Murine endothelial serine palmitoyltransferase 1 (SPTLC1) is required for vascular development and systemic sphingolipid homeostasis

  1. Andrew Kuo
  2. Antonio Checa
  3. Colin Niaudet
  4. Bongnam Jung
  5. Zhongjie Fu
  6. Craig E Wheelock
  7. Sasha A Singh
  8. Masanori Aikawa
  9. Lois E Smith
  10. Richard L Proia
  11. Timothy Hla

  • Curated by eLife

    eLife logo

    Evaluation Summary:

    This study reveals the importance of sphingolipids in endothelial cell biology. The authors have examined the role of the Sptlc1 gene in retinal injury as well as in the production of sphingolipid metabolites. These studies provide key insight into how endothelial cell production of sphingolipids alters vascular repair and systemic metabolism.

    (This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. Reviewer #1 agreed to share their name with the authors.)

Reviewed by

Read the full article

Listed in

Log in to save this article

Abstract

Serine palmitoyl transferase (SPT), the rate-limiting enzyme in the de novo synthesis of sphingolipids (SL), is needed for embryonic development, physiological homeostasis, and response to stress. The functions of de novo SL synthesis in vascular endothelial cells (EC), which line the entire circulatory system, are not well understood. Here, we show that the de novo SL synthesis in EC not only regulates vascular development but also maintains circulatory and peripheral organ SL levels. Mice with an endothelial-specific gene knockout of SPTLC1 ( Sptlc1 ECKO), an essential subunit of the SPT complex, exhibited reduced EC proliferation and tip/stalk cell differentiation, resulting in delayed retinal vascular development. In addition, Sptlc1 ECKO mice had reduced retinal neovascularization in the oxygen-induced retinopathy model. Mechanistic studies suggest that EC SL produced from the de novo pathway are needed for lipid raft formation and efficient VEGF signaling. Post-natal deletion of the EC Sptlc1 also showed rapid reduction of several SL metabolites in plasma, red blood cells, and peripheral organs (lung and liver) but not in the retina, part of the central nervous system (CNS). In the liver, EC de novo SL synthesis was important for acetaminophen-induced rapid ceramide elevation and hepatotoxicity. These results suggest that EC-derived SL metabolites are in constant flux between the vasculature, circulatory elements, and parenchymal cells of non-CNS organs. Taken together, our data point to the central role of the endothelial SL biosynthesis in maintaining vascular development, neovascular proliferation, non-CNS tissue metabolic homeostasis, and hepatocyte response to stress.

Article activity feed

  1. Version published to 10.7554/elife.78861 on eLife
  2. eLife

    Author Response

    Reviewer #1 (Public Review):

    “The synthesis and metabolism of sphingolipid (SL) are involved in wide range of biological processes. In the present study, the authors investigate the role of SPTLC1, one of the essential subunits of serine palmitoyl transferase complex, in both physiological and pathophysiological angiogenesis, via using inducible endothelial-specific SPTLC1 knockout mice. They found SPTLC1 deficiency in ECs inhibited retinal angiogenesis along with reducing several SL metabolites in plasma, red blood cells, and peripheral organs. In addition, the authors found SPTLC1 EC-KO mice are resistant to APAP-induced liver injury. Overall, the in vivo findings in the present study are of potential interest and the authors have given clear evidence that endothelial SPTLC1 is critical to retinal angiogenesis. However, the underlying mechanisms are completely lacking in the present study. Most of the evidence provided is circumstantial, associative, and indirect.”

    We appreciate the positive comments of the reviewer. We have addressed the reviewer’s concern regarding underlying mechanisms as detailed below.

    “To be specific,

    1. The authors found endothelial SPTLC1 is important to both angiogenesis and the plasma lipid profile. However, the authors did not present the data to demonstrate the relationship between them. The in vivo findings about the phenotype and the plasma lipid profile might be true and unrelated. It would be important to know whether supplementing the reduced lipid induced by SPTLC1 KO could rescue the angiogenesis related phenotype in mice, or, whether the alternative way to inhibit the SL synthesis could mimic the phenotype of KO mice.”

    In the manuscript, we discussed the possibility whether S1P is involved, since it is one of the most down-regulated SL in the plasma and a major regulator of angiogenesis. We think it is unlikely that reduced plasma S1P is responsible for the phenotype. First, the retinal angiogenesis defect in Sptlc1 ECKO mice is the opposite of S1pr1 ECKO as we have published previously (PMID: 22975328, PMID: 32059774). Moreover, deletion of sphingosine kinase, the enzyme produces S1P, in the endothelium does not influence retinal angiogenesis at P6 (Figure 3 Supplement 2 A and B). Loss of S1P chaperone ApoM- i.e., Apom KO, which exhibits 50% reduction of plasma S1P, does not show change in retinal vascular development (Figure 3 Supplement 2 C and D). Taken together, our results strongly suggest that reduction in plasma S1P is not the cause of vascular defect in Sptlc1 ECKO retinas.

    Based on our results in the manuscript, loss of SPT enzyme activity in endothelial cells reduced SL species in the endothelial cells and the plasma. Our in vitro and VEGF intraocular injection experiments (new data) suggests that the angiogenic defects seen in Sptlc1 ECKO mice is due to cell intrinsic defects in VEGF signaling and not due to changes in plasma SL levels. We have edited the discussion section to address this issue.

    “2. A major issue is that the present study did not reveal is a real downstream target. It is possible that VEGF signaling might be impaired by SPTLC1 knockout as discussed by the authors. However, the authors did not demonstrate this point with data. Including both in vivo and in vitro data to evaluate the effects of SPTLC1 deficiency on VEGF signaling might further strengthen the hypothesis. Besides, with in vitro experiments, the authors might further find the critical metabolite(s) involved in VEGF signaling and angiogenesis.”

    As discussed above, we agree with the review’s critique and have addressed this essential point with new experiments (both in vitro and in vivo) in Figure 5. Our new data shows that SPT pathway supplies the glycosphingolipid GM1, which is needed for efficient VEGF-induced ERK phosphorylation and tip cell formation.

    Reviewer #2 (Public Review):

    “Andrew Kuo et al. investigated the role of endothelial de novo sphingolipids (SL) synthesis using endothelial cell specific SPTLC1 knockout (ECKO) mice. They showed that these mice exhibited low concentration of various SL species in not only ECs but also RBC, circulation, and other non-EC tissues. They also showed that ECKO mice exhibited impaired angiogenesis in normal and oxygen-induced retinopathy models, consistent with the decrease of endothelial proliferation and tip cell formation. They finally revealed that these mice were resistant to acetaminophen-induced acute liver injury in early phase. The experiments were well-designed, and the results were clear and convincing. The authors concluded that endothelial cells were the major source of SL in circulation and various organs (liver and lung) other than retina (and probably brain). The weakness of the current version of the manuscript is that the authors did not elucidate the mechanisms underlying the observed phenomena.

    1. The authors showed impaired angiogenesis in ECKO mice using neonatal retina model. Based on the fact that this phenotype was similar to that in endothelial VEGFR2 deficient mice, they suggested that VEGF responsiveness is altered in ECKO mice. Although this hypothesis is plausible, the authors would need to prove it by evaluating VEGFR signaling (VEGFR phosphorylation, Akt activation etc.) in ECKO mice.”

    We thank the reviewer for positive comments. As for the weakness identified, we have addressed this point by conducting new in vitro and in vivo experiments (detailed above). The new Figure 5 addresses this issue directly.

    “2) The acetaminophen-induced liver injury was reduced in ECKO mice in early phase. However, it is still unclear whether SL production itself affects liver injury. The authors discussed the possibility that gene deficiency increases unconsumed serine resulting in GSH increase, but it is essentially independent to SL. If possible, it would be good if the authors could investigate the effect of SL administration on the liver injury progression.”

    We appreciate the reviewer’s concern about liver injury model in the Sptlc1 ECKO mice. Our data suggests that SL species supplied from EC impacts hepatocyte response to stress. Since the acetaminophen induced liver injury is highly dependent on reactive oxygen species, our finding that increased glutathione levels in the Sptlc1 ECKO mice may be involved in the phenotype. However, we are simply considering them as biochemical markers of liver injury. This has been addressed in the discussion.

    “3) This paper showed the impaired cell proliferation in Sptlc1 KO EC mice, and discussed it. Authors described that this phenotype was similar to that of Nos3 KO mice, but its inconsistency with Sptlc2 ECKO adult mice was only justified by a word "isoform-selective function". Authors could quantify eNOS expressions in Sptlc1 KO mice, compared results and then discuss this matter. “

    In figure 1C, we used eNOS as an EC marker to show purity during our EC isolation process. In fact, we did not observe change of eNOS expression in Sptlc1 ECKO. We also did not detect elevated phospho-eNOS in Sptl1c ECKO in contrast to Sptlc2 ECKO adult mice (Figure1 supplement 4). Additionally, our work in the retina was performed in postnatal-genedeletion pups from P6-P17 which is different from the published Sptlc2 ECKO study. The differences in gene deletion strategy (early postnatal vs. adult) could result in differences in eNOS expression . We have added discussion about this issue.

  3. eLife

    Evaluation Summary:

    This study reveals the importance of sphingolipids in endothelial cell biology. The authors have examined the role of the Sptlc1 gene in retinal injury as well as in the production of sphingolipid metabolites. These studies provide key insight into how endothelial cell production of sphingolipids alters vascular repair and systemic metabolism.

    (This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. Reviewer #1 agreed to share their name with the authors.)

  4. eLife

    Reviewer #1 (Public Review):

    The synthesis and metabolism of sphingolipid (SL) are involved in wide range of biological processes. In the present study, the authors investigate the role of SPTLC1, one of the essential subunits of serine palmitoyl transferase complex, in both physiological and pathophysiological angiogenesis, via using inducible endothelial-specific SPTLC1 knockout mice. They found SPTLC1 deficiency in ECs inhibited retinal angiogenesis along with reducing several SL metabolites in plasma, red blood cells, and peripheral organs. In addition, the authors found SPTLC1 EC-KO mice are resistant to APAP-induced liver injury. Overall, the in vivo findings in the present study are of potential interest and the authors have given clear evidence that endothelial SPTLC1 is critical to retinal angiogenesis. However, the underlying mechanisms are completely lacking in the present study. Most of the evidence provided is circumstantial, associative, and indirect. To be specific,

    1. The authors found endothelial SPTLC1 is important to both angiogenesis and the plasma lipid profile. However, the authors did not present the data to demonstrate the relationship between them. The in vivo findings about the phenotype and the plasma lipid profile might be true and unrelated. It would be important to know whether supplementing the reduced lipid induced by SPTLC1 KO could rescue the angiogenesis related phenotype in mice, or, whether the alternative way to inhibit the SL synthesis could mimic the phenotype of KO mice.

    2. A major issue is that the present study did not reveal is a real downstream target. It is possible that VEGF signaling might be impaired by SPTLC1 knockout as discussed by the authors. However, the authors did not demonstrate this point with data. Including both in vivo and in vitro data to evaluate the effects of SPTLC1 deficiency on VEGF signaling might further strengthen the hypothesis. Besides, with in vitro experiments, the authors might further find the critical metabolite(s) involved in VEGF signaling and angiogenesis.

  5. eLife

    Reviewer #2 (Public Review):

    Andrew Kuo et al. investigated the role of endothelial de novo sphingolipids (SL) synthesis using endothelial cell specific SPTLC1 knockout (ECKO) mice. They showed that these mice exhibited low concentration of various SL species in not only ECs but also RBC, circulation, and other non-EC tissues. They also showed that ECKO mice exhibited impaired angiogenesis in normal and oxygen-induced retinopathy models, consistent with the decrease of endothelial proliferation and tip cell formation. They finally revealed that these mice were resistant to acetaminophen-induced acute liver injury in early phase. The experiments were well-designed, and the results were clear and convincing. The authors concluded that endothelial cells were the major source of SL in circulation and various organs (liver and lung) other than retina (and probably brain). The weakness of the current version of the manuscript is that the authors did not elucidate the mechanisms underlying the observed phenomena.

    1. The authors showed impaired angiogenesis in ECKO mice using neonatal retina model. Based on the fact that this phenotype was similar to that in endothelial VEGFR2 deficient mice, they suggested that VEGF responsiveness is altered in ECKO mice. Although this hypothesis is plausible, the authors would need to prove it by evaluating VEGFR signaling (VEGFR phosphorylation, Akt activation etc.) in ECKO mice.
    2. The acetaminophen-induced liver injury was reduced in ECKO mice in early phase. However, it is still unclear whether SL production itself affects liver injury. The authors discussed the possibility that gene deficiency increases unconsumed serine resulting in GSH increase, but it is essentially independent to SL. If possible, it would be good if the authors could investigate the effect of SL administration on the liver injury progression.
    3. This paper showed the impaired cell proliferation in Sptlc1 KO EC mice, and discussed it. Authors described that this phenotype was similar to that of Nos3 KO mice, but its inconsistency with Sptlc2 ECKO adult mice was only justified by a word "isoform-selective function". Authors could quantify eNOS expressions in Sptlc1 KO mice, compared results and then discuss this matter.
  6. Version published to 10.1101/2022.03.22.485296v1 on bioRxiv