Fig 1.
Distribution of the Ndg protein.
(A-F) Confocal images showing embryonic (A-D), wing imaginal disc (E), and ovarian tissues (F) stained with anti-Ndg (A-C, and E-F) or anti-GFP (Ndg.sGFP, D). (A, B) In stage 16 wild-type embryos, Ndg (red) is found in the BMs surrounding most tissues, including muscles (A, A’), gut (B, B’ (arrow) and VNC (B, B’, arrowhead) and in chordotonal organs (asterisk). (C, C’) Lateral view of a stage 13 embryo showing Ndg (red) accumulation around caudal visceral mesodermal cells visualized with the marker crocLacZ (green, arrow in C’). (D) Ndg is found in embryonic macrophages (arrow). (E) Ndg (red) is found at the basal surface of wing imaginal disc epithelial cells (arrow) and cells of the peripodial membrane (arrowhead). (F) Ndg (red) accumulates in the basement membrane (BM) around the follicular epithelium (arrow). Scale bars represent 20μm (A-F).
Fig 2.
Fat body adipocytes and blood cells are the main source of Ndg in larval BMs.
(A) Schematic representation of the in vivo sGFP interference approach (isGFPi). (B) Confocal images showing the localization of functional Ndg.sGFP fusion protein expressed under control of the endogenous promoter in different tissues of the 3rd instar larva. Images compare tissues from control larvae (+) and larvae where Ndg.sGFP expression has been knocked down in fat body adipose tissue and blood cells (Cg>isGFPi). Ndg.sGFP signal is not observed in the BMs of Cg>isGFPi larvae, except in wing disc myoblasts (asterisk) and partially in the imaginal ring of the salivary gland, body wall muscles and VNC (purple arrows). Note that Ndg.sGFP signal disappears from myoblasts when isGFPi is driven with myoblast driver Mef2-GAL4 (Mef2>isGFPi). Arrow and circles in the fat body panel indicate BM and CIVICs (Collagen IV Intercellular Concentrations), respectively. In tracheal images, apical cuticle autofluorescence is observed in the tracheal lumen (downward-pointing yellow arrow). Nuclei stained with DAPI (white).
Fig 3.
CRISPR/Cas9-generated Nidogen mutants are viable.
(A) Schematic representation of the Ndg locus (2nd chromosome), Ndg mutants generated, sgRNAs used for generation of mutants (green boxes, 1–4), sequence targeted by Ndg RNAi construct (purple box, Ndgi), epitope recognized by the rabbit anti-Ndg antibody generated in this study (black box) and qRT-PCR primers used to molecularly characterize the Ndg1 mutant (A-D). (B) Expression of Ndg mRNA in wild type (w1118) and Ndg1 homozygous mutant larvae, assessed by qRT-PCR. Error bars represent 95% confidence intervals from three repeats. (C, D) Confocal images of larval fat body (C, left panels), wing imaginal discs (C, right panels) and embryos (D) from wild type (upper panels), Ndg1 (C, lower panels) and NdgΔRodG3-1 (D, lower panel) mutant animals stained with anti-Ndg antibody (yellow). (C) Nuclei stained with DAPI (blue). (E) Homozygous Ndg1 and Ndg2 mutants are viable and show no obvious morphological abnormalities.
Fig 4.
Ndg mutants show discontinuous adipose tissue BMs.
(A) Confocal images of the larvae fat body BM showing localization of Laminin (LanB1.sGFP, cyan), Collagen IV (Vkg.GFP, green) and Perlecan (Trol.YFP, yellow) from control (upper panels), Ndg1 mutant (middle panels) and larvae where Ndg has been knocked down under control of Cg-GAL4 (Ndgi, lower panels). Loss of Ndg causes discontinuity of adipose tissue BMs. (B) Discontinuous BMs (Vkg.GFP, green) in the larval fat body of transheterozygotes Ndg1/Ndg2 (upper panel), Ndg1/NdgΔRod-G3.1 (middle panel) and Ndg1/NdgDf(2R)BSC281 (lower panel). (C) Fat body BM (Cg25C.RFP, red) in wild type, Ndg1 mutant and Ndg1 mutant rescued with Ndg.sGFP. (D) Lipid droplets (neutral lipid dye BODIPY, green) in fat body of wild type (upper panel), Ndg1 mutant (middle panel) and Ndg2 mutant (lower panel) larvae. Asterisks point to cells with reduced content of lipid droplets. Nuclei stained with DAPI (blue). Scale bar represents 50μm (A-D). (E) Quantification of lipid droplet diameter in 16 cells from wild type control, Ndg1 and Ndg2 mutants. Each dot represents a single droplet. Particles smaller than 3μm in diameter were excluded from the analysis. Horizontal lines indicate the mean value and error bars represent ±SD. Difference with the wild type are significant in non-parametric Mann-Whitney tests (****: p<0.0001).
Fig 5.
Rod domain is necessary but not sufficient for Nidogen localization to BMs.
(A) Schematic depiction of Ndg full length (FL) and different Ndg deletions tested in this study. All Ndg versions are tagged with GFP in the C-terminal. SP = Signal Peptide. G1/2/3 = Globular domains 1/2/3. (B) Confocal images of wing discs from larvae expressing the indicated versions of Ndg under control of Cg-GAL4. Correct BM localization, detected by GFP signal (green, arrows), is observed only for Ndg versions containing the Rod domain (NdgFL, NdgRodG3, NdgΔG3, NdgΔG2, NdgΔG1 and NdgG2Rod) but neither NdgRod nor any other single domain can localize to the BM by itself. Nuclei stained with DAPI (blue). (C) Model of domain requirements for proper Ndg localization to BMs.
Fig 6.
The G3 domain is essential for Ndg function.
(A) Confocal images showing localization in fat body BM of the indicated GFP-tagged versions of Ndg, detected by GFP signal (white), in wild type (left column) and upon knock down of laminin (LanAi, middle column) and Collagen IV (Cg25Ci, right column). (B, C) Quantification of GFP signal intensity of indicated GFP-tagged Ndg versions in the BM of wild type control, LanAi (B) and Cg25Ci (C) fat body. Each dot represents a single measurement of intensity inside a 500μm2 square. Horizontal lines indicate the mean value (****: p<0.0001; **:p<0.01; n.s.: not significant). See Materials and Methods for statistical testing details. (D) Confocal images of the larvae fat body BM (Cg25C.RFP in red) in control, Ndg1 mutant, and Ndg1 mutant expressing NdgΔG1, NdgΔG2 or NdgΔG3. Integrity of the BM is restored by NdgΔG1 and NdgΔG2 but not NdgΔG3. (E) Confocal images showing uncoupling of Collagen IV (Cg25C.RFP in red, middle panels) and Laminin (LanB1.sGFP in cyan, right panels) in Ndg1 mutant (lower panels) and wild type fat body (upper panels). Merged channels are shown in left panels. (F) Confocal images of Collagen IV (Cg25C.RFP in red, middle panels) and Perlecan (Trol.GFP in green, right panels) in Ndg1 mutant (lower panels) and wild type fat body (upper panels). Merged channels are shown in left panels. (G) Quantification of Pearson’s colocalization coefficient after Costes thresholding (R coloc) of Collagen IV with Laminin (E) and Collagen IV with Perlecan (F). (H) Model for the role of the G1, G2 and G3 domains on Ndg binding to Laminin and Collagen IV.
Fig 7.
Laminins and Col IV are required for proper Ndg incorporation into adipose tissue BMs.
(A-D’) Confocal images showing stage 16 embryos stained with anti-Ndg. Images compare control embryos (A) with embryos depleted of the different BM components (B-D). (A-A’) In control embryos, Ndg localizes to the BM of muscles (A), gut (A’, arrowhead), VNC, midline pores (A’, arrow) and in chordotonal organs (A, asterisk). (B) In Laminin-depleted embryos, Ndg is strongly reduced in the BM of muscles (B), gut (B’, hollow arrowhead) and VNC (B’, hollow arrow) and not affected in chordotonal organs (B, asterisk). (C-D) In contrast, Ndg deposition in these embryonic BMs is not affected in either Col IV (C-C’) or Perlecan (D-D’) mutant embryos, except for a reduction in Ndg in the midline pores in Col IV mutants (C’, hollow arrow). Scale bar represents 20μm (A-D). (E) Confocal images of the fat body (adipose tissue) BM showing localization of Ndg (Ndg.sGFP, green), Laminin (LanB1.sGFP, cyan), Collagen IV (Vkg.GFP, green) and Perlecan (Trol.YFP, yellow). Images show fat body from wild type larvae (upper panels) and larvae where LanB1, vkg or trol have been knocked down using Cg-GAL4 (lower panels). (F) Table summarizing effects of the absence of each of the four major BM components on the other components in the BM of the fat body. (G) Model for the mutual relations of Laminin, Nidogen, Collagen IV and Perlecan in the BM of the fat body.
Fig 8.
Ndg loss enhances Laminin and Perlecan loss phenotypes.
(A) Larval ventral nerve cord (VNC) in LanA216/LanA160, Ndg1/Ndg1, LanA216/+ and Ndg1/Ndg1; LanA216/+ larvae. (B) Quantification of VNC length in Ndg1, LanA216/+, Ndg1; LanA216/+, LanA160/+ and Ndg1; LanA160/+. Each dot represents an individual VNC measurement. Horizontal lines indicate the mean value. Differences with the control were significant in non-parametric Mann-Whitney tests (**; p<0.01). (C) Images of pupae where Ndg, trol or both Ndg and trol have been knocked down using act-GAL4 at either 25°C or 30°C. (D) Quantification of pupal length in wild type control, act>Ndgi, act>troli and act>Ndgi+troli pupae. Each dot represents a single pupa measurement. Horizontal lines indicate the mean value. Differences with the control were significant in non-parametric Mann-Whitney tests (****; p<0.0001).