Modelling disrupted-in-schizophrenia 1 loss of function in human neural progenitor cells: tools for molecular studies of human neurodevelopment and neuropsychiatric disorders

Human neural progenitor cells (NPCs) offer a new strategy for de novo modelling of neurodevelopment and neuropsychiatric disease.1 They can be derived from numerous stem cell types including embryonic and adult stem cells, or isolated from neurogenic tissues of the human central nervous system. As lineage-restricted cells, NPCs are self-renewing and easily differentiated to neurons, astrocytes and oligodendrocytes using methods that emulate in vivo development and neural cell replacement. Genetically modified cells can be generated to exhibit disease-related phenotypes and aberrant mechanisms for cellular and molecular analysis. Furthermore, they have the potential to be used as tools for early phase drug discovery by predicting in vivo drug response through in vitro cell-based assayology of drug candidates.

Disrupted-in-schizophrenia 1 (DISC1) is a risk factor for a variety of psychiatric and mental disorders, with mutant forms likely causing an overall loss of function in schizophenia.2, 3, 4 Non-human animal and cell-based studies indicate that DISC1 is a hub protein involved in multiple biochemical pathways, is highly expressed in the developing hippocampus (HP) and is localized to several subcellular domains including the centrosome, mitochondria, growth cone and synapse.5, 6, 7, 8, 9 DISC1 knockdown in maturing granule cells of the mouse HP causes aberrant cell migration, dendrite formation and axon termination.8, 9, 10 Reduced expression of DISC1 mRNA has been reported in lymphoblastoid cells of patients diagnosed with schizophrenia or chronic psychiatric illness.11 Taken together, cytogenetic and biological studies of DISC1 support the developmental hypothesis of schizophrenia.

Despite the remarkable evidence for a role of DISC1 in brain development and major mental illness, its mechanism and function must be elucidated using biologically relevant methods. In this study, we report the development of novel in vitro cell-based models of DISC1 loss of function using human embryonic stem (ES) cells and human brain-derived NPCs with RNA interference. Brain-derived NPCs represent region-specific cell lines, isolated from fetal HP, subventricular zone (SVZ) and cortex (ReNcell CX). Genetic manipulation of the lines was preceded by comprehensive characterization of wild-type stem cell and/or NPC phenotypes and differentiation potential to neuronal cell derivatives. All NPC lines can be cultured long-term for scale up (HP: >20, SVZ: >30, ReNcell CX: >20 and ES-NPCs: >30 passages) before transfection and/or differentiation, with HP-NPCs stably expressing the proliferation marker Ki-67 (>86%) and neural stem/progenitor markers nestin (>95%), sox2 (>99%) and vimentin (>99%) (Figure 1a). SVZ-NPCs and ReNcell CX exhibited similar expression profiles (Supplementary Figure 1a, b). Human ES cell-derived NPCs (ES-NPCs) also displayed a high level of Ki-67 (>95%), nestin (>98%), sox2 (>96%) and vimentin (>99%) (Figure 1b).

Figure 1

(a, b) Representative confocal photomicrographs and flow cytometry graphs of neural progenitor cells (NPCs) derived from (a) hippocampus (HP) and (b) human embryonic stem (ES) cells, showing robust expression of the proliferation marker Ki-67 and the NPC markers nestin, sox2 and vimentin. Scale=30 μm. (c) A representative confocal photomicrograph of HP-NPCs differentiated to βIII-tublin (red) and prox1 (green) positive cells, reminiscent of dentate granule cells. Scale=20 μm. (d) A representative confocal photomicrograph of differentiated human ES-NPCs expressing the neuronal marker MAP2 (green). Scale=50 μm. (e) A partial vector map of pLentiLox 3.7 showing a multiple cloning site where DISC1 targeting 19mers is inserted (short hairpin RNA 1 (shRNA1) or shRNA2). (f) Quantitative real-time-PCR analysis of DISC1 mRNA level after pLentiLox 3.7-mediated RNA interference. A significant reduction of DISC1 mRNA expression (60%) was observed in HP-NPCs transduced with shRNA1 compared with shRNA control (**; P<0.01). Similarly, after a single round of fluorescent-activated cell sorting selecting for green fluorescent protein (GFP), human ES-derived NPCs transduced with shRNA1 exhibited a significant reduction in DISC1 mRNA level compared with the shRNA control (**; P<0.01). (g) Effect of DISC1 silencing on NPC migration depicted by representative fluorescent photomicrographs. Transduced SVZ-NPCs were seeded onto an Oris migration detection plate for 6 days and processed for GFP immunocytochemistry. Quantitative fluorescent image analysis showed that DSC1 knockdown lines (shRNA1 and shRNA2) displayed repressed migration (2.42±1.36 and 2.65±0.54%, respectively) compared with the shRNA control line (4.98±0.74%) (***, P<0.001). Scale=200 μm.

Differentiation of HP-NPCs generated prox1/βIII-tubulin-positive neurons reminiscent of HP granule cells (Figure 1c). Similarly, neuronal derivatives of SVZ-NPCs were primarily γ-aminobutyric acid (GABA)/calbindin-D28k (Supplementary Figure 1c) or GABA/parvalbumin-positive neurons (data not shown), representing major sub-populations of central nervous system interneurons. Differentiated ReNcell CXs were by-and-large vesicular glutamate transporter-1/βIII-tubulin (Supplementary Figure 1d) or GABA/βIII-tubulin-positive cells (data not shown). Under the same neural differentiation condition as brain-derived NPCs, differentiated ES-derived NPCs generated MAP2-positive neurons (Figure 1d).

Transgenic NPC lines were generated using a novel lentivirus vector expressing short hairpin shRNA against a sequence targeting all isoforms of DISC1 mRNA (Figure 1e). Transduction efficiency was initially estimated by flow cytometry of a green fluorescent reporter protein. Approximately 70–80% of total inoculated HP-NPCs, SVZ-NPCs and ReNcell CXs were transduced after single inoculation of the lentiviral vector (data not shown). Other approaches to RNA interference using conventional liposome-mediated transfection, electroporation or nano-particles resulted in a less than optimal transduction efficiency of 30 % (data not shown). At 72 h post-inoculation, quantitative real-time-PCR confirmed significant knockdown (70-80%) of DISC1 mRNA for all brain-derived NPC lines compared with shRNA control (Figure 1f, Supplementary Figure 1e). Inoculated human ES-NPCs were subjected to a single round of fluorescence-activated cell sorting to pool green fluorescent reporter protein-expressing/transduced cells for subsequent quantitative real-time-PCR of DISC1 mRNA expression leading to 60% reduction in sorted DISC1 knockdown line compared with shRNA control line (Figure 1f). Importantly, DISC1 silencing persisted for multi-passage subculture for all transgenic NPC lines, indicating constitutive and stable knockdown (Supplementary Figure 1f). Finally, we have confirmed an effect of gene silencing on cell phenotype by measuring reduced migration of SVZ-NPCs, consistent with in vivo findings for DISC1 knockdown in mouse cortex (Figure 1g).8

The present letter describes the generation of novel human cell-based models of DISC1 loss of function. We propose that the cell lines produced (and human neurogenic stem and progenitor cell lines in general) offer a biologically relevant strategy to studying human neurogenesis, brain development and neuropsychiatric disease. Further characterization of the morphological, biochemical and physiological properties of current lines, combined with the generation and study of different lines with a focus on other susceptibility genes will be necessary and may advance understanding of schizophrenia, other human brain disorders and pharmacotherapeutics. Accordingly, the models represent tools that are complimentary to in vivo experimentation and may be useful for preclinical phase drug discovery as predictors of new compound efficacy.


  1. 1

    Crook JM, Kobayashi NR . J Cell Biochem 2008; 105: 1361–1366.

  2. 2

    Blackwood DH, Fordyce A, Walker MT, St Clair DM, Porteous DJ, Muir WJ . Am J Hum Genet 2001; 69: 428–433.

  3. 3

    Chubb JE, Bradshaw NJ, Soares DC, Porteous DJ, Millar JK . Mol Psychiatry 2008; 13: 36–64.

  4. 4

    Ishizuka K, Kamiya A, Sawa A . Biol Psychiatry 2006; 59: 1189–1197.

  5. 5

    Pletnikov MV, Xu Y, Ovanesov MV, Kamiya A, Sawa A, Ross CA . Neurosci Res 2007; 58: 234–244.

  6. 6

    James R, Adams RR, Christie S, Buchanan SR, Porteous DJ, Millar JK . Mol Cell Neurosci 2004; 26: 112–122.

  7. 7

    Morris JA, Kandpal G, Ma L, Austin CP . Hum Mol Genet 2003; 12: 1591–1608.

  8. 8

    Kamiya A, Kubo K, Tomoda T, Takaki M, Youn R, Ozeki Y et al. Nat Cell Biol 2005; 7: 1167–1178.

  9. 9

    Mao Y, Ge X, Frank CL, Madison JM, Koehler AN, Doud MK et al. Cell 2009; 136: 1017–1031.

  10. 10

    Kim JY, Duan X, Liu CY, Jang MH, Guo JU, Pow-Anpongkul N et al. Neuron 2009; 63: 761–773.

  11. 11

    Millar JK, Pickard BS, Mackie S, James R, Christie S, Buchanan SR et al. Science 2005; 310: 1187–1191.

Download references

Author information

Correspondence to J M Crook.

Ethics declarations

Competing interests

The authors declare no conflict of interest.

Additional information

Supplementary Information accompanies the paper on the Molecular Psychiatry website

Supplementary information

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Kobayashi, N., Sui, L., Tan, P. et al. Modelling disrupted-in-schizophrenia 1 loss of function in human neural progenitor cells: tools for molecular studies of human neurodevelopment and neuropsychiatric disorders. Mol Psychiatry 15, 672–675 (2010) doi:10.1038/mp.2009.131

Download citation

Further reading