Cell source

We generated iPSCs from the peripheral blood mononuclear cells (PBMCs) of an ASD pedigree and control individuals. Specifically, we used two clones (PC1, PC2) derived from the male ASD patient, two clones (HC1, HC2) from his unaffected father, and two clones (Ctrl1, Ctrl2) from a healthy male individual who is not related to the ASD pedigree.

Mutation identification

Whole genome sequencing (WGS) was performed to screen for any autism-related mutation in autism patient in comparison with his unaffected parents. Based on a previous sequencing panel [14], our analysis revealed a single de novo heterogenous stop-gain mutation Y29X (C to G) in SHANK2 gene (NM_133266.5) and no other stop-gain mutations or frameshift indels were identified upon manual inspection. Furthermore, no other de novo SNV mutations were found to be related to ASD.

Sanger sequencing

Cell suspension was prepared by digesting the iPSCs with EDTA and centrifuged at 12000 rpm for 1 min. The supernatant was discarded, and 450 μl lysis buffer (1 M Tris-HCl, 0.5 M EDTA, 5 M NaCl, 10% SDS) with 50 μl proteinase K (1 mg/ml, Merck, 39450-01-6) were added. The cells were then lysed at 55 °C for 2 h. Extracted DNA were sent to TIANYIHUIYUAN Company for Sanger sequencing. The genotypes of iPSCs cell lines were identified through the DNA sequencing analysis (Chromas).

Genetic editing technology

To establish the Iso1 cell line model, SHANK2 point mutation Y29X (C to G) was edited by Biocytogen Extreme Genome Editing System (EGE) based on CRISPR/Cas9 and developed by Biocytogen. The candidate sgRNAs, located in the SHANK2 intron1, were screened by the CRISPR design tool (http://www.sanger.ac.uk/htgt/wge/). The selected RNAs with high specificity were then determined according to on-target activity using UCATM (Universal CRISPR Activity Assay). The sgRNA-mediated CRISPR/Cas9 system-induced Nuclease-induced double-strand breaks (DSBs) can be homology-directed repair (HDR) pathway when endogenous donor DNA with homologous arms is present. The donor plasmid containing the 5′ (1.3 KB) and 3′ (0.9 KB) homologous arms, as well as the insertion of gene fragment with the point mutation c. 87G>C and a resistance cassette (puroDeltatk) were co-electoporated into the Ctrl1 cells to establish Iso1 cell line.

Both PC and Iso1 cell line were heterozygous, carrying Y29X (C to G) mutation in one allele. Only PC1-res cell line was homozygous, as the Y29X mutation had been repaired (G to C).

Cell cultureiPSCs culture

Neuronal induction and differentiation of iPSCs clones was performed according to our previous protocol [11, 15]. Briefly, cells were cultured in six-well plates coated with Matrigel (BD MatrigelTM, hESC-qualified Matrix, 354277), and grown in mTeSRTM medium (Stemcell Technologies, 85850). The cells were digested using EDTA (5 × 10−4 mol/L, Thermo Fisher Scientific, 25200056) and passaged manually, every 3–4 days for 2–3 times. OCT4 and SSEA4 were used to assess the pluripotency of cells.

Neuronal induction

Neural progenitor cells (NPCs) were generated through a neuronal induction process lasting for more than 15 days, starting from iPSCs. NPCs were derived using a dual SMAD and Wnt inhibition method, with manual isolation of neuronal rosettes. During the initial 8 days of differentiation, the cells were cultured in the N2B27 medium (Thermo Fisher Scientific, 17502048, 17504044) supplemented with SB431542 (5 µmol/L, Selleck, S1067) and dorsomorphin (5 µmol/L, Selleck, S7840), referred to as the SMAD neural proliferation system I. The medium was changed every 2 days. After 8 days, neural rosettes were carefully scraped from the culture surface and transferred onto a Matrigel-coated plate. They were then maintained in N2B27 medium for an additional 8 days. On the 16th day of proliferation, rosettes were mechanically scraped into floating fragments and plated into non-coated T25 flask for floating culture. The neural proliferation system II was used, which consisted of N2B27 medium supplemented with 20 ng/mL bFGF (Thermo Fisher Scientific, PHG0266) and 20 ng/mL EGF (Thermo Fisher Scientific, PHG0315). To expand NPCs, they were dissociated with Accutase (StemPro Accutase cell dissociation reagent, Thermo Fisher Scientific, A1110501) and passaged for 3 times. NPCs were characterized by expression of NPCs marker NESTIN and SOX2, which are indicative of their neural progenitor identity.

Neuronal differentiation

For neuronal differentiation, NPCs were expanded for 3 passages and then dissociated into single cells. These cells were plated directly on glass coverslips coated with a glial cell feeder layer which was prepared from neonatal mouse astrocyte (P0–P3), at a density of 105 cells per well in 24-well plates for molecular assays. The cells were then fed with a neuronal differentiation medium (N2B27) supplemented with 1 µM Dibutyryl cyclic-AMP (Sigma, S7858) and 20 ng/μL BDNF (PeproTech, 450-02-10). The medium was changed every other day. Following that, the cells were collected from the coverslips for immunocytochemistry. The differentiated cells were positive for various neuronal markers, including tyrosine hydroxylase (TH), vesicular glutamate transporter 1 (VGLUT1), and vesicular GABA transporter (VGAT). These results indicate that the NPCs were capable of differentiating into various types of neurons.

Neuron induction with SHH

The iPSCs were cultured as described above. When the confluence of iPSCs reached 95%, the cells were detached with EDTA (5×10−4 mol/L) and replated into 12-well plate. After 1–2 days, when the cells reached full confluence, the medium was changed to neural proliferation system I to initiate neuronal differentiation. Following 8 days of neuronal induction, cells were mechanically scraped into floating fragments and replanted into two wells of a 6-well Matrigel-coated plate. The cells were then cultured in N2B27 medium supplemented with sonic hedgehog (SHH, 100 ng/mL, R&D system, 464-SH-025) and fibroblast growth factor-8 (FGF8, 100 ng/mL, R&D system, 423-F8-025). On the 8th day of proliferation, cell clones from one well of the 6-well plate were mechanically scraped into floating fragments and then replated into non-coated T25 flask for floating culture with neural proliferation system III (N2B27 + 20 ng/mL bFGF + 20 ng/mL EGF + SHH + FGF8). Typically, neurospheres usually formed on the second day of culture in neural proliferation system III. Following formation of neurospheres, Accutase was then used to detach them, and the single cells were plated in Matrigel-coated dishes with neuronal differentiation medium (N2, Thermo Fisher Scientific; B27, Thermo Fisher Scientific; 1 µM dibutyryl-cAMP, Sigma; 20 ng BDNF, PeproTech; 20 ng/mL GDNF; 200 mM AA).

Virus transduction

NPCs were digested into single cells and infected with lentivirus (CAMKII, LV-gene (44213-1), Genecham, GOSL0188572, 6 × 108 TU/mL) packed with a tetracycline-controlled red fluorescent protein (RFP) expression construct. This allowed for visualization of fluorescent images and subsequent morphological reconstruction of glutamertagic neurons. Additionally, NPCs were digested into single cells and infected with lentivirus (expressed under the control of DAT promoter) to label dopaminergic neurons for electrophysiological recordings. After 6 h of infection, the cells were centrifuged, the medium was changed. The following day, the cells exhibiting RFP signals were cultured on coverslips coated with a glial cell feeder layer at a density of 6 × 104 cells per well in a 24-well plate to promote neuronal development.

Immunocytochemistry (ICC)

Cells grown on coverslips on 9th day of neural induction from NPCs were fixed using 4% paraformaldehyde (PFA) for 15 min, washed 3 times for 5 min each with PBS, permeabilized with PBS containing 0.3% Tween-20 (Solarbio, T8220) (PBST), and blocked by PBST containing 3% bovine serum albumin (BSA, VETECTM) at room temperature (25 ± 1 °C) for 1 h. After blocking, the cells were incubated with the primary antibodies (Table S1) diluted in PBST containing 1% BSA overnight at 4 °C. The next day, cells were washed 3 times and 10 min each with PBS and incubated with secondary antibodies diluted in PBS at room temperature for 1 h. To analyze the neuronal morphology, images of neurons were acquired using 20×, 40× objective (Imager Z2, Zeiss) and digitized them using a Zeiss camera (Axiocam 506 mono, Zeiss), axon, dendrite, and soma were traced, and morphology, including dendrite length and number of dendrite branches were analyzed using neurolucida 360 for Sholl analysis.

Quantitative real-time PCR (qRT-PCR)

mRNA extracted from neurons harvested on 9th day of neural induction from NPCs or mouse brain tissues were converted to cDNA using PrimeScript RT Reagent Kit with gDNA Eraser (#RR047A, Takara). SHANK2E and SHANK2A transcripts were produced from primer “Human long SHANK2 transcripts”. Primer “Mouse Shank2A” was designed to target exon 1 and 2 of Shank2A (NM_001113373.3), primer “Mouse Shank2B” was designed to target exon 1 and 2 of Shank2B (NM_001081370.3), primer “Mouse Shank2E” was designed to target exon 9 and 10 of Shank2E (XP_006508583.1). Primers used were listed in supplementary table (Table S2). qPCR reactions were performed using TB Green® Premix Ex Taq™ II (Tli RNaseH Plus) (#RR820A, Takara) in triplicates for each sample. GAPDH was chosen as reference gene. Initial denaturation was 95 °C for 30 s, the amplification cycles were 95 °C for 5 s, 60 °C for 30 s, for 40 repeats. All qRT-PCR reactions were performed and analyzed using the StepOne Plus Real-Time PCR System (Applied Biosystems) with the comparative Cq method.

Western blot assay

Proteins were extracted from neurons harvested on 9th day of neural induction from NPCs or brain tissues from 8-week-old mice by adding appropriate amount of RIPA lysis buffer (Thermo Fisher Scientific, 89900) containing 1% protease inhibitor (Merck, 539137-10VL). After sonication and centrifugation, extracted protein were obtained for analysis. Protein concentrations were measured using PierceTM BCA Protein Assay Kit (Thermo Fisher Scientific, 23209). Protein were mixed with 5× loading buffer (Beyotime, P0015) and heated at 95 °C for 8 min, then separated on SDS-PAGE gel (Beyotime, P0012A) and transferred to a polyvinylidene difluoride (PVDF) membrane (Millipore, IPVH00010). Membranes were blocked in TBS with 0.2% Tween-20 (TBST) (Solarbio, T8220) containing 5% bovine serum antigen (BSA) at room temperature for 1.5 h and incubated with primary antibody (Table. S1) at 4 °C overnight. On the following day, the membranes were rinsed by TBST for 5 times for 8 min each and incubated with anti-Rabbit IgG, HRP-linked antibody (1:4000, 7074P2, Cell Signaling) at room temperature for 1.5 h, then rinsed 5 times for 8 min again. Protein signals were developed using ChemiDoc Touch Imaging System (Bio-RAD, USA).

Electrophysiological recordings

For electrophysiological recordings, iPSCs-derived neurons were used after 4-5 weeks of culture. Patch-clamp recordings were performed using an IR-DIC microscope (Nikon Eclipse FN-1 microscope) and an amplifier (MultiClamp 700B, Molecular Devices). Cells were incubated at room temperature in the artificial cerebrospinal fluid (ACSF, in mM: 126 NaCl, 26 NaHCO3, 10 D-glucose, 2.5 KCl, 2 CaCl2, 2 MgCl2, and 1.25 NaH2PO4, pH 7.4) that was oxygenated with 95% O2 and 5% CO2. Borosilicate microelectrodes with a resistance of 4-8 MΩ were pulled using a pipette puller (Narishige PC10). For recordings on intrinsic membrane properties, the pipettes were filled with a solution (pH 7.3) containing: 4 mM KCl, 126 mM K-gluconate, 10 mM HEPES, 0.3 mM Na2-ATP, 4 mM Mg-GTP, 10 mM phosphocreatine. Cell slide was placed in a recording tank containing ACSF with oxygen supplementation. Neurons were visualized with a 40x water objective on an IR-DIC microscope and recorded using the amplifier. The recorded signals were filtered with a low pass frequency of 3 kHz and digitized at a frequency of 20 kHz (DigiData 1550 A, Molecular Devices). In the current-clamp mode, the input resistance (Rin) and positive resting membrane potential (RMP) were recorded. Neurons were subjected to positive and negative current stimuli of −10 pA and 2 pA per second, respectively, and the slope of the linear fit of the current-voltage curve was used to determine the input resistance (Rin). The RMP was recorded as the uncorrected fluid-contact potential. The neurons were then recorded in voltage-clamp mode, in which voltage was maintained at −70 mV. A sodium current was induced by increasing the voltage by 5 mV every 300 ms from −20 to 50 mV, and the peak value and threshold of sodium current were recorded. The neurons were then recorded in current-clamp mode. The current was increased by 10 pA every 1000 ms from −30 pA to 80 pA, and the peak and frequency of the action potential were recorded. Spontaneous excitatory postsynaptic current (sEPSC) and spontaneous inhibitory postsynaptic current (sIPSC) were recorded at −70 mV and 0 mV potentials, respectively. The collected electrophysiology data were analyzed with pClamp10 (Molecular Devices).

Bulk RNA-sequencing

We conducted an analysis of the transcriptome of neurons derived from PC1 and HC1 clone on 9th day of neural induction from NPCs. To ensure data quality, we utilized cutadapt (version 2.6) for the processing of raw reads. This involved the removal of adapter sequences and elimination of low-quality reads. The resulting RNA-seq clean reads were then aligned to the human reference genome (GRCh38) using STAR (version 2.7.10a). For gene expression quantification, we employed RSEM (v1.2.28) with GENCODE annotation (human release 31). Specifically, we extracted the Transcripts Per Million (TPM) values of the SHANK2 gene, as well as its SHANK2E, SHANK2A, and SHANK2B transcripts, from the RSEM results. Finally, we utilized the ggplot2 package to visualize the obtained data.

BaseScope

Probe design was conducted by Bio ACD company. SHANK2E: length: 1849 amino acid (aa), targeting 496-716 of NM_012309.4; SHANK2A: length: 1470 aa, targeting 36-111 of ENST00000656230.1; SHANK2B: 1261 aa, targeting 2–78 of NM_133266.5. For assessing the expression of various transcripts, HC1, PC1 and PC1-res1 NPCs were digested into single cells. Cells were inoculated onto a chamber culture slide. After 5 days culture, the neurons were pretreated according to protocol (Advanced Cell Diagnostics, 322381). For SHANK2A expression, iPSCs from HC1 and PC1 were inoculated onto a chamber culture slide 1 day before further treatment. Briefly, cells were gradiently dehydrated and rehydration on slides before being treated with H2O2 and protease III from RNAscope® H2O2 and Protease Reagents (Advanced Cell Diagnostics, 322381) at room temperature for 10 min. Probe hybridization and signal amplification were performed according to the manufacturer’s instructions. Briefly, cells were separately detected the fast red detection of C2 signal and green detection of the C1 signal, using the BaseScopeTM Duplex Detection Reagents (Advanced Cell Diagnostics, 323800). Subsequently, the same slides were hybridized and combined with TH (Millipore, AB152, 1:1000) immunohistochemistry (IHC) staining. The secondary antibody, donkey anti-rabbit Alexa-488 (InvitrogenTM, A21206, 1:500) was used. Finally, the slides were dried at 60 °C for 15 min and mounted using VectaMount (Vector Labs Vectamount, 321584). The probe signals were detected under standard bright field (Zeiss).

Generation of Shank2 KI mice

All animal experimental protocols have been approved by the Ethics Committee of Experimental Animals of Jinan University in accordance with IACUC guidelines for animal research. The sample size was determined based on literatures from the same field, and in consideration of animal size limit due to animal welfare requirement and animal experimental ethical code. Randomization methods were not used in this study.

The 29th amino acid (Y29; codons TAC) in the human SHANK2 gene corresponds to the conserved 28th amino acid (Y28; codons TAC) in the mouse SHANK2 gene. The 28th amino acid residue Tyr was located within the exon 16 (ENSMUSE00000756726) of the mouse SHANK2 gene. This design was based on mouse SHANK2 transcript-003 (NM_001081370) and completed by Biocytogen. Using CRISPR/Cas9 technology, human-carried SHANK2BY29X mutation was introduced into C57BL/6 mice, which replaces the codon encoding Tyr with a stop condon to obtain heterozygous SHANK2 (SHANK2BY29X) mice. At least 5 times backcrossing were done to remove off-target effects brought by CRISPR procedure. Subsequently, heterozygous mice were interbred to obtain homozygous (SHANK2 KI) mice. If not otherwise indicated, behavioral and molecular assays were performed on male wild-type (WT) and homozygous SHANK2 KI mice at 8-9 weeks of age.

Genotype identification

For genotype identification, gingeli size ear or tail tissue of mice was obtained 3 weeks after birth. 300 μl lysis buffer (1 M Tris-HCl, 0.5 M EDTA, 5 M NaCl, 10% SDS) containing 6 μl proteinase K (Merck, 39450-01-6) was added to the tissues and then transferred to a 55 °C oven overnight in order to digest the tissue completely. DNA was extracted by phenol chloroform (Solarbio, P1013) and purified by absolute ethanol then 70% ethanol. Extracted DNA was dried at 37 °C overnight and then dissolved by 200 μl ddH2O and stored at 4 °C until use. In the Hot Start enzyme PCR reaction program, DNA template, two primers (Primer F: 5′-TGCCTGGATCTGAGCTGGGTGATTA-3′; Primer R: 5′-CATGCCTGCAGGAAGCCTAGATGTT-3′), Taq enzyme (TAKARA, RR003A), ddH2O, were mixed and followed by the PCR program (Biometra, Germany). DNA samples were sent to TIANYIHUIYUAN Company for Sanger sequencing. The genotypes of mice were identified using the DNA sequencing analysis software (Chromas).

DAB staining

Mice were anesthetized with isoflurane and perfused with PBS, followed by 4% PFA. The brain was isolated and post-fixed in PFA solution overnight, and gradiently dehydrated in 20% and 30% sucrose at 4 °C. Brain slides in 40 μm thickness were obtained using a freezing microtome (CM1950, Leica, Germany). After being rinsed by PBS for 3 times and 10 min each, brain slides were immersed in H2O2 for 30 min, followed by another 3 times rising by PBS. Slides were then blocked in 5 ml PBS containing 5% BSA and 0.5% normal goat serum (Sigma, 566380) (blocking buffer) at room temperature for 1 h. The primary antibody (anti-TH, Millipore, AB152, 1:300) diluted in blocking buffer was added, and incubated at 4 °C for 24 h. After being rinsed by PBS for 3 times, the slides were incubated with the secondary antibody coupled to peroxidase (1:500) at 4 °C overnight, followed by 3 times rising by PBS. The DAB staining solution was prepared according to DAB staining kit (Cell signaling, 8059S). Brain slices were immersed in 2 ml DAB solution until appropriate staining was observed under an optical microscope and DAB solution was replaced by PBS to terminate staining. After air dried, brain slides were gradiently dehydrated with 70% ethanol, 95% ethanol, absolute ethanol, each for 10 min and finally with xylene. Once removed from xylene, slides were mounted, completely dried and stored at room temperature until use.

Immunofluorescence (IF) staining

Brain slides in 30 μm thickness were obtained for IF staining. After rinsed by PBS for 3 times, slides were blocked in PBS containing 0.3% TritonX-100 and 5% donkey serum (Solarbio, SL050) at room temperature for 2 h. Slides were then incubated with the primary antibodies (Table S1) diluted in PBS at 4 °C for 24 h. After being rinsed by PBS for 3 times, sliced were incubated with the secondary antibody diluted in PBS at room temperature for 1.5 h. Slices were then rinsed in PBS, mounted, covered and visualized with laser scanning confocal microscope with ×20, ×40, ×63 magnification equipped with 405/488/546/647 nm lasers (Axiocam 506 mono, Zeiss, Germany). For mean fluorescence intensity analysis of TH on NAc brain region, five 250μm × 250μm subregions from one image were obtained to get an average of fluorescence intensity for one image field. n ≥ 10 image fields from 3 WT mice and 3 KI mice were included in analysis. The intensity of TH was normalized to the average of WT group for all analyzed brain regions.

Open field test

Mice were placed in the central area of the apparatus (50 × 50 × 40 cm, Noldus, Netherlands) at the beginning and allowed to explore the arena for 15 min.

Elevated plus-maze

The elevated plus-maze (Noldus, Netherlands) is consisted of a central area, two open arms, and two closed arms, which is 50 cm height to the ground. Mice were placed in the central area heading to the open arm at the beginning of the assay and allowed to explore the arena for 5 min.

Morris water maze test

In the Morris water maze test, mice underwent a training trial for five consecutive days. During each training session, mice were trained four times a day to locate a hidden platform (10 × 10 cm) in the circular apparatus (120 cm in diameter) with visual cues on the wall. If the test mouse failed to find the platform within 60 s, it was guided to the platform and allowed to stay on it for 10 s. Escape latency to platform was recorded once the mouse found and stayed on the platform for 10 s. On the 6th day, a probe trial was conducted in which the platform was removed, and test mouse was allowed to explore the arena for 1 min. Escape latency to platform, time spent in target quadrant (where the platform was originally locate) were recorded and used to assess learning and memory function.

Social interaction assay and repetitive behavior in home cages

For the social interaction assay and repetitive behavior recording, a plastic apparatus (15 × 8 × 12 cm) with beddings was used as the home cage. Test mice were habituated in the plastic apparatus for 2 days before the assay. In the beginning of the assay, the empty cylindrical apparatus (5 cm in diameter) was placed at one side of the home cage. In the first phase of the social interaction assay, a test mouse was allowed to freely move around the home cage and explore the cylinder for 10 min. In the second phase, a stranger mouse of the same gender and smaller in size as the test mouse was placed into the cylinder, and the test mouse was allowed to explore the stranger mouse for 10 min. The duration spent on sniffing on the cylinder in each phase were recorded to demonstrate the social preference. For repetitive behavior, mouse was allowed to freely explore the home cage for 5 min, the duration spent on jumping, grooming and digging were manually recorded by one experimenter.

Three-chamber social assay

The three-chamber social assay takes place in a transparent apparatus (62 cm × 40 cm × 40 cm, Noldus, Netherlands) that is divided into three chambers (left, middle, right) by thin transparent wall, with dim light in the behavioral room. In the first phase, inanimate objects (O1 and O) was placed in two small cylindrical containers in the left and right chamber, respectively. The test mouse was initially placed in the middle chamber and allowed to freely explore all three chambers for 10 min to acclimate to the environment. In the second phase, the test mouse was gently guided back to the center chamber, and the two entrances to the side chambers were blocked. One of the containers (O1) was replaced with a strange mouse (Stranger 1, S1), while the other container retained the inanimate object (O). Then, the two entrances were opened to allow the mouse to explore the chamerbs for 10 min. In the third phase, the test mouse was gently guided back to the center chamber again, with the entrances to the side chambers blocked. The remaining container with the inanimate object (O) was replaced with Stranger 2 (S2), representing an unfamiliar novel mouse. The test mouse was allowed to explore and interact with either Stranger 1 or 2 for 10 min. The duration and frequency of sniffing on the containers were manually calculated by two experimenters. Preference ratio is calculated by sniffing duration (S1 − O)/(S1 + O) in second phase, and (S2 − S1)/(S2 + S1) in third phase, to represent social preference and social memory recognition, respectively.

Rotarod test

The Rotarod apparatus (47700(rat)/600(mouse)) was used in the test. As the test started, the rod accelerated from 0 rpm to 80 rpm in 5 min. The standard motor learning task was performed as three trials per day for three consecutive days. Mice were allowed to take a 5 min rest between trial. The duration that each mouse stays on the rotating rod in each trial was recorded as the latency to fall.

In vivo fastscan cyclic voltammetry (FSCV)

FSCV was used to measure dopamine fluctuations in the nucleus accumbens (NAc). Mice were anesthetized with urethane (1.5 g/kg, i.p.) and placed in a stereotaxic frame equipped with a heating pad to maintain body temperature (Harvard Apparatus, Holliston, MA, USA). A bipolar, stainless-steel stimulating electrode, coupled with a 26-gauge infusion cannula (Plastics One, Roanoke, VA, USA), was placed into the VTA (AP − 5.2 mm, ML + 0.5 to −1.5 mm, DV between −7.4 and −8.1 mm), a carbon fiber microelectrode was implanted in the NAc core (AP  + 1.2 mm, ML − 1.4 mm, DV from −6.0 to 7.0 mm) and an Ag/Ag Cl reference electrode was placed in the contralateral hemisphere. A low-pass filtered (2 kHz) triangular waveform (−0.4 to +1.3 V and back to −0.4 V at a rate of 400 V/s, repeated at 100 ms intervals) was applied to the carbon fiber microelectrode. Data were digitized and processed using NI-6711 and NI-6251 DAC/ADC cards (National Instruments, Austin, TX, USA) and Demon Voltammetry and Analysis Software (Wake Forest Baptist Medical Center, Winston-Salem, NC). Background-subtracted cyclic voltammograms (CVs) were obtained by digitally subtracting stable background currents to resolve CVs for dopamine. Electrical pulses delivered to the VTA were computer-generated with a 6711 PCI card (National Instruments, Austin, TX, USA) and were optically isolated from the electrochemical system (NL 800 A, Neurolog, Digitimer Ltd, Hertfordshire, UK). Electrical stimulation for phasic evoked DA release consisted of biphasic square wave pulses (300 μA, 24 pulses, with a 4 ms pulse width, applied at 60 Hz). For tonic-evoked DA release, all parameters remained the same except that the frequency was changed to 5 Hz. VTA stimulation, and subsequent dopamine current recordings detected by electrode in the NAc, were performed with a 3 min interval between each stimulation to allow releasable DA stores to baseline. For the FSCV experiment, at baseline, 6 evoked-DA samples were first collected.

Whole-cell patch clamp recording

Mice were sacrificed at postnatal 38–40 days, coronal slices containing the VTA were used for recording miniature excitatory postsynaptic current (mEPSC). Slices were immersed in ice-cold ACSF saturated with 95% O2 and 5% CO2 and incubated at 32 °C for 30 min before being kept at room temperature. The slices were then switched to the recording chamber circulated with ACSF. The recording electrode contains a potassium-based solution (in mM: 130 K-gluconate, 4 KCl, 2 NaCl, 10 HEPES, 4 ATP-Mg, 0.3 GTP-Na, 1 EGTA, and 14 phosphocreatine, pH 7.2, 295 mOsm), and has an electrical resistance of 4 ~ 6 MΩ. Neurons were voltage-clamped at −70 mV. VTA neurons were patch clamped with an electrode containing a potassium-based internal solution that contains 0.25% biocytin.

Morphological reconstruction and analysis of VTA TH+ neurons

When the whole-cell patch clamp recordings were completed, the cell received 500 pA of current to allow the exchange of the cell fluid and electrode fluid, and biotin in the electrode fluid was exchanged into the cell for cell morphology construction. ACSF was discarded, the recorded brain slices were transferred to 24-well plates containing 4% PFA at 4 °C overnight, then rinsed by PBS for 3 times and 10 min for each. The brain slices were blocked in blocking buffer (5% donkey serum and 5% goat serum in PBS containing 0.3% TritonX-100) at room temperature for 1 h and stained with the primary antibody diluted in blocking buffer at 4 °C overnight. After 3 times rinse by PBS for 10 min each, slices were incubated in mixed secondary antibody, including both sheep-rabbit immunofluorescence secondary antibody and anti-biotin avidin coupled secondary antibody, at 4 °C overnight, followed by rinsed by PBS. Brain slices were visualized under laser scanning confocal microscope and dendritic spines of TH positive dopamine neurons with biotin staining were analyzed. Mushroom and stubyby protrusion from the dendritic branch were identified as spine. At least three dendrite segments in each neuron were included in analysis, each analyzed dot represented one neuron. Data came from more than 8 mice in each group.

Single Cell RNA-Seq (scRNA-Seq)

WT and KI mouse midbrain tissues were obtained on embryonic day 13.5, and subjected to single-cell RNA-Seq. To achieve a homogenous single-cell suspension, the Worthington Papain Dissociation System kit (Worthington, LK003153) was used following the manufacturer’s recommended protocol. Briefly, the midbrain tissues were washed with DMEM/F12 and then dissected into small pieces in Papain/DNase solution. The tissue suspension was subsequently transferred to an orbital shaker set at 60 rpm/min and maintained at 37 °C for 30 min. Gentle trituration was performed to obtain a single-cell suspension, and the papain was inactivated using an albumin-ovomucoid inhibitor solution. Cells were pelleted at 300 g, resuspended in PBS containing 0.04% BSA (Sigma, A2153), and adjusted to a concentration of 500–1000 cells/μl.

For the scRNA-Seq, a 3′-library was constructed utilizing the Chromium Single Cell Reagent Kits v3 (10X GENOMICS) according to the manufacturer’s user guide. The amplified and purified library was quantified using the Quant-iT dsDNA Assay Kit, high sensitivity (Thermo Fisher) on Qubit 2.0, and subsequently sequenced with PE 150 (paired-end sequencing, 150 bp reads).

To align the clean reads obtained from the scRNA-Seq to the mouse reference genome (mm10), the STAR (v.2.7) software was employed. The resulting BAM files were sorted and indexed using the sort and index functions provided by SAMtools (v1.9). To visualize the aligned reads, the BAM files were viewed using the Integrative Genomics Viewer (IGV).

SMART-Seq single-cell data analysis

To assess expression of SHANK2 isoform in the midbrain, we re-analyzed SMART-Seq single-cell data of the human ventral midbrain from La Manno et al. [16]. First, the single cell RNA-seq reads were aligned to the human reference genome (hg38) using Hisat2 (version 2.1.0). The output of hisat2 is a SAM file. Then we transformed from SAM files to BAM files and sort the BAM files using Samtools (version 1.9). Quantification of isoform expression was performed using StringTie (v1.3.3b) with GENCODE annotation (human release 31). We extracted read count information directly from the files generated by StringTie using prepDE.py mentioned in the Stringtie manual. The count value of Transcript for each cell was normalized using CPM (Counts Per Million) and then log transformed.

We acquired processed data including the gene counts matrix and cell meta data. UMI counts were normalized for each cell by the total expression, multiplied by 10,000 scaling factors, and log-transformed. The highly-variable genes (HVGs) were identified using the function highly_variable_genes with min_mean = 0.0125, max_mean = 3, min_disp = 0.5. The scale function was used to regress out variation due to differences in total UMIs per cell. We chose terminally differentiated cell types (hDA1, hDA, hEndo, hGaba, hMgl, hOMTN, hOPC, hPeric, hSert) for analysis. To reduce noise and the gene expression space further, we then performed principal component analysis (PCA) on the scaled data for the variable genes. The neighborhood graph of cells was computed using the top 30 PCA vectors space, and the number of neighboring data points was set to 10. t-SNE (t-distributed Stochastic Neighbor Embedding) was used for visualization.

Finally, we employed heatmaps to visualize the expression of marker genes across different cell types, as well as the three transcripts of SHANK2: SHANK2A, SHANK2B and SHANK2E within each cell type. Gene expression values and isoform expression values were standardized between 0 and 1.

Statistical analysis

Except RNA-Seq, all statistical analyses were performed using GraphPad Prism8 (USA). All data presented in the study are mean ± standard error (SEM) unless otherwise noted. No data were excluded from this study. Data were firstly tested for normality. If not otherwise indicated, parametric data were compared using unpaired Student’s t-test, Two-way ANOVA followed by Šídák’s multiple comparisons test, or One-way ANOVA followed by Tukey’s post-hoc comparison test. Non-parametric data were compared using Mann–Whitney test and Wilcoxon test. All reported p values are two-sided. All parametric tests were made based on the non-equal variance assumption. p < 0.05 was regarded as significance. BaseScope on cell line, behavioral tests and molecular assays on animals were analyzed by persons who were blinded to the grouping. Other analysis were analyzed by persons who were not blinded to the grouping.