In the present study, we have addressed these questions. In doing so, we provide support for the notion that ARHGAP11B is sufficient to increase BP proliferation and abundance to a human-like level in chimpanzee cerebral organoids. Conversely, we find that dominant-negative inhibition of ARHGAP11B's function by ARHGAP11A220 reduces cycling BP abundance in human cerebral organoids to the chimpanzee level. Finally, by subjecting ARHGAP11A plus ARHGAP11B double-knockout human forebrain organoids to either ARHGAP11A or ARHGAP11B rescue, we find that ARHGAP11B is essential to maintain the level of basal (or outer) radial glia (bRG), the BP type of particular relevance for neocortex expansion. Together, these findings provide direct evidence in support of an indispensable role of ARHGAP11B in neocortical expansion during human development and evolution.

Yet, a key question regarding ARHGAP11B 's role in human neocortex expansion remains unanswered: Can the human-specific ARHGAP11B gene increase the proliferation and abundance of BPs when expressed in cerebral organoids of the chimpanzee, our closest living relative? And, conversely, regarding human neocortical development: Is ARHGAP11B required to maintain the full level of BP proliferation and abundance in human cerebral organoids?

Considering these sets of findings together, the question arises to which extent ARHGAP11B contributes to the increase in cycling BPs in the context of the expansion of the neocortex in the course of human evolution. A first clue in this regard was obtained by the observation that a truncated form of the ARHGAP11A protein, ARHGAP11A220, which acts in a dominant-negative manner on ARHGAP11B's ability to amplify BPs in embryonic mouse neocortex, reduces the abundance of cycling BPs in fetal human neocortical tissue ex vivo (Namba et al , 2020 ).

ARHGAP11B is a human-specific gene (Sudmant et al , 2010 ; Dennis et al , 2017 ) and the first such gene to have been implicated in human neocortical development and evolution (Florio et al , 2015 , 2016 ; Kalebic et al , 2018 ; Heide et al , 2020 ; Xing et al , 2021 ). In fetal human neocortex, ARHGAP11B is preferentially expressed in cNPCs (Florio et al , 2015 , 2018 ). When (over)expressed in embryonic mouse and ferret neocortex, ARHGAP11B has been found to increase the proliferation and abundance of BPs (Florio et al , 2015 ; Kalebic et al , 2018 ; Xing et al , 2021 ), the cNPC class implicated in neocortical expansion during human development and evolution (Lui et al , 2011 ; Borrell & Götz, 2014 ; Florio & Huttner, 2014 ; Dehay et al , 2015 ). Moreover, a recent study in which ARHGAP11B was expressed under the control of its own promoter to physiological levels in the fetal neocortex of the common marmoset has demonstrated that this human-specific gene can indeed induce the hallmarks of neocortical expansion in this non-human primate, increasing neocortex size, folding, BP levels, and upper-layer neuron numbers (Heide et al , 2020 ). Consistent with this, physiological ARHGAP11B expression in a transgenic mouse line not only resulted in increased neocortical size and upper-layer neuron numbers that persist into adulthood, but also in increased cognitive abilities (Xing et al , 2021 ). Importantly, the ability of ARHGAP11B to increase the proliferation and abundance of BPs has been attributed not to the gene as it arose ≈ 5 mya by partial duplication of the widespread gene ARHGAP11A (Sudmant et al , 2010 ; Dennis et al , 2017 ), referred to as ancestral ARHGAP11B , but to an ARHGAP11B gene that subsequently underwent a point mutation, referred to as modern ARHGAP11B (Florio et al , 2016 ). These studies therefore establish (i) that modern ARHGAP11B is sufficient to expand BPs, including in primates, and (ii) that the resulting neocortex expansion and increase in upper-layer neuron numbers are associated with an increase in cognitive abilities.

In light of these findings, cerebral organoids have emerged as a promising primate model system to study cortical development and evolution. In addition, cerebral organoids offer the opportunity of extrinsic genetic manipulation (Fischer et al , 2019 ). This is particularly relevant in the case of human-specific genes that in fetal human neocortex are preferentially expressed in cNPCs and hence have been implicated in human-specific features of neocortical development (Florio et al , 2015 , 2018 ; Fiddes et al , 2018 ; Suzuki et al , 2018 ; Heide & Huttner, 2021 ). Examining such genes for their function in, and effects on, cNPC proliferation in cerebral organoids of human and chimpanzee, respectively, could not only provide corroborating evidence in support of their presumptive role in neocortical development during human evolution, but also provide further insights into their action and effects.

Thus, cerebral organoids have been shown to exhibit several (albeit not all) hallmarks of developing neocortical tissue, including a ventricular zone (VZ) and subventricular zone (SVZ) as well as the two major classes of cNPCs therein, the apical progenitors (APs) and the basal progenitors (BPs) (Kadoshima et al , 2013 ; Lancaster et al , 2013 ; Qian et al , 2016 ; Quadrato et al , 2017 ; Heide et al , 2018 ). Cerebral organoids also exhibit a cortical plate-like region with neuronal layers (NLs) containing the various types of cortical neurons (Kadoshima et al , 2013 ; Lancaster et al , 2013 , 2017 ; Qian et al , 2016 ; Quadrato et al , 2017 ; Heide et al , 2018 ; Velasco et al , 2019 ). Moreover, human cerebral organoids have been shown to recapitulate gene expression programs of fetal human neocortex development (Camp et al , 2015 ; Velasco et al , 2019 ; Bhaduri et al , 2020 ).

A way out of this dilemma has been provided by recent, seminal advances in pluripotent stem cell (PSC) research, which led to the development of the brain organoid technology (Watanabe et al , 2005 ; Eiraku et al , 2008 ; Kadoshima et al , 2013 ; Lancaster et al , 2013 , 2017 ; Pasca et al , 2015 ; Qian et al , 2016 ; Quadrato et al , 2017 ; Karzbrun et al , 2018 ; Giandomenico et al , 2019 ). A specific subtype of brain organoids, the cerebral organoids are relatively small (a few mm in diameter), three-dimensional (3D) structured cell assemblies that can be grown from embryonic stem cells (ESCs) (in the case of human) or induced pluripotent stem cells (iPSCs) (in the case of human and chimpanzee) and that emulate cerebral tissue (Lancaster et al , 2013 , 2017 ; Kelava & Lancaster, 2016 ; Di Lullo & Kriegstein, 2017 ; Arlotta, 2018 ; Heide et al , 2018 ; Fischer et al , 2019 ).

Over the past 8 years, genes have been identified that specifically evolved in the human lineage, that are preferentially expressed in cNPCs, and that promote cNPC proliferation (Florio et al , 2015 , 2018 ; Fiddes et al , 2018 ; Suzuki et al , 2018 ). Such genes have therefore been implicated in human-specific features of neocortical development (Florio et al , 2015 , 2018 ; Fiddes et al , 2018 ; Suzuki et al , 2018 ; Heide & Huttner, 2021 ). However, a human–chimpanzee comparison to explore whether such human-specific genes are responsible for a human-like cNPC proliferative capacity has not yet been carried out, mainly for the following reason. Whereas tissue of developing human neocortex can, in principle, be obtained and subjected to experimental studies, this is not the case for tissue of developing chimpanzee neocortex.

The neocortex, the evolutionarily youngest part of the brain, is the seat of our higher cognitive abilities. It is therefore of crucial importance to investigate the development of the neocortex. This has been done in several model systems and has provided pivotal insight (Rakic, 2009 ; Lui et al , 2011 ; Florio & Huttner, 2014 ; Sun & Hevner, 2014 ; Dehay et al , 2015 ; Molnar et al , 2019 ; Silver et al , 2019 ). Identifying the features that characterize the development specifically of the human neocortex is, however, a fundamental challenge. The human neocortex exhibits an increase in size and in the numbers of neurons compared with non-human primates. This increase is thought to reflect a greater proliferative capacity of the cortical stem and progenitor cells (collectively referred to as cortical neural progenitor cells (cNPCs)) in human (Fish et al , 2008 ; Lui et al , 2011 ; Florio & Huttner, 2014 ; Sun & Hevner, 2014 ; Dehay et al , 2015 ).

Results

Human and chimpanzee cerebral organoids as a test system for gene function For most of the data presented in this study, human and chimpanzee cerebral organoids were grown from human iPSCs of the line SC102A1 (Camp et al, 2015; Mora-Bermudez et al, 2016; Kanton et al, 2019) and chimpanzee iPSCs of the line Sandra A (Mora-Bermudez et al, 2016; Kanton et al, 2019), respectively (for the iPSC line used to generate knockout forebrain organoids, see below). Cerebral organoids were generated according to an established protocol (Lancaster et al, 2013; Lancaster & Knoblich, 2014; Camp et al, 2015; Mora-Bermudez et al, 2016; Kanton et al, 2019). In brief, iPSCs were aggregated to form embryoid bodies followed by their transformation into 3D cerebral tissue exhibiting numerous ventricular structures (Fig 1A). Cerebral organoids were subjected to manipulations (see below) between day 51 and 55 (see Fig 1A). This time window was chosen based on the known time courses of VZ formation, SVZ formation, and deep- and upper-layer neuron generation. In contrast to macaque organoids, these time courses are roughly comparable between human and chimpanzee cerebral organoids (Mora-Bermudez et al, 2016; Otani et al, 2016; Kanton et al, 2019). After 51 or 55 days of organoid culture, various mixtures of DNA constructs, consisting of a cytosolic-GFP expression vector and either an expression vector with the cDNA of interest or the corresponding control vector, were then microinjected into the lumen of the larger ventricle-like structures within the cerebral organoids, followed by electroporation to transfect the cNPCs in the VZ (Fig 1A and B; Lancaster et al, 2013; Li et al, 2017; Fischer et al, 2019; Giandomenico et al, 2019). Depending on the specific scientific question asked, organoid culture was continued for 2–15 days after electroporation followed by fixation of the cerebral organoids, in the case of 2 days with addition of BrdU 1 h prior to fixation as indicated (Fig 1A). Fixed cerebral organoids were subjected to immunohistochemical analyses, using GFP immunofluorescence to identify the targeted cNPCs and their progeny (Fig 1A and B). These organoids were mainly of telencephalic identity as indicated by the expression of the telencephalic marker FOXG1 (Fig 1C). Figure 1. Experimental protocol of cerebral organoid production and time points of electroporation and analyses A. Timeline of cerebral organoid production detailing media as well as time points of electroporation (beginning of green bars), duration of vector expression (lengths of green bars; 2, 4, 10 and 15 days), and time points of fixation and analysis (end of green bars) of cerebral organoids.

B. Left: Cartoon depicting the microinjection and electroporation of a ventricle-like structure of a cerebral organoid; Right: Immunofluorescence for GFP (green), combined with DAPI staining (white), of a 57-day-old chimpanzee cerebral organoid 2 days after electroporation with GFP expression plasmid plus control plasmid. Scale bar, 500 μm.

C. Double immunofluorescence for GFP (green) and the telencephalic marker FOXG1 (yellow), combined with DAPI staining (white), of a 57-day-old chimpanzee cerebral organoid 2 days after electroporation with GFP expression plasmid plus control plasmid. Scale bar, 150 μm.

D. Double immunofluorescence for SOX2 (magenta) and GFP (green), combined with DAPI staining (white), of a 57-day-old chimpanzee cerebral organoid 2 days after electroporation with GFP expression plasmid plus control plasmid (first row), of a 59-day-old chimpanzee cerebral organoid 4 days after electroporation with GFP expression plasmid plus control plasmid (second row), and a 61-day-old chimpanzee cerebral organoid 10 days after electroporation with GFP expression plasmid plus control plasmid (third row). Tick marks indicate the border between VZ and SVZ/NL. Scale bars, 50 μm. Representative examples of control vector-transfected chimpanzee cerebral organoids 2, 4, and 10 days after electroporation are presented in Fig 1D and Appendix Figs S1–S3. These images show that depending on the time after electroporation, different cell populations in distinct zones of the developing cerebral organoid wall contain GFP-positive cells. This GFP expression reveals the transfected APs and their progeny and hence indicates the cells that, either directly or by inheritance, would be affected by a given electroporated DNA. Two days after electroporation, the majority of the GFP-positive cells was still observed in the VZ, colocalizing with a marker of proliferating cNPCs, SOX2 (Fig 1D). A minority of the GFP-positive cells was already observed basal to the VZ in the SVZ and neuronal layers (NL), colocalizing with a marker of basal progenitors, TBR2 (Englund et al, 2005; Sessa et al, 2008; Appendix Fig S1), but barely colocalizing with a marker of deep-layer neurons, CTIP2 (Arlotta et al, 2005; Molyneaux et al, 2007; Appendix Fig S2), and not colocalizing with a marker of upper-layer neurons, SATB2 (Alcamo et al, 2008; Britanova et al, 2008; Appendix Fig S3). These data are consistent with the length of the cell cycle of APs observed in chimpanzee cerebral organoids of ≈ 2 days (Mora-Bermudez et al, 2016) and suggest that the GFP-positive cells observed in the VZ 2 days after electroporation were either targeted APs, daughter APs of targeted APs, newborn BPs derived from targeted APs, or (few) newborn deep-layer neurons derived from targeted progenitors. Four days after electroporation, GFP-positive cells were observed in the basal region of the VZ, at the boundary between VZ and SVZ, and in the SVZ and NL, largely colocalizing with either SOX2 (Fig 1D), TBR2 (Appendix Fig S1) or CTIP2 (Appendix Fig S2) but not with SATB2 (Appendix Fig S3). This suggests that the GFP-positive cells observed in the VZ 4 days after electroporation were daughter APs of targeted APs, BPs derived from targeted APs, or newborn deep-layer neurons derived from targeted progenitors. Ten days after electroporation, GFP-positive cells were observed mostly in the basal SVZ and NL, colocalizing rarely with SOX2 (Fig 1D), still somewhat with TBR2 (Appendix Fig S1), mostly with CTIP2 (Appendix Fig S2), but not yet often with SATB2 (Appendix Fig S3). This is consistent with the notion that this longer period after electroporation should allow the targeted APs to carry out multiple rounds of BP-generating cell divisions, with the resulting BPs carrying out neuron-generating divisions. Accordingly, after the 10-day period following electroporation, a greater proportion of the targeted AP progeny is neurons, mostly of the deep-layer type, than after the 4-day period following electroporation (Appendix Figs S2 and S3).

Expression of human-specific ARHGAP11B in chimpanzee cerebral organoids increases the abundance of cycling BPs Having established transfection of cerebral organoids as our test system, we first investigated whether, similar to the other non-human model systems of neocortex development previously used to study the effects of ARHGAP11B (Florio et al, 2015, 2016; Kalebic et al, 2018; Heide et al, 2020; Xing et al, 2021), ARHGAP11B would increase BP proliferation and abundance when expressed in chimpanzee cerebral organoids. For this purpose, we employed a previously used construct leading to ectopic expression of ARHGAP11B under the constitutive CAGGS promoter (pCAGGS-ARHGAP11B; Florio et al, 2015). (For details of this overexpression construct, and the justification and appropriateness of its use, please see Materials and Methods). We used the experimental protocol described above and in Fig 1 to determine the co-electroporation efficiency of the pCAGGS-EGFP and the pCAGGs-ARHGAP11B vectors in chimpanzee cerebral organoids. We found that ≥ 90% of the GFP-positive progeny of the targeted APs was also positive for ARHGAP11B by immunofluorescence at 2, 4, and 10 days after electroporation (Fig EV1A and B), indicating a high electroporation efficiency. Hence, we used this experimental protocol with a 2-day period between the electroporation of chimpanzee cerebral organoids with the ARHGAP11B expression vector and analysis of the transfected organoids by immunofluorescence for the BP marker TBR2. We found a marked, twofold increase in the proportion of the GFP-positive progeny of the targeted APs that were TBR2-positive, and hence newborn BPs, in the ARHGAP11B-transfected chimpanzee organoids in comparison to control-transfected organoids (Fig 2A and B). This increase by ≈ 10% points of the total GFP+ cells likely occurred at the expense of the APs, as ARHGAP11B has previously been shown to induce symmetric, consumptive BP-genic divisions of these cNPCs. These data therefore indicate that ARHGAP11B increases the abundance of BPs upon expression in chimpanzee cerebral organoids. Figure 2. Expression of ARHGAP11B in chimpanzee cerebral organoids increases the abundance of cycling BPs A. Double immunofluorescence for GFP (green) and the BP marker TBR2 (yellow), combined with DAPI staining (white), of a 57-day-old chimpanzee cerebral organoid 2 days after electroporation with GFP expression plasmid plus either control plasmid (top) or ARHGAP11B expression plasmid (bottom). Tick marks indicate the borders of the VZ and SVZ/NL; arrowheads indicate examples of GFP+ and TBR2+ double-positive cells. Scale bars, 50 μm.

B. Quantification of the proportion of GFP+ cells that are TBR2+ in 57-day-old chimpanzee cerebral organoids 2 days after electroporation with GFP expression plasmid plus either control plasmid (dark red) or ARHGAP11B expression plasmid (light red). Data are the mean of nine control and nine ARHGAP11B -transfected cerebral organoids of two independent batches each; error bars indicate SD; *** P < 0.001 (two-sided Student's t -test).

C. Triple immunofluorescence for GFP (green), the cycling cell marker Ki67 (magenta), and TBR2 (yellow), combined with DAPI staining (white), of a 59-day-old chimpanzee cerebral organoid 4 days after electroporation with GFP expression plasmid plus either control plasmid (top) or ARHGAP11B expression plasmid (bottom). Tick marks indicate the borders of the VZ and SVZ/NL; arrowheads indicate examples of GFP+, Ki67+ and TBR2+ triple-positive cells. Scale bars, 50 μm.

D. Quantification of the proportion of GFP+ cells in the SVZ/NL that are Ki67+ and TBR2+ double-positive in 59-day-old chimpanzee cerebral organoids 4 days after electroporation with GFP expression plasmid plus either control plasmid (dark red) or ARHGAP11B expression plasmid (light red). Data are the mean of eight control and eight ARHGAP11B-transfected cerebral organoids of four independent batches each; error bars indicate SD; *P < 0.05 (one-sided Wilcoxon rank sum test). Click here to expand this figure. Figure EV1. GFP and ARHGAP11B are co-expressed when co-electroporated in chimpanzee cerebral organoids A. Double immunofluorescence for GFP (green) and ARHGAP11B (magenta) of a 59-day-old chimpanzee cerebral organoid 4 days after electroporation with GFP expression plasmid plus ARHGAP11B expression plasmid. Note that GFP and ARHGAP11B immunofluorescence signals do not completely overlap, as GFP is localized in the cytoplasm, whereas ARHGAP11B is localized in mitochondria. Scale bar, 50 μm.

B. Quantification of the percentage of GFP+ cells that are ARHGAP11B+ in 57-, 59-, and 61-day-old chimpanzee cerebral organoids 2, 4, and 10 days after electroporation with GFP expression plasmid plus ARHGAP11B expression plasmid. Data are the mean of five, six, or seven ventricle-like structures 2, 4, or 10 days after electroporation of three ARHGAP11B-transfected cerebral organoids each; error bars indicate SD. To further analyze the effect of ARHGAP11B on BP abundance in chimpanzee cerebral organoids, we used the same experimental protocol as described above and in Fig 1, but this time with a 4-day period between electroporation and analysis. Compared with control, transfection of the chimpanzee cerebral organoids with ARHGAP11B caused, again, an increase in the proportion of the GFP-positive progeny of the targeted APs in the SVZ and NL that were Ki67+ and TBR2+ double-positive, that is, cycling BPs (Fig 2C and D). In this case, the increase in Ki67+ TBR2+ cells most likely occurred at the expense of newborn neurons in the SVZ and NL (see below). Taken together, these results indicate that, similar to results previously obtained in embryonic mouse (Florio et al, 2015; Xing et al, 2021), embryonic ferret (Kalebic et al, 2018), and fetal marmoset (Heide et al, 2020) neocortex, the human-specific gene ARHGAP11B can substantially increase the abundance of cycling BPs in developing cerebral cortex-like tissue of our closest living relative, the chimpanzee.

ARHGAP11B expression in chimpanzee cerebral organoids increases the abundance of bRG In principle, two different types of BPs can be distinguished, i.e., basal intermediate progenitors (bIPs) and basal radial glia (bRG, also called outer radial glia; Haubensak et al, 2004; Miyata et al, 2004; Noctor et al, 2004; Fietz et al, 2010; Hansen et al, 2010; Reillo et al, 2011; Betizeau et al, 2013). bRG, in particular, are thought to be key for mammalian neocortex evolution and to drive expansion and folding of the human neocortex (Lui et al, 2011; Borrell & Götz, 2014; Florio & Huttner, 2014; Dehay et al, 2015; Fernandez et al, 2016; Llinares-Benadero & Borrell, 2019). In light of the increase in BP abundance upon ARHGAP11B expression in chimpanzee cerebral organoids (Fig 2), we next asked whether this increase applied to bRG. For this purpose, we used immunofluorescence for HOPX, a marker of radial glia (Pollen et al, 2015; Vaid et al, 2018), and quantified, specifically in regions basal to the VZ, i.e., in the SVZ and NL, the proportion of the GFP-positive progeny of the targeted APs that were HOPX-positive, which would be indicative of bRG. Four days after electroporation, we found almost a doubling in this proportion in the ARHGAP11B-transfected chimpanzee cerebral organoids in comparison to control-transfected organoids (Fig 3A and B). This increase by ≈ 10% points presumably occurred, again, at the expense of newborn neurons in the SVZ and NL (see below). Figure 3. Expression of ARHGAP11B in chimpanzee cerebral organoids increases the abundance of HOPX-positive cells in the SVZ A. Double immunofluorescence for GFP (green) and the radial glia marker HOPX (yellow), combined with DAPI staining (white), of a 59-day-old chimpanzee cerebral organoid 4 days after electroporation with GFP expression plasmid plus either control plasmid (top) or ARHGAP11B expression plasmid (bottom). Tick marks indicate the borders of the VZ and SVZ/NL; arrowheads indicate examples of GFP+ and HOPX+ double-positive cells. Scale bars, 50 μm.

B. Quantification of the proportion of GFP+ cells in the SVZ/NL that are HOPX+ in 59-day-old chimpanzee cerebral organoids 4 days after electroporation with GFP expression plasmid plus either control plasmid (dark red) or ARHGAP11B expression plasmid (light red). Data are the mean of 10 control and 10 ARHGAP11B -transfected cerebral organoids of three independent batches each; error bars indicate SD; ** P < 0.01 (one-sided Wilcoxon rank sum test).

C. Double immunofluorescence for GFP (green) and HOPX (yellow), combined with DAPI staining (white), of a 61-day-old chimpanzee cerebral organoid 10 days after electroporation with GFP expression plasmid plus either control plasmid (top) or ARHGAP11B expression plasmid (bottom). Tick marks indicate the borders of the VZ and SVZ/NL; arrowheads indicate examples of GFP+ and HOPX+ double-positive cells. Scale bars, 50 μm.

D. Quantification of the proportion of GFP+ cells in the SVZ/NL that are HOPX+ in 61-day-old chimpanzee cerebral organoids 10 days after electroporation with GFP expression plasmid plus either control plasmid (dark red) or ARHGAP11B expression plasmid (light red). Data are the mean of five control and seven ARHGAP11B-transfected cerebral organoids of three independent batches each; error bars indicate SD; *P < 0.05 (two-sided Student's t-test). To corroborate this result, we used a 10-day period between electroporation and analysis to allow a larger proportion of the GFP-positive progeny to migrate to the SVZ and NL (see Fig 1A and D, and Appendix Figs S1–S3). We first analyzed the transfected chimpanzee cerebral organoids by PCNA immunofluorescence to quantify cycling cells in the SVZ. Compared with control, transfection of the chimpanzee cerebral organoids with ARHGAP11B caused a doubling in the proportion of the GFP-positive progeny of the targeted APs in the SVZ that were PCNA-positive (Fig EV2A and B), consistent with a doubling of cycling BPs also after this longer post-electroporation period. We next performed HOPX immunofluorescence and found a marked, threefold increase in the proportion of the GFP-positive progeny of the targeted APs in the SVZ that were HOPX-positive in the ARHGAP11B-transfected chimpanzee cerebral organoids in comparison to control-transfected organoids (Fig 3C and D). Figure 4. Expression of ARHGAP11B in chimpanzee cerebral organoids differentially affects the generation of deep-layer vs. upper-layer cortical neurons A. ARHGAP11B expression plasmid (bottom). Tick marks indicate the borders of the VZ and SVZ/NL; arrowheads indicate examples of GFP+ and Hu+ double-positive cells. Note that the same electroporated regions are depicted in Fig Double immunofluorescence for GFP (green) and the neuron marker Hu (magenta), combined with DAPI staining (white), of a 61-day-old chimpanzee cerebral organoid 10 days after electroporation with GFP expression plasmid plus either control plasmid (top) orexpression plasmid (bottom). Tick marks indicate the borders of the VZ and SVZ/NL; arrowheads indicate examples of GFP+ and Hu+ double-positive cells. Note that the same electroporated regions are depicted in Fig EV2 with a different marker (PCNA). Scale bars, 50 μm.

B. Quantification of the proportion of GFP+ cells that are Hu+ in 61-day-old chimpanzee cerebral organoids 10 days after electroporation with GFP expression plasmid plus either control plasmid (dark red) or ARHGAP11B expression plasmid (light red). Data are the mean of seven control and 11 ARHGAP11B -transfected cerebral organoids of two independent batches each; error bars indicate SD; ** P < 0.01 (two-sided Student's t -test).

C. Double immunofluorescence for GFP (green) and the deep-layer neuron marker CTIP2 (magenta), combined with DAPI staining (white), of a 59-day-old chimpanzee cerebral organoid 4 days after electroporation with GFP expression plasmid plus either control plasmid (top) or ARHGAP11B expression plasmid (bottom). Tick marks indicate the borders of the VZ and SVZ/NL; arrowheads indicate examples of GFP+ and CTIP2+ double-positive cells. Scale bars, 50 μm.

D. Quantification of the proportion of GFP+ cells in the SVZ/NL that are CTIP2+ in 59-day-old chimpanzee cerebral organoids 4 days after electroporation with GFP expression plasmid plus either control plasmid (dark red) or ARHGAP11B expression plasmid (light red). Data are the mean of 10 control and 10 ARHGAP11B -transfected cerebral organoids of two independent batches each; error bars indicate SD; *** P < 0.001 (two-sided Student's t -test).

E. Double immunofluorescence for GFP (green) and the upper-layer neuron marker SATB2 (magenta), combined with DAPI staining (white), of a 66-day-old chimpanzee cerebral organoid 15 days after electroporation with GFP expression plasmid plus either control plasmid (top) or ARHGAP11B expression plasmid (bottom). Tick marks indicate the borders of the VZ and SVZ/NL; arrowheads indicate examples of GFP+ and SATB2+ double-positive cells. Scale bars, 50 μm.

F. Quantification of the proportion of GFP+ cells in the SVZ/NL that are SATB2+ in 66-day-old chimpanzee cerebral organoids 15 days after electroporation with GFP expression plasmid plus either control plasmid (dark red) or ARHGAP11B expression plasmid (light red). Data are the mean of six control and four ARHGAP11B-transfected cerebral organoids of two independent batches each; error bars indicate SD; **P < 0.01 (one-sided Wilcoxon rank sum test). Source data are available online for this figure. Source Data for Figure 4 [embr202254728-sup-0006-SDataFig4.zip] Click here to expand this figure. Figure EV2. Expression of ARHGAP11B in chimpanzee cerebral organoids increases the abundance of PCNA-positive cells in the SVZ A. ARHGAP11B expression plasmid (bottom). Tick marks indicate the borders of the VZ and SVZ/NL; arrowheads indicate examples of GFP+ and PCNA+ double-positive cells. Note that the same electroporated regions are depicted in Fig Double immunofluorescence for GFP (green) and the cycling cell marker PCNA (yellow), combined with DAPI staining (white), of a 61-day-old chimpanzee cerebral organoids 10 days after electroporation with GFP expression plasmid plus either control plasmid (top) orexpression plasmid (bottom). Tick marks indicate the borders of the VZ and SVZ/NL; arrowheads indicate examples of GFP+ and PCNA+ double-positive cells. Note that the same electroporated regions are depicted in Fig 4 with a different marker (Hu). Scale bars, 50 μm.

B. Quantification of the proportion of GFP+ cells in the SVZ/NL that are PCNA+ in 61-day-old chimpanzee cerebral organoids 10 days after electroporation with GFP expression plasmid plus either control plasmid (dark red) or ARHGAP11B expression plasmid (light red). Data are the mean of seven control and 10 ARHGAP11B-transfected cerebral organoids of two independent batches each; error bars indicate SD; *P < 0.05 (two-sided Student's t-test). Source data are available online for this figure. In summary, we conclude that the increase in cycling BP abundance upon expression of the human-specific gene ARHGAP11B in chimpanzee cerebral organoids includes—similar to results obtained in animal model systems (Florio et al, 2015; Kalebic et al, 2018; Heide et al, 2020; Xing et al, 2021)—an increase in bRG abundance.

Increased abundance of cycling BPs in chimpanzee cerebral organoids upon ARHGAP11B expression is initially associated with reduced deep-layer neuron generation but eventually results in increased upper-layer neuron generation If the observed increase in the abundance of cycling BPs (Fig EV2A and B) in chimpanzee cerebral organoids upon ARHGAP11B expression reflected indeed an increased proliferation of BPs, that is, BPs dividing to generate more BPs rather than neurons, one would expect a concomitant reduction in the generation of neurons from these cNPCs, as implied in our interpretation of the data in Figs 1 and 2. We explored this possibility by performing immunohistochemistry for neuronal markers 10 days after electroporation, to allow neuron generation from BPs to proceed. Compared with control, expression of ARHGAP11B in chimpanzee cerebral organoids resulted in a reduction in the proportion of the GFP-positive progeny of the targeted APs that were positive for the neuron markers Hu (Fig 4A and B) and NeuN (Fig EV3A and B). This reduction by ≈ 10% points was consistent with the increases by in BPs and bRGs observed in Figs 1 and 2. In line with this GFP-positive progeny being neurons, the majority of the Hu-positive (Fig 4A) and NeuN-positive (Fig EV3A) cells were located basal to the SVZ in the NL. We conclude that the increased abundance of cycling BPs observed after a 10-day period of expression of ARHGAP11B in chimpanzee cerebral organoids reflects an increased generation of BPs from BPs, at the expense of—and hence resulting in a reduction in—the generation of cortical neurons from BPs during this time period. Click here to expand this figure. Figure EV3. Expression of ARHGAP11B in chimpanzee cerebral organoids decreases the abundance of NeuN-positive cells A. Double immunofluorescence for GFP (green) and the neuron marker NeuN (magenta), combined with DAPI staining (white), of a 61-day-old chimpanzee cerebral organoid 10 days after electroporation with GFP expression plasmid plus either control plasmid (top) or ARHGAP11B expression plasmid (bottom). Tick marks indicate the borders of the VZ and SVZ/NL; arrowheads indicate examples of GFP+ and NeuN+ double-positive cells. Scale bars, 50 μm.

B. Quantification of the proportion of GFP+ cells that are NeuN+ in 61-day-old chimpanzee cerebral organoids 10 days after electroporation with GFP expression plasmid plus either control plasmid (dark red) or ARHGAP11B expression plasmid (light red). Data are the mean of seven control and 10 ARHGAP11B-transfected cerebral organoids of two independent batches each; error bars indicate SD; *P < 0.05 (two-sided Student's t-test). In the developing neocortex, the generation of cortical neurons begins with the production of deep-layer neurons, followed by the production of upper-layer neurons (Molyneaux et al, 2007; Cooper, 2008; Agirman et al, 2017). In light of the decrease in neuron generation upon ARHGAP11B expression in chimpanzee cerebral organoids, we investigated whether this decrease applied to the generation of deep-layer neurons, by performing immunohistochemistry for the deep-layer neuron marker CTIP2 4 days after the electroporation of the organoids. This time period should be just sufficient for two sequential rounds of cNPC division, (i) targeted APs generating GFP-positive BPs, and (ii) GFP-positive BPs generating either GFP-positive neurons (control) or GFP-positive BPs (ARHGAP11B). Quantification of the occurrence of CTIP2 in GFP+ cells indicated that compared with control, expression of ARHGAP11B in chimpanzee cerebral organoids resulted in a reduction in the proportion of the GFP-positive progeny of the targeted APs in the SVZ and NL that were positive for this deep-layer neuron marker (Fig 4C and D). We conclude that the decrease in neuron generation observed upon ARHGAP11B expression in chimpanzee cerebral organoids involves the generation of deep-layer neurons. Previous studies on ARHGAP11B showed that its expression in developing mouse (Xing et al, 2021), ferret (Kalebic et al, 2018), and marmoset (Heide et al, 2020) neocortex can increase the abundance of upper-layer neurons—the cortical neuron type that expanded disproportionally during primate evolution (Hutsler et al, 2005; Molnar et al, 2006; Fame et al, 2011). Hence, the question arises if expression of ARHGAP11B in chimpanzee cerebral organoids can eventually increase the generation of upper-layer neurons. To address this, we extended the time period between electroporation and analysis from 10 to 15 days (Fig 1A), as 10 days after electroporation GFP-positive cells were barely colocalizing with the upper-layer neuron marker SATB2 (Appendix Fig S3). The rationale for this extension was twofold. First, it can be assumed that due to the repeated divisions of cNPCs during this 15-day period, notably after 10 days (see Fig EV1), the original level of the ARHGAP11B expression plasmid introduced into the targeted APs by electroporation and inherited by the BPs generated therefrom would decline over time, resulting in progressively lower ARHGAP11B levels in the BPs. This in turn should progressively reduce the tendency of BPs to undergo symmetric proliferative rather neuron-generating divisions and increasingly promote the latter type of BP division. Second, during this 15-day period, in particular after 10 days, the BPs should either switch from deep- to upper-layer neuron production or at least start to produce also upper-layer neurons. If so, the increase in BP abundance due to ARHGAP11B's action during the early phase of this 15-day period should result in an increase in upper-layer neuron production by the end of this period. To this end, the electroporated chimpanzee organoids were analyzed by immunofluorescence for the upper-layer neuron marker SATB2. Indeed, compared with control, expression of ARHGAP11B in chimpanzee cerebral organoids resulted in an increase in the proportion of the GFP-positive progeny of the targeted APs in the SVZ and NL that were positive for SATB2 (Fig 4E and F). In line with the data shown in Fig 4D, this increase likely occurred at the expense of deep-layer neuron production. We conclude that expression of ARHGAP11B in chimpanzee cerebral organoids eventually leads to an increase in upper-layer neuron generation. We have so far implied that the observed decrease in deep-layer neuron production observed 4 days after ARHGAP11B expression and the observed increase in upper-layer neuron production 15 days after electroporation are linked to ARHGAP11B's action in BPs that fades with time. Could these changes in deep-layer vs. upper-layer neuron levels be also, or perhaps only, be due to an effect of ARHGAP11B on neuronal fate? We find this scenario unlikely, as all previous studies from our lab have shown that ARHGAP11B, consistent with its action in mitochondria to increase glutaminolysis, only affects cycling cells, not post-mitotic cells such as neurons (Florio et al, 2015, 2016; Kalebic et al, 2018; Heide et al, 2020; Namba et al, 2020; Xing et al, 2021).

Dominant-negative inhibition of ARHGAP11B's function in human cerebral organoids reduces cycling BPs to the chimpanzee level The increase in cycling BP levels upon ARHGAP11B expression in chimpanzee cerebral organoids (Figs 1 and 2) is fully consistent with previous findings in another primate model, i.e., transgenic marmoset fetuses with physiological-like ARHGAP11B expression (Heide et al, 2020), and provides strong further support for the notion that this human-specific gene is a prime candidate to have caused neocortex expansion in the course of human evolution. We chose two approaches to determine to which extent the cycling BP levels in human cerebral cortex tissue depend on ARHGAP11B and hence to gain insight into ARHGAP11B's contribution to the expansion of the human neocortex during development and evolution. First, we made use of a truncated form of the ARHGAP11A protein (ARHGAP11A220) that has previously been shown to act in a dominant-negative manner on ARHGAP11B's ability to amplify BPs (Namba et al, 2020). This dominant-negative action on ARHGAP11B and not on ARHGAP11A can be explained by the following two findings. (i) ARHGAP11A220, similar to ARHGAP11B and in contrast to full-length ARHGAP11A, localizes to mitochondria and not (like ARHGAP11A) to the nucleus (Namba et al, 2020). (ii) ARHGAP11A220, via its truncated GAP domain, can interact with the same downstream effector system in mitochondria as ARHGAP11B, however, without being able to change its activity, which requires the human-specific C-terminal domain of ARHGAP11B (Namba et al, 2020). We examined the effects of ARHGAP11A220 on cycling BP levels in human cerebral organoids 2 days after electroporation of the corresponding expression plasmid, with a 1 h BrdU pulse prior to fixation (Fig 1A). The pattern of SOX2 immunostaining was used to distinguish cNPCs in the VZ vs. SVZ and to attribute the GFP-positive progeny to either of these two germinal zones (Fig 1D). We found that compared with control, transfection of the cNPCs in the VZ of human cerebral organoids with the dominant-negative ARHGAP11A220 resulted in a marked reduction, virtually down to the level observed in chimpanzee cerebral organoids, in the proportion of the GFP-positive progeny of the targeted APs found in the SVZ that had incorporated BrdU (Fig 5A and B). As inhibition of ARHGAP11B function is not known to result in a shortening of S-phase (and hence in reduced BrdU incorporation), these data suggest that inhibition of ARHGAP11B function reduces the level of cycling BPs. We therefore conclude that ARHGAP11B is required to maintain the elevated level of cycling BPs in human cerebral organoids. Figure 5. Dominant-negative ARHGAP11A220 reduces the level of cycling BPs in ARHGAP11B-expressing human cerebral organoids to the chimpanzee level but has no effect on BP levels in chimpanzee cerebral organoids A. ARHGAP11A220 expression plasmid (lower panels). Tick marks indicate the borders of the VZ and SVZ/NL; arrowheads indicate examples of GFP+ and BrdU+ double-positive cells. Scale bar, 50 μm. Double immunofluorescence for GFP (green) and the thymidine analog BrdU (magenta, see Fig 1A ), combined with DAPI staining (white), of 57-day-old human (top panels) and chimpanzee (bottom panels) cerebral organoids 2 days after electroporation with GFP expression plasmid plus either control plasmid (upper panels) orexpression plasmid (lower panels). Tick marks indicate the borders of the VZ and SVZ/NL; arrowheads indicate examples of GFP+ and BrdU+ double-positive cells. Scale bar, 50 μm.

B. Quantification of the proportion of GFP+ cells in the SVZ/NL that are BrdU+ in 57-day-old human (blue) and chimpanzee (red) cerebral organoids 2 days after electroporation with GFP expression plasmid plus either control plasmid (dark blue and dark red) or ARHGAP11A220 expression plasmid (light blue and light red). Data are the mean of 12 human control, 12 human ARHGAP11A220 -transfected, 9 chimpanzee control, and 8 chimpanzee ARHGAP11A220 -transfected cerebral organoids (grown from one human and one chimpanzee iPSC line, respectively) of six (human) and four (chimpanzee) independent batches each; error bars indicate SD; n.s., not significant; ** P < 0.01 (Kruskal–Wallis test).

C. Double immunofluorescence for GFP (green) and TBR2 (yellow), combined with DAPI staining (white), of 57-day-old human (top panels) and chimpanzee (bottom panels) cerebral organoids 2 days after electroporation with GFP expression plasmid plus either control plasmid (upper panels) or ARHGAP11A220 expression plasmid (lower panels). Tick marks indicate the borders of the VZ and SVZ/NL; arrowheads indicate examples of GFP+ and TBR2+ double-positive cells. Scale bar, 50 μm.

D. Quantification of the proportion of GFP+ cells in the SVZ/NL that are TBR2+ in 57-day-old human (blue) and chimpanzee (red) cerebral organoids 2 days after electroporation with GFP expression plasmid plus either control plasmid (dark blue and dark red) or ARHGAP11A220 expression plasmid (light blue and light red). Data are the mean of 9 human control, 9 human ARHGAP11A220-transfected, 10 chimpanzee control, and 10 chimpanzee ARHGAP11A220-transfected cerebral organoids (grown from one human and one chimpanzee iPSC line, respectively) of four (human) and three (chimpanzee) independent batches each; error bars indicate SD; n.s., not significant; *P < 0.05 (Kruskal–Wallis test). To corroborate this conclusion, we analyzed the ARHGAP11A220-transfected human cerebral organoids for GFP-positive cells containing the BP marker TBR2, which in fetal human neocortex is typically expressed in the bIP subpopulation of BPs (Fietz et al, 2010). Transfection of the human cerebral organoids with ARHGAP11A220 caused after 2 days a reduction down to half of control in the proportion of the GFP-positive progeny of the targeted APs in the SVZ and NL that were TBR2-positive (Fig 5C and D). This decrease in the proportion of BPs among the GFP+ cells in SVZ and NL implied an apparent increase in the proportion of TBR-negative cells, which we believe reflected an actually constant proportion of neurons among the GFP+ cells in SVZ and NL due to a decrease in the absolute pool size of GFP+ cells. Interestingly, for the human and chimpanzee iPSC lines used in the present study to generate cerebral organoids, this reduction brought the level of the BrdU+ BPs (Fig 5B) and the TBR2+ BPs (Fig 5D) down to that observed in control chimpanzee cerebral organoids. We therefore conclude that ARHGAP11B is required to maintain the elevated level of cycling BPs that is characteristic of human cerebral organoids.

ARHGAP11A220 specifically inhibits ARHGAP11B function We electroporated ARHGAP11A220 into chimpanzee cerebral organoids, which lack ARHGAP11B, to determine whether the effects of ARHGAP11A220 observed in human cerebral organoids were specific for ARHGAP11B. Indeed, 2 days after transfection of the APs in the VZ of chimpanzee cerebral organoids with ARHGAP11A220 vs. control, we observed no change in the proportion of the GFP-positive progeny of the targeted APs in the SVZ, i.e., of BPs, that had incorporated BrdU (Fig 5A and B), or that were TBR2-positive (Fig 5C and D). Hence, the reduction in the level of cycling BPs observed upon transfection of human cerebral organoids with the dominant-negative ARHGAP11A220 reflected a specific inhibition of ARHGAP11B's ability to amplify BPs.