1. Neuroscience

In-situ glial cell-surface proteomics identifies pro-longevity factors in Drosophila

  1. Madeline P Marques
  2. Bo Sun
  3. Ye-Jin Park
  4. Tyler Jackson
  5. Tzu-Chiao Lu
  6. Yanyan Qi
  7. Erin Harrison
  8. Miranda C Wang
  9. Omar Moussa Pasha
  10. Amogh Varanasi
  11. Dominique Kiki Carey
  12. DR Mani
  13. Jonathan Zirin
  14. Mujeeb Qadiri
  15. Yanhui Hu
  16. Kartik Venkatachalam
  17. Norbert Perrimon
  18. Steven A Carr
  19. Namrata D Udeshi
  20. Liqun Luo
  21. Jiefu Li
  22. Hongjie Li  Is a corresponding author
  1. Department of Molecular and Human Genetics, Baylor College of Medicine, United States
  2. Huffington Center on Aging, Baylor College of Medicine, United States
  3. Department of BioSciences, Rice University, United States
  4. Department of Integrative Biology and Pharmacology, University of Texas Health Science Center at Houston, United States
  5. The Broad Institute of MIT and Harvard, United States
  6. Department of Genetics, Blavatnik Institute, Harvard Medical School, Harvard University, United States
  7. Howard Hughes Medical Institute, United States
  8. Department of Biology, Howard Hughes Medical Institute, Stanford University, United States
  9. Janelia Research Campus, HHMI, United States
https://doi.org/10.7554/eLife.109422.3
Share this article
Cite this article
  1. Madeline P Marques
  2. Bo Sun
  3. Ye-Jin Park
  4. Tyler Jackson
  5. Tzu-Chiao Lu
  6. Yanyan Qi
  7. Erin Harrison
  8. Miranda C Wang
  9. Omar Moussa Pasha
  10. Amogh Varanasi
  11. Dominique Kiki Carey
  12. DR Mani
  13. Jonathan Zirin
  14. Mujeeb Qadiri
  15. Yanhui Hu
  16. Kartik Venkatachalam
  17. Norbert Perrimon
  18. Steven A Carr
  19. Namrata D Udeshi
  20. Liqun Luo
  21. Jiefu Li
  22. Hongjie Li
(2026)
In-situ glial cell-surface proteomics identifies pro-longevity factors in Drosophila
eLife 14:RP109422.
https://doi.org/10.7554/eLife.109422.3
  2. Download BibTeX
  3. Download .RIS
5 figures, 1 table and 2 additional files

Figures

Figure 1 with 1 supplement
Download asset Open asset
Glial cell-surface proteomics of the intact brains of young and old flies.

(A) Transgenic expression of horseradish peroxidase (HRP) fused to the transmembrane protein CD2 targets CD2-HRP to the extracellular side of the glial plasma membrane. In the presence of H2O2, CD2-HRP converts BxxP substrate into phenoxyl radicals that biotinylate nearby endogenous proteins. Biotinylated proteins are then enriched from brain lysates using streptavidin beads, followed by on-bead trypsin digestion, tandem mass tag (TMT) labeling, and liquid chromatography-tandem mass spectrometry analysis (LC-MS/MS). (B) To confirm HRP-dependent biotinylation of glial cells, Repo-GAL4 were crossed with either UAS-HRP-CD2 or wild-type (w1118) flies and adult brains were stained with neutravidin. Proteins were biotinylated in the experimental group (UAS-HRP-CD2) but not in the control group (w1118). (C) Biochemical characterization by streptavidin blot showed that the glial cell-surface-targeted HRP biotinylated a wide range of proteins compared to the negative control lacking HRP-CD2 expression. (D) For each time point (5 days and 50 days), two biological replicates and two negative controls (one lacking HRP-CD2 expression and one lacking H2O2 in the reaction) were profiled, for a total of eight samples. Each sample consisted of 300 mixed male and female fly brains, for a total of 2400 brains. In total, we detected 2007 proteins with two or more unique peptides. Potential contaminants were removed by comparing differential enrichment between the experimental and negative control groups (i.e. the TMT ratio for each protein identified; see Methods). In the TMT labeling system, 126, 127 N, 127 C, 128 N, 128 C etc. were used to index and compare different samples. (E) Strong correlations between biological replicates were observed, further validating the quality of the proteome profiles. (F) The receiver operating characteristic (ROC) curve depicts the true-positive rate against the false-positive rate of detected proteins, for each biological replicate (only 126:127 C pictured). (G) Gene Ontology (GO) analysis was performed based on the 872 proteins identified following ratiometric and cutoff (see Methods). Top five GO terms on cellular compartments with the lowest false discovery rates are shown. See Figure 1—figure supplement 1.

Figure 1—figure supplement 1
Download asset Open asset
Glial cell-surface proteomics.

(A) We searched the 872 proteins included in our proteomic profiles against three databases: FlyBase, UniProt, and the Gene Ontology Data Archive (DOI 10.5281/zenodo.1205166; Version 2023-05-10; Ashburner et al., 2000) to determine which proteins had previous plasma membrane or extracellular matrix annotations. We found that 682 out of the 872 proteins could be validated by at least one database, further validating the quality of our cell-surface proteomic labeling. (B) We plotted the receiver operating characteristic (ROC) curve for each biological replicate and observed high specificity. (C) We identified the top 10 ‘hub’ genes in our protein-protein interaction (PPI) network and found that of the top 10 hub genes identified, seven had reported function in cell adhesion (klg, kst) and synaptic wiring (DIP-β, DIP-δ, DIP-γ, DIP-κ, DIP-ɑ), while the remaining three had reported function in neuron projection (CG33543), axon guidance (CG34353), and dendrite guidance (Toll-6).

Figure 2
Download asset Open asset
Quantitative comparison of young and old proteomes and selection of candidate genes.

(A) To select approximately 50 candidates for further screening, we generated a volcano plot, where the tandem mass tag (TMT) ratio (log2FoldChange) for each of the 872 proteins identified was plotted on the x-axis and the log10(Ratio P-value) was plotted on the y-axis. Genes associated with the largest change in protein levels between time points were identified using an arbitrary cutoff targeting about 50 genes for screening. In total, 48 candidate genes were selected for further screening, most of which have at least one strong human ortholog. Down-regulated candidate genes selected for functional screening are labeled in blue whereas up-regulated candidate genes are labeled in red. (B) A Gene Ontology (GO) analysis investigating proteome biological processes revealed that genes encoding the most up-regulated proteins were primarily associated with cell localization, transport, and maintaining homeostasis, while most down-regulated proteins were primarily associated with neural development, synapse organization, and neuroplasticity. (C) Genes encoding the most up- and down-regulated proteins were clustered by their reported protein-protein interactions (PPIs) and corresponding confidence scores (using a Markov clustering algorithm, inflation value set to 2.5). Each protein in the PPI plot was then assessed individually to determine if it had reported function in synapse organization, synapse regulation, cell adhesion, development and cell migration, or localization and transport.

Figure 3 with 7 supplements
Download asset Open asset
Functional screening of candidate genes predicted to be involved in normal brain aging.

(A) Confocal images showing that the glial-GeneSwitch (GS) line only induced transgene expression in the presence of the inducible drug Ru486. (B) Forty eight candidate genes were selected for functional screening. For each of the 19 up-regulated genes, two RNAi lines were used to control for possible off-target effects. Of the 29 down-regulated genes selected, we were able to acquire and screen lines for 19 genes. (C) No obvious lifespan effect was observed in control lines: a w1118 wild-type control, a UAS-luciferase RNAi control, a UAS-lacZ control, and a flySAM control. (D) Pie charts summarizing the results of our broad-scale lifespan screen for male and female flies. (E) Of the 19 up-regulated candidate genes screened, RNAi knockdown of four genes led to consistent lifespan decreases for males. Alternatively, RNAi knockdown of three genes led to consistent lifespan increases for females. (F) Of the 19 down-regulated candidate genes screened, overexpression of two genes led to lifespan increases for males, whereas overexpression of three genes led to lifespan decreases. In parallel, overexpression of 10 genes led to lifespan increases for females, whereas overexpression of one gene led to decreased lifespan. (G) Summary of lifespan effect for both up- and down-regulated genes. See Figure 3—figure supplements 1–7.

Figure 3—figure supplement 1
Download asset Open asset
Functional screening of candidate genes.

(A–B) In our initial functional lifespan screen, glial overexpression of DIP-β was found to significantly extend lifespan for both male and female flies, compared to the negative controls (only the luciferase RNAi and flySAM controls are pictured here).

Figure 3—figure supplement 2
Download asset Open asset
Functional screening of up-regulated candidate genes.

(A–E) We selected 19 up-regulated candidate genes for functional screening, which we selectively knocked down in glia to examine their influence on lifespan. For each up-regulated gene, we screened two RNAi lines (to control for possible off-target effects). Of the 19 up-regulated candidates screened, we found that RNAi knockdown of four genes led to consistent lifespan decreases for males. Alternatively, for female flies, RNAi knockdown of three genes led to consistent lifespan increases. The result of the first five genes are shown.

Figure 3—figure supplement 3
Download asset Open asset
Functional screening of up-regulated candidate genes.

(A–E) We selected 19 up-regulated candidate genes for functional screening, which we selectively knocked down in glia to examine their influence on lifespan. For each up-regulated gene, we screened two RNAi lines (to control for possible off-target effects). Of the 19 up-regulated candidates screened, we found that RNAi knockdown of four genes led to consistent lifespan decreases for males. Alternatively, for female flies, RNAi knockdown of three genes led to consistent lifespan increases. The result of the second five genes is shown.

Figure 3—figure supplement 4
Download asset Open asset
Functional screening of up-regulated candidate genes.

(A–E) We selected 19 up-regulated candidate genes for functional screening, which we selectively knocked down in glia to examine their influence on lifespan. For each up-regulated gene, we screened two RNAi lines (to control for possible off-target effects). Of the 19 up-regulated candidates screened, we found that RNAi knockdown of four genes led to consistent lifespan decreases for males. Alternatively, for female flies, RNAi knockdown of three genes led to consistent lifespan increases. The result of the third five genes is shown.

Figure 3—figure supplement 5
Download asset Open asset
Functional screening of up-regulated candidate genes.

(A–E) We selected 19 up-regulated candidate genes for functional screening, which we selectively knocked down in glia to examine their influence on lifespan. For each up-regulated gene, we screened two RNAi lines (to control for possible off-target effects). Of the 19 up-regulated candidates screened, we found that RNAi knockdown of four genes led to consistent lifespan decreases for males. Alternatively, for female flies, RNAi knockdown of three genes led to consistent lifespan increases. The result of the last four genes is shown.

Figure 3—figure supplement 6
Download asset Open asset
Functional screening of down-regulated candidate genes.

(A–J) We selected 29 down-regulated candidate genes for functional screening, which we selectively overexpressed in glia to examine their influence on lifespan. For each of the 29 down-regulated genes, we screened one overexpression line (using either UAS-cDNA lines or flySAM2.0 lines), when available. In total, we were able to acquire and screen lines for 19 genes. Of these candidates, overexpression of two genes led to lifespan increases for males, whereas overexpression of three genes led to lifespan decreases. Alternatively, overexpression of 10 genes led to lifespan increases for females, whereas overexpression of one gene led to decreased lifespan. We found that constitutive overexpression of DIP-β in glia significantly increased lifespan in both males and females compared to controls. The result of the first 10 genes are shown.

Figure 3—figure supplement 7
Download asset Open asset
Functional screening of down-regulated candidate genes.

(A–I) We selected 29 down-regulated candidate genes for functional screening, which we selectively overexpressed in glia to examine their influence on lifespan. For each of the 29 down-regulated genes, we screened one overexpression line (using either UAS-cDNA lines or flySAM2.0 lines), when available. In total, we were able to acquire and screen lines for 19 genes. Of these candidates, overexpression of two genes led to lifespan increases for males, whereas overexpression of three genes led to lifespan decreases. Alternatively, overexpression of 10 genes led to lifespan increases for females, whereas overexpression of one gene led to decreased lifespan. We found that constitutive overexpression of DIP-β in glia significantly increased lifespan in both males and females compared to controls. The result of the remaining nine genes is shown.

Figure 4
Download asset Open asset
Validating that pan-glial overexpression of DIP-β increases lifespan and serves a protective function during aging.

(A–D) Glial overexpression of DIP-β (using both a flySAM2.0 and a UAS overexpression line) significantly extends lifespan for both male and female flies, compared to the negative controls (only the w1118 and UAS-lacZ controls are shown here). (E) To further validate that glial overexpression of DIP-β is serving a protective function during brain aging, climbing (i.e. negative geotaxis) assays were conducted. Glial overexpression of DIP-β was also associated with improved climbing and motor control ability in aged flies.

Figure 5 with 1 supplement
Download asset Open asset
Single-nucleus RNA-sequencing - investigating the downstream transcriptomic effects of glial overexpression of DIP-β during aging.

(A) Two whole-head samples (20 heads each) were prepared for snRNA-seq: one where DIP-β had been overexpressed in glia from day 5–50 and one control. Single-nucleus RNA-seq libraries were prepared using the 10 X Genomics Chromium Single Cell 3’ HT v3.1 protocol. (B) In total, approximately 49,000 high-quality cells were obtained after quality control and filtering from our fly head samples, which consist of 8 broad cell types. (C) Broad cell-type cluster compositions were largely consistent between the glial DIP-β overexpression treatment (Ru486+) and the control (Ru486-) groups. (D) A broad analysis of differentially expressed genes (DEGs) revealed that the glial cell-type cluster showed the second highest number of DEGs between the glial DIP-β overexpression treatment (Ru486+) and the negative control (Ru486-) groups. (E) A more detailed analysis investigating DEGs across 90 different cell subtypes revealed that of the glial cell subtypes, cortex glia (i.e. ‘adult brain cell body glia’) and ensheathing glia showed the highest number of DEGs between the DIP-β overexpression treatment (Ru486+) and the control (Ru486-) groups. (F) Of the 252 DEGs identified for glial cells, 186 were unique to cortex glia, 27 were unique to ensheathing glia, and 40 were found to be differentially expressed in both cortex and ensheathing glia. (G) As fat cells exhibited the highest number of DEGs between the DIP-β overexpression and control groups, we conducted a Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis to evaluate the biological functions of these genes. Most fat cell DEGs were up-regulated, and the top altered metabolic and signaling pathways for these genes were found to be related to homeostasis, signal integration, stress resistance, and longevity. (H) Cell-cell communication (CCC) analysis was performed using FlyPhoneDB2. DIP-β overexpression was generally associated with an increase in cell-cell communication, particularly glia-neuron and glia-fat cell communication. The top altered glial signaling pathways are shown. (I) Our proposed model: Glial overexpression of DIP-β leads to improved glia-neuron and glia-fat cell communication as flies age, which may contribute to the observed lifespan extensions. See Figure 5—figure supplement 1.

Figure 5—figure supplement 1
Download asset Open asset
Single-nucleus RNA-sequencing - investigating the downstream transcriptomic effects of pan-glial DIP-β overexpression.

(A) In total, we sequenced approximately 49,000 cells from our fly head samples, with approximately 23,000 from the glial DIP-β overexpression (i.e. Ru486+) sample and approximately 26,000 from the negative control (i.e. Ru486-) sample. (B) Detailed cell subtype cluster composition annotations. (C) A detailed analysis investigating differentially expressed genes (DEGs) across cell subtypes.

Tables

Key resources table
Reagent type (species) or resourceDesignationSource or referenceIdentifiersAdditional information
Strain, strain background (Drosophila)UAS-HRP-CD2Liqun Luo lab Li et al., 2020
Strain, strain background (Drosophila)GSG3285-1 (pan-glial GeneSwitch)Linda Partridge labBDSC 59948Nicholson et al., 2008
Strain, strain background (Drosophila)DIP-β flySAMNorbert Perrimon labBDSC 98929Jia et al., 2018
Strain, strain background (Drosophila)UAS-DIP-βChris Doe lab 
Chemical compound, drugRu486 drug (Mifepristone)Cayman Chemical100063174 mM concentration
Chemical compound, drugNeutravidinThermo Fisher310001:250 dilution
Antibodychicken polyclonal, anti-GFPAves LabsNC95105981:1000 dilution
Antibodymouse monoclonal, anti-RepoDSHB8D12 anti-Repo1:100 dilution
Antibodyrat monoclonal, anti-NcadDSHBDN-Ex #81:40 dilution
AntibodyRabbit polyclonal, anti-DsRedTakara Bio6324961:200 dilution
AntibodyChicken Alexa Fluor 488, secondary antibodyJackson ImmunoResearch703-545-1551:250 dilution
AntibodyMouse Alexa Fluor 568, secondary antibodyInvitrogenA100371:250 dilution
AntibodyRat Alexa Fluor 647, secondary antibodyJackson ImmunoResearch112-605-1671:250 dilution
otherSlowfade mount bufferInvitrogenS36936See Methods/
Immunohistochemistry
Commercial assay, kitPierce Streptavidin Magnetic BeadsThermo Fisher88817NA
Commercial assay, kitTMT10plex Isobaric Label Reagents and KitsThermo Scientific90110NA

Additional files

Supplementary file 1

Original proteomic data, candidate genes and fly lines.

(A) Original proteomic data prior to analyses. (B) Candidate genes selected for further investigation. (a) In total, we selected 29 down-regulated candidate genes for functional screening. Candidate genes were selected for further screening only if they were known (or predicted) to be located either on the cell surface or in the extracellular space. (b) In total, we selected 19 up-regulated candidate genes for functional screening. Candidate genes were selected for further screening only if they were known (or predicted) to be located either on the cell surface or extracellular space. Most candidate genes selected for further screening were found to have at least one strong human ortholog (Chen et al., 2017; Wang et al., 2019a; Wang et al., 2019b). (C) RNAi lines used to screen up-regulated candidate genes and OE lines used to screen down-regulated candidate genes. For each of the 19 up-regulated candidate genes, we screened two RNAi transgenic lines to control for potential off-target effects. For each of the 29 down-regulated genes, we screened one overexpression line (using either UAS-cDNA lines or flySAM2.0 lines), when available. Of the 29 down-regulated genes selected, we were able to acquire and screen lines for 19 genes.

https://cdn.elifesciences.org/articles/109422/elife-109422-supp1-v1.xlsx
MDAR checklist
https://cdn.elifesciences.org/articles/109422/elife-109422-mdarchecklist1-v1.docx

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Madeline P Marques
  2. Bo Sun
  3. Ye-Jin Park
  4. Tyler Jackson
  5. Tzu-Chiao Lu
  6. Yanyan Qi
  7. Erin Harrison
  8. Miranda C Wang
  9. Omar Moussa Pasha
  10. Amogh Varanasi
  11. Dominique Kiki Carey
  12. DR Mani
  13. Jonathan Zirin
  14. Mujeeb Qadiri
  15. Yanhui Hu
  16. Kartik Venkatachalam
  17. Norbert Perrimon
  18. Steven A Carr
  19. Namrata D Udeshi
  20. Liqun Luo
  21. Jiefu Li
  22. Hongjie Li
(2026)
In-situ glial cell-surface proteomics identifies pro-longevity factors in Drosophila
eLife 14:RP109422.
https://doi.org/10.7554/eLife.109422.3