Fig 3. Silencing ird1 in hemocytes reduces the fraction of Hml^Δ>GFP–positive cells, while increasing total blood cell numbers and rescuing lymph gland morphology in Tl^10b larvae.

A. Circulating blood cells are visualized by Hml^Δ-GAL4-driven UAS-GFP (green) and eater-DsRed (red) plasmatocyte reporter. Overexpression of UAS-GFP alone (+), or together with UAS-ird1^IR (>ird1^IR) in wild-type (+) or Tl^10b genetic background. B. Dissected lymph glands from larvae of the indicated genotypes (the anterior end to the left). Nuclear Hoechst staining (blue) and eater-DsRed expression visualize the entire organ, whereas the Hml^Δ>GFP signal is seen mostly in primary lobes. Between 8–10 lymph glands per genotype were analyzed. C, D. Number of eater-DsRed expressing sessile plasmatocytes in a single body segment (C) or extracted from circulation of larvae (D), carrying Hml^Δ- or FB-GAL4 alone (+), together with either UAS-ird1^IR (>ird1^IR), Tl^10b, ird1^Δvps15 or both of the latter. Plasmatocytes of the same blood cell populations were also counted for larvae carrying UAS-Tl^10b alone (>Tl^10b) or in combination with Hml^Δ-GAL4. Box-plots in C depict medians and quartiles for number of sessile cells counted automatically using imaging software in microscopic images of 12–15 individual larvae per genotype. For D, 30 larvae per genotype were bled and plasmatocyte numbers assessed using a hemocytometer. Values for significant differences from pairwise comparisons by post-hoc tests after two-way ANOVA are depicted above the whiskers in C and D. Hemocytes in A were imaged by fluorescence microscopy, whereas lymph gland images in B are maximum intensity projections of z-stacks generated by confocal microscopy.