Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2022 May 19;13(1):2760.
doi: 10.1038/s41467-022-30490-6.

Lipid droplet degradation by autophagy connects mitochondria metabolism to Prox1-driven expression of lymphatic genes and lymphangiogenesis

Odeta Meçe #  1   2 , Diede Houbaert #  1   2 , Maria-Livia Sassano  1   2 , Tania Durré  3 , Hannelore Maes  1 , Marco Schaaf  1   2 , Sanket More  1   2 , Maarten Ganne  1   2 , Melissa García-Caballero  4   5 , Mila Borri  4   5 , Jelle Verhoeven  1   2 , Madhur Agrawal  1   2 , Kathryn Jacobs  1   6   7 , Gabriele Bergers  6   7 , Silvia Blacher  3 , Bart Ghesquière  8 , Mieke Dewerchin  4   5 , Johan V Swinnen  9 , Stefan Vinckier  4   5 , María S Soengas  10 , Peter Carmeliet  4   5 , Agnès Noël  3 , Patrizia Agostinis  11   12
Affiliations

Lipid droplet degradation by autophagy connects mitochondria metabolism to Prox1-driven expression of lymphatic genes and lymphangiogenesis

Odeta Meçe et al. Nat Commun. .

Abstract

Autophagy has vasculoprotective roles, but whether and how it regulates lymphatic endothelial cells (LEC) homeostasis and lymphangiogenesis is unknown. Here, we show that genetic deficiency of autophagy in LEC impairs responses to VEGF-C and injury-driven corneal lymphangiogenesis. Autophagy loss in LEC compromises the expression of main effectors of LEC identity, like VEGFR3, affects mitochondrial dynamics and causes an accumulation of lipid droplets (LDs) in vitro and in vivo. When lipophagy is impaired, mitochondrial ATP production, fatty acid oxidation, acetyl-CoA/CoA ratio and expression of lymphangiogenic PROX1 target genes are dwindled. Enforcing mitochondria fusion by silencing dynamin-related-protein 1 (DRP1) in autophagy-deficient LEC fails to restore LDs turnover and lymphatic gene expression, whereas supplementing the fatty acid precursor acetate rescues VEGFR3 levels and signaling, and lymphangiogenesis in LEC-Atg5-/- mice. Our findings reveal that lipophagy in LEC by supporting FAO, preserves a mitochondrial-PROX1 gene expression circuit that safeguards LEC responsiveness to lymphangiogenic mediators and lymphangiogenesis.

PubMed Disclaimer

Conflict of interest statement

The authors declare no competing interests.

Figures

Fig. 1
Fig. 1. Autophagy is essential for lymphatic endothelial cell homeostasis.
a Representative blots for indicated proteins of si CTRL, si ATG5, and chloroquine (CQ 25 µM, 48 h) treated LEC. Densitometric quantification is indicated beneath the blots. Mean ± SD, N = 3 biological replicates analyzed using one-way ANOVA, with Tukey’s test for multiple comparisons *p < 0.05 and ***p < 0.001 vs si CTRL. Representative images (b) and quantification (c) of si CTRL, si ATG5 and CQ-treated LEC after GFP-LC3 transfection. Nuclei are stained with DAPI. Scale bars represent 10 µm. Mean ± SD, N = 3 biological replicates analyzed by one-way ANOVA, with Tukey’s test for multiple comparisons, *p = 0.02 and ***p = 0.0002 vs si CTRL. Mean represents mean per independent experiment, with a minimum of 19 cells analyzed per condition per independent experiment. Representative images (d) and quantification (e) of si CTRL and si ATG5 LEC spheroids upon stimulation with VEGF-C (100 ng/mL, 48 h). Scale bar represents 100 µm. Mean ± SD, N = 4 independent donors analyzed using unpaired Student’s t test (two-tailed), **p = 0.004. Mean represents mean per independent experiment, with a minimum of 10 spheroids analyzed per condition per independent experiment. f [3H]Thymidine incorporation assay into DNA of si CTRL and si ATG5 LEC upon stimulation with VEGF-C. Mean ± SD, N = 3 biological replicates analyzed using unpaired Student’s t test (two-tailed), *p = 0.01. g Quantification of scratch wound healing for si CTRL and si ATG5 LEC monolayers in the presence of 500 μg/ml Mitomycin C. Mean ± SD, N = 3 analyzed using unpaired Student’s t test (two-tailed), ** p = 0.001. h Flow cytometry analysis of cell death by propidium iodide (PI) of si CTRL and si ATG5 LEC. Mean ± SD, N = 3 biological replicates analyzed using unpaired Student’s t test (two-tailed), p = 0.97. i Representative images of mixed spheroids containing equal amount of fluorescently green labeled si CTRL and red labeled si ATG5 LEC. Scale bar represents 100 µm. j Representative immunofluorescent images of whole corneal mounts (dashed lines) of wild type (WT) and LEC-Atg5 knock out mice (LEC-Atg5−/−) stained for LYVE1, 8 days post corneal cauterization. Scale bar represents 1 mm. k Representative immunofluorescent images of corneal sections dissected from WT and LEC-Atg5−/− mice stained for LYVE1 and CD31, 8 days post corneal cauterization. Scale bars represent 100 µm. l Quantification for the number of end points, number of branch points, average of cumulative length, and surface density for LYVE1+ lymphatic vessels. Mean ± SD, N = 5 corneas for WT and LEC-Atg5−/− analyzed using unpaired Student’s t test (two-tailed), **p < 0.01. m Quantification for the number of end points, number of branch points, average of cumulative length, and surface density for CD31+ blood vessels. Mean ± SD. N = 5 corneas for WT and LEC-Atg5−/− analyzed using unpaired Student’s t test (two-tailed), p = not significant.
Fig. 2
Fig. 2. Autophagy deficiency impairs expression of lymphatic markers.
a, b Representative blots for the indicated proteins in si CTRL, si ATG5 or CQ-(25 µM, 48 h) treated LEC. Densitometric quantification is indicated beneath the blots. Mean ± SD, N ≥ 3 biological replicates analyzed by one-way ANOVA, with Tukey’s test for multiple comparisons, *p < 0.05, **p < 0.01, ***p < 0.001 and ****p < 0.0001 vs si CTRL. c Representative blot for the indicated proteins in si CTRL and si ULK1 LEC. Densitometric quantification is indicated beneath the blots. Mean ± SD, N ≥ 3 biological replicates analyzed using unpaired Student’s t test, *p < 0.05, **p < 0.01 and ***p < 0.001. RT-qPCR analysis of si CTRL, si ATG5, or CQ-treated LEC. mRNA expression of VEGFR3, LYVE1, PROX1, and NR2F2 (relative to HPRT). Mean ± SD, N = 3 (d, g), N = 4 (e) or N = 6 (f) biological replicates analyzed by one-way ANOVA, with Tukey’s test for multiple comparisons, *p < 0.05, **p < 0.01, ***p < 0.001 and ****p < 0001 vs si CTRL. h–k RT-qPCR analysis of si CTRL and si ULK1 LEC. mRNA expression of VEGFR3, LYVE1, PROX1, and NR2F2 (relative to HPRT). Mean ± SD, N = 3 (i, j), N = 4 (h, k) biological replicates analyzed by unpaired Student’s t test (two-tailed), *p < 0.05, **p < 0.01, ****p < 0.0001. l Representative immunofluorescent images of LYVE1+ lymphatic vessels (dashed lines) from corneal sections dissected from wild type (WT) and LEC-Atg5 knock out mice (LEC-Atg5−/−) stained for-VEGFR3. Nuclei are stained with DAPI. Scale bar represents 100 µm. m Quantification of VEGFR3 fluorescent intensity (arb. unit, arbitrary unit) per vessel area. Mean ± SD, N = 3 corneas with a minimum of 24 images analyzed per condition, analyzed by unpaired Student’s t test (two-tailed), ****p < 0.0001.
Fig. 3
Fig. 3. Autophagy regulates lipid droplets homeostasis in lymphatic endothelial cells.
a Representative images of BODIPY 493/503 staining of lipid droplets in si CTRL, si ATG5, and CQ-(25 µM, 48 h) treated LEC. Nuclei are stained with DAPI. Scale bars represent 10 µm. b Quantification of lipid droplet number per cell is shown. Mean ± SD, N = 3 biological replicates analyzed by one-way ANOVA, with Tukey’s test for multiple comparisons, *p = 0.04, **p = 0.008 vs si CTRL. Mean represents mean per independent experiment, with a minimum of 52 cells analyzed per condition per independent experiment. c Representative images of BODIPY 493/503 staining of lipid droplets in untreated (CTRL) or CQ-treated LEC transfected with a mCherry-LC3 plasmid. Nuclei are stained with DAPI. Scale bars represent 10 µm. d Co-localization of BODIPY 493/503 and mcherry-LC3 in untreated (CTRL) or CQ-treated LEC (PC = Pearson coefficient). Mean ± SD, N = 3 biological replicates with a minimum of 40 cells per condition in total analyzed by Mann–Whitney test (two-tailed), ****p < 0.0001. e Triglycerides bar plots calculated as an average of the lipid subspecies analyzed across si CTRL and si ATG5 LEC. Mean ± SD, N = 4 independent LEC donors analyzed by unpaired Student’s t test (two-tailed), *p = 0.03. f Representative confocal images of BODIPY 493/503 fluorescence staining (green) of LYVE1+ lymphatic vessels (dashed lines) from corneal sections dissected from wild type (WT) and LEC-Atg5 knock out mice (LEC-Atg5−/−). A mask following the shape of the lymphatic vessel (LYVE1+ staining) was applied to isolate the lipid droplets (LDs) signal (BODIPY 493/503) coming from the LYVE1+ lymphatic vessel (dashed lines). Merged images of LYVE1 (magenta), BODIPY 493/503 (green), and DAPI (blue) represent masked images of the three channels. Nuclei are stained with DAPI. Scale bar represents 100 µm. g Quantification of BODIPY 493/503 fluorescent intensity calculated per LYVE1+ positive area (representative of lymphatic vessel area) (arb. unit, arbitrary unit). Mean ± SD, N = 6 corneas analyzed using unpaired Student’s t test (two-tailed), *p = 0.04. h Representative super resolution AiryScan images of BODIPY 493/503 fluorescence staining (green) of LYVE1+ lymphatic vessels (dashed lines) from corneal sections dissected from wild type (WT) and LEC-Atg5 knock out mice (LEC-Atg5−/−). Nuclei are stained with DAPI. Scale bars represent 10 µm.
Fig. 4
Fig. 4. LEC-autophagy is required for the traffic and release of lipids for mitochondria fatty acid oxidation.
Representative super resolution AiryScan live images (a) and quantification of colocalization of fluorescently labeled (C16:0) palmitate (BODIPY-PAL) in combination with NILE RED (Manders M1 coefficient) (b) and quantification of lipid droplets (c) in si CTRL and si ATG5 LEC pulsed with BODIPY-PAL (overnight incubation) and imaged at 0 h and after 24 h chase period. Zoomed boxed areas show BODIPY-PAL (green) colocalization with NILE RED-stained lipid droplets (magenta). Mean ± SD, N = 3 biological replicates with a minimum of 19 cells per condition in total analyzed using one-way ANOVA with Tukey’s test for multiple comparisons (b) and Kruskal–Wallis test with Dunn’s multiple comparisons test (c), **p < 0.01, ***p < 0.001 and ****p < 0.0001 vs si CTRL (0 h). Scale bars represent 10 μm. d–j Representative super resolution AiryScan live images (d) of si CTRL and si ATG5 LEC pulsed with BODIPY-PAL (overnight incubation) and imaged after 6 h and 24 h chase period. Mitochondria were labeled with MitoTracker deep red (30 min), before imaging (e, h), mitochondrial circularity (f, i), and index of mitochondrial fragmentation (number of mitochondria/ total mitochondrial area) (g, j). Zoomed boxed areas show colocalization of the mitochondria (magenta) with BODIPY-PAL-signals (green). Mean ± SD, N = 3 biological replicates with a minimum of 23 cells per condition in total analyzed using unpaired two-tailed Student’s t test (e, f, h) or two-tailed Mann-Whitney test (g, i, j), **p < 0.01 and ****p < 0.0001 vs si CTRL. Scale bars represent 10 μm. k Oxygen consumption rate (OCR) of LEC isolated from WT or LEC-Atg5−/− mice. l ATP-linked respiration. m Maximal respiration and n Spare respiratory capacity from (k). Mean ± SD, N = 3 biological respiration using unpaired Student’s t test (two-tailed), ****p < 0.0001. o OCR of si CTRL or si ATG5 LEC measured in ß-oxidation assay medium. Etomoxir (Eto, 40 μM) was added 15 min prior to baseline measurement in ß-oxidation assay medium. p Maximal respiration and q Spare respiratory capacity from OCR in (o). Mean ± SD, N = 4 biological replicates analyzed using one-way ANOVA with Tukey’s test for multiple comparisons, *p < 0.05 and **p < 0.01 vs si CTRL. r Fatty acid oxidation (FAO) assay of si CTRL and si ATG5 LEC supplemented with 2 mCi/mL [9,10-3H]-palmitic acid for 18 h. Mean ± SD, N = 4 biological replicates analyzed using one sample t test (two-tailed), **p = 0.003.
Fig. 5
Fig. 5. Autophagy regulates CPT1A expression and acetyl-CoA levels.
Representative blots for indicated proteins (a) and densitometric quantification (b) of CPT1A levels in si CTRL or si ATG5 LEC. Mean ± SD, N = 3 biological replicates analyzed using unpaired Student’s t test (two-tailed), ***p = 0.0001. c RT-qPCR analysis of si CTRL and si ATG5 LEC. mRNA expression of CPT1A (relative to HPRT). Mean ± SD, N = 3 biological replicates analyzed using one sample t test (two-tailed), *p = 0.03. d Relative Citric acid levels in si CTRL and si ATG5 LEC measured via mass spectrometry. Mean ± SD, N = 3 biological replicates analyzed using one sample t test (two-tailed), *p = 0.03. e Relative acetyl CoA/CoA ratio measured via mass spectrometry in si CTRL, si ATG5, or etomoxir (Eto, 100 µM, 48 h) treated LEC. Mean ± SD, N = 3 biological replicates analyzed using one way ANOVA with Tukey’s test for multiple comparisons, *p < 0.05 vs si CTRL. Representative blots (f) and densitometric quantification (g) for H3K9 ac levels relative to total H3 of si CTRL and si ATG5 LEC. Mean ± SD, N = 5 biological replicates analyzed using unpaired Student’s t test (two-tailed), *p = 0.02. h Representative blots for indicated proteins with GAPDH as loading control of si CTRL, si ATG5, si ATG7, or si ULK1 LEC. Densitometric quantification is indicated beneath the blots (H3K9 ac levels relative to total H3 and ATG5, ULK1, and ATG7 levels relative to GAPDH). Mean ± SD, N ≥ 3 biological replicates analyzed using one way ANOVA with Tukey’s test for multiple comparisons, *p < 0.05, **p < 0.01, ***p < 0.001 and ****p < 0.0001 vs si CTRL.
Fig. 6
Fig. 6. Acetate rescues the mitochondrial-Prox1 transcriptional circuit caused by loss of autophagy in LEC.
a–c RT-qPCR analysis of si CTRL and si ATG5 LEC treated with sodium acetate (AC, 20 mM, 48 h) or vehicle. mRNA expression of Prox1, VEGFR3, and LYVE1 (relative to HPRT). Mean ± SD, N = 3 biological replicates analyzed by one-way ANOVA, with Tukey’s test for multiple comparisons, **p < 0.01 and ***p < 0.001 vs si CTRL, ##p < 0.01 and ###p < 0.001 vs si ATG5. d Representative blots for the indicated proteins in si CTRL or si ATG5 LEC treated with AC or vehicle. Densitometric quantification is indicated beneath the blots. Mean ± SD, N ≥ 3 biological replicates analyzed by one-way ANOVA, with Tukey’s test for multiple comparisons, *p < 0.05,***p < 0.001. e Representative blots for indicated proteins in si CTRL and si ATG5 LEC treated with palmitate (PAL, 500 nM, 48 h) or BSA. Densitometric quantification is indicated beneath the blots. Mean ± SD, N ≥ 3 biological replicates analyzed by one-way ANOVA, with Tukey’s test for multiple comparisons. *p < 0.05, **p < 0.01, ***p < 0.001 and ****p < 0.0001. f Representative immunofluorescent images of si CTRL and si ATG5 treated with AC or vehicle and stained for mitochondrial protein TOMM20 and BODIPY 493/503. Nuclei are stained with DAPI. Scale bars represent 10 µm. g Quantification of mitochondrial fragmentation index (number of mitochondria/ total mitochondrial area). Mean ± SD, N = 3 biological replicates analyzed by one-way ANOVA, with Tukey’s test for multiple comparisons, *p < 0.05 vs si CTRL, #p < 0.05 vs si ATG5. Mean represents mean per independent experiment, with a minimum of 12 cells analyzed per condition per independent experiment. h Quantification of lipid droplet number per cell. Mean ± SD, N = 4 biological replicates analyzed by one-way ANOVA, with Tukey’s test for multiple comparisons, *p < 0.05 vs si CTRL. Mean represents mean per independent experiment, with a minimum of 29 cells analyzed per condition per independent experiment. i Representative super resolution AiryScan images of si CTRL and si ATG5, si CTRL + si DRP1, and si ATG5 + si DRP1 LEC stained for the mitochondrial protein TOMM20 and BODIPY 493/503. Nuclei are stained with DAPI. Scale bars represent 10 µm. j Quantification of mitochondrial index of fragmentation (number of mitochondria/ total mitochondrial area). Mean ± SD, N = 3 biological replicates analyzed by one-way ANOVA, with Tukey’s test for multiple comparisons, ****p < 0.0001 vs si CTRL, ####p < 0.0001 vs si ATG5. Mean represents mean per independent experiment, with a minimum of 35 cells analyzed per condition per independent experiment. k Quantification of lipid droplet number per cell. Mean ± SD, N = 3 biological replicates analyzed by one-way ANOVA, with Tukey’s test for multiple comparisons, **p < 0.01 vs si CTRL. Mean represents mean per independent experiment, with a minimum of 47 cells analyzed per condition per independent experiment. l–o RT-qPCR analysis in si CTRL, si ATG5, si CTRL + si DRP1 and si ATG5 + si DRP1 LEC. mRNA expression of PROX1, VEGFR3, LYVE1, and CPT1A (relative to HPRT). Mean ± SD, N = 3 (l, n, o) and N = 4 (m) biological replicates analyzed by one-way ANOVA, with Tukey’s test for multiple comparisons, **p < 0.01 vs si CTRL and p = ns vs si ATG5.
Fig. 7
Fig. 7. Acetate rescues LEC function in autophagy inhibited cells in vitro and injury-driven lymphangiogenesis in LEC-ATG5 KO mice.
a Representative spheroid images from si CTRL and si ATG5 LEC under basal conditions, stimulated with recombinant VEGF-C (100 ng/mL) alone, supplemented with sodium acetate (AC, 20 mM, 48 h) alone or in combination. Scale bars represent 100 µm. b Quantification of sprouts per spheroid. Mean ± SD, N = 3 independent donors analyzed using Two-way ANOVA, with Tukey’s test for multiple comparisons **p < 0.05 vs si CTRL, ####p < 0.0001 vs si ATG5. Mean represents mean per independent experiment, with a minimum of 10 spheroids analyzed per condition per independent experiment. c Representative immunofluorescent images of whole corneal mounts (dashed lines) of wild type (WT) and LEC-Atg5 knock out mice (LEC-Atg5−/−) treated with vehicle or sodium acetate (AC, 400 µl 0.5 M i.p daily) stained for LYVE1, 8 days post corneal cauterization. Scale bar represents 1 mm. d–g Quantification of the number of end points, number of branch points, average of cumulative length, and surface density for LYVE1+ lymphatic vessels. Mean ± SD, N = 21 corneas for WT, N = 24 corneas for LEC-Atg5−/−, N = 15 corneas for WT + AC and N = 27 corneas for LEC-Atg5−/− + AC analyzed using one-Way ANOVA corrected for multiple comparisons using Tukey’s test *p < 0.05, ***p < 0.001 vs WT, #p < 0.05 ## p < 0.01 vs LEC-Atg5−/−. h Representative immunofluorescent images of corneal sections dissected from WT and LEC-Atg5−/− treated with vehicle or sodium acetate (AC), stained for LYVE1 and CD31, 8 days post corneal cauterization. Scale bar 100 µm. i–l Quantification for the number of end points, number of branch points, average of cumulative length, and surface density for the CD31+ blood vessels. Mean ± SD, N = 8 corneas for WT, N = 5 corneas for LEC-Atg5−/−, N = 8 corneas for WT + AC, and N = 7 corneas for LEC-Atg5−/− + AC analyzed using one-Way ANOVA corrected for multiple comparisons using Tukey’s test. p = not significant.
Fig. 8
Fig. 8. Lipophagy maintains lymphatic identity and lymphangiogenesis by supplying fatty acids to the mitochondria for fatty acid oxidation and PROX1-mediated gene expression.
The figure depicts the main phenotypes of (left to right) autophagy-proficient LEC (autophagy ON), autophagy-deficient LEC (autophagy OFF), and their rescue with acetate, described in the study. Under homeostatic conditions (left panel), constitutive levels of autophagy favor the turnover of lipid droplets (LDs) in order to supply free fatty acids to the mitochondria and foster fatty acid oxidation (FAO). The actively respiring elongated mitochondria replete acetyl-CoA pools, which is used by the acetyltransferase p300 to acetylate histones at PROX1-target genes. This autophagy-regulated mechanism maintains PROX1-driven expression of lymphatic markers and VEGFR3-mediated lymphangiogenic signaling. When autophagy is compromised (center panel), LDs turnover is impaired causing accumulation of LDS in the cytosol. Under these conditions, the ability of mitochondria to perform FAO is compromised, the mitochondrial network is fragmented and acetyl-CoA levels drop to levels unable to support transcription of PROX1-driven lymphatic gene networks and lymphangiogenesis. This phenotype can be rescued by feeding LEC with the acetyl-CoA precursor acetate (right panel). Acetate rescues mitochondrial morphology and FAO and recovers PROX1-mediated expression of lymphatic genes, thus bypassing the lymphangiogenesis defects caused by genetic loss of autophagy. Image created with Biorender.com.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

See all "Cited by" articles

References

    1. Alitalo K. The lymphatic vasculature in disease. Nat. Med. 2011;17:1371–1380. doi: 10.1038/nm.2545. - DOI - PubMed
    1. Stacker SA, et al. Lymphangiogenesis and lymphatic vessel remodelling in cancer. Nat. Rev. Cancer. 2014;14:159–172. doi: 10.1038/nrc3677. - DOI - PubMed
    1. François M, et al. Sox18 induces development of the lymphatic vasculature in mice. Nature. 2008;456:643–647. doi: 10.1038/nature07391. - DOI - PubMed
    1. Srinivasan RS, et al. The nuclear hormone receptor Coup-TFII is required for the initiation and early maintenance of Prox1 expression in lymphatic endothelial cells. Genes Dev. 2010;24:696–707. doi: 10.1101/gad.1859310. - DOI - PMC - PubMed
    1. Srinivasan RS, et al. The Prox1-Vegfr3 feedback loop maintains the identity and the number of lymphatic endothelial cell progenitors. Genes Dev. 2014;28:2175–2187. doi: 10.1101/gad.216226.113. - DOI - PMC - PubMed

Publication types