DEV028209 Supplementary Material

Files in this Data Supplement:

• Supplemental Figure S1 -

  Fig. S1. Babo inactivation in single neurons results in MB lobe fusions in the entire brain. (A-C′) MARCM analysis of single-cell αβ wild-type (A), babo⁵² (B) or OK107>babo_a; babo⁵² clones (C). Additional panels (A′,B′,C′) show the corresponding Fas2-positive (mostly heterozygote) projections; Fas2 labels all γ axons (weakly) and αβ axons (strongly). Note the heterozygous medial lobe fusions when babo⁵² single-cell clones were fused (B′, compare with the wild type A′), which were absent when babo clones were rescued by UAS-babo_a or babo_b expression (C′; data not shown; quantified in D). Scale bar: 20 µm. (D) Quantification of the β lobe fusion phenotype in the indicated clones (black bars), and in the corresponding heterozygote (grey bars) genotype. Where heterozygote projections were analysed, the corresponding genotype and phenotype of the babo single-cell clones are indicated in parenthesis. Owing to the frequency of clonal generation, most brains contained single-cell (or multiple single-cell) clones in both hemispheres. The numbers (n) quantified represent the number of brains with single-cell clones in both hemispheres.

• Supplemental Figure S2 -

  Fig. S2. CA Babo expression also leads to axon guidance defects, whereas ectopic wild-type Babo expression does not lead to CA Babo defects. (A,B) CA babo misexpression also resulted in axon guidance defects (yellow arrows). Fas2-stained (magenta; appearing white in overlap with the mCD8::GFP marker, green) αβ axons appear highly fasciculated and misguided. Two examples are shown. In B, bundled αβ axons form a ball-like structure close to cell bodies. (C-F) Babo immunostaining revealing ectopic levels of CA Babo (C), wild-type Babo-a (D) or Babo-b (E). Additional panels (C′-E′) show the corresponding Babo staining (magenta) overlapped with CD8::GFP expression (green). Using one (D) or two (E) copies of UAS transgenes, despite high levels of ectopic wild-type Babo, only CA Babo (C) reveals a dominant axon phenotype. The same confocal settings were used to image these samples. UAS transgenic lines used: C, CA Babo, two copies of lines 1B and 9B; D, Babo-a, one copy of line X; E, Babo-b, two copies of lines X and 2D2. Scale bar: 20 µm.

• Supplemental Figure S3 -

  Fig. S3. The developmental origin of CA babo phenotypes. Developmental staging of control (mCD8::GFP expression alone) (A-E) and CA babo-misexpressing (F-J) animals, analysed at the indicated times during pupal development. During the onset of puparium formation (0 hours after puparium formation, APF), the majority of MB neurons consist of γ and α′β′ neurons (A). A phase of neurogenesis occurs at this period and MB neuroblasts give rise to αβ neurons, which become visible 24 hours APF (B). As axon collaterals continue to grow, dorsal (α) and medial (β) lobes become more prominent between 30 and 48 hours APF (C-E). After 48 hours APF, these lobes appear almost indistinguishable from adult MB (E, compare with Fig. 1B). In CA babo neurons, MB lobes were not significantly different from wild-type neurons at 0-24 hours APF (F,G, compare with the wild type A,B). However, at 30-48 hours APF, the β lobes appear truncated (H-J). These results suggest that CA babo misexpression predominantly results in axon extension defects. White arrowheads indicate normal axon projections at the indicated stages of development. Open red arrows indicate β lobe truncations, consistent with axon extension defects. Cell body sections (E,I,J) have been removed to clearly reveal MB axons. Scale bars: 20 µm.

• Supplemental Figure S4 -

  Fig. S4. Loss of one copy of LIMK1 results in β lobe overextension phenotypes in many CA Babo-misexpressing animals (15 out of 17 brains). Scale bar: 20 µm

• Supplemental Figure S5 -

  Fig. S5. Babo regulates OL axon targeting. Images of OL axons from the y-projection stack of serial confocal sections. Viewed from the dorsal-ventral axis, wild-type OL axons (A) appear to extend a principal axon to the anterior-most part of the OL (white arrow), followed by fine branching along the anterior-posterior (a-p) axis. Loss of babo (B) results in less branching along the a-p axis, particularly in the distal (lateral) section of the OL (open blue arrowheads). These phenotypes were not observed in Smad2 or Med clones (C; data not shown). However, partial disruptions were occasionally observed in the distal sections of Smad2 clones. See also Fig. 8 and Results. OL axons of the indicated genotype (A-C) are identical to those in Fig. 8F-H. Lateral is to the right and medial to the left of each panel. Blue arrowheads, wild-type axon termination zones. Open blue arrowheads, targeting defects in terminal zones ('gaps'). Scale bar (x-axis): 20 µm.

• Supplemental Figure S6 -

  Fig. S6. LIMK1, but not CA Babo, expression leads to high F-actin levels in MB neurons. (A-E′) F-actin staining in control wild-type (y,w) (A) or LIMK1 (B,C) or CA babo (D,E) misexpressing brains. Additional panels (A′-E′) indicate the corresponding mCD8::GFP expression, which labels MB lobes, as indicated. Note the high levels of F-actin in LIMK1-misexpressing neurons (B,C, compare with the wild type A). This was especially notable when LIMK1 was misexpressed at high levels (line m6, B), but was also detectable using an intermediate expression line F4 (D). Despite exhibiting similar axonal phenotypes (yellow arrows in B′ and D′, and red arrows in C′ and E′), different levels of CA Babo expression (high, four UAS copies using lines 1B and 9B; intermediate, two copies of lines 1B and 9B) did not result in similar levels of F-actin accumulation (D,E). White arrowheads indicate normal axon projections. Yellow arrows indicate axon guidance defects. Open red arrows indicate truncation defects. Cell body sections (C′) have been removed to clearly reveal MB axons. Similar confocal settings were used to image the F-actin staining in all samples, except in B, where owing to high F-actin levels, reduced gain settings were used. Scale bar: 20 µm.