Skip to main content

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

  • Article
  • Published:

The role of fatty acid β-oxidation in lymphangiogenesis

Abstract

Lymphatic vessels are lined by lymphatic endothelial cells (LECs), and are critical for health. However, the role of metabolism in lymphatic development has not yet been elucidated. Here we report that in transgenic mouse models, LEC-specific loss of CPT1A, a rate-controlling enzyme in fatty acid β-oxidation, impairs lymphatic development. LECs use fatty acid β-oxidation to proliferate and for epigenetic regulation of lymphatic marker expression during LEC differentiation. Mechanistically, the transcription factor PROX1 upregulates CPT1A expression, which increases acetyl coenzyme A production dependent on fatty acid β-oxidation. Acetyl coenzyme A is used by the histone acetyltransferase p300 to acetylate histones at lymphangiogenic genes. PROX1–p300 interaction facilitates preferential histone acetylation at PROX1-target genes. Through this metabolism-dependent mechanism, PROX1 mediates epigenetic changes that promote lymphangiogenesis. Notably, blockade of CPT1 enzymes inhibits injury-induced lymphangiogenesis, and replenishing acetyl coenzyme A by supplementing acetate rescues this process in vivo.

This is a preview of subscription content, access via your institution

Access options

Rent or buy this article

Prices vary by article type

from$1.95

to$39.95

Prices may be subject to local taxes which are calculated during checkout

Figure 1: CPT1A expression and FAO flux are higher in LECs.
Figure 2: Inhibition of CPT1A impairs lymphatic development.
Figure 3: CPT1A regulates VEC-to-LEC differentiation.
Figure 4: PROX1-mediated FAO produces acetyl-CoA for histone H3 acetylation.
Figure 5: FAO enhances histone acetylation at lymphatic genes.
Figure 6: Acetate supplementation recovers lymphangiogenesis.

Similar content being viewed by others

Accession codes

Primary accessions

Gene Expression Omnibus

References

  1. Christiansen, A. & Detmar, M. Lymphangiogenesis and cancer. Genes Cancer 2, 1146–1158 (2011)

    Article  Google Scholar 

  2. Kim, H., Kataru, R. P. & Koh, G. Y. Regulation and implications of inflammatory lymphangiogenesis. Trends Immunol. 33, 350–356 (2012)

    Article  CAS  Google Scholar 

  3. Alitalo, K. The lymphatic vasculature in disease. Nat. Med. 17, 1371–1380 (2011)

    Article  CAS  Google Scholar 

  4. Ghesquière, B., et al. Metabolism of stromal and immune cells in health and disease. Nature 511, 167–176 (2014)

    Article  ADS  Google Scholar 

  5. Petrova, T. V. et al. Lymphatic endothelial reprogramming of vascular endothelial cells by the Prox-1 homeobox transcription factor. EMBO J. 21, 4593–4599 (2002)

    Article  CAS  Google Scholar 

  6. Wigle, J. T. et al. An essential role for Prox1 in the induction of the lymphatic endothelial cell phenotype. EMBO J. 21, 1505–1513 (2002)

    Article  CAS  Google Scholar 

  7. Yang, Y. & Oliver, G. Development of the mammalian lymphatic vasculature. J. Clin. Invest. 124, 888–897 (2014)

    Article  CAS  Google Scholar 

  8. Aranguren, X. L. et al. COUP-TFII orchestrates venous and lymphatic endothelial identity by homo- or hetero-dimerisation with PROX1. J. Cell Sci. 126, 1164–1175 (2013)

    Article  CAS  Google Scholar 

  9. Srinivasan, R. S. et al. Lineage tracing demonstrates the venous origin of the mammalian lymphatic vasculature. Genes Dev. 21, 2422–2432 (2007)

    Article  CAS  Google Scholar 

  10. Srinivasan, R. S. et al. The Prox1–Vegfr3 feedback loop maintains the identity and the number of lymphatic endothelial cell progenitors. Genes Dev. 28, 2175–2187 (2014)

    Article  CAS  Google Scholar 

  11. Schoors, S. et al. Fatty acid carbon is essential for dNTP synthesis in endothelial cells. Nature 520, 192–197 (2015)

    Article  CAS  ADS  Google Scholar 

  12. Karkkainen, M. J. et al. Vascular endothelial growth factor C is required for sprouting of the first lymphatic vessels from embryonic veins. Nat. Immunol. 5, 74–80 (2004)

    Article  CAS  Google Scholar 

  13. Hägerling, R. et al. A novel multistep mechanism for initial lymphangiogenesis in mouse embryos based on ultramicroscopy. EMBO J. 32, 629–644 (2013)

    Article  Google Scholar 

  14. Bazigou, E. et al. Genes regulating lymphangiogenesis control venous valve formation and maintenance in mice. J. Clin. Invest. 121, 2984–2992 (2011)

    Article  CAS  Google Scholar 

  15. Lin, F. J. et al. Direct transcriptional regulation of neuropilin-2 by COUP-TFII modulates multiple steps in murine lymphatic vessel development. J. Clin. Invest. 120, 1694–1707 (2010)

    Article  CAS  Google Scholar 

  16. Hirashima, M. et al. Lymphatic vessel assembly is impaired in Aspp1-deficient mouse embryos. Dev. Biol. 316, 149–159 (2008)

    Article  CAS  Google Scholar 

  17. Mouta-Bellum, C. et al. Organ-specific lymphangiectasia, arrested lymphatic sprouting, and maturation defects resulting from gene-targeting of the PI3K regulatory isoforms p85α, p55α, and p50α. Dev. Dyn. 238, 2670–2679 (2009)

    Article  CAS  Google Scholar 

  18. Stanczuk, L. et al. cKit Lineage Hemogenic Endothelium-Derived Cells Contribute to Mesenteric Lymphatic Vessels. Cell Rep. S2211-1247(15)00172-2 (2015)

  19. Hong, Y. K. et al. Prox1 is a master control gene in the program specifying lymphatic endothelial cell fate. Dev. Dyn. 225, 351–357 (2002)

    Article  CAS  Google Scholar 

  20. Johnson, N. C. et al. Lymphatic endothelial cell identity is reversible and its maintenance requires Prox1 activity. Genes Dev. 22, 3282–3291 (2008)

    Article  CAS  Google Scholar 

  21. Ichise, T., Yoshida, N. & Ichise, H. Ras/MAPK signaling modulates VEGFR-3 expression through Ets-mediated p300 recruitment and histone acetylation on the Vegfr3 gene in lymphatic endothelial cells. PLoS One 7, e51639 (2012)

    Article  CAS  ADS  Google Scholar 

  22. Goldman, J. et al. Cooperative and redundant roles of VEGFR-2 and VEGFR-3 signaling in adult lymphangiogenesis. FASEB J. 21, 1003–1012 (2007)

    Article  CAS  Google Scholar 

  23. Hay, D. A. et al. Discovery and optimization of small-molecule ligands for the CBP/p300 bromodomains. J. Am. Chem. Soc. 136, 9308–9319 (2014)

    Article  CAS  Google Scholar 

  24. Lee, J. V. et al. Akt-dependent metabolic reprogramming regulates tumor cell histone acetylation. Cell Metab. 20, 306–319 (2014)

    Article  CAS  Google Scholar 

  25. Henry, R. A., Kuo, Y. M. & Andrews, A. J. Differences in specificity and selectivity between CBP and p300 acetylation of histone H3 and H3/H4. Biochemistry 52, 5746–5759 (2013)

    Article  CAS  Google Scholar 

  26. Xu, Y. M., Du, J. Y. & Lau, A. T. Posttranslational modifications of human histone H3: an update. Proteomics 14, 2047–2060 (2014)

    Article  Google Scholar 

  27. Saharinen, P., Tammela, T., Karkkainen, M. J. & Alitalo, K. Lymphatic vasculature: development, molecular regulation and role in tumor metastasis and inflammation. Trends Immunol. 25, 387–395 (2004)

    Article  CAS  Google Scholar 

  28. Detry, B. et al. Sunitinib inhibits inflammatory corneal lymphangiogenesis. Invest. Ophthalmol. Vis. Sci. 54, 3082–3093 (2013)

    Article  CAS  Google Scholar 

  29. Wellen, K. E. et al. ATP-citrate lyase links cellular metabolism to histone acetylation. Science 324, 1076–1080 (2009)

    Article  CAS  ADS  Google Scholar 

  30. Stacker, S. A. et al. Lymphangiogenesis and lymphatic vessel remodelling in cancer. Nat. Rev. Cancer 14, 159–172 (2014)

    Article  CAS  Google Scholar 

  31. Bazigou, E. et al. Integrin-α9 is required for fibronectin matrix assembly during lymphatic valve morphogenesis. Dev. Cell 17, 175–186 (2009)

    Article  CAS  Google Scholar 

  32. Jaffe, E. A., Nachman, R. L., Becker, C. G. & Minick, C. R. Culture of human endothelial cells derived from umbilical veins. Identification by morphologic and immunologic criteria. J. Clin. Invest. 52, 2745–2756 (1973)

    Article  CAS  Google Scholar 

  33. Kriehuber, E. et al. Isolation and characterization of dermal lymphatic and blood endothelial cells reveal stable and functionally specialized cell lineages. J. Exp. Med. 194, 797–808 (2001)

    Article  CAS  Google Scholar 

  34. Michieli, P. et al. Targeting the tumor and its microenvironment by a dual-function decoy Met receptor. Cancer Cell 6, 61–73 (2004)

    Article  CAS  Google Scholar 

  35. Carlotti, F. et al. Lentiviral vectors efficiently transduce quiescent mature 3T3-L1 adipocytes. Mol. Ther. 9, 209–217 (2004)

    Article  CAS  Google Scholar 

  36. Korff, T., Krauss, T. & Augustin, H. G. Three-dimensional spheroidal culture of cytotrophoblast cells mimics the phenotype and differentiation of cytotrophoblasts from normal and preeclamptic pregnancies. Exp. Cell. Res. 297, 415–423 (2004)

    Article  CAS  Google Scholar 

  37. Blacher, S., Detry, B., Bruyère, F., Foidart, J. M. & Noël, A. Additional parameters for the morphometry of angiogenesis and lymphangiogenesis in corneal flat mounts. Exp. Eye Res. 89, 274–276 (2009)

    Article  CAS  Google Scholar 

  38. Zhang, Y. et al. Model-based analysis of ChIP-Seq (MACS). Genome Biol. 9, R137 (2008)

    Article  Google Scholar 

  39. Xu, H. et al. A signal-noise model for significance analysis of ChIP-seq with negative control. Bioinformatics 26, 1199–1204 (2010)

    Article  CAS  Google Scholar 

  40. Robinson, M.D., McCarthy, D. J. & Smyth, G. K. edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics 26, 139–140 (2010)

    Article  CAS  Google Scholar 

  41. Basu, S. S. & Blair, I. A. SILEC: a protocol for generating and using isotopically labeled coenzyme A mass spectrometry standards. Nat. Protocols 7, 1–12 (2011)

    Article  Google Scholar 

  42. Culic, O., Gruwel, M. L. & Schrader, J. Energy turnover of vascular endothelial cells. Am. J. Physiol. 273, C205–C213 (1997)

    Article  CAS  Google Scholar 

  43. Ataullakhanov, F. I. & Vitvitsky, V. M. What determines the intracellular ATP concentration. Biosci. Rep. 22, 501–511 (2002)

    Article  CAS  Google Scholar 

  44. Huang, H. et al. Physiological levels of ATP negatively regulate proteasome function. Cell Res. 20, 1372–1385 (2010)

    Article  CAS  Google Scholar 

  45. Atkinson, D. E. & Walton, G. M. Adenosine triphosphate conservation in metabolic regulation. Rat liver citrate cleavage enzyme. J. Biol. Chem. 242, 3239–3241 (1967)

    CAS  PubMed  Google Scholar 

  46. Aragones, J. et al. Deficiency or inhibition of oxygen sensor Phd1 induces hypoxia tolerance by reprogramming basal metabolism. Nat. Genet. 40, 170–180 2008)

    Article  CAS  Google Scholar 

  47. McCloy, R. A. et al. Partial inhibition of Cdk1 in G2 phase overrides the SAC and decouples mitotic events. Cell Cycle 13, 1400–1412 (2014)

    Article  CAS  Google Scholar 

Download references

Acknowledgements

We thank R. Adams and T. Mäkinen for providing the VE-cadherin(PAC)-creERT2 and Prox1-creERT2 mice, respectively. This work was supported by: fellowships from FWO (to B.W.W., X.W., B.T., J.K., R.M., U.B., J.G., B.G.), Marie Curie (to B.W.W., M.G., U.B.), EMBO (to H.H.) and LE&RN/FDRS (A.Z.); and supporting grants from IUAP P7/03 (P.C.), Methusalem funding by the Flemish Government (P.C.), FWO (G.0598.12, G.0532.10, G.0817.11, G.0834.13, 1.5.202.10.N Krediet aan navorsers, to P.C.), Leducq Transatlantic Network Artemis (P.C.), AXA Research Fund (1465, to P.C.), Foundation against Cancer (P.C.), ERC Advanced Research Grant (EU-ERC269073, to P.C.), ERC Starting Grant (IMAGINED-201293, to A.L.), German Research Foundation (D.F.G.) Grants (CRC629, CRC656, to F.K.), and co-funding by KU Leuven Methusalem (S.M.F.). We would like to thank A. Bouché, A. Carton, A. Manderveld, K. Peeters, N. Dai, J. Souffreau, A. van Nuffelen, B. Tembuyser, A. Van Den Eynde, S. Christen, K. Feyen, W. Martens, K. Brepoels, P.J. Coolen, M. Nijs, P. Vanwesemael, B. Verherstraeten, G. Dubois, E. Van Dyck, E. Gils, B. Vanwetswinkel, D. Smeets, G. Peuteman, T. Van Brussel, B. Boeckx, A. Acosta Sanchez and D. Verdegem for their technical assistance, and various laboratory members for their feedback and discussions. We thank S. Aerts, H. Gerhardt and M. Mazzone for critical suggestions and discussions.

Author information

Authors and Affiliations

Authors

Contributions

B.W.W., X.W., A.Z., B.T., I.C., J.K., M.G., R.M., H.H., U.B., S.B., S.V., J.G., H.Z., C.D., C.S., R.H., V.M., S.W., M.L., B.G., L.S., M.D., G.E. performed research and/or analysed the data; B.W.W., X.W., A.Z., B.T., I.C., M.D., D.L., P.C. designed experiments; M.K., S.J., F.K., S.M.F. provided mice and/or advice; A.L., A.N., L.M., D.L. provided reagents and discussed results; B.W.W., A.Z. made the figures; B.W.W., P.C. wrote the paper; P.C. conceptualized the study. All authors discussed the results and commented on the manuscript.

Corresponding author

Correspondence to Peter Carmeliet.

Ethics declarations

Competing interests

P.C. declares to be named as inventor on patent applications, claiming subject matter related to the results described in this paper. The other authors declare no competing financial interests.

Extended data figures and tables

Extended Data Figure 1 LECs have higher expression of fatty acid β-oxidation and transport proteins.

a, Quantitative RT–PCR analysis of the mRNA levels of carnitine palmitoyltransferase CPT1A, CPT1B, CPT1C and CPT2 relative to the housekeeping gene HPRT1 in VECs and LECs (n = 3). bf, Fold change in the mRNA levels of fatty acid binding protein FABP4 (b), FABP5 (c), fatty acid transfer protein FATP3 (d), FATP4 (e) and CD36 (fatty acid translocase, FAT) (f) in LECs versus VECs (n = 3). g, Fold change in PROX1 mRNA expression in LECs silenced for PROX1 (siPROX1) versus control silenced LECs (ctrl) (n = 3). Mean ± s.e.m. of n independent experiments. Statistical test: t-test was used for comparison of two groups. *P < 0.05; NS, not statistically significant.

Extended Data Figure 2 Cpt1a gene excision in Prox1∆CPT1A embryos and additional phenotypic information and kinetics of loss of CPT1A in vivo and in vitro.

a, Genotyping for the presence of the Cpt1a floxed (2,200 bp) and excision band (300 bp) in E14.5 Prox1∆CPT1A embryos. b, Cartoon of anatomical orientation of the dorsal aorta (DA) and jugular vein (JV) with respect to the jugular lymphatic sacs (JLS). The red dotted lines in the embryo drawing indicate the jugular region of focus in wild-type E14.5 mouse embryos. SC, spinal cord; VB, vertebral body. c, d, Staining for CPT1A (magenta) and VEGFR3 (green) in wild-type (Prox1) and Prox1∆CPT1A embryos at E14.5. CPT1A immunoreactivity (magenta) is seen within VEGFR3+ LECs (green) of the JLS in wild-type embryos (c). In the JLS of the E14.5 Prox1∆CPT1A embryo (d), CPT1A immunoreactivity is undetectable. For the CPT1A signal channel only (left panels), a higher magnification is shown than for the VEGFR3 signal channel (top right panels) or the merged signal (bottom right panels). The white dashed lines denote endothelium in the JLS. Arrows denote LECs without apparent CPT1A immunoreactivity. Scale bars, 5 μm. e, Genotyping for the presence of the Cpt1a floxed (2,200 bp) and excision band (300 bp) in E9.5 Prox1FK∆CPT1A embryos. f, g, Representative micrographs of staining for CPT1A (magenta) and VEGFR3 (green) in LECs within the JLS at E11.5 in a wild-type (f) and Prox1∆CPT1A (g) embryo. Top panels show the single channel CPT1A signal. The insets represent an overlay of CPT1A and VEGFR3 signals (bottom left f, g) and the single channel signal for VEGFR3 (bottom right f, g). Dashed white lines outline the endothelium of the JLS, as identified by VEGFR3 staining. Arrows highlight areas within the lymphatic endothelium with strong CPT1A immunoreactivity in VEGFR3+ cells within the JLS of a wild-type embryo (f) and weaker but still detectable CPT1A-immunoreactive signal in VEGFR3+ cells within the JLS in a Prox1∆CPT1A embryo (g). h, Primary antibody omission for CPT1A and VEGFR3. Isolectin B4 (blue) was used to highlight vascular structures in the negative control (wild-type E14.5 embryo), and the JLS characteristically was more weakly stained compared to the jugular vein (upper panel). As expected, primary antibody omission (VEGFR3 (488, middle panel); CPT1A (Cy5, lower panel)) revealed no aspecific immunoreactivity. i, j, Representative micrographs of staining for CPT1A (magenta) and VEGFR3 (green) in LECs within the JLS at E11.5 in Prox1FK∆CPT1A (i) and Flt4∆CPT1A (j) embryo. The left panels show the single channel CPT1A signal. The insets represent an overlay of CPT1A and VEGFR3 signals (upper i, j), the single channel signal for VEGFR3 (bottom left i, j) and the primary antibody omission for CPT1A (bottom right i, j). As expected, primary antibody omission revealed no aspecific immunoreactivity. Dashed white lines outline the endothelium of the JLS, as identified by VEGFR3 staining. Arrows highlight areas within the lymphatic endothelium with still detectable CPT1A-immunoreactive signal in VEGFR3+ cells within the JLS in a Prox1FK∆CPT1A (i) and Flt4∆CPT1A (j) embryo. Scale bars, 10 μm. k, Genotyping for the presence of the Cpt1a floxed (2,200 bp) and excision band (300 bp) in E9.5 Prox1∆CPT1A embryos. l, Quantitative RT–PCR analysis of CPT1A mRNA expression at 1–3 days after CPT1AKD in VECs (n = 3). m, Densitometric quantification of immunoblotted CPT1A protein relative to β-ACTIN at 1–4 days after CPT1AKD in VECs, illustrating that, probably owing to the CPT1A protein stability, as much as 35–40% residual CPT1A protein levels were still detectable at 3 days after silencing CPT1A (n = 3). n, Representative stereomicroscope images of E10.5 wild-type (left) and VEcad∆CPT1A (right) embryos after tamoxifen administration from E7.5–E9.5, showing severe developmental and vascular abnormalities. o, p, Representative micrograph of staining for CPT1A (magenta) and VEGFR3 (green) in an E16.5 Prox1FK∆CPT1A (o) and Flt4∆CPT1A (p) embryo. Compared to CPT1A immunoreactivity seen within the VEGFR3+ LECs of the JLS in wild-type embryos (Extended Data Fig. 2c), CPT1A immunoreactivity is undetectable in LECs of E16.5 Prox1FK∆CPT1a (o) and Flt4∆CPT1A (p) embryos (left panels). The right panels represent the single channel signal for VEGFR3 (top right) and the merged signal for CPT1A and VEGFR3 (bottom right), highlighting the lack of detectable CPT1A immunoreactivity in mutant embryos. The white dashed lines denote the lymphatic endothelium in the JLS. Arrows denote LECs without apparent CPT1A immunoreactivity. Of note, owing to our TSA-based amplification method for all of our immunostaining protocols of embryo sections (see Methods), the VEGFR3 immunoreactivity seen in o and p was maximally enhanced, which is why VEGFR3 immunoreactivity does not appear lower for the gene-deficient embryos compared to wild type shown in Extended Data Fig. 2c. Scale bars, 5 μm. Mean ± s.e.m. Statistical test: t-test was used for comparison of two groups. *P < 0.05; NS, not statistically significant.

Extended Data Figure 3 Early and later-stage lymphatic-specific loss of CPT1A results in defects in lymphatic development.

a, Representative stereomicrographs of an E14.5 Prox1 (left) and a Prox1∆CPT1A embryo with deletion of Cpt1a in PROX1-positive endothelial cells (right). Asterisks denote subdermal oedema. Arrows highlight subdermal bleeding spots. b, Zoomed in micrograph of the Prox1∆CPT1A embryo shown in a (right). Arrows denote bleeding spots. c, Representative micrograph of TER119 immunostaining (green) of erythroid precursors (arrows) present within the JLS of a Prox1∆CPT1A embryo, as highlighted by LYVE1 staining (red). Dotted lines denote endothelium in lymphatic structures. Scale bar, 100 μm. d, Zoomed in micrograph of lymphatic capillaries of an E16.5 Prox1∆CPT1A embryo, visualized by Evans blue dye injection (blue) in the forepaws. Arrows indicate blood-filled blunt-ended dermal lymphatic structures. eo, Experiments involving the anterior dorsal dermal lymphatic vessels of E15.5 wild-type (Prox1) and Prox1∆CPT1A embryos. e, Immunostaining for LYVE1+ lymphatic vessels. White dashed lines: left and right lymphatic front. Lower panels: high magnification images of red, boxed regions. Scale bars, 500 μm. f, Quantitation of the lymphatic vessel outgrowth towards the midline in the anterior dorsal back skin, as measured by the average remaining gap between left and right flank lymphatics (n = 6 embryos for wild type; n = 9 embryos for Prox1∆CPT1A; 3 litters). gi, Quantification of the total vessel length relative to the total field area (g), number of branch points (corrected for lymphatic length) (h) and average vessel width (i) (n = 6 embryos for wild type; n = 9 embryos for Prox1∆CPT1A; 3 litters). j, LYVE1+ dermal lymphatic vessels in an E15.5 Prox1∆CPT1A embryo or an E16.5 Prox1FK∆CPT1A embryo highlighting the presence of lymphatic islands (white arrows). Scale bars, 500 μm. km, Experiments involving the anterior dorsal dermal lymphatic vessels of E15.5 wild-type (Prox1) and Prox1∆CPT1A embryos treated with tamoxifen from E10.5–E12.5 to inactivate Cpt1a in LECs after initial lymphatic structure formation has occurred. k, Immunostaining for LYVE1+ lymphatic vessels. White dashed lines, left and right lymphatic front. Lower panels, high magnification images of red boxed regions. Scale bars, 500 μm. l, Quantitation of the lymphatic vessel outgrowth towards the midline in the anterior dorsal back skin, as measured by the average remaining gap between left and right flank lymphatics (n = 5 embryos for wild type; n = 6 embryos for Prox1∆CPT1A; 2 litters). m, Quantification of the total vessel length relative to the total field area (n = 5 embryos for wild type; n = 6 embryos for Prox1∆CPT1A; 2 litters). np, Experiments involving the anterior dorsal dermal lymphatic network in E15.5 wild-type (Prox1) and Prox1∆CPT1A embryos. n, High magnification images of VEGFR3+ (green) lymphatic sprouts. The white arrows indicate filopodia. Note the weaker VEGFR3 immunoreactive signal in the Prox1∆CPT1A embryo (right panel). Inset panels represent LYVE1 staining, confirming the identification of filopodia by another marker (note that LYVE1 also stains macrophages, which are sometimes in the proximity of lymphatic sprouts, but clearly distinguishable based on their morphology). Scale bars, 10 μm. o, Number of filopodia per lymphatic sprout tip (n = 6 embryos for wild type; n = 9 embryos for Prox1∆CPT1A; 3 litters). p, VEGFR3 (white) immunoreactivity in the growing dorsal dermal lymphatic vessels. Note the weaker VEGFR3 immunoreactive signal in the Prox1∆CPT1A embryo (right panel). Scale bars, 100 μm. Mean ± s.e.m. Statistical test: t-test was used for comparison of two groups. *P < 0.05.

Extended Data Figure 4 Lymphatic-specific loss of CPT1A results in defects in tracheal lymphatic development, VEGFR3-driven loss of CPT1A phenocopies Prox1-driven excision and pharmacological inhibition of FAO induces lymphatic defects at later stages of embryonic development.

ac, Investigation of lymphatic growth in the trachea of E16.5 wild-type (denoted as Prox1) and Prox1∆CPT1A embryos. a, LYVE1+ lymphatic structures in the trachea. Scale bars, 200 μm. Bottom panels, high magnification images of regions denoted by the white boxes. In the trachea, lymphatic vessels first grow horizontally, after which sprouts branch off vertically (white arrows). Note the underdevelopment of lymphatic growth in the mutant embryo (yellow arrowheads). b, c, Quantitation of the average length of vertical connections (b) and average number of branch points (c) (n = 6 embryos for wild type; n = 5 embryos for Prox1∆CPT1A; 2 litters). di, Experiments in E16.5 wild-type control (denoted as Flt4) and Flt4∆CPT1A embryos. d, Genotyping for the presence of the Cpt1a floxed (2,200 bp) and excision band (300 bp) at E11.5. e, LYVE1+ dorsal dermal lymphatic vessels. White dashed lines: left and right lymphatic front. Bottom panels: high magnification insets of red boxed region. Note that LYVE1 also stains macrophages. Scale bars, 500 μm. f, Quantitation of the lymphatic vessel outgrowth towards the midline, as measured by the average remaining gap between left and right flank lymphatics (n = 7 embryos for Flt4; n = 8 embryos for Flt4∆CPT1A; 2 litters). gi, Quantitation of the total vessel length (normalized per total area) (g), number of branch points (corrected for lymphatic length) (h) and average vessel width (i) (n = 7 embryos for Flt4; n = 8 embryos for Flt4∆CPT1A; 2 litters). jx, Embryonic phenotype of vehicle- (ctrl) or etomoxir-treated (eto) embryos. j, Representative stereomicrographs at E14.5. Asterisks denote subdermal oedema. k, VEGFR3 immunostaining in the jugular region of E14.5 embryos. Solid white line, jugular vein (JV); dashed white line, JLS. Scale bars, 50 μm. l, JLS area in E14.5 embryos (n = 8 embryos for ctrl; n = 5 embryos for eto; 1 litter for each group). ms, Analysis of E16.5 ctrl- or eto-treated embryos. m, LYVE1+ dermal lymphatic vessels. White dashed lines, indicate the left and right lymphatic front. Scale bars, 500 μm. nq, Lymphatic vessel outgrowth (n), total lymphatic vessel length (normalized per total image area; o), number of lymphatic branch points (corrected for lymphatic length; p), and average lymphatic vessel width (q) (n = 6 embryos for ctrl; n = 6 embryos for eto; 2 litters). r, Quantitation of the number of filopodia per lymphatic sprout tip (n = 5 embryos for ctrl; n = 6 embryos for eto; 2 litters per group). s, The number of proliferating BrdU+ LECs in E15.5 ctrl and etomoxir (eto) embryos (n = 4 embryos for ctrl; n = 6 embryos for eto; 1 litter per group). t, Quantitation of the vascular area in primary head vein (left and right side) and the internal carotid artery (rostral extension of the dorsal aorta; left and right side) in E10.5 ctrl- and eto-treated embryos. ux, Quantitation of the lymphatic vessel outgrowth (u), total lymphatic vessel length (v), the number of lymphatic branch points (w), and average lymphatic vessel width (x) in the dorsal dermal lymphatic network in E15.5 embryos treated from E8.5–E14.5 with vehicle (ctrlE8.5–E14.5) or etomoxir (etoE8.5–E14.5) (n = 6 embryos for ctrlE8.5–E14.5; n = 5 embryos for etoE8.5–E14.5). Mean ± s.e.m. Statistical test, ANOVA and Bonferroni post-hoc test were used in multiple group comparisons. t-test was used for comparison of two groups. *P < 0.05; NS, not statistically significant.

Extended Data Figure 5 Confirmation of CPT1AKD using a second shRNA, additional effects of CPT1AKD, silencing another FAO-related gene in LECs, CPT1A overexpression does not induce the expression of LEC markers and CPT1AKD in VECs does not change the expression of blood endothelial cell markers.

Unless otherwise listed, CPT1AKD refers to knockdown 1 (KD1) of the 2 non-overlapping shRNAs tested. a, Representative immunoblot for CPT1A and β-ACTIN in LECs after silencing of CPT1A with two non-overlapping shRNAs (abbreviated as KD1 and KD2, respectively). b, Relative FAO flux in LECs at baseline and upon CPT1AKD versus VECs (n = 3). c, Relative [3H]thymidine incorporation in DNA in CPT1AKD LECs (using KD2) versus control LECs (n = 3). The same control samples were used from Fig. 3a. d, Total sprout length in CPT1AKD LEC spheroids (using KD2) versus control LEC spheroids (n = 3). The same control samples were used as Fig. 3c. e, Quantitation of the number of sprouts per spheroid in control LECs and upon CPT1AKD (using KD1 or KD2) (n = 3). f, Measurement of lactate dehydrogenase (LDH) in the cell culture media as a measure of cell death in control and CPT1AKD LECs (using KD1 or KD2) (n = 3). g, Relative FAO flux in pLECs at baseline and upon CPT1AKD versus VECs (n = 3). hj, Fold change in ITGA9 (h), CCL21 (i) and PROX1 (j) mRNA expression in pLECs at baseline and upon CPT1AKD versus VECs (n = 3). k, Fold change in VEGFR3 mRNA expression in pLECs at baseline and after treatment with etomoxir versus VECs (n = 3). l, Fold change in VLCAD mRNA expression in pLECs at baseline and upon VLCADKD (n = 3). m, Relative FAO flux in LECs at baseline and upon VLCADKD (n = 3). n, Fold change in VEGFR3 mRNA expression in pLECs at baseline and upon VLCADKD (n = 3). o, Relative FAO flux in VECs overexpressing CPT1A (CPT1AOE) versus control-transduced VECs (ctrl) (n = 3). pr, Fold change in CPT1A (p), PROX1 (q) and VEGFR3 (r) mRNA expression in CPT1AOE VECs versus ctrl VECs (n = 3). sv, Fold change in CPT1A (s), KDR (t), EFNB2 (u) and EPHB4 (v) mRNA expression in CPT1AKD VECs versus ctrl VECs (n = 3). Mean ± s.e.m. Statistical test: ANOVA and Bonferroni post-hoc test were used in multiple group comparisons. t-test was used for comparison of two groups. *P < 0.05; NS, not statistically significant.

Extended Data Figure 6 Metabolic effects of CPT1AKD in LECs.

ae, In vitro experiments with pLECs at baseline (ctrl) and with silencing of CPT1A (CPT1AKD) versus VECs. a, Energy charge (n = 5). b, Percentage of oxidized glutathione (GSSG) per total glutathione (GSSG+GSH) (n = 3). c, Relative NADP+/NADPH ratio, expressed relative to ctrl VECs (n = 3). d, CM-H2DCFDA (intracellular ROS) signal (n = 3). e, H2O2 levels, as detected with Amplex Red reagent (n = 3). fk, In vitro experiments with CPT1AKD pLECs versus ctrl pLECs. f, Percentage total contribution of [U-13C]palmitate carbons to citrate (Cit), α-ketoglutarate (αKG), fumarate (Fum) and malate (Mal) (n = 3). g, Cellular pool size of citrate), α-ketoglutarate, fumarate and malate, measured as arbitrary units (AU) of the metabolite of interest normalized to the protein content (n = 3). h, Relative incorporation of [U-14C]-palmitate carbons into DNA (n = 3). i, Percentage total contribution of [U-13C]palmitate carbons to aspartate (Asp), glutamate (Glu), UTP and CTP (n = 3). j, Cellular pool size of aspartate, glutamate, UTP, CTP, ATP and GTP, measured as arbitrary units (AU) of the metabolite of interest normalized to the protein content (n = 3). k, Total sprout length from spheroid sprouting assay in pLECs upon CPT1AKD, without (ctrl) and with deoxyribonucleotide (dNTP) supplementation versus pLECs without CPT1AKD (ctrl). Note, the first two bars are the same data as Fig. 6e, performed in the same independent experiment (n = 3). Mean ± s.e.m. of n independent experiments. Statistical test: ANOVA and Bonferroni post-hoc test were used in multiple group comparisons. t-test was used for comparison of two groups. *P < 0.05; NS, not statistically significant.

Extended Data Figure 7 Additional data regarding PROX1-mediated CPT1A upregulation, the effect of targeting p300 on VEC-to-LEC differentiation in vitro and targeting SLC25A1 or ACLY phenocopies CPT1AKD in inhibiting VEC-to-LEC differentiation in vitro.

a, Fold change in CPT1A mRNA expression in VECs overexpressing native PROX1 (PROX1) or a DNA-binding PROX1 mutant (PROX1mut) versus control VECs (n = 3). b, PROX1 enrichment (using anti-PROX1 antibody) around the human CPT1A gene (green), as compared to input chromatin (black), as determined by ChIP-seq analysis of VECs transduced with a lentiviral vector expressing PROX1. Reads are binned at 100 bp intervals, and expressed per million reads (y axis). Displayed are averages of 2 independent ChIP-seq experiments. Top line, position of genes on chromosome 11; grey blocks denote CPT1A exons; black blocks denote exons of adjacent genes. Bottom line: chromosomal location (genome build hg19). The red arrowheads (A, B) above the chromosome location refers to statistically significant regions of PROX1 enrichment (q < 0.1; q value is the FDR corrected P value), corresponding to the regions identified in Fig. 3g. c, Quantitation of PCR of regions ‘A’ and ‘B’ from g for the CPT1A gene in PROX1 ChIP samples from VECs (empty–FLAG in Fig. 3g) and pLECs (PROX1–Flag in Fig. 3g), expressed as percentage of input (n = 3). df, Fold change in VEGFR3 (d), ITGA9 (e) and CCL21 (f) mRNA expression in pLECs treated with vehicle (ctrl) or p300 inhibitor SGC-CBP30 (SGC) versus VECs (n = 3). g, Fold change in EP300 mRNA expression (gene encoding p300 protein) in pLECs with p300 silencing (p300KD) versus pLECs with control silencing (ctrl) (n = 3). hj, Fold change in VEGFR3 (h), ITGA9 (i) and CCL21 (j) mRNA expression in pLECs without (ctrl) and with p300 silencing versus VECs (n = 3). Note, as the level of p300 silencing was only approximately 50%, the degree of VEGFR3, ITGA9 and CCL21 reduction compared to control pLECs was less pronounced than upon treatment with the p300 inhibitor SGC-CBP30 (df). k, l, Fold change in ITGA9 (k) and CCL21 (l) mRNA expression in pLECs treated with vehicle (ctrl) or the SLC25A1 inhibitor 1,2,3-benzenetricarboxylate (BTC) versus VECs (n = 3). m, n, Fold change in ITGA9 (m) and CCL21 (n) mRNA expression in pLECs without (ctrl) and with ATP-citrate lyase (ACLY) silencing (ACLYKD; lowering ACLY mRNA levels by ± 75%) versus VECs (n = 3). oq, Fold change in ACLY (o), CPT1A (p) and VEGFR3 (q) mRNA expression in pLECs at baseline (ctrl) and upon silencing of ACLY (ACLYKD) (n = 5). r, Relative FAO flux in pLECs at baseline (ctrl) and upon silencing of ACLY (ACLYKD) (n = 4). s, Intracellular pools of malonyl-CoA determined by LC-MS in pLECs at baseline (ctrl) and upon silencing of ACLY (ACLYKD), expressed as relative malonyl-CoA/acetyl-CoA/CoA ratios (n = 3). Mean ± s.e.m. of n independent experiments. Statistical test: ANOVA and Bonferroni post-hoc test were used in multiple group comparisons. t-test was used for comparison of two groups. Statistical analysis for ChIP-seq described in Methods. *P < 0.05; NS, not statistically significant.

Extended Data Figure 8 Dynamic regulation of H3K9ac at LEC genes, and PROX1 does not interact with the KDR gene or affect H3K9 acetylation at the KDR gene.

a, Proportion of LEC versus non-LEC genes (refers to all annotated genes not included in LEC ontology) with enhanced H3K9ac (red portion) in pLECs versus VECs. The green portion denotes genes without significantly enhanced H3K9ac levels. b, Proportion of Prox1-target genes versus unrelated genes with enhanced H3K9ac (red portion) in pLECs versus VECs. The green portion denotes genes without significantly enhanced H3K9ac levels. cf, H3K9ac enrichment around the PROX1 (c) and SOX18 (e) genes in VECs (blue) and pLECs without (green) and with CPT1AKD (dark blue). Top line above each snapshot: scheme of the gene; grey blocks denote the exons of the gene of interest; black blocks denote exons of adjacent genes. Red arrowheads indicate sites of PROX1 enrichment in the PROX1 (c) or SOX18 (e) genes. Bottom line under each snapshot: chromosomal location of the gene (human genome build GRCh37/hg19). Grey boxes at the chromosomal locations denote regions with H3K9ac peak areas as determined using MACS. When the top half of the box is green, it denotes regions of significantly enhanced H3K9ac in pLECs versus VECs. When the bottom half of the box is red, it denotes regions of significantly reduced H3K9ac in pLECs with CPT1AKD versus pLECs. q value is the FDR corrected P value. Quantitative PCR verification of H3K9ac ChIP-seq results for the sequences of PROX1 (d) and SOX18 (f) (asterisks in c, e, respectively) on H3K9ac-ChIP samples from pLECs without (ctrl) and with CPT1A silencing (CPT1AKD) versus VECs, expressed as relative to input (n = 3). g, p300 enrichment around the VEGFR3 gene in VECs (blue) and pLECs (green). Top line: scheme of the gene; grey blocks denote the exons of the gene of interest; black blocks denote exons of adjacent genes. Red arrowheads indicate PROX1-binding sites. Bottom line, chromosomal location of the gene (human genome build GRCh37/hg19); grey boxes at the chromosomal locations, regions with H3K9ac peak areas. Asterisk denotes region confirmed by qPCR in h. h, Quantitative PCR verification of p300 ChIP-seq results for the sequence of VEGFR3 on p300 ChIP samples from pLECs versus VECs as a percentage of input, expressed as relative to VEC. i, PROX1–Flag enrichment around the human KDR (VEGFR2) gene (PROX1–Flag, green) versus empty-Flag control (empty–Flag, blue) in VECs transduced with a lentiviral vector expressing Flag-tagged PROX1 (PROX1–Flag) or empty-Flag, and PROX1 enrichment (using anti-PROX1 antibody ChIP) in pLECs (PROX1, green) and VECs (GFP, blue), as compared to the input chromatin (purple) and determined by ChIP-seq analysis. Top line, scheme of the KDR gene; grey blocks denote KDR exons and black blocks denote exons of adjacent genes. Bottom line: chromosomal location (human genome build GRCh37/hg19) (q < 0.1; q value is the FDR corrected P value). j, H3K9ac enrichment around the human KDR gene in VECs (blue) and pLECs (green). Top line, scheme of the KDR gene; grey blocks denote KDR exons and black blocks denote exons of adjacent genes. Bottom line: chromosomal location (human genome build GRCh37/hg19). q value is the FDR corrected P value. Mean ± s.e.m. of n independent experiments. Statistical test, Chi-square test was used to compare gene ontology subsets for presence of increased H3K9ac. ANOVA and Bonferroni post-hoc test were used in multiple group comparisons. Statistical analysis for ChIP-seq is described in Methods. *P < 0.05.

Extended Data Figure 9 PROX1 overexpression increases p300 binding and H3K9 acetylation at the CPT1A gene.

a, p300 enrichment around the human CPT1A gene in VECs (blue) and pLECs (green), as determined by ChIP-seq analysis of VECs transduced with a lentiviral vector expressing a control vector (GFP) or PROX1, respectively. Reads are binned as 100 bp intervals, and expressed per million reads (y axis). Displayed are averages of 2 independent ChIP-seq experiments. Top line: scheme of the CPT1A gene; grey blocks denote CPT1A exons and black blocks denote exons of adjacent genes. Bottom line: chromosomal location (human genome build GRCh37/hg19). The red arrowheads (A, B) above the chromosome location refer to statistically significant regions of PROX1 enrichment from Fig. 3g. b, c, p300 ChIP and qPCR for the PROX1-binding site at CPT1A ‘A’ (b) or CPT1A ‘B’ (c) in pLECs without (ctrl) and with CPT1A silencing (CPT1AKD) versus VECs, values expressed as a percentage of input and relative to VECs (n = 3). d, H3K9ac enrichment around the human CPT1A gene in VECs (blue) and pLECs (green), as determined by ChIP-seq analysis of VECs transduced with a lentiviral vector expressing a control vector (GFP) or PROX1, respectively. Top line, scheme of the CPT1A gene; grey blocks denote CPT1A exons and black blocks denote exons of adjacent genes. Bottom line: chromosomal location (genome build hg19). The red arrowheads (A, B) above the chromosome location refer to statistically significant regions of PROX1 enrichment from Fig. 3g. q value is the FDR corrected P value. e, Quantitative PCR of the PROX1-binding regions ‘A’ and ‘B’ of the CPT1A gene (illustrated in d) on H3K9ac-ChIP samples from pLECs versus VECs, calculated as relative to the input and expressed relative to the values obtained for VECs (n = 3). f, g, Quantitation of the densitometry from PROX1-ChIP and PCR for the PROX1-binding site at CPT1A ‘A’ (f) or CPT1A ‘B’ (g) (identified in Fig. 3g) in pLECs without (ctrl) and with CPT1A silencing (CPT1AKD) versus VECs, expressed as relative to input (n = 3). h, Quantitation of the densitometry from PROX1-ChIP and PCR for the PROX1-binding site at VEGFR3 in pLECs without (ctrl) and with CPT1A silencing (CPT1AKD) versus VECs, expressed as relative to input (n = 3). i, p300 ChIP and qPCR for the PROX1-binding site at VEGFR3 in pLECs without (ctrl) and with CPT1A silencing (CPT1AKD) versus VECs, values expressed as a percentage of input and relative to VECs (n = 3). Mean ± s.e.m. of n independent experiments. Statistical test: ANOVA and Bonferroni post-hoc test were used in multiple group comparisons. t-test was used for comparison of two groups. *P < 0.05; NS, not statistically significant.

Extended Data Figure 10 Pharmacological inhibition of FAO results in defects in angiogenesis.

a, b, Quantitation of lymphatic vessel end-point density (number of vessel extremities per corneal area; a) and lymphatic vessel length density (cumulative length of the lymphatic vessel network; b), both parameters per total corneal area, in wild-type mice subjected to corneal cauterization-induced injury treated daily with vehicle (ctrl) or etomoxir in Fig. 5e–k (n = 5 animals per group). c, Representative images of CD31+ blood vessel outgrowth 9 days after corneal cauterization-induced injury in adult wild-type mice treated daily with vehicle (ctrl) or etomoxir. Scale bars, 250 μm. dg, Quantitation of percentage blood vessel outgrowth area relative to total area (d), branch point density (e), end-point density (f) and length density (g) in wild-type mice subjected to corneal cauterization-induced injury treated daily with vehicle (ctrl) or etomoxir from the corneal cauterization experiments in Fig. 5e–k (n = 5 animals per group). h, Measurement of plasma acetate concentrations in mice treated with PBS (ctrl) or 400 μl of 0.5 M acetate (ac) for 9 days (n = 6 mice per group). i, j, Quantitation of lymphatic vessel end-point density (number of vessel extremities per corneal area; i) and lymphatic vessel length density (cumulative length of the lymphatic vessel network) (j), both parameters per total corneal area, in wild-type mice subjected to corneal cauterization-induced injury treated daily with vehicle (ctrl), acetate only (ac), etomoxir only (eto) or etomoxir plus acetate (eto+ac) from the corneal cauterization experiments in Fig. 6f–j (n = 10 animals per group). k, Traditional and adapted model of the mechanism of PROX1-induced lymphatic formation. Traditional model: PROX1 binds to the promoter region of lymphangiogenic genes, including VEGFR3, enhancing their transcription. Adapted model: by binding to the CPT1A promoter, PROX1 induces CPT1A expression, enhancing the production of FAO-derived acetyl-CoA. In a dual manner, FAO (red) thereby regulates: (1) LEC proliferation through providing acetyl-CoA, which helps to sustain the Krebs cycle and deoxyribonucleotide (dNTP) synthesis in conjunction with an anaplerotic substrate during lymphangiogenesis (purple); and (2) lymphangiogenic gene expression through supplying acetyl-CoA (AcCoA) used for histone acetylation (Ac) at lymphangiogenic genes, thereby modulating LEC differentiation, proliferation and migration (green). PROX1 promotes histone acetylation preferentially at lymphatic genes by enriching p300 at PROX1-binding sites, and by supplying more acetyl-CoA (through FAO induction). Thus, in the adapted model, PROX1 still functions as a transcription factor (as proposed in the traditional model), but its role is extended by using FAO metabolism to enhance its own transcriptional activity. Mean ± s.e.m.; (n = 5 animals per group from 1 experiment in a, b, dg), (n = 6 animals per group from 1 experiment in h), (n = 6 animals per group from 2 experiments in i, j). Statistical test: ANOVA and Bonferroni post-hoc test were used in multiple group comparisons. t-test was used for comparison of two groups. *P < 0.05.

Supplementary information

Supplementary Figure

Uncropped scans with size marker indication for Fig. 1d, Fig. 3e, Fig. 3i, Fig. 4d, Fig. 6b, Fig. 6c and Extended Data Fig. 5a. (PDF 128 kb)

Supplementary Information

This file contains Supplementary Discussion 1-4 and an additional reference. (PDF 572 kb)

Supplementary Table 1

This file shows peaks from Prox1-FLAG Chip-seq and mapping of Prox1, H3K9ac and P300 Chip-seq data. (XLSX 377 kb)

Supplementary Table 2

This fie shows peaks from H3K9ac Chip-seq in pLECs versus VECs and pLECs upon CPT1aKD versus pLECs at baseline. (XLSX 6518 kb)

PowerPoint slides

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Wong, B., Wang, X., Zecchin, A. et al. The role of fatty acid β-oxidation in lymphangiogenesis. Nature 542, 49–54 (2017). https://doi.org/10.1038/nature21028

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1038/nature21028

This article is cited by

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Search

Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing