ABSTRACT
Cellular migration is a ubiquitous feature of development that brings brain cells into appropriate spatial relationships. Cortical thinning has been reported in post-mortem brain samples of patients with bipolar disorder (BD). It could be that defective cellular migration during brain development is one of the contributing mechanisms in BD pathogenesis, leading to abnormalities reported at post-mortem, and during brain imaging. To probe the role of cellular migration in BD, we conducted time-lapse analysis of migration of neural precursor cells (NPCs) previously generated in our laboratory. Two NPC lines (B1 and B2) and one control line (C1) were used for the cell migration experiments. Time-lapse images were recorded every 15 min, for 15 hours, and analysed for speed and direction of cellular migration. Abnormalities in cellular migration was a common feature observed in both patient-derived NPCs. Consequently, we investigated underlying mechanistic irregularities in B1 and B2 lines that may contribute to the observed phenotype. To this end, transcriptional changes were compared between the patient-derived cells with the control line. Additionally, we examined exonic variations in genes related to cellular migration or motility. Our analysis showed downregulation of multiple genes that are all part of the EGF/ERBB signaling pathway. Collective dysregulation may be producing the defective cellular phenotype. These cellular migration abnormalities may be linked to structural changes in the brain reported in bipolar disorder.
Introduction
The precise cytoarchitecture of the brain is vital for its functioning. One of the processes that makes the unique structure of the human brain possible is cellular migration (Yun kee et al., 2007, Metin et al., 2007). During brain development, neurons are guided to their final destinations by crawling along glia which acts as a cellular guide (Nadarajah and Parnavelas, 2002). The migrating neurons later mature and integrate into the neuronal circuits which eventually helps in the development of the brain. Defects in neuronal migration could contribute to neurodevelopmental disorders like bipolar disorder (BD) (Uribe and Wix., 2012).
The World Health Organization has ranked BD among the leading causes of disease-related morbidity and lifetime disabilities (Alonso et al., 2011; Tondo et al., 2016). People with BD are at a very high risk of death by suicide, more than 20 times higher than in the general population (Schaffer et al., 2015). A smaller brain size and reduced cortical gray and white matter was also reported in BD patients (Vonk et al., 2014). This is further supported by the MRI studies that have revealed structural brain abnormalities in the cortical region with reduction in white matter volume (Ching R.K. et al., 2020; De Peri et al., 2012). Postmortem analysis shows cortical thinning (Hibar et al., 2018; Knöchel et al., 2016), and decreased interneurons in the cerebral cortex and hippocampus (Rajkowska et al., 2001). This may be due to abnormalities in proliferation, differentiation, or migration of the neuronal cells. We have specifically investigated the role of cellular migration in BD using patient derived induced pluripotent stem cells (IPSCs).
IPSCs derived from patients have been used as a model for neuropsychiatric disorders like schizophrenia, Timothy syndrome, autism, Rett syndrome and BD (Noh et al., 2017; Yazawa and Dolmetsch, 2013; Marchetto et al., 2010; O’Shea and McInnis, 2015, 2016).Our previous studies in the same model system have revealed the involvement of mitochondrial-, cell cycle- and cell death-related phenotypes in neural precursor cells (NPCs) derived from BD patients (Paul et al. 2020). Here, we report that these patient NPCs also exhibit abnormal cell migration. Further, we identify specific molecular targets and pathways related to this cellular abnormality.
Materials and methods
Details of clinical recruitment (Viswanath et al. 2018), generation of NPCs (Mukherjee et al., 2019) and cellular characterization (Paul et al. 2020) have been described in earlier publications. The study protocol has been approved by the ethics committee of the National Institute of Mental Health and Neurosciences, Bengaluru, India. Clinical protocol for this study conforms to the ethical norms and standards in the Declaration of Helsinki. The NPCs from BD-patient lines (B1, B2) (Paul et al., 2020) and 1 control line (C1) were used for experiments reported here.
Exome sequencing
Variants identified from exome sequencing were prioritised by a three step filtering process: prediction of deleteriousness by five in silico algorithms (SIFT, Polyphen-2, HDIV, Mutation taster 2, Mutation assessor and LRT); rarity in south Asian population estimated using the Exome Aggregation Consortium data; and absence of these variants in control individuals from the same genetic background (Ganesh et al.,2018).
Transcriptome data analysis
RNA-Seq was performed on the Illumina® Hi-Seq platform. Genes which showed >1-fold difference with FDR adjusted P-value < 0.05 were considered differentially expressed (Paul et al., 2020). Subsequently, we aimed to identify genes among these that are related to cellular migration by comparing the data with the ConsensusDB for the processes related to migration or motility. In addition, the data from published references on genes related migration were included in our study (Bavamian S et al., 2015, Madison JM et al., 2015, Buchsbaum and Cappello., 2019). From the analysis, 60 and 58 genes specific to cellular migration were identified in B1 and B2 lines respectively. A Venn diagram was plotted (http://bioinformatics.psb.ugent.be/webtools/Venn/) to identify genes (29) that were common in both patient-derived lines, out of which 25 genes showed similar expression changes. An interactome of these 25 genes was created with the STRING database (https://string-db.org/) to identify any functional interaction between expressed proteins.
Neural precursor cell culture
Details of NPC generation and characterization have been published previously (Paul et al. 2020). Care was taken to carry out all experiments with cells at similar densities and passage numbers ranging from P10-P19.
Cell migration assay
For cell migration experiments, cells were seeded at a density of 15000-20000 cells, in Ibidi culture inserts (Cat. no. 81176) placed in two-well Ibidi micro-slides. After cells adhered to the matrigel-coated surface, migration assay was commenced with the removal of inserts. Movement of cells across the 500-µm gap generated by the insert was recorded in time-lapse images every 15 min, for 15 hours with Olympus microscope equipped with a camera (Hamamatsu) and using a 10X dry objective. The experiments were done in biological triplicates. The Ibidi dishes with cells were incubated at 37°C in a humidified incubator, with 5% CO2. The images were analyzed for speed and direction of cellular migration using Image J (Version IJ1.46r) with the plugin ‘Manual Tracking’.
Mathematical modelling of NPC migration
The path of the cells migrated for the NPCs were tracked with the help of X and Y coordinates from the images taken. From the trajectories of each cell (i), we estimated the displacement d between the initial location and final location (after 15 hours). Using these measures, the box-plot of displacements of 100 cells for each cell line were plotted.
From the trajectories of each cell i, we also computed the mean-squared displacement (MSD) as a function of time separation between any two locations in the cell trajectory. The Levy exponent, which determines the functional relationship between MSD and the time lag (tau), given by MSD (tau) ∼ tau^alpha, was also calculated. When alpha = 1, the cells are performing the classic random walk or the Brownian motion. A value of alpha = 2 implies that cells are performing a highly directed or ballistic motion.
Results
Aberrant cellular migration in patient-derived neural precursors
Assessment of migration of NPCs from B1 and B2 line showed an aberrant migration pattern and higher speed when compared to the C1 control line. While the control cells showed directed migration patterns, with an average velocity of 0.045 um/min, the patient lines showed abnormal migration, with a faster average velocity of 0.095 um/min for B1 and 0.087 um/min for B2 (Fig 1b,c). Most of the control cells migrated in a directed manner; whereas the B1 and B2 line showed three types of migration: cells migrating in a directed pattern, in circular motion and of cells involved a back and forth motion (Fig 1d). The quantification of the distance showed that 75% of control cells migrate less than 50um; whereas the majority of cells from B1 (100%) and B2 (82%) migrate distance greater than 50um. A proportion of cells from B1(35%) and B2 (25%) migrate distances greater than 100um. (Fig 1e). The computational modelling of the migratory path showed the directional pattern of migration for control lines whereas the patient lines showed irregular patterns of migration (Fig 1f).
Abnormal migration phenotype in patient-derived neural precursors
A. Brain MRI data for B1 and B2. B1 show an decrease in brain volume and B2 show an increase in WM volume compared to control Indian brain template (GM-Grey matter, WM-White matter, TIV: Total intracranial volume.)
B. Migration of C1, B1 and B2 cells at different time points. Cells of C1 line alone, close the gap after 12 hours of migration.
C. Comparison on speed of migration for control, B1 and B2 cells. B1 and B2 cells migrate at a speed >0.08um/min***; whereas control cells at 0.045.Â
D. Direction of migration. Control cells migrate in a directional pattern; B1 and B2 cells migrate in different directions. Both BD lines show back and forth movement, directional and circular pattern of migration.
E. Comparison of distance moved by migrating cells. Control cells migrate less than 50um (75%); whereas B1 (100%) and B2 (82%) migrate distance greater than 50um.
F. Computational modelling for migratory path for Control, B1 and B2 lines.
G. Box plots of displacement of the migrating cells after 15 hours time point (n = 100 cells per line)
H. Mean square displacement (MSD) as a function of time for cell migration. Alpha is the slope of the relationship on log-log axis. Values of alpha ∼ 1 correspond to Brownian motion whereas alpha ∼ 2 correspond to ballistic motion (n = 100 cells per line).
The X and Y coordinates were used to analyse the net displacement of the cells over a 15 hour time period. We find that although cell lines B1 and B2 show larger variation among cells in the net displacement, there is no statistically significant difference with the control line (Fig 1g). Furthermore, the mean square displacement for the three cell lines was calculated; we estimate the value of alpha to be 1.64, which suggests that control cells move at a much more ‘directed’ fashion than random walks. In contrast, we find values of alpha to be 1.19 and 1.29 for the B1 and B2 lines, respectively. This is strongly suggestive that B1/B2 cell lines show a much more jagged unpredictable trajectory (Fig 1h).
Transcriptome analysis revealed differential expression of RNA transcripts related to cell migration
Transcriptomic analysis was carried out to further explore the basis of the differential cellular phenotype. We identified 61 differentially expressed transcripts in B1 and 58 in B2, related to cell migration and motility in comparison with C1. 29 transcripts were commonly dysregulated across both patient lines (Fig 2b,c). Functional interaction between the genes thus shortlisted was explored using STRING (Fig 2d).
RNA-seq analysis shows dysregulated EGF/ERBB signaling pathway
A. Transcriptome analysis pipeline
B. Transcripts common and unique to B1 and B2 patient-derived lines
C. Comparison of fold change for common transcripts among the two patient-derived neural precursor lines.
D. Interactions among proteins shortlisted from ‘C’, to include only those where direction of change of expression was similar in the two patient-derived lines, plotted using STRING online database (https://string-db.org/)
Exome analysis showed LAMA1, RAPFGEF4, PLCG2 and NRG2 as genes related to migration
In order to identify the gene variants related to migration in the B1 and B2 lines we analyzed whole exome exome sequencing data. We identified LAMA1 (rs142063208), RAPFGEF4 (rs777158731) common for B1 and B2; PLCG2 showed two different variants for B1 (rs370477719) and for B2 (rs138637229). NRG2 (rs148371256) was unique to B1.The variants were identified to be “rare” as depicted by the ExAC_SAS score of 0.007 for LAMA1, 0.0005 for NRG2, 0.0004 for PLCG2 and 0.0001 for RAPGEF4. Furthermore, LAMA1 variant was identified as “deleterious” by all the five prediction tools and NRG2,PLCG2 and RAPGEF4 variants by three of the tools-SIFT, PolyPhen and Mutation taster 2. Interestingly, these genes were all differentially expressed in the BD cell lines, in comparison with control NPCs.
Discussion
We provide evidence indicating defective migration in NPCs from patients with BD. We also identify differentially expressed genes and rare damaging exonic variants related to cellular migration in these patient lines.
The BD patient NPCs migrated faster but lacked directionality in their movement. Consistent with previous reports (Tanaka et al. 2006; Yanagida et al. 2012), quantitative analysis in the present study also showed multi-directional movement of migrating control cells. The STRING protein interaction analysis exhibit EGFR as the core of the interactome map and is shown to interact directly with other proteins in the cluster like EGF, ERBB3, ERBB4, NRG1, PLCG2, PRKCA, PTK2B, PIK3CB, HBEGF and ITK. The molecule EGF signals through the EGFR/ERBB pathway (Citri & Yarden., 2016) to signal through neuregulins NRG1 and NRG2 (Iwakura and Nawa., 2013). The Rat Genome Database (RGD) depicts PLCG2 as an important member of the EGFR/Neuregulin signaling pathway, although no direct evidence is available. EGF stimulation increased the phospholipase activity of PLCG2 in fibroblast COS7 cells (Magno et al., 2019). Calcium signaling is well studied in cytoskeletal reorganisation and cellular migration (Tsai et al., 2015; Minton 2014); EGFR and PRKCA are identified as central regulators of the calcium signaling pathway (Buttner et al., 2018). Laminins have EGF like domains which may interact with the EGF receptor to induce the phosphorylation of the EGFR and activate its pathway (Schenk et al., 2003). Furthermore, attachment of cells to laminin was found to stimulate consistently higher levels of EGFR phosphorylation (Yarwood and Woodgett., 2001). Other molecules which signal through the EGFR/ERBB pathway are PTK2B (Block et al., 2010) PIK3CB (Ciraolo et al., 2008) and HBEGF (Elenius et al.,1997) and may contribute to cellular migration (Block et al., 2010; Zhang et al., 2018; Mine et al., 2005)
We investigated the genetic basis to abnormal neuronal migration patterns in patient-derived NPCs. We found rare, damaging exonic variants in LAMA1, RAPGEF4 and PLCG2 genes in both BD patients. These genes also showed differential expression patterns in the B1 and B2 transcriptome data. The expression of all these genes were found to be downregulated in both cell lines which can contribute to the abnormal migration phenotype of these NPCs. From the STRING protein interactome data, RAPGEF4 did not depict any direct interaction with the EGF/ERBB signaling pathway. However, activation of RAPGEF4 inhibits EGF signaling (Kiermayer et al., 2005). The NRG2 variant was unique to B1 NPC, which may add to the severity of the migration-related defects in B1.
Laminins are structural proteins essential for cerebellar development (Heng et al., 2011). Deficiency of Laminin Subunit Alpha 1 (LAMA1) results in altered cytoskeletal dynamics resulting in defective migration and contributes to neurodevelopmental disorders like cerebellar dysplasia (Aldinger et al., 2014), laminopathy (Vilboux et al., 2016) and mutations are reported in patients with intellectual disability (Kashevarova et al., 2018).
RAPGEF4 or EPAC2 plays an important role in neural cell migration as the gene is expressed in migrating NPCs and is involved in neurogenesis (Hyunhyo Seo & Kyungmin Lee., 2019). In addition, EPAC2 regulates neurotransmitter release, neuronal differentiation, neurite growth. Deleterious mutations in EPAC2 affect dendrite spine morphology (Schmidt et al., 2013, Jones et al., 2019). PLCG2 encodes for phospholipase C gamma2 and is an important member of the transmembrane signaling cascade and is associated with Alzheimer’s disease (Van der Lee et al., 2019). PLCG2 regulates epithelial cell migration (Gaudet et al., 2011); however no direct evidence for migration is reported with neuronal cells. NRG2 belongs to the neuregulin family, critical in cellular migration; and has been associated with schizophrenia, ADHD and related psychiatric disorders (Benzel et al., 2007; Yan et al., 2018). NRG2 has roles in cell survival, proliferation, synaptogenesis and dendritic outgrowth which suggest its significance in brain development (Lee et al., 2015; Nakano et al., 2016; Ghashghaei et al., 2006).
In conclusion, the genes LAMA1, RAPGEF4, PLCG2 and NRG2 may influence the EGF/ ERBB signaling pathway and induce abnormal neuronal migration patterns with defects in directionality of patient-derived NPCs. These cellular migration abnormalities may be linked to structural changes in the brain reported in BD.
Data Availability
Available upon request
Funding support
This work was supported by a grant from the Department of Biotechnology (COE), India funded grants-“Targeted generation and interrogation of cellular models and networks in neuro-psychiatric disorders using candidate genes” (BT/01/CEIB/11/VI/11/2012) and “Accelerating program for discovery in brain disorders using stem cells” (BT/PR17316/MED/ 31/326/2015) (ADBS); the Department of Science and Technology, India funded grants-“Imaging-genomics approach to identify molecular markers of Lithium response in Bipolar disorder” through the DST-INSPIRE Faculty Fellowship awarded to Dr. Biju Viswanath (Project number 00671, Code: IFA-12-LSBM-44) and SCIENCE& ENGINEERING RESEARCH BOARD (SERB) project “Dissecting the biology of lithium response in human induced pluripotent stem cell derived neurons from patients with bipolar affective disorder” (FILE NO. ECR/2016/002076).
References
