Drosophila Aurora-A kinase inhibits neuroblast self-renewal by regulating aPKC/Numb cortical polarity and spindle orientation

Abstract

Regulation of stem cell self-renewal versus differentiation is critical for embryonic development and adult tissue homeostasis. Drosophila larval neuroblasts divide asymmetrically to self-renew, and are a model system for studying stem cell self-renewal. Here we identify three mutations showing increased brain neuroblast numbers that map to the aurora-A gene, which encodes a conserved kinase implicated in human cancer. Clonal analysis and time-lapse imaging in aurora-A mutants show single neuroblasts generate multiple neuroblasts (ectopic self-renewal). This phenotype is due to two independent neuroblast defects: abnormal atypical protein kinase C (aPKC)/Numb cortical polarity and failure to align the mitotic spindle with the cortical polarity axis. numb mutant clones have ectopic neuroblasts, and Numb overexpression partially suppresses aurora-A neuroblast overgrowth (but not spindle misalignment). Conversely, mutations that disrupt spindle alignment but not cortical polarity have increased neuroblasts. We conclude that Aurora-A and Numb are novel inhibitors of neuroblast self-renewal and that spindle orientation regulates neuroblast self-renewal.

Figure 1.

aurora-A mutants have too many larval neuroblasts. (A,B) Larval brains at 96 h ALH showing increased Mira⁺ neuroblasts in the mutant central brain (boxed) and thorax (T). (A′,B′) Pupal brains at 12–18 h after pupariation showing increased numbers of proliferative Dpn⁺ neuroblasts and expanded CNS size. (C) Quantification of brain neuroblast numbers (using Dpn or Mira markers). (D) Molecular lesions in aurA alleles superimposed on the crystal structure of human Aurora-A (Cheetham et al. 2002; Nowakowski et al. 2002). Lightly shaded region is absent in aurA⁸⁸³⁹. (E,F) aurA mutant neuroblasts are correctly specified. Wild-type (E) and aurA¹⁴⁶⁴¹ (F) neuroblasts are Dpn⁺ Elav⁻ and incorporate BrdU, showing that they are specified correctly and are proliferative. (G–J) aurA mutant neuroblasts can proliferate and generate neuronal progeny. Wild-type (G,H) and aurA¹⁴⁶⁴¹ (I, J) 96-h-ALH larvae incorporate BrdU into neuroblasts following a 4-h BrdU pulse (G,I), and “chase” the BrdU into post-mitotic neurons following a 24-h BrdU-free chase (H,J). (K–M) aurA mutant neuroblasts generate multiple neuroblasts. Single-neuroblast clones identified by βgal expression (circled with the dotted line). (K) In wild type, clones contain one Dpn⁺ neuroblast and many nuclear Pros⁺ progeny. (L) In aurA¹⁴⁶⁴¹, clones contain multiple Dpn⁺ neuroblasts and nuclear Pros⁺ progeny. (M) Summary. (Green) Neuroblasts; (red) GMCs.

Figure 2.

aurora-A neuroblasts have ectopic aPKC and delocalized Numb. (A–C,H–K) Wild-type mitotic neuroblasts at 120 h ALH show normal apical protein localization (A–C) and basal protein localization (H–K). (D–G,L–P) aurA⁸⁸³⁹ mutant mitotic neuroblasts at 120 h ALH show delocalization of aPKC (26%, n = 47) and Numb (34%, n = 69). Baz, Pins, Mira, Pros, Brat (B,C,I–K) and Gαi, Insc, and Pon (data not shown) have essentially normal localization.

Figure 3.

Numb acts downstream from AurA to inhibit neuroblast self-renewal. (A,B) Numb inhibits neuroblast self-renewal. (A) Wild-type neuroblast clones always contain a single neuroblast and multiple differentiated progeny. (B) numb² mutant clones contain multiple neuroblasts. Clones were induced by standard methods at 0–24 h ALH and processed at 72 h ALH for mCD8 (clone marker), Mira (neuroblast marker), and Pros (GMC/neuron marker). (C–E) Numb overexpression can suppress supernumerary neuroblast formation in aurA mutants. (C) Quantification of Mira⁺ neuroblast numbers for a 120-h-ALH brain lobe of the indicated genotypes. (D,E) Single optical section 8 μm below the dorsal surface of a 120-h-ALH brain lobe stained for the Mira neuroblast marker in aurA¹⁴⁶⁴¹ (D) or aurA¹⁴⁶⁴¹ wor-gal4 UAS-numb (E) larvae.

Figure 4.

aurora-A neuroblasts and SOPs fail to align the mitotic spindle with the cortical polarity axis. (A–F) aurA mutant neuroblasts fail to align the mitotic spindle with the cortical polarity axis. (A) Wild-type metaphase (A) or telophase (E) 120-h-ALH neuroblasts showing tight alignment of the mitotic spindle (αtub, α-tubulin) with the apical/basal cortical polarity axis (Mira, Miranda basal marker). (B,C,F) aurA⁸⁸³⁹ mutant larval metaphase (B,C) or telophase (F) 120-h-ALH neuroblasts showing misalignment of the mitotic spindle relative to the apical/basal cortical polarity axis. Quantification of metaphase spindle orientation is shown at right (black, >40% within the 15° sector; gray, 1%–39%; white, 0%). (G,H) The Mud spindle anchoring protein is cortical and centrosomal in both wild-type (G) and aurA¹⁴⁶⁴¹ (H) metaphase neuroblasts, although aurA¹⁴⁶⁴¹ mutants have increased Mud staining on the spindle. (Right panel) Summary. (I,J) Wild-type and aurA¹⁴⁶⁴¹ mutant mitotic SOPs at 16–20 h APF; anterior up, labeled for the indicated cortical polarity marker (magenta) and the α-tubulin spindle marker (green). The spindle is often orthogonal to the cortical polarity axis. Quantification of metaphase spindle orientation is shown at right.

Figure 5.

Figure 5. aurora-A neuroblasts undergo symmetric cell division to expand the neuroblast population. (A–D) aurA mutant neuroblasts (NB) can divide symmetrically. (A) Wild-type neuroblasts divide asymmetrically. (B) The majority of aurA mutant neuroblasts divide asymmetrically. (C) A minority of aurA mutant neuroblasts undergo symmetric cell division. Bar, 5 μm. (D) Quantification of neuroblast sibling cell size and nuclear size in wild type and aurA mutants. Data are derived from movies of wild-type and aurA¹⁴⁶⁴¹/aurA¹⁷⁹⁶¹ neuroblasts expressing histone:GFP (His2AvD:GFP) under its native promoter (Clarkson and Saint 1999). Neuroblasts and GMCs are outlined in the first and last panel. 0:00:00 (hours:minutes:seconds) indicates nuclear envelope breakdown; colored boxes indicate cell cycle stages. (E) Example of an aurA neuroblast lineage tree. The neuroblast divides twice symmetrically and then undergoes a fully asymmetric cell division. White numbers are the ratio of sibling cell diameters; 1.0–1.25, symmetric; >1.5, asymmetric. (Gray arrow) Neuroblast initiated division in the last frame of the movie; (black arrow) neuroblast remained in interphase. (F) aurA mutants have delays in cell cycle progression. aurA neuroblasts dividing symmetrically or asymmetrically have similar delays and are pooled for analysis. (Blue) Prometaphase; (yellow) metaphase. Numbers are in minutes:seconds (mean values); n ≥ 11 for each genotype.

Figure 6.

Figure 6. AurA inhibits neuroblast self-renewal. (A) AurA inhibits aPKC basal localization and promotes Numb basal localization; Numb inhibits neuroblast self-renewal. AurA independently promotes spindle–cortex alignment. Although aPKC mutants affect all known basal proteins (Numb, Mira, Brat, and Pros), aurA mutants only show detectable changes in Numb localization. (B) Schematic illustrating the consequences of losing one or both AurA pathways for inhibiting neuroblast self-renewal. (Top line) aurA mutants have defects in both basal Numb localization and spindle alignment; loss of both pathways leads to massive increases in neuroblast numbers. (Middle line) aurA mutants overexpressing Numb have defects only in the spindle alignment pathway and have a smaller increase in neuroblast numbers (comparable to mud or cnn mutants) (Fig. 3C; Siller et al. 2006; Bowman et al. 2006). (Bottom line) Wild-type neuroblast numbers. Time axis depicts 0–120 h ALH.