe Library (UPL) (Roche). The Roche UPL Assay design centre was used to design intron-spanning primers with a specific UPL probe for each gene (TBP F-gaacatcatggatcagaacaaca R-atagggattccgggagtcat Probe 87; NANOG F-tctccaacatcctgaacctca R-ttgctattcttcggccagtt Probe 87; OCT4 F-tgccgtgaaactggagaag, R-gcttggcaaattgttcgagt Probe 78; DAT F-agactgcccgaagtgtgc, R-gcagtttcccgttacaccaa Probe 14; NURR1 F-atttcctcgaaaacgcctgt, R-catactgcgcctgaacacaa Probe 41; SOX6 F-gcttctggactcagcccttta, R-ggccctttagcctttggtta Probe 50; VMAT2 F-cgggattctgcatcatgttt, R-tggcaatcagcaggaagg Probe 67). Reactions (10 μl) containing cDNA, primers, UPL Probe, LightCycler ® 480 Probes Master mix (Roche) and PCR water were performed in 386-well plates as described in the manufacturer's instructions. The data were normalized to levels of TATA-binding protein (TBP) and the data of technical replicates were plotted.

| Flow cytometry
For flow cytometry, cells were harvested with Accutase and washed with NDM. Cells were re-suspended with flow cytometry buffer, which consisted of DPBS (Sigma) + 2% fetal calf serum (Thermo Fisher Scientific), and stained with rat anti-CORIN antibody (R&D Systems,1:200). Cells were washed with flow cytometry buffer and stained with donkey anti-rat IgG Alexa Fluor-488 (Thermo Fisher Scientific). Flow cytometry was performed on BD FACS Calibur (BD Biosciences) and analysed on the FlowJo software program (FlowJo, LLC).
Protein was analysed by HuPage gel electrophoresis and a single protein species was observed (Supporting Information Figure S5A), and electrospray mass spectrometry (ESI+) gave a deconvoluted mw of 14502 Da (mw calc. 14501 Da for N-acetylated species). Size exclusion chromatography with multi-angle static light scattering (SEC-MALS) (data not shown), JC-1 fluorescence (Lee et al., 2009), and attenuated total reflectance Fourier-transform infrared (ATR-FTIR) spectrometry analysis were used to confirm that the recombinant human α-synuclein was monomeric (Supporting Information Figure S5B,C). Endotoxin level was measured by Limulus Amebocyte Lysate test (PTS20F, Charles River Laboratories). The protein was not phosphorylated at position serine-129 as determined by the pS129-αSyn antibody (Abcam).

| PFF seeding and pS129-αSyn quantification
On day 44 of mDA differentiation, neurons were exposed to α-synuclein PFFs or monomers (5 μg/ml) by replacing consumed media with feeding media containing the α-synuclein protein. Half of the consumed media were replaced weekly until 5 weeks after seeding. After immunostaining with pS129-αSyn and β-III tubulin, neurons were imaged with Axio Observer (Zeiss) or Eclipse Ti-E (Nikon) and deconvoluted with Huygens (Scientific Volume Imaging). Images were processed with Fiji (National Institutes of Health): Z-project images of the two channels were created after setting brightness and contrast, followed by converting to binary images. In order to only quantify pS129-αSyn structures within β-III tubulin-positive neurons, a Boolean "AND" function of the two binary images was performed on Fiji. The area of pS129-αSyn immunostaining divided by the total area of β-III tubulin was used for quantification. Statistical analysis was performed in PrismGraphPad software to compare experimental conditions. A two-tailed, unpaired nonparametric analysis was performed using the Mann-Whitney U test.

| CRISPR-mediated deletion of SNCA in hESCs
A series of guideRNAs (gRNAs) flanking exon 2, the first coding exon, of human SNCA were designed using the Zhang Lab CRISPR DESIGN webtool (http://crispr.mit. edu). The oligonucleotides encoding for the top 2 pairs of gRNAs at the 5′ end of SNCA exon 2 and the top 2 pairs of gRNAs at the 3′ end of SNCA exon 2 were cloned in the pSpCas9n-2A-Puro plasmid (Ran et al., 2013) and tested for deletion efficiency in HEK 293 cells (data not shown). We employed the nickase mutant of Cas9, Cas9-D10A (Cas9n), to avoid off-target double-strand breaks in the cells (Ran et al., 2013). Based on cutting efficiency in HEK 293 cells, we decided upon 5G1 and 3G1 gRNAs to target the 5′ end of SNCA exon 2, and 5G4 and 3G4 gRNAs to target the 3′ end ( Figure 1a). We designed a PCR screening strategy to detect hESC clones with a complete deletion of exon 2. We used RC17 hESCs as the parental line, since it has all the regulatory paperwork in place to be used as a cell therapy in clinical trials (De Sousa et al., 2016). Here, hESCs were electroporated with four gRNA Cas9n plasmids and clones selected in puromycin. After replica plating of the clonal hESC lines, genomic DNA was isolated from 51 clones and analysed for exon 2 deletion by PCR (Figure 1a). Nine putative SNCA −/− clones and four putative SNCA +/− clones were identified by this method. The wild-type PCR band is expected to be ~450 bp and the knock-out SNCA alleles were typically between 400 bp and 200 bp in size for this initial PCR screening (Figure 1b). A collection of putative knock-out clones were investigated further with long-range PCR genotyping strategies to determine if larger deletions could be detected (Figure 1c,d). This revealed that clone M1-4 harboured two unique knock-out alleles, only the first of which could be detected by the original PCR genotyping method. Clone L5-3 appeared to have only a single deletion allele, and no wildtype alleles (Figure 1c,d). The PCR products of four hESC clones were TOPO-cloned and sent for Sanger sequencing to determine the precise deletion events (Supporting Information Figures S1-S3). Clones 4-4 and L5-4 were putative SNCA +/− cell lines. Clone 4-4 had one intact allele, and a 210-bp deletion in the second allele removing most of exon 2 including the ATG start site. Clone L5-4 harboured a 175-bp deletion in one allele removing the ATG, and a 12-bp insertion in the 5′ UTR and 28-bp deletion in intron 2 of the second allele (Figure 1e, Supporting Information Figure S2). Clones M1-4 and L5-3 were putative SNCA −/− cell lines. Clone M1-4 had a 214-bp deletion in one allele, and a 428-bp deletion in the second allele, both of which remove exon 2 and the ATG start codon (Figure 1e, Supporting Information Figure S3). Clone L5-3 harboured a 250-bp deletion removing exon 2 and an absence of any wild-type allele. All three PCR genotyping reactions only detected a single allele in Clone L5-3 suggesting the deletions are nearly identical on both alleles, or that the second allele has a deletion beyond the detection of the ~1.9 kb PCR reaction (Figure 1d). We investigated the four most likely off-target cleavage sites in the genome using a T7 endonuclease I (T7E1) assay. We did not observe any offtarget cleavage at these sites as anticipated from using the Cas9n system (Supporting Information Figure S4). We then checked α-synuclein protein expression in these clones and others by western blotting of undifferentiated hESC lysates. Clones M1-4, L5-3, and other clones were confirmed to lack any detectable α-synuclein expression, while putative SNCA +/− clones, 4-4 and L5-4 had reduced expression (Figure 1f). This suggested that the small insertion on the 5′UTR and 28-bp deletion in intron 2 of Clone L5-4 did not affect α-synuclein expression.

| Normal midbrain dopaminergic differentiation of SNCA targeted hESCs
Reduction or deletion of the SNCA gene may affect differentiation efficiency into mDA neurons. In order to address this question, we differentiated two SNCA +/− hESC lines and two SNCA −/− lines, as well as the wild-type RC17 parental hESC line, using a modified version of an established mDA neuronal protocol (Figure 2a; . FACS analysis for the floor plate marker, CORIN, at day 16 of differentiation showed that all cell lines produced a highly pure population of floor plate cells regardless of SNCA genotype (Figure 2b). Gene expression analysis at day 44 of differentiation showed a downregulation of the pluripotent markers OCT4 and NANOG in all genotypes of cell line, and an up-regulation of the dopaminergic markers NURR1, SOX6, DAT, and VMAT2 in all cell lines regardless of SNCA genotype (Figure 2c). This suggests deletion of the SNCA gene does not have a major effect on the differentiation of hESCs into mDA neurons. Immunostaining for the pan-neuronal marker, β-III tubulin, the dopaminergic enzyme tyrosine hydroxylase (TH), and the midbrain marker, FOXA2, further confirmed there were no gross differences in the differentiation potential of SNCA −/− hESCs to produce dopaminergic neurons (Figure 3a,b). Wild-type and both SNCA +/− hESC lines exhibited nuclear and cytoplasmic α-synuclein expression, while mDA neurons differentiated from both SNCA −/− cell lines lacked this expression (Figure 4), in agreement with the western blotting data. The localization of α-synuclein is not synaptic in these neuronal cultures at day 44 because they were not sufficiently mature yet. This is in agreement with the observation that α-synuclein is not redistributed from the cell body to synapses until 18 weeks of gestation (Galvin, Schuck, Lee, & Trojanowski, 2001).

| α-Synuclein-null mDA neurons are resistant to acquiring Lewy-like pathology
We induced synucleinopathy in human neurons by using an established α-synuclein preformed fibrils (PFFs) model (Volpicelli-Daley et al., 2011). Human α-synuclein protein was produced using a HEK 293 mammalian expression system. Monomeric protein was agitated for 10 days to form fibrils that were quality controlled by JC-1 fluorescence (Lee et al., 2009), and Fourier-transform infrared (FTIR) spectrometry analysis (Kaylor et al., 2005) (Supporting Information Figure S5). The fibrillar protein was then sonicated on ice to produce PFFs. Recombinant α-synuclein PFFs, or monomers, were added to the culture medium of day 44 mDA neurons at a concentration of 5 μg/ml. After 5 weeks of additional culture and maturation of neurons, the cells were immunostained for β-III tubulin and phosphoSer-129α-synuclein (pS129-αSyn) (Figure 5a), a posttranslational modification that occurs in Lewy bodies and is a reliable biomarker of aggregated α-synuclein (Fujiwara et al., 2002). mDA neurons exposed to α-synuclein PFFs exhibited numerous pS129-αSyn structures within neurons after 5 weeks, while α-synuclein monomers were unable to induce the formation of these structures (Figure 5b). To quantify the extent of Lewy-like molecular pathology in neurons, the immunostaining of pS129-αSyn and β-III tubulin were converted to binary images and the percentage of β-III tubulin-positive space (i.e. neurons) occupied by pS129-αSyn signal was calculated ( Figure 5b). We then compared wild-type, SNCA +/− , and SNCA −/− mDA neurons for their susceptibility to form phosphorylated α-synuclein structures in response to pathology-inducing PFFs in the same time-frame (5 weeks). Wild-type mDA neurons exhibited extensive pS129-αSyn structures due to α-synuclein PFF seeding (Figure 6a). Heterozygous SNCA +/− mDA neurons exhibited much less pS129-αSyn immunostaining, but a small number of neurons showed punctate pS129-αSyn-positive structures along axons (Figure 6a). Knock-out SNCA −/− mDA neurons did not possess any pS129-αSyn-positive structures that followed axonal tracks or co-localised with β-III tubulin-positive neuronal cell bodies ( Figure 6a). Quantification of the pS129-αSyn structures with respect to the area of β-III tubulin-positive neurons showed that SNCA +/− and SNCA −/− mDA neurons were significantly less susceptible to forming Lewy-like structures than wild-type neurons when challenged with α-synuclein PFFs (Figure 6b). Control monomeric α-synuclein at the same concentration or vehicle, PBS, were unable to initiate the formation of pS129-αSyn structures in any of the mDA neuronal cultures (Figure 6b, Supporting Information Figure  S6). To confirm these results with an independent cell line, we used the same strategy to target exon 2 of the SNCA gene in AST18 iPSCs, which harbour a triplication containing the SNCA locus (Devine et al., 2011). We obtained four clonal iPSC lines that lacked α-synuclein protein expression based on western blotting (Figure 7a). AST18 (SNCA Trip ) and AST18-7B (SNCA null ) iPSCs were differentiated into mDA neurons and treated with α-synuclein PFFs using the same conditions for hESC-derived neurons. Upon fixing and immunostaining for pS129-αSyn it was observed that SNCA Trip neurons exhibited numerous pS129-αSyn structures along neurites, while these structures were virtually absent in the SNCA null neurons (Figure 7b). Image quantification of the pS129-αSyn area demonstrated a significant reduction of the Lewy-like structures in SNCA null neurons (Figure 7c), which is in agreement with the observation in hESC-derived neurons.

| DISCUSSION
A dopaminergic cell replacement therapy for Parkinson's is rapidly approaching clinical trials (Barker et al., 2017). We know from clinical trials with fetal mesencephalic grafts that significant dopaminergic re-innervation can be achieved, and this is accompanied by significant clinical benefit (Lindvall et al., 1990(Lindvall et al., , 1994Ma et al., 2010). However, the cells transplanted into the Parkinson's striatum are susceptible to synucleinopathy and the formation of Lewy bodies (Kordower et al., 2008;Li et al., 2008;Mendez et al., 2008). Over time this reduces the clinical benefit of the graft and a re-emergence of motor symptoms (Li et al., 2016). The ongoing synucleinopathy to other parts of the brain, such as the cortex, will likely contribute to a worsening of symptoms as well. Here, we have engineered human pluripotent stem cells with CRISPR/Cas9 nickase technology to reduce or remove the SNCA alleles encoding for α-synuclein. Using Cas9n with F I G U R E 5 α-Synuclein pre-formed fibrils (PFFs) seeding and the process of phosphorylated Serine 129 α-synuclein (pS129-αSyn) quantification. (a) Time line of PFFs seeding and pS129-αSyn quantification during mDA differentiation. (b) WT mDA neurons seeded with PFFs for 5 weeks, immunostained with pS129-αSyn (green) and β-III tubulin (red): colour images (left panel) and respective binary images (right panel), with partially 2.8× enlarged images and percentage of overlapping pS129-αSyn staining normalized to β-III tubulin immunostaining four gRNAs, we were able to successfully remove the first coding exon, exon 2, of the human SNCA gene in clinicalgrade RC17 hESCs to produce a number of SNCA +/− and SNCA −/− clonal cell lines, as well as deleting all four alleles of SNCA from the α-synuclein triplication iPSC line, AST18. All genotypes of cell line efficiently produced mDA neurons in a modified floor plate protocol suggesting that removal of the α-synuclein protein does not grossly affect the differentiation propensity of pluripotent stem cells into this neuronal subtype. More detailed transcriptomic analysis and functional studies of dopamine release, for example, will be required to determine if more subtle differences exist between wild-type human neurons and the ones that lack α-synuclein, especially since Snca −/− mice have reduced dopamine content in the striatum despite normal numbers of dopaminergic neurons in the substantia nigra (Abeliovich et al., 2000;Al-Wandi et al., 2010). Analysis of synaptic vesicle mobilisation in mature mDA neurons will also be important, since mice lacking α-synuclein exhibit a deficit in replenishing pools of presynaptic vesicles after depletion in hippocampal neurons (Cabin et al., 2002), and this phenotype was also observed in the striatum of triple knock-out mice lacking all three synuclein genes, Snca, Sncb, Sncg (Anwar et al., 2011). It is interesting to note that Snca −/− mice exhibit a degree of resistance to F I G U R E 6 pS129-αSyn quantification of SNCA +/− , SNCA −/− or WT mDA neurons seeded with PFFs or monomers. (a) SNCA +/− , SNCA −/− and WT mDA neurons seeded with PFFs for 5 weeks, immunostained with pS129-αSyn (green) and β-III tubulin (red): colour images (left panel) and respective binary images (right panel). (b) Quantification of overlapping pS129-αSyn immunostaining in PFF and monomer-seeded mDA neurons normalized to β-III tubulin immunostaining presented as box plots. (PFF data: SNCA +/+ , n = 6, 18 images; SNCA +/− , n = 6, 16 images; SNCA −/− , n = 4, 12 images. Monomer data: SNCA +/+ , n = 3, 6 images; SNCA +/− , n = 4, 11 images; SNCA −/− , n = 1, 2 images) ****p < 0.0001, *p < 0.05: Mann-Whitney U test the mitochondrial toxin MPTP (Dauer et al., 2002;Robertson et al., 2004), suggesting α-synuclein-null grafts may have some protection against environmental toxins. However, Snca −/− mice are highly susceptible to neuroinvasive viruses, such as West Nile virus, while heterozygous Snca +/− mice do not display this phenotype (Beatman et al., 2015).
In agreement with our data, the loss of α-synuclein in mice confers complete resistance to experimentally induced synucleinopathy (Luk et al., 2012), which is analogous to the finding that the endogenous PrP gene is critical to be susceptible to prion infection (Mallucci et al., 2003). When we challenged wild-type, SNCA +/− and SNCA −/− mDA neurons with recombinant α-synuclein PFFs, we observed a significant number of punctate pS129-αSyn structures in wild-type neurons, but very few or none in SNCA +/− and SNCA −/− neurons. Exposing mDA neurons to monomeric α-synuclein protein at the same concentration was unable to produce pS129-αSyn structures. The lack of pS129-αSyn structures in SNCA −/− neurons was expected from the data published in rodent neurons and mice (Luk et al., 2012;Volpicelli-Daley et al., 2011). However, the significantly reduced pS129-αSyn structures in heterozygous SNCA +/− neurons were somewhat surprising. This is most likely due to the length of time of the assay (5 weeks). Since these neurons express about half the level of α-synuclein protein, they would require longer time periods to accumulate and mature the pS129-αSyn structures. This has implications for the longevity of heterozygous SNCA +/− grafts in patients, given the time (>10 years) it took wild-type fetal grafts to acquire Lewy pathology and begin to fail. Reducing the α-synuclein protein level by 50% could significantly increase the functional longevity of the graft by slowing and reducing Lewy body burden, while avoiding the potential issues of synaptic mobilisation defects or increased susceptibility to neuroinvasive viruses. Preclinical testing of SNCA +/− and SNCA −/− mDA cells will require an animal model of host-to-graft α-synuclein spread. A 6-hydroxydopamine (6-OHDA) lesion model with AAV6-human α-synuclein administration in the rat has been reported (Kordower et al., 2011). Wild-type rat ventral mesencephalon (VM) was used to graft the animals, and between 2% to 15% of the dopaminergic neurons acquired human αsynuclein from the host (Kordower et al., 2011). A mouse model of host-to-graft spread has also been reported where embryonic mouse VM is transplanted on transgenic mice expressing human α-synuclein driven by the mouse SNCA promoter (Hansen et al., 2011). Demonstration of diseaseresistance of human SNCA +/− and SNCA −/− mDA cells in one of these models, or similar models, will be required prior to clinical translation.
Genome editing with CRISPR/Cas9 technology is being applied to address diverse medical problems, including cancer, HIV infection, and monogenic diseases (Hsu, Lander, & Zhang, 2014;Knott & Doudna, 2018). Genomeedited human pluripotent stem cells for clinical trials have yet to be reported, but is set to follow the work of genome editing of other cell types, such as T cells, for medical applications. The use of Cas9 nickase technology (Ran et al., 2013), or next-generation Cas9 enzymes (Kleinstiver et al., 2016), have significantly reduced or eliminated offtarget disruption of the genome and dramatically improved the safety profile of genome-engineered cells for clinical applications.
Here, we have shown that SNCA +/− and SNCA −/− mDA neurons can be generated by a straight-forward genome engineering method via hESCs or iPSCs, and they present a F I G U R E 7 pS129-αSyn pathology model is reproducible in hiPSCs. (a) Western blotting of triplication SNCA AST18 iPSC clones targeted for deletion of exon 2 of SNCA gene. (b) SNCA Trip and SNCA null (7B clone) mDA neurons seeded with α-Syn PFFs or monomers, immunostained with pS129-αSyn (green) and β-III tubulin (red). Three representative images are shown for each condition. (c) Quantification of pS129-αSyn immunostaining in β-III tubulin-positive domain for PFF and monomer-seeded mDA neurons presented as box plots. **p < 0.01: Mann-Whitney U test