Recapitulating cortical development with organoid culture in vitro and modeling abnormal spindle-like (ASPM related primary) microcephaly disease
The development of a cerebral organoid culture in vitro offers an opportunity to generate human brain-like organs to investigate mechanisms of human disease that are specific to the neurogenesis of radial glial (RG) and outer radial glial (oRG) cells in the ventricular zone (VZ) and subventricular zone (SVZ) of the developing neocortex. Modeling neuronal progenitors and the organization that produces mature subcortical neuron subtypes during early stages of development is essential for studying human brain developmental diseases. Several previous efforts have shown to grow neural organoid in culture dishes successfully, however we demonstrate a new paradigm that recapitulates neocortical development process with VZ, OSVZ formation and the lamination organization of cortical layer structure. In addition, using patient-specific induced pluripotent stem cells (iPSCs) with dysfunction of the Aspm gene from a primary microcephaly patient, we demonstrate neurogenesis defects result in defective neuronal activity in patient organoids, suggesting a new strategy to study human developmental diseases in central nerve system.
Keywordsneocortical development cerebral organoid microcephaly ASPM
The most highly evolved structure in the human brain is the neocortex, which is responsible for the higher cognitive functions unique to humans (Rakic, 2009; Borrell and Reillo, 2012). The dramatic human neocortical expansion and gyri formation underline the precise orchestration of neurogenesis and neuronal migration in the embryonic stage, especially during the peak of neurogenesis, which occurs during the second trimester of gestation in humans (Kriegstein et al., 2006; Lui et al., 2011; de Graaf-Peters and Hadders-Algra, 2006). Radial glial (RG) cells constitute a major population of neural progenitor cells and are referred to vRG cells (ventricular radial glial cells); these cells occupy the proliferative ventricular zone (VZ) (Noctor et al., 2002; Hartfuss et al., 2001; Anthony et al., 2004). vRG cells display interkinetic nuclear migration (INM) behavior, proliferate extensively at the luminal surface of the VZ and predominantly undergo asymmetric division to self-renew while simultaneously giving rise either directly to neurons or to an IPC that subsequently divides symmetrically to produce neurons (Kriegstein et al., 2006; Noctor et al., 2001; Fishell and Kriegstein, 2003; Huttner and Kosodo, 2005). A new population of neural stem/progenitor cells has recently been reported to occupy the greatly expanded subventricular zone (SVZ), which is composed of intermediate progenitor cells (IPCs) and outer radial glial (oRG) cells and can be divided into the inner (ISVZ) and outer SVZ (OSVZ) by the inner fiber layer (IFL) in primates (Dehay and Kennedy, 2007; Hansen et al., 2010; LaMonica et al., 2012; Garcia-Moreno et al., 2012). Unlike vRG cells, oRG cells show distinctive mitotic somal translocation (MST) behavior, giving rise to IPCs in humans (Hansen et al., 2010; LaMonica et al., 2012; Ostrem et al., 2014; Taverna and Huttner, 2010; Gertz et al., 2014) instead of neurons, as in mice (Wang et al., 2011), and serve as transit amplifying machinery for proliferative pool expansion (Lui et al., 2011; Fietz and Huttner, 2011; Pontious et al., 2008). An enlarged OSVZ, accompanied by an increase in both oRG cells and IPCs, may account for neuronal amplification and the tangential dispersion of neurons that contributes to cortical expansion and gyrification (Borrell and Reillo, 2012; LaMonica et al., 2012; Reillo et al., 2011; Nonaka-Kinoshita et al., 2013).
Extensive studies have demonstrated that the neocortex is a six-layered laminated structure generated by a specific pattern of neuronal production and migration. Newborn neurons migrate successively following the guidance of radial glial fibers over the existing early-born neurons during cortical development and occupy superficial layers in the cortical plate (CP), creating cortical layers (CL). Cortical lamination provides the basic framework for neocortical function (Gao et al., 2013; Douglas and Martin, 2004). Recent achievements suggest that the developing human neocortex differs significantly from the rodent model; however, due to limitations on the experimental manipulation of human tissues and on the monitoring of long-term development in patients, much of the recent progress in understanding cortical development has come from rodent studies (Wang et al., 2011; Caviness and Rakic, 1978; D’Arcangelo, 2006). Therefore, new research strategies need to be established to study human cortical development.
Several previous efforts to grow neural tissues in culture dishes, such as an optic cup (Eiraku et al., 2011), a pituitary gland (Suga et al., 2011), cortex-like cell layers (Nasu et al., 2012; Kadoshima et al., 2013; Eiraku et al., 2008), some brain-region-specific organoids (Qian et al., 2016; Pasca et al., 2015), entire brain-like organoids (Lancaster et al., 2013; Camp et al., 2015), and even organoids with surface folding (Li et al., 2017) have been reported. Also, organoids have been proven to be a remarkable model to study diseases, such as ZIKA virus (Qian et al., 2016; Cugola et al., 2016; Garcez et al., 2016; Nowakowski et al., 2016; Tang et al., 2016; Xu et al., 2016) and neural system disease (Lancaster et al., 2013; Bershteyn and, 2017). Therefore, organoid culture system is suitable for modeling neurodevelopmental diseases and related cognition defects.
To address this issue, we have developed a new three-dimensional culture system for a cortex-like tissue derived from induced pluripotent stem cells (iPSCs) in vitro. Here, we show that these cultured organoids recapitulate the lamination organization of the layered neocortex generated from vRG, oRG, and IP cells and eventually mimic the cortical development in vivo. Finally, in an investigation of the ASPM-dependent pathogenesis of microcephaly, we report that the cortical organoids generated from primary microcephaly patient-derived iPSCs were unable to form normal cortical lamination resulting in defective neuronal activity.
Self-organized organoids mimics early cerebral development
To examine RG identity during the early development of these cerebral-like 3D tissues, we performed staining for phospho-vimentin (P-vim) and Sox2 on day 36 (Fig. S1A). We observed vRG for Sox2+ cells along the apical side with apical anchor and extended basal cellular processes toward the outer surface of these tissues in cerebral organoids. Since vRG cells typically divided on the lumen surface, we observed a series of mitotic cells at different dividing stages in organoids, such as prophase, metaphase, anaphase and telophase (Fig. S1A), resembling innate vRGs features in early human neocortical development. In primates, the population of oRGs is significantly larger than in rodents, and these oRGs play critical roles in neocortical neurogenesis. Sox2+/P-vim+ cells above VZ region exhibited long basal processes (Fig. S1A), suggesting that oRGs also exist in organoids. In consistent with previous proliferating marker staining results, we observed 19.55% and 6.39% mitotic vRGs and oRGs in VZ/SVZ area respectively (Fig. S1A).
In mammals, vRG cells initially divide symmetrically to enlarge the progenitor pool and then divide asymmetrically to maintain one daughter cell as vRG and the other as a progenitor cell, such as an oRG cell, an IP cell, or a newborn neuron (Hansen et al., 2010; Wang et al., 2011); hence the cell division behavior is a determinant for neural progenitor pool and cortical neurogenesis. To confirm division behavior of vRGs in cerebral-like tissue, we assessed the angle between cleavage plane and ventricle surface (Fig. S1B). In our hiPSC culture, proliferating apical progenitors preferentially divided from a “vertical” cleavage plane (60–90°, 57.0%; Fig. S1C) on day 36, indicating that vRGs primarily undergo symmetrically division to enrich vRG pool and also contribute to progenitor or neuron production at early culture time, which is similar to early cortical development. Additionally, we use time lapse to observe the activity of these vRG cells, and found typically interkinetic nucleus movement (INM) like vRG in vivo (Fig. S1D and Movie S1). These results indicate vRGs in organoids mimicked cellular activity in vivo.
To determine whether features of human brain development with oRG and IP cell existence were recapitulated in cerebral organoids, we analyzed the organoids on day 65 (Fig. 1E). The distribution of Sox2+ progenitors were examined and we observed a population concentrated close to the apical surface (VZ) that was thought to be vRG cells, as well as another population displaced from the apical surface (OSVZ), that was consistent with oRG identity in organoids (Fig. 1E and 1F). In addition, IP cells identified by Tbr2 staining primarily localized outside of but close to the VZ on day 65 (Fig. 1E and 1F). The presence and distribution of RGs and IP cells demonstrated that this new 3D culture system could mimic early development of brain with all types of known neural progenitors in dish.
Spatial neuron lamination of organoids
The mammalian cortical plate consists of six distinct layers. To determine whether organoids could recapitulate the cortical spatial lamination, we stained for cortical layer markers on day 65. In hiPSC-derived neocortex-like tissues, Tbr2 positive subventricular zone (SVZ) was adjacent to the ventricular zone (VZ), which indicated by Sox2+ region, this result was consist with structures in vivo (Figs. 2A and S2A). Also, cortical plate (CP) indicated by Satb2+ neuron located outside of Pax6+ and Tbr2+ regions (Fig. 2B and 2C). Furthermore, we observed Reelin+ neurons along the basal surface (Fig. 2D), suggesting the presence of Cajal-Retzius like cells, which regulates the generation of cortical plate laminar architecture. Finally, in organoid of day 90, we could observe that later-born Satb2+ neurons localized more superficially to the early-born Ctip2+ neurons (Fig. 2E). This lamination was similar to human developing neocortex (Fig. S2B). Thus, this distribution pattern of dorsal cortical subtype-specific neurons resembles in vivo regional sub-specification, indicating these organoids derived from our culture system could efficiently model neocortex organization in vivo.
Electrophysiological properties of neurons in organoids
Organoids recapitulate microcephaly with Aspm mutation
To test the proliferation defect, we independently performed RNAi knockdown of Aspm by electroporating GFP expressing shRNA in organoids. The endogenous Aspm RNAi efficiency was first verified by RT-qPCR (Fig. S5C). Knockdown of Aspm in hiPSC-control organoids led to a decrease of Ki67+ and pH3+ cells close to ventricle (Fig. S5D–F), consistent with the defective phenotype observed in ASPM patient organoids. These findings support the conclusion that ASPM mutant could induce deficiency in proliferation of progenitors. These patient-derived cerebral organoids provided a unique opportunity to examine the hypoplasia observed in microcephaly.
Since organoids generated from hiPSC-ASPM-2 exhibited the more condensed cell mass and intact morphology than organoids derived from hiPSC-ASPM-1 or hiPSC-ASPM-3 (Figs. 4A and S4A), and all of organoids showed similar developing phenotypes, we utilized hiPSC-ASPM-2 as a representative for more detailed studies. Consistent with observations at early culture time, organized lumen structures were hardly maintained in mutants on day 65 (Fig. 4E). We could observe dividing Sox2+/P-vim+ vRG cells along the apical side in hiPSC-control tissues but not hiPSC-ASPM organoids (Fig. 4E). The amount of vRG and oRG cells in hiPSC-ASPM organoids was significantly reduced compared to control (Fig. 4E). On day 65, Tbr2+ IPCs primarily localized outside of but close to the VZ, showing a pattern similar to that of the human developing cortex (Fig. 1E). Unsurprisingly, Tbr2+ cells were discrete and disorganized in mutant organoids (Fig. 4E). In summary, the loss of lumen structure and neural progenitor cells in hiPSC-ASPM organoids might mimic the pathogenesis of primary microcephaly in human fetal brain development. Although hiPSC-ASPM organoids developed towards telencephalon revealed by FoxG1 staining, the elaborative lamination was not observed in Aspm mutant 3D aggregations, suggesting that the culture system was able to model cortical development diseases.
Calcium imaging data of control and mutant organoids
In this report, we have established an approach to generate cerebral organoids from hESCs or hiPSCs following dissociation, self-aggregation, and differentiation with minimal extrinsic signals added to the culture system (Fig. 1A). This three-dimensional culture method recapitulates in vivo cortical development with formation of a well-polarized ventricle neuroepithelial structure; vRG, oRG and IP cells; and production of mature neurons within layers (Figs. 1–3).
Cerebral organoid culture has been developed and applied widely in cortical developmental diseases. Primary microcephaly with CDK5RAP2 mutation and Zika virus infection were studied by cerebral organoid culture suggests that 3D culture could recapitulate real in vivo organ development to some extent (Lancaster et al., 2013; Tang et al., 2016). Since this system displays fundamental characteristics of human telencephalon cortical development, it is a promising approach to study the pathogenesis of neurodevelopmental disorders. Here, we have modeled microcephaly by generating organoids from the iPSCs of ASPM-mutation patients. Mutations in the Aspm gene at the MCPH5 locus are the most common cause of human primary microcephaly leading to a reduction of cerebral cortex size, congenital failure and mental retardation (Bond et al., 2002, 2003). We have observed defects of early development in aggregates from hiPSC-ASPM cells characterized by a less organized neuroepithelium and fewer vRG, oRG, and defective layer lamination (Figs. 4 and S4). Studies in Drosophila suggest that ASPM may interact with myosin II in regulating spindle orientation and interkinetic nuclear migration (INM) in neural progenitor cells (Rujano et al., 2013), suggesting that vRG cells with Aspm mutation possibly experienced defective INM, resulting in reduction of oRG cells in the late developmental stages in our culture system (Fig. 4E). This observation could explain why no sufficient ASPM-deletion mouse models that recapitulate microcephaly phenotypes have been developed. Because oRG cells are considered to play a major role in cortical expansion in primates; however, this cell type is limited in rodents (Wang et al., 2011). Cerebral cortical development is an elaborative temporal-spatial progress. Using calcium imaging, we found that fewer neurons had matured and less synchronized neuronal activities had been detected at the late stage of development in the Aspm mutants (Fig. 5), which could be the reason that ASPM-related microcephaly patients present congenital failure and mental retardation (Bond et al., 2002, 2003; Shen et al., 2005).
The availability of our novel culture system offers a resource for pathological studies of neurodevelopmental diseases, furthermore, this hESC/hiPSC-derived organoid culture may also be used for various future biomedical applications, including cell therapies, drug discovery and toxicology studies.
Materials and Methods
Human induced pluripotent stem cell culture
Cell maintenance and 3D culture procedure are described in SI Materials and Methods.
Karyotyping and RT-PCR
Karyotypes were determined with standard G-banding chromosome analysis in the Oakland Children Hospital. RT-PCR procedure and primer sequences are in SI Materials and Methods.
Cell aggregates were fixed in cold 4% PFA for 30 min at 4°C followed by incubated in 15% (w/v) sucrose solution overnight at 4°C and then embedded in O.C.T. compound, frozen at −20°C and cryosectioned at 10 μm. Immunostaining and antibody information is available in SI Materials and Methods.
Cell aggregates were fixed in cold 70% alcohol for 30 min at 4°C overnight followed by two washes with PBS. Aggregates were incubated in PBS including Propidium Iodide (PI, 50 μg/mL) and RNase A (50 μg/mL) for 30 min at 4°C. Signals were detected with a FACS Calibur flow cytometer (BD Biosciences). ModFit was used to analysis cell cycle.
Electrophysiology and calcium imaging
Detail materials and methods are available in SI Materials and Methods.
All data are presented as mean values and standard error of mean (s.e.m.), and two-tailed Student’s t-tests were used to determine statistical significance. Cell numbers were counted with Imaris (Bitplane) software and ImageJ. Detail quantification methods are available in SI Materials and Methods.
We thank A. Alvarez-Buylla, A.R. Kriegstein, R.Blelloch, and S. Wang for comments on the manuscript. We also thank C. Nicolas (Kriegstein lab) and K. Tomoda (Yamanaka lab) for suggestions and members of the Wang laboratory for discussion. This work was supported by National Basic Research Program of China (Nos. 2014CB964600, 2015CB964800 and 2017YFA0102601), the Strategic Priority Research Program of the Chinese Academy of Sciences (XDA01020309), the National Natural Science Foundation of China (NSFC) 31371100, 91332105 to X.W., 31400937, 31676072 to Q. W., 31400939 to L.S. and 31600840 to R.L.
CL, cortical layers; CP, cortical plate; EBs, embryonic bodies; INM, interkinetic nuclear migration; iPSCs, induced pluripotent stem cells; MST, mitotic somal translocation; oRG, outer radial glial; RG, radial glial; SVZ, subventricular zone; VZ, ventricular zone
Compliance with ethics guidelines
All procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation (Institute of Biophysics, Chinese Academy of Science, committee on biosafety and ethics) and with the Helsinki Declaration of 1975, as revised in 2000 (5). Informed consent was obtained from all patients for being included in the study. Rui Li, Le Sun, Ai Fang, Peng Li, Qian Wu, and Xiaoqun Wang declare that they have no conflict of interest.
Q.W. and X.W. conceptualized and designed the experiments. R.L., Q.W., L.S., A. F. and P. L. performed all the experiments, analyzed and interpreted the data. Q.W., R.L., L.S. and X.W. wrote the manuscript, which all authors edited and approved.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.