A novel mutation in intron 1 of Wnt1 causes developmental loss of dopaminergic neurons in midbrain and ASD-like behaviors in rats

Autism spectrum disorder (ASD) is a group of neurodevelopmental disorders with a strong genetic liability. Despite extensive studies, however, the underlying pathogenic mechanism still remains elusive. In the present study, we identified a homozygous mutation in the intron 1 of Wnt1 via large-scale screening of ASD risk/causative genes and verified that this mutation created a new splicing donor site in the intron 1, and consequently, a decrease of WNT1 expression. Interestingly, humanized rat models harboring this mutation exhibited robust ASD-like behaviors including impaired ultrasonic vocalization (USV), decreased social interactions, and restricted and repetitive behaviors. Moreover, in the substantia nigra compacta (SNpc) and the ventral tegmental area (VTA) of mutant rats, dopaminergic (DAergic) neurons were dramatically lost, together with a comparable decrease in striatal DAergic fibers. Furthermore, using single-cell RNA sequencing, we demonstrated that the decreased DAergic neurons in these midbrain areas might attribute to a shift of the boundary of the local pool of progenitor cells from the hypothalamic floor plate to the midbrain floor plate during the early embryonic stage. Moreover, treatments of mutant rats with levodopa could attenuate the impaired USV and social interactions almost completely, but not the restricted and repetitive behaviors. Our results for the first time documented that the developmental loss of DAergic neurons in the midbrain underlies the pathogenesis of ASD, and that the abnormal progenitor cell patterning is a cellular underpinning for this developmental DAergic neuronal loss. Importantly, the effective dopamine therapy suggests a translational significance in the treatment of ASD.

Guide RNA (gRNA) targeting vector with targeting sequence, flanked by 120 bp homologous sequence combined on both sides is designed.The c.104+1G>A (TGG>TGA) in donor oligo will be introduced into exon1 by homology-directed repair.The mixture of gRNA, Cas9 mRNA generated by in vitro transcription and donor oligo were microinjected into fertilized eggs isolated from SD rats.The pups then were genotyped by PCR followed by sequence analysis.A female founder (F0) was generated, and then backcrossed to SD rats for at least three generations to avoid potential off-target mutations.The sequences used for CRISPR/Cas9 editing and the primers used for genotyping of the mutant site are listed in Supplementary Table1.All rats were raised in an SPF animal facility at GWCMC under the standard conditions:12 h of light/dark cycle, the temperature of 20-22 ℃, the humidity of 60%, food and water ad libitum.Both male and female mutant rats, together with their littermate controls, were used throughout the experiments.The sample sizes of animals were determined based on the rules of 3R (reducing, reusing, recycling), and an adequate statistical power was guaranteed at the same time.Usually, tissue samples for molecular biological experiment were at least 3 in each group, animals for behavioral tests were at least 10 in each group, except in a few cases that the number of animals was below 10.

In vitro assay
EGFP genetic sequence, with Kpn1 and Bgl2 restriction enzyme cutting site in N and C terminal respectively, was amplified.The above EGFP sequence and pBudCE4.1 plasmid were double digested by Kpn1 and Bgl2 respectively and then ligated, thus pBudCE4.1-EGFPplasmid was generated.The human genomic DNAs were extracted from the peripheral blood samples from a healthy people and the ASD patient who harbors the Wnt1 sp/sp mutation.Wnt1 WT and mutant genomic sequence was amplified, with ScaI and XbaI restriction enzyme cutting site in N and C terminal respectively.Then, the above Wnt1-nor and Wnt1-mutant PCR fragments were inserted into the intermediate vector pEGM-Teasy, name pEGM-Teasy-Wnt1-WT and pEGM-Teasy-Wnt1-mutant.The pBudCE4.1-EGFP, pEGM-Teasy-Wnt1-WT and pEGM-Teasy-Wnt1-mutant plasmids were double digested by ScaI and XbaI respectively.The enzyme-digested products of Wnt1-nor-myc and Wnt1-mutant-myc were separately cloned into pBudCE4.1-EGFPplasmids, and these plasmids were respectively named as Wnt1-normal-myc plasmid and Wnt1 sp -myc plasmid.All the elements of the plasmids were confirmed by sequencing.The pBudCE4.1-EGFP plasmid was used as the control.PCR primers used for these plasmids above are listed in Supplementary Table 2.The same amount in each plasmid above was transfected into 293T cells (purchased from and authenticated by Procell) using the lipo3000 system (Thermo Fisher Scientific).Transfected cells were harvested after culturing in DMEM (Procell) plus 10% FBS (Gibco) for 48 hours.

Rapid amplification of cDNA ends (5 'RACE)
A Roche 5'/3 'RACE kit (2nd, Cat.No. 03353621001) was used for 5 'RACE following the manufacturer's instructions.Briefly, the total RNA of cells or tissues were extracted by Trizol and chloroform.RNA was precipitated by isopropanol and then dissolved in DEPC-treated distilled water.A human/rat Wnt1-specific primer 1 was used for a reverse transcription and the synthesized cDNA was purified by using a commercial kit (GenStar D206-01).A poly-A tail signaling was added to the 5' end of the cDNA, and then amplified by PCR with a human/rat Wnt1-specific primer 2 coupled with an oligo dT-anchor.A human/rat Wnt1-specific primer 3 and internal primers PCR-anchor were used for nested PCR.The PCR products were subjected to agarose electrophoresis and sequencing analysis.The primer sequences were listed in Supplementary Table 3.The Wnt1 DNA and cDNA sequence were obtained from the PubMed (Human Wnt1 cDNA sequence: NM_005430.4;Rat Wnt1 cDNA sequence: NM_001105714.1;Human Wnt1 genomic sequence: NG_033141.1;Rat Wnt1 genomic sequence: NC_051342.1).

Real-time quantitative PCR (RT-qPCR)
The procedures for the total RNA extraction were described in 5'RACE.After removal of the genomic DNA, reverse transcription (RT) was conducted following the instructions of a RT kit (TaKaRa RR047A-1).The qPCR was performed according to the instructions of a qPCR kit (CWBIO CW3008S).Primer sequences are listed in Supplementary Table 4.

Chromatin co-immunoprecipitation (ChIP)
The experimental procedures were conducted according to the instruction of Magna ChIP G Tissue Kit (Millipore 17-10085).Briefly, 3-4 brain tissues except forebrain from E12.5 embryos were dissected and pooled together.Tissue stabilizing solution and 1% polyformaldehyde were used for protein-DNA fixation and crosslinking.Then tissue was homogenized in tissue lysis buffer and remove the supernatant.The precipitate was resuspended in CHIP dilution buffer and the genomic DNA was fragmented via sonication.The sample above incubated with the mixture of protein G magnetic beads and anti-ß-catenin antibody in 4℃ overnight with rotation.The above mixture was washed with low salt, high salt, LiCl, TE wash buffer in succession.
Chromatin complex was eluted from the protein G magnetic beads with CHIP elution buffer plus proteinase K.The de-crosslinked DNA was purified following the instruction of the kit (GenStar D206-01).The purified DNA was amplified by PCR with the primers followed, Otx2-F: 5'-TGT TCA AAG GCT TCG CTG GG-3'; Otx2-R: 5'-ACA CAC ACA CAC ACA CAA AAC TTC AG-3'.The PCR products were then subjected to agarose gel electrophoresis.

Chemical administration
Levodopa (Topscience T0848) and Carbidopa (Topscience T6795) powder were dissolved in PBS containing 10mg/ml Vitamin C (Sigma) to delay the oxidation reaction.Carbidopa (12.5 mg/kg) was administered (i.p.) 30 minutes before levodopa (50 mg/kg) was given (i.p.)(3).Both compounds were used once a day, for 3 consecutive days.The behavioral test was conducted 2 hours after the last injection of levodopa.For USV test, because the most obviously behavioral phenotype was observed in P11 (Fig. 1A), here we chose this time point to examine the efficacy of the DA-RT.Pups at P8 were subjected to the treatment, and behavioral test were performed at P11.

DA and its metabolites
DA and its metabolites were detected by using HPLC-MS, and the procedures were described previously.Briefly, a total of 10 mg of striatal tissue was homogenized in 20-fold volume of 2% formic acid solution on ice.A total of 10 μl of the mixture containing isoproterenol (concentration of 40 ng/ml), 10 μl 0.1% acetic acid, and 100 μl methanol was used as the internal standard.The tissue samples were centrifuged, and the supernatants were collected, and then subjected to HPLC-MS.The standard substances of DA, homoprotocatechuic acid (DOPAC) and homovanillic acid (HVA), were purchased from Sigma Aldrich.Data were collected by using Vanquish ultra-high performance liquid chromatography and TSQ Quantis triple quadrupole tandem mass spectrometer (ThermoFisher Company of the United States).The chromatographic column adopted X Select CSH C18 (100mm × 2.1mm, 2.5μm, Waters, the United States).

Immunofluorescent staining and Immunohistochemistry
The procedures for both immunofluorescent staining and immunohistochemistry were described previously (4).Briefly, rats were anesthetized with ketamine (100 mg/kg)/xylazine (20 mg/kg) mixture and were perfused transcardially with 0.9% saline, followed by 4% paraformaldehyde (PFA).Brains were post-fixed in 4% PFA overnight and then equilibrated with 30% sucrose in phosphate buffer overnight.Coronal brain sections (20 µm) of rats at P28, or 10 µm in thickness for rats at E12.5, were made with a Cryostat (Leica 3050S).For immunofluorescent staining, sections were permeabilized in PBS with 0.3% Triton X-100, and were blocked with 5% goat serum in PBS with 0.1% Triton X-100 for 1 hour.The sections were then incubated in PBS with a primary antibody overnight at 4°C, and followed by a fluorescent secondary antibody for 1 hour at room temperature.After washing, the slides were mounted with an anti-fade mounting medium (Electron Microscopy Sciences), and the fluorescent images were captured by a confocal microscopy (Leica SP8).The primary antibodies used included: TH (Millipore AB152) at 1:1000, Aldh1a1 (Proteintech 15910-1-AP) at 1:200, Dlk1 (Santa Cruze sc376755) at 1:200, and Foxa1 (Abcam 170933) at 1:400 fold-dilution.The secondary antibodies used were: Goat anti-rabbit IgG (Alexa Fluor 488; Abcam 150081) at 1:1000, Goat anti-mouse IgG (Alexa Fluor 555; Abcam 150114) at 1:1000 fold-dilution.For immunohistochemistry, the endogenous peroxidase was inactivated by using 3% hydrogen peroxide in PBS, and the sections were permeabilized in PBS with 0.3% Triton X-100, and blocked with 5% goat blocking serum in TBS with 0.1 Triton X-100 for 1 hour.After all these, the sections were incubated with primary antibody in PBS overnight at 4°C, followed by an HRP-linked rabbit/mouse secondary antibody (Dako REAL EnVision Detection System, k5007) for 1 hour at RT.Then, the Dako REAL Substrate Buffer containing hydrogen peroxide combined with DAB was used for chromogenic reaction.Finally, slides were dehydrated with alcohol gradient and mounted with neutral resin.The primary antibodies were: TH (Millipore AB152) at 1:3000, Nkx2-1 (Abcam ab76013) at 1:500, Foxa1 (Abcam ab170933) at 1:500, and Otx2 (Proteintech 13497-1-AP) at1:500.Images were captured by an inverted microscope (Leica DMi8).All captured images were analysed with Image-J software (win 64).

Behavioral tests
All rats in each genotype were randomly numbered and grouped by experimenters who would not conduct the behavioral tests.Experimenters who conducted the behavioral tests were blind to the genotype of each animal until data analysis.In order to avoid any potential effect from the behavioral test, animals used for behavioral tests only for one test.Tests were conducted between 9:00-18:00 in a sound-and light-proved behavioral room.

1) Ultrasonic vocalization (USV) test.
A protocol of isolation-induced USV in pups was used, and the procedures were described in our previous publication (5).Briefly, neonates were examined following a brief maternal separation on P2, P5, P8, P11, and P14.USVs from individually isolated pups were recorded using an externally polarized condenser microphone with a frequency range of 30 to 300 kHz that was attached 15 to 20 cm above the floor of a housed chamber.The output information from the microphone was transferred into an Avisoft-Ultrasound Gate recording system (Avisoft Bioacoustics), and the pup-emitted calls were recorded to WAV sound files using parameters optimized for rats.Pups were individually placed in the sound-proof chambers, and calls were recorded for 300 seconds.Data were analyzed using a generalized linear model with a negative binomial distribution and a log-link function.Data were exported and processed by SAS Lab Pro (Version 5.2.10;Avisoft Bioacoustics, Germany).
2) Three-chamber test.The procedures for this test were described in our previous publication (5).Briefly, test rats at the age of 4 weeks were used to assess sociability and preference for social novelty.Demo subjects (stranger 1 and stranger 2), at the same age, were first habituated in a cylindrical cage of the device for three days, where the stranger 1 and stranger 2 would be place there during the data collecting stage.Test rats were placed in the testing room for 1 day prior to the data collecting stage.This stage began with a 10-min freely-moving phase in three chambers of the device, in order to habituate to the environment.For the sociability test, a demo rat (stranger 1) was randomly put into one of the two wire cages, and then a test rat was introduced to the middle chamber, and allowed to freely explore the environment for 10 min.
Following this, another demo rat (stranger 2) was introduced into the other wire cage, and the test animal was allowed to freely explore for another period of 10 min.Parameters including the time spent in each chamber was recorded.All these parameters, together with the track maps were analyzed by using an automated SMART software.
3) Open-field test.The procedures for this test were described in our previous publication (6).

Single-cell RNA sequencing (scRNA-seq)
We conducted scRNA-seq in Guangzhou Yuanxin Biotechnology Company, China.The steps are briefly described as follows: 1) Dissection of Embryonic 11.5 (E11.5)brains.Following a timed mating, each pregnant dam at E11.5 was anesthetized with 10% chloral hydrate, and the entire uterus was dissected out and immediately placed into ice-cold PBS.Individual embryos were carefully collected, and were placed into ice-cold PBS.Brain tissues including the hypothalamus, midbrain, and hindbrain but not the forebrain, were dissected under a stereomicroscope.Tissues were quickly stored in liquid nitrogen, and single-cell isolation was performed within 24 hrs.
2) Nuclear extraction.The procedures for nuclear extraction were followed up by the manufacturer's instruction.Briefly, tissue was homogenized in Hibernate E containing B27 and GlutaMAX by using a Pasteur pipette on ice, and cell debris were first removed by filtering with a 30 μm MACS SmartStrainer.The suspension was then centrifuged at 500 rcf for 5 min, 4 o C.
After discarding the supernatant, the nuclei were washed and resuspended with a resuspension buffer gently.Then, Myelin was removed using Myelin Removal Beads II and a single LS column.Sucrose Cushion Buffer I was added to the nuclei followed by density gradient centrifugation, so that the final density is 700-1200 nuclei/μl.
3) Preparation of cDNA libraries and sequencing.The nuclei extracted above were processed following the instructions of Chromium Next GEM Single Cell 3ʹ Reagent Kits v3.1 (10× GENOMICS) to generate cDNA libraries.Briefly, each nuclear suspension was adjusted to 1000 nuclei/μl, and was loaded onto the Chromium Controller instrument to generate single-cell gel bead in emulsions (GEMs), and individual nuclear was then separated into droplets along with gel beads coated with cell barcode tags (10× cell barcode), Unique Molecular Identification Tag (UMI) and poly (dT) sequences.GEM reverse transcription was performed on a Veriti 96-well thermal cycler (Thermo Fisher Scientific, Waltham, MA, USA).The cDNA library was then amplified using primers for the R1 and P5 arms after a reverse transcription.The cDNA library was fragmented, end repaired, and poly A-tailed.Adapters were then ligated after selecting fragments of appropriate length.Sample index PCR was performed and the SPRI selection beads were used for a final purification.Finally, the sequencing library was sequenced by using an Illumina NovaSeq 6000 platform.

4) Read alignment and quality control.
Raw reads of each sample were aligned to the rat genome (Rnor_6.0), and the gene expression matrices were generated for each sample by the cellranger (v3.1.0)count function with default parameters.To determine the fraction of ambient RNA in each single cell, the SoupX package (v1.5.2) was used with default parameters.The RNA expression in each cell was then corrected using the ambient mRNA expression profile and estimated contamination.After this correction, we used DoubletFinder (v2.0.3) to identify doublet cells.After removing candidate doublets, we discarded cells with low quality, based on the criteria: unique molecular identifiers (UMI) lower than 800 or upper than 4000.To exclude ribosomal genes from contributing to the clustering, we removed ribosomal genes from the expression matrix.

5) Cell clustering and cell-type annotation.
The R package Seurat (v3.0.2) was used to cluster the cells in the merged matrix of data.From the filtered cells, the gene expression matrixes were normalized to the total UMI counts per cell and transformed to the natural log scale.To correct the batch effects, we integrated different samples using reciprocal PCA (rPCA) implemented in Seurat.We used the FindVariableFeatures function to obtain the top 2000 highly variable genes (HVGs) of each sample, and as the input dataset for batch effect correction.Using the FindIntegrationAnchors function the default dimensions (1:20), we found a set of pairwise correspondences between individual cells.These anchors are used for downstream integration of the objects.We used the IntegrateData function with the previously computed anchor set as a parameter to integrate all sample Seurat object.The default dimensions parameters (1:20) was used for the anchor-weighting procedure.The integrated dataset on all cells were then used to scale and center the genes, compute the principal components (PCs).After PCA to reduce dimensionality and build k-nearest neighbor graphs (k =20) of the cells with the function FindNeighbors based on the Euclidean distance in the 50-dimensional PC space, the main cell cluster was identified using the Louvain-Jaccard graph-based method.For classifying all filtered cells, we set the clustering parameter resolution to 0.3 with the function FindClusters in Seurat.Next, the function RunUMAP with dimensions parameters (1:20) in Seurat was used to reduce high-dimension into twodimension (2D) for visualization.Lastly, we run the Seurat FindAllMarkers function with the default parameters to identify the genes specifically expressed in each cluster.The significance of the differences in gene expression was determined using the Wilcoxon rank sum test with Bonferroni correction, and cell types were manually annotated based on the cluster markers.A marker-cluster heatmap was generated with the R pheatmap (v1.0.12) package.To calculate the sample composition based on cell type, the number of cells for each cell type from each sample were counted.The counts were then divided by the total number of cells for each sample and scaled to 100% for each cell type.

6) Analyses of differential expression of genes (DEGs).
DEGs of two groups with each cell type were performed using the FindMarkers function in Seurat.The significance of the differences in gene expression was determined by using the Wilcoxon rank sum test with Bonferroni correction.The different genes of two groups in each subcluster were determined based on following criteria: 1) expressed in more than 10% of the cells within either or both two groups; 2) |log2FC| > 0.25; and 3) Wilcoxon rank sum test adjusted p-value < 0.05.

Statistical analysis
Before statistical analyses, all data were subjected to a Shapiro-Wilk test and F test to show distribution pattern, homogeneity of variance, as well as to determine statistical methods that will be used.A comparison between two groups was performed by using unpaired Student's t-test.
Multiple comparisons were performed by using one-way ANOVA, followed by post-hoc Duncan's test.For the results of Western blot and RT-qPCR, we first adjusted the variation of signaling intensity from different batch of experiments, and the expression level in each experiment was then normalized by an average level from all WT samples.Statistical analyses were conducted by using the Graphpad Prism (Version 8.0).All data are presented as mean ± SEM, and p < 0.05 is considered as a significant difference.The sequence in pink is exon 1, followed by intron 1, sequence in cyan is the retained intron 1, sequence in dark blue is exon 2, and red arrow indicates the the new splicing site.the sequence alignment of rat Wnt1 cDNA, rat Wnt1-WT 5'RACE products, rat Wnt1 sp/sp 5'RACE products, rat Wnt1 gene.The sequence in pink is exon 1, followed by intron 1, sequence in cyan is the retained intron 1, the sequence in dark blue is exon 2, and the red arrow points to the new splicing site (n = 3).degraded via the "destruction complex", which consists of Axin, APC, CK1α, GSK3β.The increased ß-catenin is then translocated into the nucleus from cytoplasm, and activates the expression of midbrain-related marker genes, such as En1, Lmx1a, Foxa1, and Otx2, but suppresses the expression of hypothalamic relative marker genes, such as Dlk1.The number between MHFP progenitor cells and ZHFP progenitor cells is in a balance (the picture of floor plate is from( 7)), and thus, a normal boundary between them is formed.Ultimately, the dopaminergic neurons which differentiated from the MHFP progenitor cells are developed normally.In Wnt1 sp/sp condition, however, the expression of WNT1 is reduced, which in turn leads to the degradation of ß-catenin.Consequently, the expression of midbrain-related marker genes is suppressed while the expression of hypothalamic-related marker genes is activated.At the same time, the cells with an increased Dlk1 expression level may inhibit the neighbor cells to be specialized into midbrain floor plate progenitor cells.Accordingly, the number of the midbrain progenitor cells is decreased, and the number of the hypothalamic progenitor cells is increased.The boundary between them shifts to the side of midbrain.Finally, the mature dopaminergic neurons are dramatically reduced, leading to the occurrence of ASD.
Briefly, an automatic-recording open-field working station (MED Associates) was used.The open-field box (50×50×30 cm high) was divided into 16 identical squares by invisible but computer-detectable lines, and the open-field was illuminated by a dim light (20 lux).The central 9000 cm 3 (30 cm * 30 cm * 30 cm) was defined as the central area.Two sets of 16 pulsemodulated infrared photobeam were placed on opposite walls 2.5 cm apart from the wall to record X-Y ambulatory movements.Exploratory behavior in the box was computer-interfaced at a sampling rate of 100-ms resolution.Rats at 6 weeks in age were transported to the behavioral room to adapt the environment for at least 1 hour before the experiment.Behavioral indices including total distance traveled, ambulation counts, and number of rearing were recorded automatically by the scanning system for 30 minutes.Panlab Smart 3.0 software recorded the distance, speed, and time in the central and peripheral area.Numbers of rotation, grooming were manually counted based on videos.

Fig. S3 .
Fig. S3.The effect of the mutation on the expression of WNT1 in vitro.A. A 5'-RACE

Fig. S4 .
Fig. S4.Genotyping the rats.A. Sequencing the PCR products from the genomic DNA of WT,

Fig. S5 .
Fig. S5.Sequence alignment of 5' RACE products of brain tissue in WT and Wnt1 sp/sp 322

Fig. S6 .
Fig. S6.WNT1 expression level in vivo. A. The specificity of WNT1 antibody was verified by

Fig. S8 .
Fig. S8.The loss of Th + dopaminergic neurons in mutant rats was not further enhanced 344

Fig. S10 .
Fig. S10.A changed patterning of progenitor cells in the midbrain was observed in Wnt1 sp/sp

Fig. S11 .
Fig. S11.Effect of DA-RT on three chamber social preference test and motor activity.A.

Fig. S13 .
Fig. S13.Compensatory changes of upstream molecules of WNT signaling pathway in 396