Subject selection and characterization, identification of de novo mutations
We used cell lines derived from 4 human subjects (SZ-HU-PROB, SZ-HU-MO, SZ-HU-FA, and 62F), a schizophrenia patient and his parents, i.e. a case-parent trio, and an unrelated healthy control (Table 1 and Figure 1a) for all experiments. (The patient’s two siblings, an unaffected brother and a sister diagnosed with bipolar affective disorder, were not included neither in the genetic analysis, nor the subsequent cell reprogramming.) The patient diagnosed with schizophrenia was selected from 16 similar trios, based on de novo mutations (DNMs) identified by exome sequencing, i.e. single nucleotide variants found in the patient, but not in the parents. The ExomeSeq analysis was carried out after 100 bp paired ended sequencing that was run on the Illumina HiScan(TM)SQ platform and resulted in 50 M reads on average per sample. Results were analyzed by a standard bioinformatics pipeline. The identified DNMs were validated by Sanger sequencing. The schizophrenia patient (SZ-HU-PROB) is a carrier of 3 missense DNMs in genes KHSRP (19:6416869C>A), LRRC7 (1:70505093G>A), and KIR2DL1 (19:55286658A>T). Of these, SIFT mutation algorithm predicts the KHSRP mutation as deleterious (Figures 1b and Table 2). None of these DNMs have been reported previously to the ClinVar database (25).
Table 1. Demographic and clinical data of the investigated case-parent trio, the extended family, and the unrelated healthy control individual.
Subject
|
Sex
|
Age
|
Medical history
|
Reprogramming method
|
Code
|
Father
|
M
|
59
|
No psychiatric treatment or other major somatic disorder.
|
Sendai virus reprogramming from blood PBMCs.
|
IPSC-SZ-HU-FA
|
Mother
|
F
|
55
|
No psychiatric treatment or other major somatic disorder.
|
Sendai virus reprogramming from blood PBMCs.
|
IPSC-SZ-HU-MO
|
Proband (son)
|
M
|
24
|
Diagnosed with schizophrenia at the age of 17. During the past 10 years had 7 hospitalizations, receives clozapine treatment. Predominantly negative symptoms, as measured by PANSS.
|
Sendai virus reprogramming from blood PBMCs.
|
IPSC-SZ-HU-PROB
|
Unaffected older sibling
|
M
|
28
|
No psychiatric treatment or other major somatic disorder.
|
-
|
-
|
Younger sibling
|
F
|
21
|
Diagnosed with bipolar affective disorder at age of 18 after suicidal attempt. Receives lithium and olanzapine treatment.
|
-
|
-
|
Unrelated healthy control
|
M
|
35
|
No psychiatric treatment or other major somatic disorder.
|
Sendai virus reprogramming from skin fibroblasts.
|
IPSC-62F
|
Table 2. Description of the identified DNMs in the schizophrenia proband.
Genomic position
|
1:70505093G>A
|
19:6416869C>A
|
19:55286658A>T
|
Gene
|
LRRC7
|
KHSRP
|
KIR2DL1
|
Variation type
|
missense
|
missense
|
missense
|
Amino acid change
|
Val1158Ile
|
Gly403Cys
|
Thr138Ser
|
Conservation of nucleotide
|
weak
|
high
|
not conserved
|
Conservation of amino acid
|
high
|
moderate
|
weak
|
SIFT
|
tolerated (score: 0.37, median: 4.32)
|
deleterious (score: 0.03, median: 3.54)
|
tolerated (score: 0.73, median: 3.01)
|
Mutation taster
|
polymorphism (p-value: 0.999)
|
disease causing (p-value: 0.996)
|
polymorphism (p-value: 1)
|
Generation and characterization of iPSC lines
Blood samples were obtained from trio members after written informed consent. The reprogramming and the study were approved by the Human Reproduction Committee of the Hungarian Health Science Council (ETT HRB). Blood was collected directly to cell preparation tubes with sodium heparin (BD Vacutainer CPT) to isolate peripheral mononuclear cells (PBMCs) from all samples. PBMCs were cultured for 4 days with daily medium changes at a density of 5 × 105 cells/ml in StemPro®-34 (Thermo Fisher) hematopoietic medium supplemented with 2 mM L-Glutamine and cytokines at the following final concentrations (SCF 100 ng/mL, FLT-3 100 ng/mL, IL-3 20 ng/mL, IL-6 20 ng/mL, all from Peprotech). On day 4, PBMCs were transduced with Sendai virus (Fusaki et al., 2009) particles (Thermo Fisher Cytotune 2.0) carrying KOS (hKlf4, hOct3/4, hSox2), hc-Myc, and hKlf4 at MOIs of 5, 5 and 3, respectively. After addition of the virus particles, samples were centrifuged in a 12 well plate for 90 minutes at 2250 rpm and incubated at 37°C overnight. After changing the medium, the cells were maintained for additional 2 days, then transferred to culturing dishes previously seeded with mouse embryonic fibroblasts (MEFs) and cultured in StemPro®-34 medium without cytokines. Six days after transduction, the culturing medium over the cells was gradually changed to HUES medium (KO-DMEM, supplemented with 15% KO Serum Replacement (Thermo Fisher), 100 mM glutamine, 1% nonessential amino acids, 0.1 mM β-Mercaptoethanol and 4 ng/ml recombinant human basic fibroblast growth factor, bFGF).
14-18 days after transduction, individual iPSC colonies emerging were mechanically isolated and transferred to MEFs to generate clones. To ensure virus clearance and monitor stability, the clones were repeatedly passaged and expanded up to p10 using trypsin. Heat treatment at 38.5°C was used between passage p4-p8 to take advantage of the heat sensitivity of virus particles. iPSC 62F was generated from fibroblasts using Sendai virus reprogramming, independently from the other trio iPSC lines.
The pluripotent state and spontaneous differentiation capacity of iPSCs were tested by real-time quantitative PCR and ICC (Figure 1 c,d) as described previously (26). The investigated KHSRP and LRRC7 DNMs were cross-checked in the iPSCs with Sanger sequencing (Figure 1e).
Cell culturing and neuronal differentiation
Neural progenitor cells (NPCs) were differentiated from iPSC lines SZ-HU-PROB, SZ-HU-MO, SZ-HU-FA, and 62F (unrelated healthy control) as described previously (27, 28). Briefly, before starting differentiation, iPSC cells were transferred to Matrigel (Corning, New York, USA) coated plates in mTeSR medium (Stemcell Technologies, Vancouver, Canada) and were cultured to high density. On day 1 the cells were detached with collagenase (Thermo Fisher Scientific, Massachusetts, USA) and transferred to ultra-low attachment plates (Nalgene Nunc International, New York, USA). After embryoid body (EB) formation, the medium was changed to DMEM/F-12, GlutaMAX™ (Thermo Fisher Scientific, Massachusetts, USA) medium supplemented with N2/B27 (Thermo Fisher Scientific, Massachusetts, USA) and anticaudalizing agents (Noggin (Thermo Fisher Scientific, Massachusetts, USA), DKK1 (PeproTech, New Jersey, USA), Cyclopamine (Merck, Darmstadt, Germany), and SB431542 (Sigma, Missouri, USA)). The treatment was repeated every other day. On day 20 the EBs were moved to a poly-ornithine (Sigma, Missouri, USA)/laminin (Thermo Fisher Scientific, Massachusetts, USA) coated plate to support further differentiation in adherent conditions. On Day 27 or later, manually picked rosettes were dissociated by Accutase (Thermo Fisher Scientific, Massachusetts, USA) and re-seeded onto a new poly-ornithine/laminin coated plate in DMEM/F-12, GlutaMAX™ N2/B27 medium containing FGF2 (Thermo Fisher Scientific, Massachusetts, USA) and laminin. The attached neural progenitor cells (NPCs) showed uniform morphology after 5 passages. The NPCs between passage p5 and p15 were used for the experiments. For neuronal differentiation experiments NPCs were further differentiated NPCs were seeded onto poly-ornithine/laminin coated, in eight-well Nunc Lab-Tek II Chambered Coverglass (Nalgene Nunc International, New York, USA) with 1.5x103 density in N2/B27 medium supplemented with ascorbic acid (Sigma, Missouri, USA), BDNF (PeproTech, New Jersey, USA), cAMP (Sigma, Missouri, USA), Laminin and Wnt3A (Research and Diagnostic Systems Inc, Minnesota, USA). After 3 weeks Wnt3A was omitted from the medium. The medium was changed every other day.
Molecular characterization of neuronal progenitors and neurons
For immunofluorescence staining of NPCs and NPC-derived neurons from the case-parent trio and the unrelated control cell line, NPCs were seeded onto poly-ornithine/laminin coated, eight-well chambers, and differentiated into DG neurons for 5 weeks as described above. The cells were fixed with 4% paraformaldehyde (Thermo Fisher Scientific, Waltham, Massachusetts, USA) in Dulbecco's modified PBS (DPBS) (Sigma, Missouri, USA) for 15 min at room temperature. Following washing with DPBS, the samples were blocked for 1h at room temperature in DPBS containing 2 mg/ml bovine serum albumin (Sigma, Missouri, USA), 1% fish gelatin (Sigma, Missouri, USA), 5% goat serum (Sigma, Missouri, USA) and 0.1% Triton-X 100 (Sigma, Missouri, USA). The samples were then incubated for 1h at room temperature with the following antibodies: anti-SOX2 (monoclonal/mouse, 1:20 dilution; MAB2018, R&D Systems, Minneapolis, USA), anti-Nestin (polyclonal/rabbit, 1:250 dilution; ab92391, Abcam, Cambridge, UK) or for overnight at 4oC with the following antibodies: anti-PROX1 (polyclonal/rabbit, 1:500 dilution; ab101851, Abcam, Cambridge, UK), anti-MAP2 (monoclonal/mouse, 1:500 dilution; M1406, Sigma/Merck, Darmstadt, Germany or polyclonal/rabbit, 1:1000 dilution; ab5622, Millipore, Massachusetts, USA). The proteins encoded by the genes of interest harboring DNMs were also investigated by immunofluorescence staining, using anti-KHSRP (1:1000 ab140648, Abcam, Cambridge, UK) and anti-LRRC7 (1:500, HPA005625, Sigma/Merck, Darmstadt, Germany) antibodies. After washing with DPBS, the cells were incubated for 1h at room temperature with appropriate secondary antibodies; Alexa Fluor 633-conjugated goat anti-mouse IgG or Alexa Fluor 547-conjugated goat anti-rabbit IgG (Thermo Fisher Scientific, Massachusetts, USA). The nuclei were counterstained with DAPI (Thermo Fisher Scientific, Massachusetts, USA). The stained samples were examined by a Zeiss LSM 900 confocal laser scanning microscope. Fluorescence images were analyzed with the ZEN 3.1 blue edition software. In the confocal images pseudo-color coding was used for better visualization.
Gene expression analysis
Total RNA was isolated from NPCs and NPC-derived neurons using TRIzol™ reagent following the manufacturer’s instructions (Thermo Fisher Scientific, Massachusetts, USA). cDNA samples were prepared from 0.2 μg total RNA using the Promega Reverse Transcription System Kit (Promega, Wisconsin, USA) as specified by the manufacturer. For real-time quantitative PCR the following Pre-Developed TaqMan® assays were purchased from Thermo Fisher Scientific, Massachusetts, USA: NANOG as undifferentiated stem cell marker; Pax6 , and SOX2 as markers of NPC state, as well as NeuroD1, FoxG1, and PROX1, GRIA1 as neuronal differentiation specific markers; P0 ribosomal protein as endogenous control for quantification. KHSRP and LRRC7 expression were examined as well. RT-PCR analyses were carried out using the StepOnePlus™ Real-Time PCR System (Thermo Fisher Scientific, Massachusetts, USA), according to the manufacturer’s instructions. The changes in mRNA levels between the examined and control cells were determined by the 2-ΔCt method using RPLP0 (P0) as endogenous control gene. Relative mRNA levels were presented as heatmaps using the mean values of 3 independent experiments.
RNA-Sequencing experiments
RNASeq experiments were carried out at the NPC stage for each cell line in quadruplicates. RNA was isolated at 4 different passages from each NPC line between p7-12 to ensure biological replicates. After library preparation samples were sequenced with the Illumina HiSeq technology using 2x150 bp paired-end reads resulting in 40 million read-pairs (12 Gb sequence) coverage per sample.
Mapping to the Homo sapiens (hg19) reference genome was performed using Hisat2 software, after quality control of raw reads with FASTQC and trimming with Trimgalore. Mapped reads were annotated to genes using featureCounts function of the Subread R package. All X and Y chromosome genes were excluded from downstream analyses. To compare transcriptomic differences we performed differential gene expression (DE) analysis, principal component analysis, and cluster analysis using the DESeq2 R package, PCAExplorer, Cluster 3.0, and Treeview, respectively. Differentially expressed genes between two NPC lines were selected if the logarithmic of fold change was greater than 1 and adjusted p-value smaller than 0.1. Gene Ontology (GO) and Pathway analysis of DE genes were performed using DAVID. DE genes were significantly enriched in GO and PATHWAY terms when their FDR was less than 0.1.
Ca2+ signal measurements in NPC cultures
Before the Ca2+ measurements NPCs were seeded for two days onto eight-well chambers previously coated with poly-ornithine/laminin. NPCs cells were subjected to 1.0 μM Fluo-4 AM (Thermo Fisher Scientific, Massachusetts, USA) in a serum free culture medium for 30 min at 37oC. Extracellular Fluo-4 AM was removed by changing the medium to Hanks’ balanced salt solution (Thermo Fisher Scientific, Massachusetts, USA) supplemented with 20 mM Hepes (pH=7.4) (Thermo Fisher Scientific, Massachusetts, USA) and 0.9 mM MgCl2 (Sigma, Missouri, USA) (HBSS). All measurements were performed in HBSS at room temperature. The ligand concentrations were chosen according to literature: glutamate (50 μM) (Sigma, Missouri, USA), KCl (50 mM) and ionomycin (5 μM) (Thermo Fisher Scientific, Massachusetts, USA). All experiments were performed at room temperature.
Ca2+ signal measurements were carried out by acquisition of time lapse sequences of fluorescence images with an Olympus laser scanning confocal microscope and using the FluoViewTiempo (v4.3, Olympus, http://www.olympusmicro.com) software as described earlier (29). Fluorescence images were acquired between 505 and 525 nm at 488 nm excitation. Image analysis was carried out with the same software. The cellular fluorescence signals measured in regions of interest (ROIs) placed over the cells were converted to percentage of activation based on (F-Fmin)/(Fmax-Fmin) *100, where F is the Ca2+ signal intensity at a given time point, Fmin is the minimum of the first fifteen time points and Fmax is the maximum after the administration of ionomycin. For further analysis we used those cells, which reached at least 10% activity after glutamate administration. We used Chi-square statistics to compare the percentage of activated cells between subjects. Furthermore, in activated cells we derived F/F0, where F0 is the minimum of baseline activity before glutamate administration (5 time points), and F is the maximum Ca2+ signal intensity of 50 time points after glutamate administration. Data are means (±SD) of all the individual cells (typically 150-250 cells for each subject and each parallel experiment). For statistical comparison between subjects (mother, father and patients with schizophrenia) we conducted ANOVA for the dependent variable (F/F0) and used the 3 parallel experiments as covariates. Statistical analysis was conducted with the SPSS software.
Functional phenotyping of NPC lines (cell proliferation, neuronal migration, neurite outgrowth)
Based on the RNASeq results, we performed targeted functional assessment of NPCs, namely cell proliferation, neuronal migration, and neurite outgrowth tests. For the 96-hour proliferation assays, 35.000 NPCs were plated per well in a 24-well plate using triplicates for each time point. NPCs were harvested and labeled with the viability marker propidium iodide then measured in a total volume of 120 µl by Attune flow cytometry at 24 and 48 hours after. For assessment of NPC migration, we used the scratch or „wound healing” test, as described earlier (30). 450.000 NPCs were seeded onto poly-ornithine/laminin coated, six-well plates. After reaching confluence scars were inflicted manually in each well in triplicates. NPC migration was measured at 24 and 48 hours on bright field micrographs. For neurite outgrowth experiments we used NPCs treated with calcein (10mM). 3.500 NPC-s were seeded per well in 96-well plates, after 2 hours the cells were treated with para-nitroblebbistatin (10mM), a known inductor of neurite outgrowth (31) or DMSO. Neurite outgrowth was visualized and measured by fluorescence microscopy using the automated image acquisition and analysis system of the ImageXpress High content screening system. Data were evaluated and significance levels determined using the GraphPad Prism 8.3.0 software.
Oxidative stress and mitochondrial tests
Oxidative stress tolerance was investigated in NPC lines by treatment with CoCl2, a well-known hypoxia inducer and reoxygenation (32). Cell lines were plated at a number of 35.000 cells/well in poly-ornithine/laminin coated 48-well plates using triplicates. Medium was replaced after reaching confluence with CoCl2 containing media at two different concentrations (control=0µM, 125µM, 250 µM). After 24 hours, the medium was changed to normal medium (reoxygenation). Two days after reoxygenation, viability was measured using the PrestoBlue dye and the fluorescent signal intensity was measured by Enspire Multimode plate reader (Perkin Elmer). NPCs were characterized by their baseline levels of reactive oxygen species (ROS) using the CellROX kit (Thermo Fischer Scientific, Cat. Number: C10444). NPCs were dissociated by accutase and then 200.000 cells were incubated/treated with CellROX at 500x dilution for 30 minutes at 37°C. After the incubation period cells were washed three times with 1x PBS. Subsequently, the mean fluorescence signal of labeled cells was measured and compared by flow cytometry (Attune NxT Flow Cytometer, ThermoFisher), and the dead cells were gated out by using propidium iodide staining.
Mitochondrial function in NPCs was examined by Mitotracker staining and transmission electron microscopy. The Mitotracker Red dye (MitoTracker ™ Red CMXRos, cat. Number: M7512) was applied on NPC cultures after reaching confluence at a concentration of 250 nM. After 30 min incubation at 37°C the dye was washed with DPBS, and the NPCs were fixed by 4% PFA for 15 minutes at RT. Fluorescent signals were acquired by confocal microscopy (Zeiss LSM 900). Images were evaluated with the NIH Image J software by comparing total fluorescence intensity per nucleus. For transmission electron microscopy, briefly, NPCs were dissociated manually or enzymatically with Accutase, the samples were fixed with solution containing 3.2% PFA, 0.2% glutaraldehyde, 1% sucrose, 40mM CaCl2 in 0.1M cacodylate buffer. Next, samples were post-fixed with 1% ferrocyanide-reduced osmium tetroxide and dehydrated using a graded ethanol series, and then embedded in Spurr low viscosity epoxy resin medium. Ultrathin sections were collected on Formvar (Agar Sci., Essex, UK) coated copper slot grids, counterstained with uranyl acetate and Reynolds's lead citrate, and examined on a JEOL JEM 1011 transmission electron microscope equipped with a Morada 11-megapixel camera (Olympus) using iTEM software.
Statistical analysis
All experiments were carried out in triplicates. Values are presented as mean ± SD for all experimental data. Comparisons between groups were performed by using one-way ANOVA followed by Dunn’s post hoc testing. Outliers were removed by Grubb’s test. The P-value < 0.05 was considered to indicate a significant difference. If otherwise not indicated, statistical analyses and curve-fitting were performed in GraphPad Prism 8.3.0 software.