Proteomics Analysis Identifies Orthologs of Human Chitinase-Like Proteins as Inducers of Tube Morphogenesis Defects in Drosophila melanogaster

Abstract

Elevated levels of human chitinase-like proteins (CLPs) are associated with numerous chronic inflammatory diseases and several cancers, often correlating with poor prognosis. Nevertheless, there is scant knowledge of their function. The CLPs normally mediate immune responses and wound healing and, when upregulated, they can promote disease progression by remodeling tissue, activating signaling cascades, stimulating proliferation and migration, and by regulating adhesion. We identified Imaginal disc growth factors (Idgfs), orthologs of human CLPs CHI3L1, CHI3L2, and OVGP1, in a proteomics analysis designed to discover factors that regulate tube morphogenesis in a Drosophila melanogaster model of tube formation. We implemented a novel approach that uses magnetic beads to isolate a small population of specialized ovarian cells, cells that nonautonomously regulate morphogenesis of epithelial tubes that form and secrete eggshell structures called dorsal appendages (DAs). Differential mass spectrometry analysis of these cells detected elevated levels of four of the six Idgf family members (Idgf1, Idgf2, Idgf4, and Idgf6) in flies mutant for bullwinkle (bwk), which encodes a transcription factor and is a known regulator of DA-tube morphogenesis. We show that, during oogenesis, dysregulation of Idgfs (either gain or loss of function) disrupts the formation of the DA tubes. Previous studies demonstrate roles for Drosophila Idgfs in innate immunity, wound healing, and cell proliferation and motility in cell culture. Here, we identify a novel role for Idgfs in both normal and aberrant tubulogenesis processes.

Keywords: chitinase-like proteins; dorsal appendage; imaginal disc growth factors; morphogenesis; proteomics.

Figure 1

Distinct cell types interact to create dorsal appendage (DA)-forming tubes. (A) Laid eggs from wild-type (Canton S) and bwk-mutant females. (B) Squamous somatic follicle cells (“stretch” cells, green) surround 15 germline nurse cells (purple). Columnar somatic follicle cells (white, red, and blue cells) surround the oocyte. The red and blue cells form the tube, then elongate by migrating over the stretch cells. Tube cells secrete egg-shell proteins into the tube lumens to form each DA. A, anterior; P, posterior. (C) Regulation of DA morphogenesis by Bullwinkle (Bwk) from the nurse cells, and Shark and Src42A from the stretch cells. Shark and Src42A act downstream of Bwk. Signals between the cell types are unknown. Bar, 100 µm. (B and C) adapted from Dorman et al. (2004) with permission.

Figure 2

Cell separation and mass spectrometry. (A) Cell isolation. (B) Workflow for LC-MS/MS. Data-dependent acquisition (“shotgun”) identified > 1000 proteins. Targeted mass spectrometry (selected reaction monitoring) estimated relative abundance of candidate proteins. (C) Chromatograms: relative abundance (intensities) and retention times of fragment ions. Bar graphs: total area under the fragment-ion peaks indicates higher relative abundance of Idgf peptides in stretch cells from bwk mutant females. y ions: y5 = 5-residue C-terminal fragment ion, etc. Peak areas were calculated by Skyline (MacLean et al. 2010) by integrating the area under the curves from the chromatograms. bwk, bullwinkle; Idgf, Imaginal disc growth factor; LC-MS/MS, liquid-chromatography tandem mass spectrometry; wt, wild-type.

Figure 3

Elevated levels of Idgfs contribute to DA defects. (A) Stretch cell-specific Gal4-driven Idgf1 or Idgf3 overexpression induces DA defects. Representative phenotypic categories. Graph: proportions of moderate or severe DA phenotypes. Control: Gal4-driven GFP expression (GFP.nls for c415-GAL4 Exp 2, mCD8::GFP for the rest). N for each genotype is given in Table S2 in File S1. ***P < 0.001, χ² test for trend in proportions (Rosner 2000). Significance was determined using the Benjamini–Hochberg procedure with FDR = 0.05. (B–C’) Stills from movies in File S2 (B and B’) and File S4 (C and C’) at stage 13. Anterior is up and to the left. Asterisk indicates the dorsal midline. Ecadherin::GFP marks the adherens junctions of the follicle cells. (B) Genotype: w¹¹¹⁸; Ecadherin::GFP; +/+. The tube-forming cells have migrated over the stretch cells, forming a narrow stalk from the base (arrow) and a slightly widened paddle near the tip (arrowhead) (see movie in File S2). (B’) Projection of interior optical planes shows the tube lumen underneath the roof cells. The base (arrow) and the tip (arrowhead) of the long, narrow tube are indicated. (C) Genotype: w¹¹¹⁸; Ecadherin::GFP; c415-Gal4/UAS-Idgf3. Migration of the tube-forming cells over the stretch cells is substantially impaired, resulting in abnormally short appendage primordia, even though the sample was allowed to develop hours longer than needed for WT tube development. Both the base (arrow) and tip (arrowhead) are abnormally wide (see movies in File S4 and File S3 for a second example). (C’) Projection of interior optical planes shows the tube lumen underneath the roof cells. The base (arrow) and the tip (arrowhead) of the abnormally short, wide tube are indicated. (D) Suppression of bwk phenotype by RNAi. Stretch cell-specific RNAi knockdown of Idgf5 or Idgf6 in a bwk background (UAS-Idgf RNAi/+; GAL4 bwk¹⁵¹/bwk⁰⁸⁴⁸²) suppresses the short-DA phenotype. Control: +/+ or CyO; GAL4 bwk¹⁵¹/bwk⁰⁸⁴⁸² (bwk background with no RNAi). Boxplots show distribution of DA lengths. Whisker bars: range, hinges: 25th and 75th quantiles, red line = median, gray line = median of control, circles: outliers. N for each genotype is given in Table S3 in File S1 as well as data for additional RNAi lines. *P < 0.05, **P < 0.01, and ***P < 0.001, Welch’s t-test. Significance was determined using the Benjamini–Hochberg procedure with a FDR of 0.05. Bar, 100 μm (A and D) and 50 µm (B–C’). bwk, bullwinkle; DA, dorsal appendage; Exp, experiment; FDR, false discovery rate; Idgf, Imaginal disc growth factor; ns, not significant; RNAi, RNA interference; WT, wild-type.

Figure 4

Wild-type cells express Idgf transcripts. Stretch cells in wild-type egg chambers, delineated by Spectrin protein (red), express RNA for all six Idgfs in a punctate pattern (green). DAPI (blue) marks the nuclei, which are highly polyploid in nurse cells (∼2000 C) and less so in follicle cells (32–64 C). The dotted line indicates the location of the cross section shown in the inset. The solid white line indicates the dorsal midline. Punctate expression also occurs in the columnar cells. Nurse-cell expression differs between Idgfs. Idgf2 has an unusual enrichment in the two dorsally-located nurse cells directly next to the oocyte. Idgf4 and Idgf6 have relatively high nurse-cell expression overall, and Idgfs 2, 4, and 6 transcripts appear to be transferred into the oocyte. Idgf3 is distinctive with expression mostly if not entirely restricted to the stretch cells. Idgfs 1, 4, 5, and 6 localize at or near the oocyte cortex. Bar, 50 µm. Idgf, Imaginal disc growth factor.

Figure 5

Reduction of Idgf expression produces DA defects. Stretch cell-specific knockdown of Idgfs (PG150-GAL4>Idgf RNAi) induces DA defects. Control: PG150-GAL4>mCherry RNAi. Left: Images show representative phenotypes. Right: Graph plots the proportions of DA phenotypes. *P < 0.05, **P < 0.01, χ² test for trend in proportions, N = 218 (control), 230 (Idgf1), 142 (Idgf2), 261 (Idgf3), 334 (Idgf4), 351 (Idgf5), and 247 (Idgf6). Significance was determined using Benjamini–Hochberg procedure with FDR of 0.05. Bar, 100 μm; DA, dorsal appendage; FDR, false discovery rate; Idgf, Imaginal disc growth factor; RNAi, RNA interference.

Figure 6

Hypotheses for Idgf function in DA morphogenesis. (A) Alternative signaling pathways downstream of Bwk transcriptional targets. A signal from the nurse cells to the stretch cells may regulate Idgf levels via Shark and Src42. Alternatively, Bwk may regulate levels of Idgfs through a pathway that is independent of Shark and Src42A. (B) Hypothesis 1: Idgfs are secreted from the stretch cells (green) and signal through a receptor in the tube-forming cells (blue and magenta), polarizing the tube-forming cells and guiding tube elongation. (C) Hypothesis 2: Idgfs are secreted into the extracellular matrix to regulate tube morphogenesis and cell migration through cell adhesion and signaling. (D) Hypothesis 3: Stretch cells secrete Idgfs and receive an autocrine signal that influences stretch cell tensile properties and adhesion to provide an optimal substrate for DA-cell migration. Bwk, Bullwinkle; DA, dorsal appendage; ECM, extracellular matrix; Idgf, Imaginal disc growth factor.