Aspm knockout ferret reveals an evolutionary mechanism governing cerebral cortical size

Abstract

The human cerebral cortex is distinguished by its large size and abundant gyrification, or folding. However, the evolutionary mechanisms that drive cortical size and structure are unknown. Although genes that are essential for cortical developmental expansion have been identified from the genetics of human primary microcephaly (a disorder associated with reduced brain size and intellectual disability) ¹ , studies of these genes in mice, which have a smooth cortex that is one thousand times smaller than the cortex of humans, have provided limited insight. Mutations in abnormal spindle-like microcephaly-associated (ASPM), the most common recessive microcephaly gene, reduce cortical volume by at least 50% in humans^2-4, but have little effect on the brains of mice^5-9; this probably reflects evolutionarily divergent functions of ASPM^10,11. Here we used genome editing to create a germline knockout of Aspm in the ferret (Mustela putorius furo), a species with a larger, gyrified cortex and greater neural progenitor cell diversity^12-14 than mice, and closer protein sequence homology to the human ASPM protein. Aspm knockout ferrets exhibit severe microcephaly (25-40% decreases in brain weight), reflecting reduced cortical surface area without significant change in cortical thickness, as has been found in human patients^3,4, suggesting that loss of 'cortical units' has occurred. The cortex of fetal Aspm knockout ferrets displays a very large premature displacement of ventricular radial glial cells to the outer subventricular zone, where many resemble outer radial glia, a subtype of neural progenitor cells that are essentially absent in mice and have been implicated in cerebral cortical expansion in primates^12-16. These data suggest an evolutionary mechanism by which ASPM regulates cortical expansion by controlling the affinity of ventricular radial glial cells for the ventricular surface, thus modulating the ratio of ventricular radial glial cells, the most undifferentiated cell type, to outer radial glia, a more differentiated progenitor.

Extended Data Figure 1. Cytoarchitecture and neuronal subtype lamination in the mature Aspm KO ferret cortex

a, Nissl stains of P41 littermate coronal sections, as shown in Fig. 1l, with additional Aspm+/− and Aspm−/− littermates shown. b, c, P41 littermates immunostained for cortical layer-specific projection neurons including Satb2 (layer II–IV), Ctip2 (layer V), and FoxP2 (layer VI). The experiments were repeated 3 times independently with similar results. Scale bars: 2 mm (a, top), 200 μm (a, bottom), 100 μm (b–c).

Extended Data Figure 2. Sox2/Ki67 immunostaining in additional E35 and P0 littermates dorsal cortex

Extending findings of Fig. 2e, f, each set of panels is from a separate littermate, showing the high penetrance of the neural progenitor cell basal displacement phenotype. The experiments were repeated 3 times independently with similar results. Scale bar: 200 μm.

Extended Data Figure 3. Displaced progenitors in Aspm KO ferrets have basal fibers

Extending findings of Fig. 2h, i, immunostaining of Sox2, Ki67, and Vim shows that displaced neural progenitors have basal radial fibers. The experiments were repeated 3 times independently with similar results. Scale bar: 100 μm.

Extended Data Figure 4. Aspm KO mice do not demonstrate displaced progenitors in the IZ

Unlike Aspm−/− ferrets, Aspm−/− mice do not demonstrate displaced NPC in the IZ. However, they show a variable increase in IP (Pax6-/Ki67+ cells in a and Tbr2+ cells in b), which is enhanced by heterozygous, compound mutation in Wdr62, a microcephaly gene causing more severe microcephaly. The experiments were repeated 3 times independently with similar results. Scale bar: 100 μm.

Extended Data Figure 5. Modest increase in apoptosis throughout the germinal zones of the Aspm KO telencephalon

Apoptotic cells (yellow) are indicated by enzymatic fluorescent detection of double-stranded DNA damage with DAPI nuclear counterstaining (blue). The experiments were repeated 3 times independently with similar results. Scale bars: 500 μm (a, whole section) and 100 μm (b, c, cortical wall columns).

Extended Data Figure 6. Additional immunohistochemical analyses of displaced progenitors in the Aspm KO cortex

a, E35 KO cortex stained for VRG/ORG markers Sox2 and Hopx reveals extensive co-labeling in both the VZ and SVZ, including in displaced OSVZ progenitors. b, In the E35 KO OSVZ, clusters of supernumerary displaced neural progenitor cells include numerous Tbr2+ IP and are surrounded by Dcx+ newborn neurons, indicating preserved neurogenesis within the precocious OSVZ niche of the Aspm KO cortex. The experiments were repeated 3 times independently with similar results. Scale bar: 50 μm.

Extended Data Figure 7. Single-cell RNA-seq batch, sample, and cluster analyses

a, tSNE plot from Fig. 3a with cells colored by biological replicate (i.e., animal). Most clusters include cells from all samples, except for a cluster expressing blood genes and a cluster expressing choroid plexus epithelial cells that are mostly from animal WT5E. These two cell clusters were not included in the downstream analysis. b, tSNE plot from Fig. 3a with cells colored by the batch they were processed in. Clusters are composed of cells from all batches. c, Per cell gene count and unique molecular identifier (UMI) count per sample. Each violin plot is one biological replicate and each dot is one cell. Sample WT5D was not included in the analysis due to the lower gene and UMI count compared to other samples as well as the inconsistent clustering compared to other WT samples (data not shown). d, Per cell gene count and UMI count for identified clusters. Each violin plot is one cell cluster and each dot is one cell. The three clusters in grey (EN4, BL, CPE) were not included in the downstream analysis. See Methods for details. This scRNA-seq experiment was performed once with n = 22,211 cells (8,037 cells from 2 Aspm+/+ and 1 +/− ferrets, and 14,174 cells from 4 Aspm−/− ferrets). RG1, cycling radial glial progenitors; RG2, interphase radial glial progenitors; IP, intermediate progenitors; EN1, upper-layer excitatory neurons; EN2, deep-layer excitatory neurons; EN3, Cajal-Retzius cells; IN1, immature inhibitory neurons; IN2, SST+ inhibitory neurons; IN3, ventral/inhibitory progenitors; ENDO1, endothelial cells 1; ENDO2, endothelial cells 2; OPC, oligodendrocyte precursors; MG, microglia; EN4, mixed excitatory neuron identity; BL, blood cells; CPE, choroid plexus epithelial cells.

Extended Data Figure 8. Loss of Aspm disrupts centriole duplication in ferret embryonic fibroblasts

Mitotic Aspm KO FEF, identified by staining for pH3 and co-stained for the centriolar marker Centrin, display a significant loss of centrioles. The percentage of cells with an abnormal number (less than 4) of centrioles is increased 8-fold in Aspm−/− FEF compared to Aspm+/+ FEF (n = 100 cells/genotype/independent experiments; P = 0.003). The experiments were repeated 3 times independently with similar results. Statistical analysis was performed by two-tailed t-test and the graph shows mean ± s.e.m.

Figure 1. Aspm KO ferrets robustly model human microcephaly

a, NPC diversity in humans, ferrets, and mice. b, c, ASPM protein is highly similar between humans and ferrets, including the number of IQ domains (c, in parentheses). d, Ferret Aspm gene showing targeted sequences (blue highlights) and founder frameshift deletions. e, Loss of Aspm in KO embryonic fibroblasts. f, Aspm+/− and −/− littermate brains. g–k, MRI segmentations of grey and white matter (g), gyri grouped into 4 regions (h), horizontal and coronal sections (i), and quantification of volume (j) and cortical surface area (k). For abbreviations, see Methods. (*, P < 0.05; n = 3/genotype). l–p, Aspm−/− ferrets show reduced brain weight (n, **, P < 0.005; *, P < 0.01; n = 3–17/genotype/age group) but cytoarchitecture (l), laminar organization (m), cortical thickness (o, n = 6/genotype) and body weight (p, n = 3/genotype) are preserved. q, Loss of Aspm decreases outer cortical surface area in ferrets, not in mice (n = 3/genotype) (*, P = 0.0217). Except the box plot (n), graphs show mean ± s.e.m. See Methods, Extended Data Table 1, and Source Data (graphs) for statistics and reproducibility. Scale bars: e, 10 μm; f–i, 5 mm; l, 1 mm; m, 100 μm.

Figure 2. Aspm KO ferrets show displaced NPC

a–f, Nuclear staining of Aspm−/− ferrets shows a premature OSVZ-like zone (a–c, arrowheads) containing NPC that express Pax6, Sox2, and Ki67 (d–f). g–k, Displaced NPC include Sox2+/pVim+ ORG (g, arrowheads) with a basal process (g, arrows; h, i, k), and Tbr2+ IP. Abventricular pVim+ NPC are increased 3-fold in Aspm−/− ferrets (j) (*, P = 0.006 by one-tailed t-test; mean ± s.e.m; n = 3 +/− and n = 4 −/− animals). l–o, Displaced NPC express Vim, Eomes, or Ptprz1 (l, m), have Arl13b+ cilia (o), and are either Sox2+/Neurog2+/Hopx+ (filled arrowheads) or Sox2+/Neurog2+/Hopx- (open arrowheads) (n). p, q, Aspm−/− mice lack displaced NPC. See Methods for statistics and reproducibility. Scale bars: a–c, 500 μm; d–f, 50 μm; g–i, n, o, 10 μm; k–m, p, q, 100 μm.

Figure 3. Loss of Aspm changes cell type proportions but not transcriptional programs

a, scRNA-seq identifies major cell types at E35. (For abbreviations and statistics, see Extended Data Table 2.) b, Cell type markers for each cluster. c, Proportions of each cell type, with the largest changes indicated by black outlines (bootstrap FDR < 0.01). d, Aspm is enriched at the VZ apical surface; Eomes+ IP are increased in the KO SVZ. e,f, Apod+ OPC are increased in Aspm−/− ferrets (*, P = 0.012 by one-tailed t-test; n = 4 animals/genotype). See Methods for statistics and reproducibility. The graph shows mean ± s.e.m. Asterisks in d, blood vessels. Scale bars: 100 μm.

Figure 4. Aspm controls localization of apical polarity complex proteins to the centrosome

a, aPKCζ at the ventricular surface is decreased in Aspm−/− mice. b, Depletion of ASPM in H4 cells prevents recruitment of aPKCζ and PAR6α to the centrosomes. c, ASPM and aPKCζ co-immunoprecipitate from Hela cells. d, e, Loss of Aspm decreases ventricular surface staining for Ninein. f, Depletion of ASPM and aPKCζ prevents recruitment of Ninein to the centrosomes in H4 cells. g, Levels of aPKCζ, Par6α, and Ninein are unchanged in Aspm−/− mouse embryonic fibroblasts. See Methods for reproducibility. For gel source data, see Supplementary Data. Scale bars: a, 50 μm; b–f, 5 μm.