Figure 8. Tolls regulate brain size in the pupal and adult brain.

(A,B) Constant RNAi knock-down of three Tolls with Toll-2GAL4 resulted in smaller brains. Optic lobes were smaller than in controls, and so were central brains (Figure 8—figure supplement 1A). (B) Brain area box-plots: both are One Wat ANOVA p<0.01, post-hoc Dunnett test comparisons to control. (C,D) In the triple Toll RNAi knock-out, Kenyon cell clusters were disorganized, with smaller area (Figure 8—figure supplement 1B) whilst deeper along the A/P axis, and (E) had fewer cells (albeit not significantly). (D) Box-plots, Kruskal-Wallis ANOVA, p=0.0127, post-hoc Dunn. (F–H) Conditional knock-down of two Tolls with tubGAL80^ts, Toll-2GAL4 at the adult critical period disorganized Kenyon cell clusters and altered Kenyon cell number: Toll-2,–6 RNAi reduced cell number, whereas Toll-6 ,–7 RNAi increased KC depth and cell number. (G) Box-plots, Kruskal Wallis ANOVA p<0.0001, post-doc Dunn. (H) Box-plots, One Way ANOVA p=0.0567, post-hoc Dunnett. (I–K) Conditional alteration of Toll-2 function with tubGAL80^ts, MyD88GAL4 at the adult critical period (day 0–2) increased brain size: (J) Left: over-expression of Toll-2 increased brain size, using two different UAS-Toll-2 lines: EP709 (2^nd chromosome) and an insertion into attP2 (3^rd chromosome). Middle: both over-expression of wek and MyD88 RNAi increased brain size, but over-expression of both did not, possibly because this genotype can also induce cell death. Right: over-expression of yki-RNAi alone resulted in a variable phenotype, but rescued the increase in brain size caused by Toll-2 over-expression. Box-plots, dashed lines indicate median of controls: left, One Way ANOVA, p=0.0793, not significant; centre, Kruskal Wallis ANOVA p<0.0001, post-hoc Dunn; right, One Way ANOVA p<0.05, post-hoc Tukey. (K) Mean increases in brain size ranged from 5.4% to 20% compared to controls, but most remarkably there was a persistent increase in the percentage of brains larger than the control mean of 138,585μ² upon MyD88 RNAi knock-down, over-expression of Toll-2 or wek. yki RNAi knockdown fully rescued the Toll-2 gain of function phenotype. Scale bar in (A,I): 50 μ; (C,F): 25 μ.*p<0.05, **p<0.01, ***p<0.001, ****p<0.0001. Two different RNAi lines were used for each Toll. (B,D,E,G,H) Dashed lines indicate median, maximum and minimum values in controls. For genotypes, sample sizes and statistical details, see Supplementary file 2. See Figure 8—figure supplement 1.

Figure 8—figure supplement 1. Conditional knock-down of Toll-2 and −6 in the adult brain altered cell number and brain size.

(A) Triple Toll RNAi knock-down with Toll-2GAL4 reduced central brain area, meaning it affects the central brain as well as the optic lobes shown in Figure 8A,B (images shown in Figure 8A). Box-plots, One Way ANOVA p<0.01, post-hoc Dunnett test. (B) Triple Toll RNAi knock-down with Toll-2GAL4 reduced Kenyon Cell cluster area (images shown in Figure 8C). As Toll-2 alone did not affect KCs, this shows that Tolls have redundant functions in KCs. Box-plots, Kruskal Wallis ANOVA p<0.0001, post-hoc Dunn test. (C,D) Conditional double Toll-2, Toll-6 RNAi knock-down with tubGAL80^ts, MyD88 >hisYFP increased cell number in the central brain. (D) Box-plots, quantification with DeadEasy Central Brain: Student t-test: p<0.0001. Scale bar: 50μ. (E,F) Knock-down of Toll-6, Toll-7 with tubGAL80^ts, Toll-2GAL4 > hisYFP at the adult critical period did not alter brain size, but Toll-2, Toll-6 knockdown did. With Toll-2 >Toll-2,Toll-6 RNAi, most brains were smaller than control brains, but 25% of brains were bigger. This shows loss of function for multiple Tolls in the adult critical period alters brain structure. The variable penetrance indicates complex compensatory interactions between Tolls through their various potential functions. (F) Box-plots, no statistically significant differences. Scale bar: 50μ. ****p<0.0001. For statistical details, see Supplementary file 2.