Schizophrenia-associated Mitotic Arrest Deficient-1 (MAD1) regulates the polarity of migrating neurons in the developing neocortex

Although large-scale genome-wide association studies (GWAS) have identified an association between MAD1L1 (Mitotic Arrest Deficient-1 Like 1) and the pathology of schizophrenia, the molecular mechanisms underlying this association remain unclear. In the present study, we aimed to address these mechanisms by examining the role of MAD1 (the gene product of MAD1L1) in key neurodevelopmental processes in mice and human organoids. Our findings indicated that MAD1 is highly expressed during active cortical development and that MAD1 deficiency leads to impairments in neuronal migration and neurite outgrowth. We also observed that MAD1 is localized to the Golgi apparatus and regulates vesicular trafficking from the Golgi apparatus to the plasma membrane, which is required for the growth and polarity of migrating neurons. In this process, MAD1 physically interacts and collaborates with the kinesin-like protein KIFC3 (kinesin family member C3) to regulate the morphology of the Golgi apparatus and neuronal polarity, thereby ensuring proper neuronal migration and differentiation. Consequently, our findings indicate that MAD1 is an essential regulator of neuronal development and that alterations in MAD1 may underlie schizophrenia pathobiology.


RT-PCR and quantitative real-time PCR
For semi-quantitative RT-PCR, C57BL/6 mice brains (E10, E12, E14, E16, E18, P1, and adults) were isolated, and total RNAs were extracted with TRI-Solution Kit™ (Bio Science Technology) following the manufacturer's instructions. cDNA was synthesized with ImProm-II™ Reverse Transcription System (A3800) by manufacturer's guide from 1 μg RNA of each group. Equal amounts of synthesized cDNAs were used for quantitative comparison. GAPDH (Glyceraldehyde-3-phosphate dehydrogenase) was used as an internal control. The forward and reverse primer sequences for semi-quantitative RT-PCR were 5'-CTGCGGGAACAGGAGGACAG-3' and 5'-ACTGCAGCTCCGAGACACGC-3' for mouse 3

Cell line and primary culture of neurons
HEK293 and HeLa cells were cultured with DMEM (Welgene) supplemented with 10% (v/v) FBS (Fetal Bovine Serum, Gibco) and 1% Antibiotic-Antimycotic (Gibco). Cell lines were authenticated by the STR profiling method. Cell lines were tested as negative for mycoplasma contamination.
For primary neuron cultures, pregnant mice were purchased from Hyochang Science (Daegu, Korea).
Pregnant mice were sacrificed, and E15~E16 embryos were used for culture. Isolated cortical or hippocampal tissues were kept in HBSS (Gibco) on ice. Tissues were dissected with 0.25% trypsin

Immunoblotting and immunoprecipitation
Mouse brain tissues or transfected cells were washed with 1X PBS and lysed with 1X erythrocyte lysis buffer (50 mM Tris, pH 8.0, 250 mM NaCl, 0.1% NP-40, 5 mM EDTA, 2mM sodium pyrophosphate, 5 mM NaF, 2 mM Na3VO4, 1mM DTT, and protease inhibitor cocktail (Roche)). Sonicator was used for complete lysis. For immunoprecipitation, 1~2 μg of antibody was added to the lysate and incubated at 4℃ overnight on a rotator. 20 μL of Protein-A agarose beads (Roche) was added and incubated for 2~3 hr at 4℃ on a rotator. The mixtures were washed 3 to 5 times with lysis buffer. For immunoblotting, 5X SDS sampling buffer (2% SDS, 60 mM Tris pH 6.8, 24% glycerol, 0.1% bromophenol blue, 5% βmercaptoethanol) was added, and the mixtures were boiled at 100℃ for 10 min. SDS-PAGE was performed with 8 or 10% polyacrylamide gel in an electrophoresis tank (Bio-Rad) and transferred to PVDF membrane (Millipore). Transferred membranes were blocked with 5% skim milk in Tris-buffered saline (20 mM Tris pH 8.0 and 137.5 mM NaCl) with 0.25% Tween20 (TBST) for 0.5~1 hr at room temperature (RT). The primary antibody was mixed with skim milk followed by the manufacturer's guide, and membranes were incubated at 4℃ overnight. The membranes were further incubated with HRPconjugated secondary antibodies for 1 hr at RT, and protein signals were detected by ECL solutions (Bio-Rad) in the image analyzer (Azure Biosystems). For immunohistochemistry (IHC), harvested brains were washed with PBS and fixed with 4% paraformaldehyde and 4% sucrose in PBS overnight, followed by sequential changes of sucrose concentration, 10% in PBS for one day, 20% in PBS for one day, and 30% in PBS for one day. Brains were frozen with OCT solution (Leica Biosystems) and sectioned using cryostats (Leica Biosystems) with 10~50 μm thickness depending on the experimental purpose. Then, dissected tissue was fixed to Superfrost Plus microscope slides (Thermo Fisher Scientific) and dried at least for 6 hr at RT. Dissected tissues were examined with fluorescent microscopy to confirm the electroporated cells with fluorescence. Dried samples were washed with PBS and permeabilized with 0.2% Triton X-100 in PBS for 10 min. The antigen retrieval process was followed at 95℃, 10 min with citrate buffer of pH 6.0.

Immunocytochemistry and immunohistochemistry
Blocking was conducted with CAS-Block histochemical reagent (Thermo Fisher Scientific) for 1 hr at 6 RT. Samples were incubated with primary antibody mixtures at 4℃ overnight and incubated with secondary antibody mixtures for 1~2 hr at RT. For nuclear staining, samples were incubated with Hoechst solution. Finally, coverslips were placed on slide glasses and fixed with mounting media (Biomeda). Images of brain tissue samples were acquired using confocal microscopy with z-stacks of 0.5~1 μm intervals after confirming the electroporated regions by fluorescence signals. Each z-stack image and a merged z-projection image were compared to discriminate the cell periphery.
Images of brain samples were acquired using confocal microscopy with z-stacks of 0.5~1 μm intervals after examining the electroporated regions by fluorescence signals. Only transfected neurons marked by fluorescence were analyzed. Each z-stack image and a merged z-projection image were compared to discriminate the cell periphery. To minimize the sample-by-sample variation, we analyzed the samples in the comparable tissue quality, transfected region, and total transfection efficiency in each experimental set.

Neurite tracing and Sholl analysis
Primary neurons were transfected with shCTL or shRNA cloned into the pLL3.7 vector to induce knockdown. All shRNAs, including scrambled shRNA (shCTL), were cloned into the pLL3.7-EGFP or

VSVG Trafficking and Golgi morphology analysis
The temperature-sensitive mutant version of vesicular stomatitis virus G protein (VSVG) was cloned into pcDNA3.1-myc/His [6]. HeLa cells were transfected with shCTL or MAD1 shRNA on the first day.
On the second day, the VSVG construct was transfected. After 24 hr, cells were incubated at a non-8 permissive temperature (40℃) overnight to induce accumulation of unfolded VSVG proteins in ER. Cells were incubated at a permissive temperature (32℃) to induce sequential membrane trafficking from ER to Golgi and Golgi to plasma membrane.
For the analysis of post-Golgi membrane trafficking in neurons or HeLa cells, cells were incubated at 40℃ overnight after 12 hr of VSVG transfection. Cells were incubated at 20℃ for 3 hr to induce Golgi accumulation of VSVG proteins, followed by incubation at 32℃ to induce trafficking from Golgi to the plasma membrane. Cells were fixed with 4% paraformaldehyde for 10 min and blocked with PBS supplemented with 10% FBS and 0.04% sodium azide for 1 hr or overnight at 4℃. For immunostaining, antibody solution was incubated in PBS mixed with 10% FBS, 0.2% saponin, and 0.04% sodium azide.
As Golgi markers, α-Giantin was used for HeLa cells, and α-GM130 was used for mouse neurons.
Imaging was conducted using confocal microscopy (Olympus FV3000). The acquired image was analyzed using CellSens software (Olympus) to measure Mander's overlap coefficient.
For the analysis of Golgi morphology, neurons were transfected as in the neurite outgrowth assay. All shRNAs, including scrambled shRNA (shCTL), were cloned into pLL3.7-EGFP or pLL3.7-mRFP vector to discriminate transfection and cell morphology by fluorescence. On DIV3, transfection efficiency and cell condition were assessed. Cells were immunostained with human cis-Golgi marker Giantin, mouse cis-Golgi marker GM130, or mouse trans-Golgi marker TGN38. Images were acquired by z-stack from top to bottom of each cell with a 60x or 100x lens confocal microscopy. Cells were discriminated by comparing individual stack images with z-projection images. Transfected cells were discriminated by EGFP-positive signal and chosen for the imaging.
Cells were classified by morphological traits and position of Golgi traced by the Golgi marker. To exclude small vesicles nonspecifically labeled, significantly large size particles with > 2μm diameter were counted. Images were compared in the same intensity, brightness, and contrast. Golgi stacks with a continuous structure were classified as a "normal" group, and Golgi complexes dispersed as particles are classified as a "fragmented" group. In neurons, Golgi stacks are classified as "dendritic" or "nondendritic" by positioning of major Golgi fractions on the nucleus-dendrite axis. All the experiments for trafficking and Golgi morphology analyses in neurons (shCTL, shMAD1, shKIFC3, c-MAD1, shMAD1 + c-MAD1, and shKIFC3 + c-MAD1) were done in the same set for the multiple direct comparisons. All experiments were repeated at least three times. Acquired images were analyzed with Image J software (NCBI) for evaluating neurite outgrowth, neuronal migration, or densitometric quantification of western blot results. CellSens software (Olympus) was used for analyzing colocalization.

Yeast-two-hybrid screening
For bait construct, human MAD1 (hMAD1) was cloned into pPC97 vector containing the DNA-binding domain of GAL4 and Leu gene. Human fetal brain cDNA library in pPC86 vector (GibcoBRL) containing GAL4 activation domain and Trp gene was used as prey. MaV203 yeast strain was used for screening, and transformation was conducted as previously described [7]. Transformed yeasts were incubated at 30℃ for 3 days. A total 3,375,000 co-transformants grown on Leu-/ Trp-medium was initially screened with growth test on synthetic defined (SD) medium (Leu-, Trp-, His-, and Ura-containing 20 mM of 3amino-1,2,4-triazole (3-AT; Sigma-Aldrich)). 46 putative positive colonies were transferred to SD Leu-, Trp-medium for β-galactosidase activity test with X-gal (Sigma-Aldrich) solution. Finally, 15 positive colonies were selected. Prey plasmids were isolated from positive colonies by incubating on Trp-media and prepared with Lyticase (Sigma-Aldrich). Plasmids were amplified by transformation into DH5α and analyzed by DNA sequencing and database (NCBI BLAST).
For electroporation on forebrain organoids, we utilized organoids cultured for 50 days. DNA constructs (2 μg/μl concentration) combined with Fast green were injected into empty spaces of organoid rosettes. The injection was performed with a micro-injector (PLI-100 Pico-Injector, Harvard Apparatus) set as 10 psi and 5~20 msec of injection time. After that, electroporation was performed with an electroporator (Harvard apparatus, Holliston, MA, USA) set as 80 V, 50 msec duration, 1s interval, 5 pulses.