Fig 1.
Co-activation of EGFR-Ras and PI3K in Drosophila glia causes an expansion of the glial network.
Brains from third instar larvae. Glia nuclei are labeled with GFPNLS (green) driven by repo-Gal4. Each brain is composed of 2 symmetrical hemispheres. (A, B, Aʹ, Bʹ) In glioma brains (repo>dEGFRλ; dp110CAAX) (B), both brain hemispheres are enlarged, and the number of glial cells is increased relative to WT control (A–Aʹ). Higher magnifications are shown in panels A and B. The quantification of the number of glial cells is shown in (C). Arrows indicate glial nuclei. (D–E) Optical sections of larval brain to visualize the glial network, glial cell bodies, and membranes are labeled (myrRFP). (D) RFP signal (in grayscale) in control brains shows glial membranes and the network in WT brain. (E) The glioma brain shows a dramatic increase in the membrane projections and in the size of the network. (F) Glia is labeled with UAS-Ihog-RFP (red) driven by repo-Gal4 to visualize TMs in glioma cells; nuclei are marked with DAPI (blue) and neurons are stained with Hrp (green) and enwrapped by glial TMs in glioma brains (yellow arrowheads). (G) Glia is labeled with UAS-myr-RFP driven by repo-Gal4 to visualize membrane projections in glial cells (red); nuclei are marked with DAPI (blue) and neurons are stained with ELAV (green) and enwrapped by glial TMs in glioma brains (yellow arrows). (H–I and K–L) TEM images of third instar larval brains expressing HRP in the glial cells. (H–I) HRP deposits label cell membranes in dark, thus identifying glial cells. Colored images from control brains show glial cells identified by HRP staining (magenta) and HRP-negative neurons (cyan). (J) Schematic diagram: a glioma cell labeled with HRP (magenta) showing that glioma cells produce a network of TMs that grow to surround neighboring neurons (cyan). (K–N) Several magnifications of glioma brains showing TMs that grow and enwrap neighboring neurons (cyan). Detail of several layers of a glioma membrane enwrapping a neuron (M–N) and a longitudinal section of a TM, showing to glioma cells (black arrows) connected by TMs (L), yellow arrows indicate glial membranes. (O–R) Glia are labeled with UAS-ihog-RFP to visualize active cytoneme/TM structures in glial cells as part of an interconnecting network. (O) In control brains, the active glial cytonemes are shown by repo>ihog-RFP in gray or red in the merge. In glioma brains (P), the TMs grow and expand across the brain; quantification of the network volume (Q) and the network/glial cell ratio is shown in panel R. Error bars show SD; ***P < 0.0001. Scale bar size is indicated in this and all figures. The numerical data pertaining to this figure can be found in S1 Data file. Genotypes: (A) w; repo-Gal4, UAS-GFPnls/UAS-lacZ, (B) UAS-dEGFRλ, UAS-dp110CAAX; repo-Gal4, UAS-GFPnls, (D) w; Gal80ts; repo-Gal4, UAS-myrRFP/UAS-lacZ, (E, G) UAS-dEGFRλ, UAS-dp110CAAX; Gal80ts; repo-Gal4, UAS-myrRFP, (F, P) UAS-dEGFRλ, UAS-dp110CAAX; repo-Gal4, UAS-ihog-RFP, (H-I) w; UAS-HRP:CD2; repo-Gal4/UAS-lacZ, (K-N) UAS-dEGFRλ, UAS-dp110CAAX; UAS-HRP:CD2; repo-Gal4, (O) w; repo-Gal4, UAS-ihog-RFP/UAS-lacZ. EGFR, Epidermal Growth Factor Receptor; ELAV, embryonic lethal abnormal vision; GFP, green fluorescent protein; Hrp, horseradish peroxidase; PI3K, phosphatidylinositol-3 kinase; RFP, Red Fluorescent Protein; TEM, transmission electron microscopy; TM, tumor microtube; wt, wild type.
Fig 2.
Wg/Fz1 accumulate in glioma cells, and Fz1 in glia interacts with neuronal Wg.
Larval brain sections with glial cell bodies and membranes labeled in red (myrRFP) and stained with Wg antibody show homogeneous expression in the control brains (A) in gray (green in the merge). In the glioma brains, Wg accumulates in the glial transformed cells (B); the Wg average pixel intensity ratio between Glia/Neuron quantification is shown in panel C. Arrows indicate Wg staining in glial membranes. (D–G) Glial cells are labeled with UAS-Ihog-RFP to visualize the glial network (red) and stained with Fz1 (gray or green in the merge). (D) Fz1 is homogeneously distributed in control brains, with a slight accumulation in the Ihog+ structures. (E) Fz1 accumulates in the TMs and specifically in the projections that are in contact with the neuronal clusters. (F) Upon knockdown of fz1 in glioma brains, the tumoral glial network is still formed but Fz1 is not detectable. (G) Knockdown of igl in glioma brains restores a normal glial network, and Fz1 shows a homogeneous distribution along the brain section. Arrows indicate Fz1 staining in glial membranes. (H) Fz1 average pixel intensity ratio between Glia/Neuron quantification. Arrows indicate Fz1 staining in glial membranes. (I–J) GRASP technique was used, and both halves of GFP tagged with a CD4 signal to direct it to the membranes (CD4-spGFP) were expressed in neurons (elav-lexA) and glial (repo-Gal4) cells, respectively. Only upon intimate contact is GFP protein reconstituted and green fluorescent signal is visible. (I) Control brains showed a discrete GFP (green) signal corresponding to the physiological interaction between glia and neurons. (J) In glioma brains, a strong GFP signal from the GRASP reporter is detected. Arrows indicate GRASP reconstitution GFP signal. (K–L) PLAs were performed in control and glioma brains to quantify the interactions between Wg and Fz1. (K) Control brains showed a discrete number of PLA+ puncta (green) showing the physiological interactions. (L) Glioma brains showed a 5-fold increase in the number of puncta, quantified in panel M. The number of PLA+ Wg-Fz1 interactions per Repo+ cell in control and glioma brains is shown in panel N. Arrows indicate PLA+ puncta. Nuclei are marked with DAPI (blue). Error bars show SD; *P < 0.01; ***P < 0.0001. The data underlying this figure can be found in S1 Data. Genotypes: (A) w; Gal80ts; repo-Gal4, UAS-myrRFP/UAS-lacZ, (B) UAS-dEGFRλ, UAS-dp110CAAX; Gal80ts; repo-Gal4, UAS-myrRFP, (C) w;; repo-Gal4, ihog-RFP/UAS-lacZ, (D) UAS-dEGFRλ, UAS-dp110CAAX;; repo-Gal4, UAS-ihog-RFP, (E) UAS-dEGFRλ, UAS-dp110CAAX; UAS-Fz1-RNAi; repo-Gal4, UAS-ihog-RFP, (F) UAS-dEGFRλ, UAS-dp110CAAX;; repo-Gal4, UAS-ihog-RFP /UAS-igl-RNAi, (I) w;; elav-lexA; repo-Gal4, UAS-ihog-RFP/UAS-CD4-spGFP1-10, lexAop-CD4-spGFP11, (J) UAS-dEGFRλ, UAS-dp110CAAX; elav-lexA; repo-Gal4, UAS-ihog-RFP/ UAS-CD4-spGFP1-10, lexAop-CD4-spGFP11, (K) w;; repo-Gal4, ihog-RFP/UAS-lacZ, (L) UAS-dEGFRλ, UAS-dp110CAAX;; repo-Gal4, UAS-ihog-RFP. elav, embryonic lethal abnormal vision; Fz1, Frizzled1; GFP, green fluorescent protein; GRASP, GFP Reconstitution Across Synaptic Partners; igl, igloo; myrRFP, myristoilated Red Fluorescent Protein; PLA, proximity ligation assay; TM, tumor microtube; Wg, wingless.
Fig 3.
Wg signaling pathway is active in glioma cells, and the glioma inactivates it in neuronal clusters in glioma brains.
(A–C) Larval brain sections with glial membrane projections labeled in red and stained with Arm (gray or green in the merge). (A) Cyt-Arm is homogeneously distributed in control sections. (B) In glioma brains, Cyt-Arm accumulates in the neurons’ cytoplasm where it is inactive. (C) Knockdown of igl in glioma brains restores a normal glial network, and Cyt-Arm does not accumulate showing a homogeneous distribution similar to the control. Arrows indicate Cyt-Arm staining at the glia-neuron interphase. (D) Cyt-Arm average pixel intensity ratio between neuron/glia quantification showing the Wg signaling pathway silencing in neurons in a glioma brain. (E–H) Glial cell bodies and membranes are labeled with myrRFP (red). Wg signaling pathway reporters arm-GFP (E–F) and nkd-lacZ (stained with anti-bGal) (G–H) in control and glioma brains show activation of the pathway (green) in glioma cells (red) compared with the reporter activation mostly in neurons in the control brains. Arrows indicate cells with reporter activation. (I–N) Confocal immunofluorescence single plane images of S24 glioblastoma multiforme stem-like cells (GBSC) NMRI nude mice brains (glioma) and NMRI nude mice (control) brains stained with human anti-βCatenin (I–J) and WNT1 (L–M) both show in gray (red in the merged image) an increase in the glioma samples. The corresponding quantification of the pixel intensity is shown in panels K and N. Green signal from tumor cell GFP expression allows specific detection of S24 GBSC–related structures in the mouse brain (Iʹ, Lʹ). Arrows indicate glioma or control cells. Nuclei are marked with DAPI. Error bars show SD; ***P < 0.0001 and ns for nonsignificant. The data underlying this figure can be found in S1 Data. Genotypes: (A) w; repo-Gal4, ihog-RFP/UAS-lacZ, (B) UAS-dEGFRλ, UAS-dp110CAAX;; repo-Gal4, UAS-ihog-RFP, (C) UAS-dEGFRλ, UAS-dp110CAAX;; repo-Gal4, UAS-ihog-RFP /UAS-igl-RNAi, (E) w; Gal80ts; repo-Gal4, UAS-myrRFP/ arm-GFP, (F) UAS-dEGFRλ, UAS-dp110CAAX; Gal80ts; repo-Gal4, UAS-myrRFP/ arm-GFP, (G) w; Gal80ts; repo-Gal4, UAS-myrRFP/nkd04869a-lacZ, (H) UAS-dEGFRλ, UAS-dp110CAAX; Gal80ts; repo-Gal4, UAS-myrRFP/nkd04869a-lacZ. Arm, Armadillo; bGal, beta-galactosidase; Cyt-Arm, Cytoplasm-Armadillo; GBSC, glioblastoma multiforme stem-like cell; GFP, green fluorescent protein; igl, igloo; myrRFP, myristoilated Red Fluorescent Protein; nkd-lacZ, nkd-lacZ, transcriptional beta-galactosidase reporter of naked gene NLGN3; NMRI, Naval Medical Research Institute.
Fig 4.
Wg expression in glioma cells is dispensable for tumor progression because glioma depletes Wg from neuronal membrane.
(A–F) Larval brain sections with glial membrane projections labeled in gray (red in the merge) and glial cell nuclei stained with Repo (gray, green in the merge). (C) wg knockdown in glioma cells (wg-RNAi) or armS10; Fz1-RNAi (D) does not prevent glioma cell number increase nor glioma TM volume expansion quantified in (E–F). (G–Gʹ) Control brains express 1 copy of WgNRT-HA instead of endogenous secretable Wg in neurons (green, ELAV-GFP). (H–Hʹ) Glioma samples that do not express WgNRT in neurons show an overgrown TM network. (panel I in red, magnification in panel Iʹ in gray) Glioma brains expressing membrane-anchored Wg (WgNRT-HA) in neurons show that the glial network volume size is restored to control volume in these animals. Quantification of the TM network volume is shown in panel J. Yellow arrows show glial network. Error bars show SD; ***P < 0.0001; **P < 0.001; and ns for nonsignificant. The data underlying this figure can be found in S1 Data. Genotypes: (A) w;; repo-Gal4, ihog-RFP/UAS-lacZ, (B) UAS-dEGFRλ, UAS-dp110CAAX;; repo-Gal4, UAS-ihog-RFP, (C) UAS-dEGFRλ, UAS-dp110CAAX; repo-Gal4/UAS-wg-RNAi, (D) UAS-armS10/UAS-dEGFRλ, UAS-dp110CAAX; UAS-Fz1-RNAi; repo-Gal4, UAS-ihog-RFP.UAS-ihog-RFP, (G) w; >wg>wgNRT-HA, PaxRFP/ elav-lexA, lexAop-CD8-GFP; repo-Gal4, UAS-ihog-RFP/lexAop-flp, (H) UAS-dEGFRλ, UAS-dp110CAAX; >wg>wgNRT-HA, PaxRFP; repo-Gal4, UAS-ihog-RFP/lexAop-flp, (I) UAS-dEGFRλ, UAS-dp110CAAX; >wg>wgNRT-HA, PaxRFP/ elav-lexA, lexAop-CD8-GFP; repo-Gal4, UAS-ihog-RFP/lexAop-flp. ELAV-GFP, embryonic lethal abnormal vision green fluorescent protein; Fz1, Frizzled1; TM, tumor microtube; Wg, wingless.
Fig 5.
Gliomas cause neurodegeneration and restoration of the glia-neuron Wg/Fz1 signaling equilibrium inhibits glioma progression and neurodegeneration.
Neurons from the larval neuromuscular junction are stained with Nc82 showing the synaptic active zones. (A–F) Upon glioma induction (B), the number of synapses (gray) is reduced when compared with the control (A). The number of synapses is restored upon knockdown in the glioma of Fz1 (C), igl (D), or armS10; Fz1-RNAi (E). The quantification of synapse number in all genotypes is shown in panel F. (G–J) Upon knockdown of wg (I) or Fz1 in D42 neurons (H), the number of synapses (gray) is reduced when compared with the control (G). Arrows indicate synapses. The quantification of synapse number in all genotypes is shown in panel J. (K–N) Upon glioma induction the number of synapses (gray) is reduced (L) when compared with the control (K). The number of synapses is restored upon overexpression of Fz1, specifically in the neurons (M). The quantification of synapse number is shown in panel N. Yellow arrows show active zones. Error bars show SD; ***P < 0.0001, **P < 0.001, *P < 0.01, and ns for nonsignificant. The data underlying this figure can be found in S1 Data. Genotypes: (A) w;; repo-Gal4, ihog-RFP/UAS-lacZ, (B) UAS-dEGFRλ, UAS-dp110CAAX;; repo-Gal4, UAS-ihog-RFP, (C) UAS-dEGFRλ, UAS-dp110CAAX; UAS-Fz1-RNAi; repo-Gal4, UAS-ihog-RFP, (D) UAS-dEGFRλ, UAS-dp110CAAX;; repo-Gal4, UAS-ihog-RFP /UAS-igl-RNAi, (E) UAS-armS10/UAS-dEGFRλ, UAS-dp110CAAX; UAS-Fz1-RNAi; repo-Gal4, UAS-ihog-RFP.UAS-ihog-RFP, (G) w; UAS-CD8-GFP; D42-Gal4/UAS-lacZ, (H) w; UAS-CD8-GFP/Fz1-RNAi; D42-Gal4, (I) w; UAS-CD8-GFP/wg-RNAi; D42-Gal4, (K) w; repo-Gal4, ihog-RFP/UAS-lacZ, (L) UAS-dEGFRλ, UAS-dp110CAAX; lexAop-Fz1; repo-Gal4, UAS-ihog-RFP, (M) UAS-dEGFRλ, UAS-dp110CAAX; lexAop-fz1/ elav-lexA, lexAop-CD8-GFP; repo-Gal4, UAS-ihog-RFP. Fz1, Frizzled1; igl, igloo; Wg, wingless.
Fig 6.
JNK signaling pathway is activated in glioma via the TNF receptor Grindelwald which mediates GB progression.
Glial cell nuclei are labeled with UAS-GFPNLS (gray or green in the merge) driven by repo-Gal4. (A–D) JNK signaling pathway reporter TRE-RFP (gray or red in the merge) in control, glioma, glioma Gap43-RNAi, and Glioma Fz1-RNAi brain sections. The number of glial cells is quantified in panel E, and the TRE-RFP average pixel intensity per glial cells is quantified in panel F. (G–L) Glia are labeled with UAS-Ihog-RFP (red) driven by repo-Gal4 to visualize active cytonemes/TM structures in glial cells and stained with Repo (gray or green in the merge) in the following genotypes: control, glioma, grnd-RNAi, glioma grnd-RNAi, grnd−/− and Glioma grnd−/− brain sections. The number of Repo+ cells is quantified in panel M. Nuclei are marked with DAPI. Error bars show SD; *P < 0.01, **P < 0.001, ***P < 0.0001, or ns for nonsignificant. Scale bar size is indicated in this and all figures. The data underlying this figure can be found in S1 Data. Genotypes: (A) TRE-RFP; repo-Gal4, UAS-GFPNLS/UAS-lacZ, (B) UAS-dEGFRλ, UAS-dp110CAAX; TRE-RFP; repo-Gal4, UAS-GFPNLS, (C) UAS-dEGFRλ, UAS-dp110CAAX; TRE-RFP; repo-Gal4, UAS-GFPNLS/UAS-Gap43-RNAi, (D) UAS-dEGFRλ, UAS-dp110CAAX; TRE-RFP/UAS-Fz1-RNAi; repo-Gal4, UAS-GFPNLS, (G) repo-Gal4, ihog-RFP/UAS-lacZ, (H) UAS-dEGFRλ, UAS-dp110CAAX;; repo-Gal4, UAS-ihog-RFP, (I) UAS-grnd-RNAi; repo-Gal4, ihog-RFP, (J) UAS-dEGFRλ, UAS-dp110CAAX; UAS-grnd-RNAi; repo-Gal4, UAS-ihog-RFP, (K) grndMINOS/grndMINOS; repo-Gal4, ihog-RFP, (L) UAS-dEGFRλ, UAS-dp110CAAX; grndMINOS/grndMINOS; repo-Gal4, UAS-ihog-RFP. Fz1, Frizzled1; GB, glioblastoma; GFP, green fluorescent protein; grnd, Grindelwald; Ihog, interference hedgehog; JNK, cJun N-terminal kinase; TM, tumor microtube; TRE-RFP, RFP fluorescent protein regulated by a transcriptional response element of JNK pathway.
Fig 7.
JNK pathway is necessary to localize Fz1 in the TMs, grnd mediates Wg pathway activation in GB and promote GB progression.
Brains from third instar larvae displayed at the same scale. Glia are labeled with UAS-Ihog-RFP (gray or red in the merge) driven by repo-Gal4 to visualize active cytonemes/TM structures in glial cells and stained with Fz1 (gray or green in the merge, panels A–F) and Cyt-Arm (gray or green in the merge, panels G–K) in the following genotypes control, glioma, glioma grnd-RNAi, Glioma grnd−/− and Glioma BskDN brain sections. Fz1 and Cyt-Arm average pixel intensity staining quantification ratio between ihog+ and ihog–domains (F and L). (A, G) Fz1 and Cyt-Arm is homogeneously distributed in control sections. (B, H) In glioma brains, Fz1 accumulates in the TMs, and Cyt-Arm accumulates in the neurons’ cytoplasm where it is inactive. (C–D, I–J) Knockdown or knockout of grnd and BskDN in glioma brains showing a normal glial network, and Fz1 and Cyt-Arm does not accumulate showing a homogeneous distribution similar to the control. Fz1 average pixel intensity quantification for glia/neuron ratio (F) and Cyt-Arm average pixel intensity quantification for neuron/glia ratio between ihog- and ihog+ domains (L). Nuclei are marked with DAPI (blue). Error bars show SD; *P < 0.01, **P < 0.001, ***P < 0.0001, or ns for nonsignificant. Scale bar size is indicated in this and all figures. The data underlying this figure can be found in S1 Data. Genotypes: (A, G) repo-Gal4, ihog-RFP/UAS-lacZ, (B, H) UAS-dEGFRλ, UAS-dp110CAAX;; repo-Gal4, UAS-ihog-RFP, (C, I) UAS-dEGFRλ, UAS-dp110CAAX; UAS-grnd-RNAi; repo-Gal4, UAS-ihog-RFP, (D, J) UAS-dEGFRλ, UAS-dp110CAAX; grndMINOS/grndMINOS; repo-Gal4, UAS-ihog-RFP, (E, K) UAS-dEGFRλ, UAS-dp110CAAX;; repo-Gal4, UAS-ihog-RFP/ UAS-bskDN. BskDN, dominant negative form of the effector Bsk; Cyt-Arm, Cytoplasmic-Armadillo; Fz1, Frizzled1; GB, glioblastoma; grnd, Grindelwalkd; Ihog, interference hedgehog; JNK, cJun N-terminal kinase; RFP, Red Fluorescent Protein; TM, tumor microtube; Wg, wingless.
Fig 8.
MMP1 is up-regulated in GB, and the Wg pathway is necessary and sufficient to activate MMP expression in glial cells.
Brains from third instar larvae displayed at the same scale. (A) Glia are labeled with UAS-Ihog-RFP (gray or red in the merge) driven by repo-Gal4 to visualize active cytonemes/TM structures in glial cells and stained with MMP1 (gray or green in the merge) is homogeneously distributed in control sections, with a slight accumulation in the Ihog+ projections. (B) MMP1 accumulates in the TMs and specifically in the projections that are in contact with the neuronal clusters. (C) Quantification of MMP1 staining ratio between ihog+ and ihog–domains. (D) Western blot of samples extracted from control, glioma, glioma Gap43-RNAi and glioma Fz1-RNAi larvae showing changes in the amount of MMP1 or protein. (E) qPCRs with RNA extracted from control and glioma larvae showing no change in the transcription (mRNA levels) of MMP1 or MMP2. (F–G) Glial cell bodies and membranes labeled with CD8-GFP in gray (green in the merge) driven by repo-Gal4 to the glial cells and stained with MMP1 (gray or red in the merge). (F) MMP1 staining in control brain. (G) MMP1 accumulates in the glial cells upon Fz1 overexpression. (H) Quantification of MMP1 staining ratio between GFP+ and GFP–domains. Nuclei are marked with DAPI (blue). Error bars show SD; *P < 0.01, **P < 0.001, ***P < 0.0001, or ns for nonsignificant. Scale bar size is indicated in this and all figures. The data underlying this figure can be found in S1 Data. Genotypes: (A-B) repo-Gal4, ihog-RFP/UAS-lacZ, (A-C) UAS-dEGFRλ, UAS-dp110CAAX;; repo-Gal4, UAS-ihog-RFP, (A) UAS-dEGFRλ, UAS-dp110CAAX;; repo-Gal4, UAS-ihog-RFP /UAS-Gap43-RNAi, (A) UAS-dEGFRλ, UAS-dp110CAAX; /UAS-Fz1-RNAi; repo-Gal4, UAS-ihog-RFP, (F) w;; Repo-LexA, LexAop-CD2-GFP, (G) w;; Repo-LexA, LexAop-CD2-GFP/LexAop-Fz1. Fz1, Frizzled1; GB, glioblastoma; GFP, green fluorescent protein; Ihog, interference hedgehog; MMP, matrix metalloproteinase; qPCR, quantitative Polimerase Chain Reaction; RFP, Red Fluorescent Protein; TM, tumor microtube; Wg, wingless.
Fig 9.
JNK triggers MMPs expression in GB.
Brains from third instar larvae displayed at the same scale. Glia is labeled with UAS-Ihog-RFP (gray or red in the merge) driven by repo-Gal4 to visualize active cytonemes/TM structures in glial cells and stained with MMP1 (gray or green in the merge). (A) MMP1 is homogeneously distributed in control sections. (B) MMP1 accumulates in the TMs and specifically in the projections that are in contact with the neuronal clusters. (C–E) Blocking JNK pathway by using a UAS-bskDN or UAS-grnd-RNAi in glioma brains restores a normal glial network and MMP1 does not accumulate showing a homogeneous staining along the brain section. (F) Quantification of MMP1 staining ratio between ihog+ and ihog–domains. Nuclei are marked with DAPI (blue). Error bars show SD; *P < 0.01, **P < 0.001, ***P < 0.0001, or ns for nonsignificant. Scale bar size are indicated in this and all figures. The data underlying this figure can be found in S1 Data. Genotypes: (A) repo-Gal4, ihog-RFP/UAS-lacZ, (B) UAS-dEGFRλ, UAS-dp110CAAX;; repo-Gal4, UAS-ihog-RFP, (C) UAS-dEGFRλ, UAS-dp110CAAX;; repo-Gal4, UAS-ihog-RFP/ UAS-bskDN, (D) UAS-grnd-RNAi; repo-Gal4, ihog-RFP, (E) UAS-dEGFRλ, UAS-dp110CAAX; UAS-grnd-RNAi; repo-Gal4, UAS-ihog-RFP. bskDN, dominant negative form of the effector Bsk; GB, glioblastoma; grnd, Grindelwald; Ihog, inferference hedgehog; JNK, cJun N-terminal kinase; MMP, matrix metalloproteinase; repo-Gal4,; RFP, Red Fluorescent Protein; TM, tumor microtube.
Fig 10.
TM expansion and synapse loss in the neurons requires MMPs in glioma.
Brains from third instar larvae displayed at the same scale. Glia are labeled with UAS-Ihog-RFP (gray) driven by repo-Gal4 to visualize active cytonemes/TM structures in glial cells and stained with Fz1 (gray) in the following genotypes (A–F) control, glioma, MMP1-RNAi, glioma MMP1-RNAi, MMP2-RNAi, and glioma MMP2-RNAi brain sections. (G) Quantification of Fz1 average pixel intensity staining ratio between ihog+ and ihog–domains. (H) Quantification of glial/glioma network volume expansion. (I) Survival curve of adult control, glioma, glioma MMP1-RNAi, and glioma MMP2-RNAi flies after a number of days of glioma induction and progression. (J) Neurons from the larval neuromuscular junction were stained with Nc82 (brp, in gray) to reveal and quantify the synaptic active zones. Upon glioma induction, the number of synapses is reduced when compared with the control. The number of synapses is restored toward control levels upon knockdown of MMP1 or MMP2. (K) Model: Glioma cells produce a network of TMs that grow to reach neighboring neurons. Intimate membrane contact facilitates neuronal-Wg sequestering mediated by glioma Fz1 receptor. Glial cells are initially transformed into malignant GB upon EGFR and PI3K pathways constitutive activation; afterward GB cells establish a positive feedback loop including TMs, Wg signaling, JNK, and MMPs. Initial stimulation of actin cytoskeleton remodeling via EGFR/PI3K enables initial expansion of TMs; as a consequence, Fz1 receptor accumulation in TMs mediates neuronal-Wg depletion and Wg signaling up-regulation in the GB cells, which activates JNK in GB. As a consequence, MMPs are up-regulated and facilitate further TM infiltration in the brain; hence the GB TM network expands and mediate further wingless depletion to close the loop. Error bars show SD; *P < 0.01, **P < 0.001, ***P < 0.0001, or ns for nonsignificant. Scale bar size is indicated in this and all figures. The data underlying this figure can be found in S1 Data. Genotypes: (A) repo-Gal4, ihog-RFP/UAS-lacZ, (B) UAS-dEGFRλ, UAS-dp110CAAX;; repo-Gal4, UAS-ihog-RFP, (C) UAS-MMP1-RNAi; repo-Gal4, ihog-RFP, (D) UAS-dEGFRλ, UAS-dp110CAAX; UAS-MMP1-RNAi; repo-Gal4, UAS-ihog-RFP, (E) UAS-MMP2-RNAi; repo-Gal4, ihog-RFP, (F) UAS-dEGFRλ, UAS-dp110CAAX; UAS-MMP2-RNAi; repo-Gal4, UAS-ihog-RFP. EGFR, Epidermal Growth Factor Receptor; Fz1, Frizzled1; GB, glioblastoma; Ihog, interference hedgehog; JNK, cJun N-terminal kinase; MMP, matrix metalloproteinase; PI3K, phosphatidylinositol-3 kinase; RFP, Red Fluorescent Protein; TM, tumor microtube; Wg, wingless.