2D and 3D Stem Cell Models of Primate Cortical Development Identify Species-Specific Differences in Progenitor Behavior Contributing to Brain Size

Abstract

Variation in cerebral cortex size and complexity is thought to contribute to differences in cognitive ability between humans and other animals. Here we compare cortical progenitor cell output in humans and three nonhuman primates using directed differentiation of pluripotent stem cells (PSCs) in adherent two-dimensional (2D) and organoid three-dimensional (3D) culture systems. Clonal lineage analysis showed that primate cortical progenitors proliferate for a protracted period of time, during which they generate early-born neurons, in contrast to rodents, where this expansion phase largely ceases before neurogenesis begins. The extent of this additional cortical progenitor expansion differs among primates, leading to differences in the number of neurons generated by each progenitor cell. We found that this mechanism for controlling cortical size is regulated cell autonomously in culture, suggesting that primate cerebral cortex size is regulated at least in part at the level of individual cortical progenitor cell clonal output.

Graphical abstract

Figure 1

Replication of Species-Appropriate Developmental Timing of Cortical Neurogenesis In Vitro from PSCs of Multiple Primate Species (A) Schematic comparing the in vivo developing cortical neuroepithelium and in vitro, stem-cell-derived cortical neuroepithelial rosette. In a cortical rosette, the aPKC⁺ apical surface is at the rosette center, immediately surrounded by the VZ-like region containing PAX6/Vimentin⁺ RGPCs. Outside the VZ, there is no clear positional distinction between inner subventricular zone (iSVZ) and outer SVZ (oSVZ), where both TBR2⁺ intermediate progenitor cells (IPCs) and PAX6/Vimentin⁺ outer RGP-like cells are found. These progenitor cells produce cortical neurons (such as TBR1/MAP2⁺ thalamic projection neurons), which then migrate away from the rosette center. (B) Representative immunofluorescence images of cortical neuroepithelial rosettes derived from human (HS1, HS2, HS3, and HS4), chimpanzee (PT1 and PT2), and macaque (MF1, MF2, and MN1) PSCs. Antibodies used are as indicated: PAX6/Vimentin (RGPCs), aPKC (apical cell domain), TBR2/Ki67 (IPCs), or TBR1/MAP2 (layer VI cortical neurons). Scale bars, 50 μm. (C) Semiquantitative RT-PCR for the cortically expressed transcription factor (TF), FOXG1, and ventrally/caudally expressed TFs, NKX2.1, DLX1, and ISL1. All cortical progenitor cells from human and nonhuman primate PSCs are dorsal pallial in regional identity, unless treated with the Smoothened/Hedgehog agonist purmorphamine (HS1 + Pur.) during induction to ventralize the progenitor cells to noncortical identities. (D) The cerebral cortex is organized into six layers of excitatory projection neurons with defined gene-expression profiles, based on detailed studies of the mouse cerebral cortex (Molyneaux et al., 2007): thalamic projection neurons in layer VI express TBR1, subcerebral projection neurons in layer V express CTIP2, and callosal projection neurons in layer II–IV express RORB, SATB2, KCNIP2, and MDGA1. (E) Immunofluorescence images of in vitro-derived cortical neurons of human, chimpanzee, and macaque at the indicated developmental stages post-cortical induction (days 30–70). Cultures were stained for TBR1 and SATB2 to monitor differentiation of deep- and upper-layer neurons (yellow arrowheads indicate first SATB2⁺ neurons generated). Scale bars, 200 μm. (F) Quantification of the relative proportions of TBR1⁺ and SATB2⁺ neurons in human (HS1 and HS2), chimpanzee (PT2), and macaque (MF1, MF2, and MN1) cultures at the indicated developmental stages (days 30–70). (G) Semiquantitative RT-PCR of expression of CTIP2 (layer V), RORB (layer IV), KCNIP2, and MDGA1 (layer II–IV) at the indicated stages in human, chimpanzee, and macaque cortical cultures. Transcripts enriched in later-born, upper-layer neurons (RORB, KCNIP2, and MGDA1) are expressed at an earlier stage in macaque than in humans or chimpanzee.

Figure 2

Timing of Cortical Neurogenesis Is Independent of Neuronal Lamination and 3D Organization (A) Chimpanzee cerebral cortex organoids (scale bar, 200 μm). Organoids develop in vivo-like organization of VZ, with PAX6⁺/Ki67⁺ polarized (apical aPKC localization) progenitor cells within the VZ, apical mitoses (pH3⁺ cells), and IPCs at the outer margin of the VZ. Antibodies as indicated in each panel. Scale bar, 100 μm (B) Human, chimpanzee, and macaque cortical organoids undergo sequential production of TBR1⁺ deep-layer neurons and SATB2⁺ upper-layer neurons (yellow arrowheads indicate initial SATB2⁺ neurons generated). As organoids developed for longer periods, cortical neurons migrated to form cortical plate-like structures (yellow bracket) with some separation of layers of TBR1⁺ and SATB2⁺ neurons. Scale bars, 50 μm. (C) Scatterplots of positions of TBR1⁺ and SATB2⁺ neurons relative to the ventricular surface in human day 80, chimpanzee day 80, and macaque day 60 cortical organoids. Red lines represent median positions.

Figure 3

Functional Maturation of Primate Neurons Demonstrates Species-Specific Timing (A) Detection of miniature excitatory postsynaptic currents (mEPSCs) in whole-cell recordings of human (HS2), chimpanzee (PT1), and macaque (MF2) cortical neurons. Spontaneous depolarizations indicate the presence of synaptic activity. (B) Patch-clamp, single-neuron recordings of electrophysiological properties of cortical neurons at different developmental stages (days 30–70) for human (HS1 and HS2), chimpanzee (PT1), and macaque (MF1, MF2, and MN1). In response to stepwise current stimulation (−10 to 20 pA), in vitro cortical neurons fired action potentials (APs). The response to current injection evolved over time, with mature neurons firing more APs following single stimuli. Numbers represent frequencies of patterns of AP firing at each given developmental stage.

Figure 4

Clonal Analysis Reveals Marked Differences between Human and Macaque Cortical Progenitor Cell Dynamics over Developmental Time (A) Single cortical progenitor cells were labeled with low-titer, replication-incompetent lentiviruses at clonal resolution (see Supplemental Experimental Procedures for further details). Following infection at days 20, 30, or 40, progenitor cells were cultured for 2, 6, or 10 days, fixed, and immunostained for analysis. (B) Representative immunofluorescence images of clones derived from a single progenitor cell after 2, 6, or 10 day chase periods (panels as labeled) and immunostained for Ki67 (cycling progenitor cells) and βIII-tubulin (postmitotic neurons). Scale bar, 100 μm. (C) Human and macaque clone size distributions for each developmental stage (days 20, 30, and 40) at each time point postlabeling (2, 6, and 10 days). Red horizontal bars represent medians, and vertical bars indicate the interval between the first and third quartiles of the clonal distribution. Data for each species are combined from four human pluripotent cell lines (two ESCs and two iPSCs) and from three macaque ESC lines. Total number of clones analyzed for each line is as follows: HS1, n = 440; HS2, n = 43; HS3, n = 201; HS4, n = 93; MF1, n = 247; MF2, n = 469; MN1, n = 303. (D) Human and macaque average clone sizes for time points shown in (C). Significant differences between the average sizes of human and macaque clones at day 30 + 10 (p = 0.0437), day 40 + 6 (p = 0.0154), and day 40 + 10 (p = 0.205 × 10⁻²) are labeled. Error bars, SD.

Figure 5

Differences in Clone Growth between Macaque and Human Are Reflected by Differences in Progenitor Cell Proliferative Behaviors (A) Quantification of the average number of Ki67⁺ progenitors in a human or macaque clone after various chase periods (2, 6, and 10 days) following clonal labeling at days 20, 30, and 40. Data analysis for this and subsequent panels is from two human lines (HS1 and HS3) and three macaque lines (MF1, MF2, and MN3). Error bars, SD. (B) Average size of all “persisting” clones (which contain one or more Ki67⁺ progenitor cells) with different chase periods following clonal labeling at days 20, 30, and 40. The black solid line represents the theoretically predicted values for persisting clone expansion following day 40 labeling (see Supplemental Experimental Procedures for further details on the computational model). Error bars, SD. (C) Percentage of human and macaque “exited” clones (which no longer contain any Ki67⁺ progenitor cells) with different chase periods after clonal labeling at days 20, 30, and 40. Error bars, SD. (D) Human and macaque clone size distributions of total and persisting clones after clonal labeling at day 40 and analysis 2, 6, and 10 days after labeling. Red dotted lines represent theoretically predicted values (see Supplemental Experimental Procedures for details of computational modeling).

Figure 6

Testing Predictions of Progenitor Cell Proliferative Behaviors during Human and Macaque Cortical Development (A) Experimental design of BrdU/EdU double-labeling assay. From day 40, BrdU was added to human and macaque cortical cultures to cumulatively label all progenitor cells and their progeny until day 45, at which point BrdU was switched to EdU to reveal the ratio of persisting progenitor cells (BrdU⁺EdU⁺) to exited progenitor cells (BrdU⁺EdU⁻). (B) Representative scatterplot of EdU/BrdU double-labeling assay analyzed by flow cytometry. Three distinct populations of cells are evident: BrdU⁻EdU⁻ noncycling cells (which were postmitotic at the beginning of the experiment), BrdU⁺EdU⁻ exited progenitor cells, and BrdU⁺EdU⁺ persisting progenitor cells. The proportion of progenitor cells persisting after 5 day chase (BrdU⁺EdU⁺/all BrdU⁺) is higher for human than macaque (p = 0.0119). Human data average of n = 4; macaque data average of n = 5. Error bars, SD. (C) Time-lapse imaging of human and macaque cortical progenitor cell divisions. GFP-labeled progenitors in clones were followed and imaged every 12 hr over a period of 168 hr. From the sequential images, a lineage tree of clonal progenitor divisions was reconstructed. Cells were assigned as “progenitor” if they divided in the span of recording (red circle), “postmitotic” if they did not divide for more than 60 hr (equivalent to the third quartile of the distribution of all cell divisions recorded) (blue circle), or “unknown/apoptotic” if cells either disappeared from the imaging frame or were born close to the end of filming period (gray circle). (D) Representative lineage trees showing cell divisions of human (HS1 and HS2) and macaque (MF1 and MF2) progenitors, reconstructed from sequential images. (E) Bar graphs showing distributions of the lengths of cell cycles based on reconstructed lineage trees for human (blue) and macaque progenitor cells (orange). (F) Pie charts showing proportions of cell division types for human and macaque progenitors, based on reconstructed linage trees. “Proliferative” divisions are those giving rise to two progenitors, “asymmetric” divisions giving one progenitor and one postmitotic cell, and “terminal” division giving two postmitotic cells.

Figure 7

Species-Specific Cortical Progenitor Cell Proliferative Behavior and Developmental Timing Are Regulated by Cell-Autonomous Mechanisms (A) Schematic representation of the experimental design of in vitro, interspecies mixed culture assays. Cortical progenitor cells of species A were labeled with cytoplasmic GFP, delivered by high-titer lentivirus, and subsequently mixed with GFP⁻ progenitors from species B in a 1:100 or 1:1,000 ratio at day 35. Transplanted cortical progenitor cells were cultured with host cells for 2, 6, and 10 days (day 35 + 2, day 35 + 6, and day 35 + 10) before being fixed and immunostained. (B) Immunofluorescence images of GFP⁺ human and macaque clones introduced into macaque and human backgrounds, respectively. GFP⁺ cortical progenitor cells were efficiently incorporated into rosettes of host species (white dotted line). Scale bars, 50 μm. (C) The size distributions of human HS1 clones and macaque MF2 clones in either human or macaque backgrounds at day 35 + 10. Red horizontal lines indicate median clone sizes, and vertical lines show the span between 25% and 75% quartiles for each distribution. n = number of clones analyzed for each culture condition. Dilution of donor cells to host cells (1/100, 1/1,000) is as shown. (D) Representative immunofluorescence images of GFP⁺ human (HS1) or macaque (MF2) clones introduced into macaque background. Cultures were immunostained for transcription factors expressed by deep (TBR1) and upper (SATB2) cortical neurons. Yellow arrowheads indicate SATB2⁺ upper-layer neurons produced from a transplanted macaque progenitor cell. Scale bars, 100 μm. (E) Proportions of TBR1⁺ and SATB2⁺ cortical neurons generated by transplanted progenitor cells of each species in each background as indicated. Host/recipient environment does not affect cell types generated by each species. n = number of cells expressing each transcription factor. (F) Representative immunofluorescence images showing a long-term chimeric mixture of human (HS1) and macaque (MF2) neural progenitors. Single macaque progenitors were introduced into a human (HS1) or macaque (MF2) background at day 25. The mixed cultures were incubated further for 30 days and fixed and stained for the presence of upper-layer cortical neurons (SATB2⁺, yellow arrowheads). Scale bars, 150 μm.