Fig 1.
Larval expression patterns of the Gal4 drivers used in these studies.
(a-c) DIC and fluorescence images of first instar larvae expressing UAS-mCherry driven by lov91Y-Gal4 (a), cut(ue)-Gal4 (b), and btl-Gal4 (c). Identical exposure conditions were used for the fluorescence images to reveal the differences in expression levels for the three Gal4 lines. lov91Y-Gal4 shows neural expression and faint activity in the tracheal DT epithelial cells. The bright dots of fluorescence in (a) are either fusion cells of the DTs or nuclei of peripheral neurons. (d) shows lov91Y-Gal4 expression in the epithelial cells and fusion cells (FC) of a third instar larval DT, (e) shows Lov immunostaining in DT nuclei at the same larval stage, (f) shows expression of cut(ue)-Gal4 in a cluster of tracheoblast cells (TB) in a first instar larvae, (g) shows high expression of cut(ue)-Gal4 in Tr10 of the DTs, the bridge (B), and the posterior section of the DTs adjacent to the spiracles. Scale bars = 50 μm. CNS = central nervous system, DT = dorsal trunks, SG = salivary glands.
Fig 2.
Tunneling behavior of various genotypes examined in these studies.
Images of cleaned agarose plates from tunneling assays for key genotypes are shown. Tunneling was quantitated using Image J. Five control genotypes (panels (d)-(h)) show robust tunneling. lov91Y-Gal4 > RNAi larvae show comparable tunneling activity (panel (c)) but tunneling is completely absent for cut(ue)-Gal4 > RNAi and btl-Gal4 > RNAi larvae (panels (a) and (b). lov91Y-Gal4 /lov def larvae show limited tunneling behavior (panel (i)).
Fig 3.
Failure to burrow, tunnel and survive to adulthood correlate strongly with fluid accumulation in the tracheae.
(a) Quantitation of tunneling was performed as described in Material and Methods for the genotypes shown. Mean values ± SEMs for the tunneling areas were calculated and then normalized to the value for the +/lov RNAi control. P values, Student’s t-test, * = P< 0.01, ** = P<0.00001, compared to +/ lov RNAi control. b) For larvae subjected to the burrowing/tunneling assay, the average percentage that died in the food, or outside the food (an indication of failure to burrow) was determined, in addition to the average percentage that pupated and the average percentage that emerged as adults. (c) Larvae of the same genotypes as those used for burrowing/tunneling assays, were imaged daily and the images were used to assess accumulation of fluid in the DTs (see Material and Methods). The percentages of larvae with partially fluid-filled DTs (checkered bars) or completely fluid-filled DTs (solid bars) on days 0–5 post-hatching are shown.
Fig 4.
Correlation of failure to burrow, tunnel, and survive to adulthood, with tracheal fluid-filling in lov hemizygous genotypes.
Data for the experimental and control genotypes as in Fig 3 in a lov hemizygous background. All data were generated and shown as in Fig 3, except that in (a) the quantitated tunneling areas are normalized to the value of lov91Y-Gal4 /+. P values, Student’s t-test, * = P< 0.01, ** = P<0.00001, compared to the lov91Y -Gal4 /+ control. Note i) in the hemizygous condition, lov91Y -Gal4 /lov def > lov RNAi larvae fail to tunnel and ii) lov91Y -Gal4 /lov def larvae show reduced tunneling despite having air-filled tracheae.
Fig 5.
Fluid-filling of the DTs in larvae of the genotypes studied.
Bright field images of dorsal views of control (left-hand column) and experimental (right-hand column) second instar larvae for the three Gal4 > lov RNAi genotypes studied and for lov91Y -Gal4 in the hemizygous condition. The btl-Gal4 construct used has UAS-actin-GFP sequences on the same chromosome. This allowed GFP imaging of the tracheal system (shown in the (g) and (h) lower panels). Air in the tracheae produces light refraction making the outlines of the DTs and lateral connective tracheal branches visible in control larvae. Fluid in the tracheae renders the tracheae almost invisible. For btl-Gal4 > lov RNAi (h), although grossly morphologically normal (compare (g) and (h) lower panels) the entire tracheal system is fluid-filled, whereas for lov91Y -Gal4 > lov RNAi (d), the entire system is air-filled. In cut(ue)-Gal4 > lov RNAi larvae (b), region Tr10 of the DTs is always fluid-filled and is noticeably shorter. In this larva, Tr9 is also fluid-filled in one of the DTs. lov91Y -Gal4 /lov def larvae (f) show a different phenotype: the tracheae are air-filled but overall larval growth is inhibited producing convoluted tracheae in some larvae. LC = lateral connectives, SP = spiracles. Scale bars = 200 μm.
Fig 6.
Failure to burrow and tunnel mimic hypoxia-induced behaviors.
(a) cut(ue)-Gal4 > lov RNAi larvae or control larvae were buried under a pile of yeast paste and the number of larvae that crawled out of the pile was monitored over time. Control larvae stayed in the yeast but the experimental larvae left the paste in numbers and at a rate comparable to findings for hypoxic larvae [13], n = 200 in batches of 10 larvae. Error bars = +/- SEM. (b) Wandering wild-type Canton-S larvae, tunneling in plates as used for the tunneling assay, were subjected to an atmosphere of N2 gas and the number of larvae that left their tunnels and moved to the agarose surface was monitored over time. N = 120 in batches of 10 larvae. Error bars = +/-SEM.
Fig 7.
Locomotor defects of lov91Y-Gal4/lov def larvae compared to cut(ue)-Gal4>lov RNAi larvae and lov91Y-Gal4/lov def larval growth defects.
(a) Larval locomotion was analyzed as described in Material and Methods, for forward and backward strides, and head turns. N = >50 for all genotypes. P values, Student’s t-test, black *** = P 377 <0.0001, black * = P < 0.01, red ** = P < 0.001, red * = P < 0.01, all compared to cut-Gal4/+. lov91Y -Gal4/lov def larvae are more compromised than cut(ue)-Gal4 > lov RNAi larvae. (b) Larval length measurements. At least 20 larvae for each genotype were imaged as described in Material and Methods. The images were processed with Adobe Photoshop and NIH Image J to measure the body length. P value, Student’s t-test, * = P <0.001 compared to +/lov def.
Fig 8.
LDH expression is up-regulated in cut(ue)-Gal4 > lov RNAi larvae.
Late wandering third instar larvae were collected for RNA extraction. Semi-Q RT-PCR was used for transcript analysis. LDH and actin PCR products for each RNA sample were run in parallel in separate gel lanes. 1 and 2 indicate RT-PCR products from two independent RNA preparations. Four RNA preparations total gave comparable results. For quantitation of LDH upregulation in cut(ue)-Gal4 > lov RNAi larvae, band intensities for the LDH and actin PCR products were quantitated with NIH image J and used to calculate LDH/actin (L/a) ratios for control and experimental (cut(ue)-Gal4 > lov RNAi) samples. The fold increase in LDH expression in the experimental samples was then calculated as the average value for L/a experimental/L/a control.
Fig 9.
cut(ue)-Gal4 > lov RNAi larval tracheae show normal immunostaining for septate junction constituents.
Confocal images of (a) cut(ue)-Gal4/+ and (b) cut(ue)-Gal4 > lov RNAi tracheae stained with either Coracle (Cora) or Fas III antibodies. Tracheae were dissected from third instar larvae as described in Material and Methods. Staining at the plasma membranes for both proteins is similar in control and lov knockdown tracheae but the cells of Tr10 are much smaller in the cut(ue)Gal4 > lov RNAi genotype. Scale bars = 20 μm.
Fig 10.
In cut(ue)-Gal4 > lov RNAi tracheae, cells of DT Tr10 are smaller, more numerous, and show decreased DNA content compared to adjacent cells in DT Tr9.
DAPI stained nuclei in confocal sections of DT metameres Tr9 and Tr10 from (a) cut(ue)-Gal4 /+ and (b) cut(ue)-Gal4 > lov RNAi third instar tracheae. (c) After DAPI quantitation (see Material and Methods) an RTr10/Tr9 value was calculated for all control or lov knockdown tracheae examined. RTr10/Tr9 = average of DAPI intensity per nucleus (Tr10)/average of DAPI intensity per nucleus (Tr9)). Horizontal bars show average RTr10/Tr9 values. The average RTr10/Tr9 value for cut(ue)-Gal4 > lov RNAi is approximately half that of cut(ue)-Gal4 /+ larvae. (d) The cell number ratios for control and lov knockdown tracheae in Tr9 and Tr10. Cell number ratios = cell numbers in Tr10/cell numbers in Tr9 and were calculated as described in Material and Methods. An increase in cell number of ~50% is indicated in Tr10 compared to Tr9 for cut(ue)-Gal4 > lov RNAi, as compared to control. Each + represents data collected from an individual larva. Scale bars = 20 μm. P value, Student’s t-test, * = P <0.00002 for cut(ue)-Gal4 > lov RNAi compared to cut(ue)-Gal4 /+.