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
. 2023 Jan 11;15(1):e16373.
doi: 10.15252/emmm.202216373. Epub 2022 Dec 13.

PGF2α facilitates pathological retinal angiogenesis by modulating endothelial FOS-driven ELR+ CXC chemokine expression

Yan Zhao  1   2 , Yi Lei #  3 , Huying Ning #  1 , Yaqiang Zhang  4 , Guilin Chen  1 , Chenchen Wang  1   2 , Qiangyou Wan  2 , Shumin Guo  1 , Qian Liu  1 , Ruotian Xie  1 , Yujuan Zhuo  1 , Shuai Yan  2 , Jing Zhao  5 , Fengjiang Wei  5 , Lu Wang  1 , Xiaohong Wang  1 , Weidong Li  5 , Hua Yan  3 , Ying Yu  1
Affiliations

PGF2α facilitates pathological retinal angiogenesis by modulating endothelial FOS-driven ELR+ CXC chemokine expression

Yan Zhao et al. EMBO Mol Med. .

Abstract

The pathological retinal angiogenesis often causes blindness. Current anti-angiogenic therapy for proliferative retinopathy targets the vascular endothelial growth factor (VEGF), but many patients do not radically benefit from this therapy. Herein, we report that circulating prostaglandin (PG) F2α metabolites were increased in type 2 diabetic patients with proliferative retinopathy, and the PGF2α receptor (Ptgfr) was upregulated in retinal endothelial cells (ECs) from a mouse model of oxygen-induced retinopathy (OIR). Further, disruption of the PTGFR receptor in ECs attenuated OIR in mice. PGF2α promoted the proliferation and tube formation of human retinal microvascular endothelial cells (HRMECs) via the release of ELR+ CXC chemokines, such as CXCL8 and CXCL2. Mechanistically, the PGF2α /PTGFR axis potentiated ELR+ CXC chemokine expression in HRMECs through the Gq /CAMK2G/p38/ELK-1/FOS pathway. Upregulated FOS-mediated ELR+ CXC chemokine expression was observed in retinal ECs from PDR patients. Moreover, treatment with PTGFR inhibitor lessened the development of OIR in mice in a CXCR2-dependent manner. Therefore, inhibition of PTGFR may represent a new avenue for the treatment of retinal neovascularization, particularly in PDR.

Keywords: PGF2α; anti-angiogenic therapy; pathological retinal angiogenesis.

PubMed Disclaimer

Figures

Figure 1
Figure 1. Activation of the PGF2α‐PTGFR axis in pathological retinal angiogenesis

  1. Serum levels of 15‐keto‐dihydro‐PGF2α in type 2 diabetic patients without diabetic retinopathy (no DR), with non‐proliferative diabetic retinopathy (NPDR), and with proliferative diabetic retinopathy (PDR; n = 20–24).

  2. Schematic diagram of the OIR mouse model. Mice on postnatal day 7 were placed in a 75% oxygen container for 5 days and were then transferred to a normal environment on postnatal day 12 for another 5 days to induce retinal vessel proliferation.

  3. Retinal PGF2α production in OIR mice during postnatal days 12–17 (n = 4).

  4. mRNA expression of PGF2α synthases in the retinas of OIR mice during postnatal days 12–17 (n = 4).

  5. mRNA expression of Ptgfr in the retinas of OIR mice during postnatal days 12–17 (n = 6–8).

  6. mRNA expression of Ptgfr in each retinal layer from normoxic and OIR mice on postnatal day 16 (n = 6). The left‐hand image displays representative cross‐sections from OIR and normoxic retinas, with isolectin B4‐stained vessels in green, DAPI‐stained nuclei in blue, and LCM isolation locations circled by white dashed circles. INL, inner nuclear layer; ONL, outer nuclear layer; RGC, retinal ganglion cells.

  7. Ptgfr mRNA expression in retinal endothelial cells (CD31+CD45−) sorted by MACS from normoxic and OIR mice on postnatal day 16 (n = 5).

Data information: n.s. stands for “not significant.” Data were analyzed by Kruskal–Wallis test with Dunn's multiple comparisons test (A), two‐way ANOVA with Tukey's multiple comparisons test (C, D, E, F), or Mann–Whitney test (G). Scale bar: 20 μm. Data are represented as mean ± SEM. Source data are available online for this figure.
Figure EV1
Figure EV1. PGF2α and PGE2 generation in retinas from OIR mice
Retinas were pooled from six mice per group and subjected to PG analysis through LC/MS. Data were analyzed by the unpaired Student's t‐test (n = 3). n.s. stands for “not significant.” Data are represented as mean ± SEM.
Figure 2
Figure 2. Ptgfr knockout in vascular ECs attenuates angiogenesis in OIR mice

  1. Generation of endothelial cell (EC)‐specific Ptgfr knockout mice. Ptgfr‐floxed mice were crossed with Tie2‐Cre mice to create EC‐specific Ptgfr knockout mice (CKO‐T).

  2. Ptgfr mRNA levels in retinal microvessels of CKO‐T and control mice (n = 6). INL, inner nuclear layer; ONL, outer nuclear layer; RGC, retinal ganglion cells.

  3. Representative images of OIR retinas from CKO‐T and control mice on postnatal day 17. The green color shows the isolectin B4‐stained vessels. The second row panels are the enlarged images of white boxes from the first row panels. The third row images show neovascular tufts (NV, white) and the fourth row images show the vaso‐obliteration (VO, white) area.

  4. Quantitation of oxygen‐induced retinal neovascularization in CKO‐T and control mice (n = 12).

  5. Quantitation of retinal vaso‐obliteration in CKO‐T and control mice (n = 12).

Data information: n.s. stands for “not significant.” Data were analyzed by unpaired Student's t‐test (B), Mann–Whitney test (D, E). Scale bars: 500 μm (unmagnified image), 150 μm (magnified image). Data are represented as mean ± SEM. Source data are available online for this figure.
Figure EV2
Figure EV2. Ptgfr knockout in vascular ECs attenuates angiogenesis in OIR mice

  1. Generation of endothelial cell (EC)‐specific Ptgfr knockout mice. Ptgfr‐floxed mice were crossed with VE‐cadherin‐Cre mice to create EC‐specific Ptgfr knockout mice (CKO‐V).

  2. Ptgfr mRNA levels in retinal microvessels of CKO‐V and control mice (n = 4).

  3. Representative images of OIR retinas from CKO‐V and control mice on postnatal day 17. The green color shows the isolectin B4‐stained vessels. The second row panels are the enlarged images of white boxes from the first row panels. The third row images show neovascular tufts (NV, white) and the fourth row images show the vaso‐obliteration (VO, white) area.

  4. Quantitation of oxygen‐induced retinal neovascularization in CKO‐V and control mice (n = 12).

  5. Quantitation of retinal vaso‐obliteration in CKO‐V and control mice (n = 12).

Data information: n.s. stands for “not significant.” Data were analyzed by unpaired Student's t‐test (B), Mann–Whitney test (D, E). Scale bars: 500 μm (unmagnified image), 150 μm (magnified image). Data are represented as mean ± SEM. Source data are available online for this figure.
Figure 3
Figure 3. PGF2α promotes the migration, proliferation, and tube formation of human retinal microvascular endothelial cells (HRMECs) as well as aortic ring sprouting in culture
  1. A

    Representative images of scratch‐induced migration of PGF2α‐treated HRMECs. Yellow solid lines represent scratch locations, dashed lines represent post‐migration locations, and arrows indicate migration direction.

  2. B

    Quantitation of the cell migration distance in A (n = 4).

  3. C

    Effect of PGF2α treatment on HRMEC proliferation in a dose‐dependent manner (n = 6).

  4. D

    Representative images of staining for BrdU incorporation in PGF2α‐treated HRMECs.

  5. E

    Quantification of BrdU+ cells in D (n = 4).

  6. F

    Representative tube formation images of PGF2α‐or vehicle (DMSO)‐treated HRMECs.

  7. G–J

    Quantitation of total tubule length, number of junctions, number of meshes, and percentage of mesh area in F (n = 4).

  8. K

    Representative images of endothelial cell sprouting of PGF2α‐treated mouse aortic rings. Yellow lines indicate the EC sprouting area.

  9. L

    Quantification of the EC sprouting area in K (n = 6).

Data information: One data point represented the mean value of three technical replicates from one independent biological replicate. Data were analyzed by Mann–Whitney test (B, E, L), unpaired Student's t‐test (G, H, I, J) or Kruskal–Wallis test with Dunn's multiple comparisons test (C). Scale bars: 100 μm (A) and 200 μm (F, K). Data are represented as mean ± SEM. Source data are available online for this figure.
Figure 4
Figure 4. PGF2α promotes proliferation and tube formation of HRMECs through the upregulation of ELR+ CXC chemokines
  1. A

    Heat map of upregulated genes in PGF2α‐treated HRMECs (n = 3 samples).

  2. B

    GO enrichment of the top five biological processes involving the upregulated genes in PGF2α‐treated HRMECs.

  3. C

    Effect of PGF2α treatment on ELR+ CXC chemokine expression in HRMECs (n = 3).

  4. D–G

    Relative expression levels of enriched ELR+ CXC chemokines in retinal microvascular ECs from patients with PDR (n = 4) compared with those from normal subjects (n = 7; Gene Expression Omnibus Database dataset GSE94019). TPM stands for “Transcripts per kilobase million.”

  5. H

    Effect of CXCR2 inhibitor SB265610 (1 μM) on PGF2α‐stimulated HRMEC migration. The yellow solid line indicates the original position of the scratch, the dashed line indicates the position after cell migration, and the arrows indicate the migratory direction.

  6. I

    Quantitation of scratch‐induced HRMEC migration in H (n = 4).

  7. J

    Effect of SB265610 (1 μM) on PGF2α‐stimulated HRMEC proliferation (n = 4).

  8. K

    Effect of CXCR2 inhibitor SB265610 (1 μM) on tube formation in PGF2α‐treated HRMECs.

  9. L

    Statistical graph of tube formation experiments in (K), comparing the total tubule length, number of junctions, number of meshes, and percentage of mesh area (n = 4).

Data information: n.s. stands for “not significant.” Data were analyzed by unpaired Student's t‐test (C), Mann–Whitney test (D, E, F, G), or two‐way ANOVA with Tukey's multiple comparisons test (I, J, L). Data in (C, I, J, L) are represented as mean ± SEM. Data in (D–G) are presented as the mean + SD. Scale bar: 100 μm (H) and 200 μm (K). Source data are available online for this figure.
Figure 5
Figure 5. Retinal Cxcl1 overexpression abolishes decreased neovascularization in OIR retinas from CKO mice

  1. Effect of CXCR2 inhibitor SB265610 (1 μM) on PGF2α‐promoted sprouting of aortic rings. Yellow lines show the sprouting area.

  2. Quantitation of the sprouting areas in A (n = 6).

  3. mRNA expression of ELR+ CXC chemokines in retinas from OIR mice during postnatal days 12–17 (n = 4–6).

  4. Effect of PTGFR deficiency in ECs on ELR+ CXC chemokine expression in retinas from OIR mice on postnatal day 16 (n = 6).

  5. Schematic diagram of Cxcl1‐expressing lentivirus transfection in OIR mice. Neonatal mice were injected with lentivirus on postnatal day 5, followed by oxygen treatment on day 7.

  6. Retinal CXCL1 protein‐expression levels in Cxcl1‐expressing lentivirus‐injected CKO mice (n = 3).

  7. Quantitation of retinal CXCL1 protein in F (n = 3).

  8. Representative images of oxygen‐induced retinal angiogenesis in Cxcl1‐expressing lentivirus‐injected CKO mice on postnatal day 17. Green represents the isolectin B4‐stained vessels, the second row panels display the enlarged images of white boxes in the first row panels, the third row images show neovascular tufts (NV, white), and the fourth row images show the vaso‐obliteration (VO, white) area.

  9. Quantitation of oxygen‐induced retinal neovascularization in H (n = 8).

  10. Quantitation of retinal vaso‐obliteration in H (n = 8).

Data information: n.s. stands for “not significant.” Data were analyzed by two‐way ANOVA with Tukey's multiple comparisons test (B, C, D, G, I, J). Scale bar: 200 μm (A), 500 μm (H, unmagnified image), 150 μm (H, magnified image). Data are represented as mean ± SEM. Source data are available online for this figure.
Figure 6
Figure 6. PGF2α induces expression of ELR+ CXC chemokines in HRMECs through transcription factor FOS
  1. A

    Heat map of differentially expressed transcription factors in HRMECs in response to 500 nM PGF2α (n = 3).

  2. B

    Relative expression levels of FOS in retinal microvascular ECs from patients with PDR (n = 4) compared with the corresponding expression levels from normal subjects (n = 7; Gene Expression Omnibus Database dataset GSE94019). TPM stands for “Transcripts per kilobase million.”

  3. C

    Effect of FOS knockdown on 500 nM PGF2α‐induced CXCL8 expression in HRMECs (n = 4).

  4. D

    Effect of FOS knockdown on 500 nM PGF2α‐induced HRMEC CXCL8 secretion in the culture medium (n = 4).

  5. E, F

    Effect of RA (1 μM, AP‐1 dimer inhibitor), SN‐50 (100 μg/ml, NFκB inhibitor), and Inca‐6 (2.5 μM, NFAT inhibitor) on 500 nM PGF2α‐induced CXCL8 expression in HRMECs and HRMEC CXCL8 secretion in the culture medium (n = 4).

  6. G

    Representative images of scratch‐induced migration of 500 nM PGF2α‐treated HRMECs with or without FOS knockdown. The yellow solid line indicates the original position of the scratch, the dashed line indicates the position of the cells after migration, and the arrows represent the migratory direction.

  7. H

    Quantitation of the cell‐migration distance in H (n = 4).

  8. I

    Effect of FOS knockdown on 500 nM PGF2α‐induced HRMEC proliferation (n = 4).

  9. J

    Representative images of tube formation by 500 nM PGF2α‐treated HRMECs with or without FOS knockdown.

  10. K

    Quantitation of total tube length, number of junctions, number of meshes, and percentage of mesh area in J (n = 4).

  11. L

    Fos mRNA levels in retinas from OIR mice during postnatal days 12–17 (n = 3–4).

  12. M

    Effect of PTGFR deficiency in ECs on Fos expression in the retina from OIR mice on postnatal day 16 (n = 6).

  13. N

    Effect of FOS inhibitor SR11302 (1 μM) on 500 nM PGF2α‐promoted sprouting of mouse aortic rings.

  14. O

    Quantitation of sprouting areas in N (n = 6).

  15. P

    Effect of FOS inhibitor SR11302 (1 μM) on 500 nM PGF2α‐promoted CXCL1 secretion in mouse aortic ring culture medium (n = 6).

Data information: n.s. stands for “not significant.” Data were analyzed by Mann–Whitney test (B, C, D, E, F, M, P) or two‐way ANOVA with Tukey's multiple comparisons test (H, I, K, L, O). Scale bar: 100 μm (G) and 200 μm (J, N). Data in (B) are expressed as the mean + SD. Data in C, D, E, F, H, I, K, L, M, O, P are represented as mean ± SEM. Source data are available online for this figure.
Figure 7
Figure 7. PGF2α upregulates FOS expression in HRMECs through the Gq/CAMK2G/p38/ELK‐1 signaling pathway

  1. Schematic diagram of the potential PTGFR‐mediated signaling pathways. All inhibitors of targeted molecules are shown in red.

  2. Effect of U73122 (10 μM), BAPTA (50 μM), Y27632 (5 μM), and Ly294002 (15 μM) treatment on 500 nM PGF2α‐induced FOS gene expression in HRMECs (n = 4).

  3. Effect of KN93 (10 μM) and RO‐318220 (250 nM) treatment on 500 nM PGF2α‐induced FOS expression in HRMECs (n = 4).

  4. Effect of SB203580 (2 μM), PD98059 (20 μM), and SP600125 (60 nM) treatment on 500 nM PGF2α‐induced FOS gene expression in HRMECs (n = 4).

  5. Effect of U73122 (10 μM), BAPTA (50 μM), and KN93 (10 μM) treatment on 500 nM PGF2α‐induced p38 phosphorylation in HRMECs.

  6. Quantitation of the ratio of p38 phosphorylation to total p38 protein in E (n = 3).

  7. Effect of U73122 (10 μM), BAPTA (50 μM), and SB203580 (2 μM) on the phosphorylation of CAMK2 isoforms in 500 nM PGF2α‐treated HRMECs.

  8. Quantitation of the ratio of CAMK2G phosphorylation to total CAMK2G in G (n = 3).

  9. Quantitation of the ratio of phosphorylated CAMK2D to total CAMK2D in G (n = 3).

  10. Effect of CAMK2G or CAMK2D knockdown on 500 nM PGF2α‐triggered p38 phosphorylation in HRMECs.

  11. Quantitation of the ratio of p38 phosphorylation to total p38 protein in J (n = 3).

  12. Effect of CAMK2G or CAMK2D knockdown on 500 nM PGF2α‐induced FOS expression in HRMECs (n = 4).

  13. Effect of U73122 (10 μM), BAPTA (50 μM), KN93 (10 μM), and SB203580 (2 μM) on ELK‐1 phosphorylation in 500 nM PGF2α‐treated HRMECs.

  14. Quantitation of the ratio of phosphorylated ELK‐1 to total ELK‐1 in M (n = 3).

  15. Schematic diagram of the PGF2α/PTGFR axis‐mediated angiogenic signaling in retinal ECs.

Data information: n.s. stands for “not significant.” Data were analyzed by Mann–Whitney test (B, C, D, and L) or unpaired Student's t‐test (F, H, I, K, and N). Data are represented as mean ± SEM. Source data are available online for this figure.
Figure EV3
Figure EV3. Effect of different chemical inhibitors on PGF2α‐induced CXCL8 expression in HRMECs
  1. A, B

    Effect of U73122, BAPTA, Y27632, and Ly294002 treatment on PGF2α‐induced CXCL8 gene expression in HRMECs and secretion in culture medium (n = 4).

  2. C, D

    Effect of KN93 and RO‐318220 treatment on PGF2α‐induced CXCL8 gene expression in HRMECs and secretion in culture medium (n = 4).

  3. E, F

    Effect of SB203580, PD98059, and SP600125 treatment on PGF2α‐induced CXCL8 gene expression in HRMECs and secretion in culture medium (n = 4).

Data information: n.s. stands for “not significant.” Data were analyzed by the Mann–Whitney test (A, B, C, D, E, F). Data are represented as mean ± SEM.
Figure 8
Figure 8. Administration of PTGFR inhibitor AL8810 attenuates retinal angiogenesis in OIR mice

  1. Schematic diagram of AL8810 administration in OIR mice. i.p. stands for “intraperitoneally injected”.

  2. Representative images of OIR retinas in WT and Cxcr2 −/− mice with or without AL8810 treatment. The green color shows the isolectin B4‐stained vessels, the second row panels display the enlarged images of white boxes in the first row panels, the third row images show neovascular tufts (NV, white), and the fourth row images show the vaso‐obliteration (VO, white) area.

  3. Quantitation of oxygen‐induced retinal neovascularization in WT and Cxcr2 −/− mice with or without AL8810 treatment (n = 12).

  4. Quantitation of retinal vaso‐obliteration in WT and Cxcr2 −/− mice with or without AL8810 treatment (n = 12).

  5. Effect of AL8810 on retinal Fos expression in WT and Cxcr2 −/− OIR mice on postnatal day 16 (n = 6).

  6. Effect of AL8810 on retinal Cxcl1 expression in WT and Cxcr2 −/− OIR mice on postnatal day 16 (n = 6).

Data information: n.s. stands for “not significant.” Data were analyzed by two‐way ANOVA with Tukey's multiple comparisons test (C, D) or Mann–Whitney test (E, F). Scale bar: 500 μm (unmagnified image), 150 μm (magnified image). Data are represented as mean ± SEM. Source data are available online for this figure.
Figure EV4
Figure EV4. PGTFR inhibitor OBE022 attenuates retinal angiogenesis in OIR mice

  1. Schematic diagram of OBE022 administration in OIR mice. i.p. stands for intraperitoneally injected.

  2. Representative images of OIR retinas in mice with or without OBE022 treatment. The green color shows the isolectin B4‐stained vessels, the second row panels display the enlarged images of white boxes in the first row panels, the third row images show neovascular tufts (NV, white), and the fourth row images show the vaso‐obliteration (VO, white) area.

  3. Quantitation of oxygen‐induced retinal neovascularization with or without OBE022 treatment (n = 10).

  4. Quantitation of retinal vaso‐obliteration with or without OBE022 treatment (n = 10).

  5. Effect of OBE022 on retinal Fos expression in OIR mice on postnatal day 16 (n = 6).

  6. Effect of OBE022 on retinal Cxcl1 expression in OIR mice on postnatal day 16 (n = 6).

Data information: Data were analyzed by Mann–Whitney test (C, D, E, F). Scale bar: 500 μm (unmagnified image), 150 μm (magnified image). Data are represented as mean ± SEM. Source data are available online for this figure.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

References

    1. Acker G, Zollfrank J, Jelgersma C, Nieminen‐Kelhä M, Kremenetskaia I, Mueller S, Ghori A, Vajkoczy P, Brandenburg S (2020) The CXCR2/CXCL2 signalling pathway – an alternative therapeutic approach in high‐grade glioma. Eur J Cancer 126: 106–115 - PubMed
    1. Ahmad SF, Akoum A, Horne AW (2015) Selective modulation of the prostaglandin F2α pathway markedly impacts on endometriosis progression in a xenograft mouse model. Mol Hum Reprod 21: 905–916 - PubMed
    1. Alva JA, Zovein AC, Monvoisin A, Murphy T, Salazar A, Harvey NL, Carmeliet P, Iruela‐Arispe ML (2006) VE‐cadherin‐Cre‐recombinase transgenic mouse: a tool for lineage analysis and gene deletion in endothelial cells. Dev Dyn 235: 759–767 - PubMed
    1. Ashraf S, Bell S, O'Leary C, Canning P, Micu I, Fernandez JA, O'Hare M, Barabas P, McCauley H, Brazil DP et al (2019) CAMKII as a therapeutic target for growth factor‐induced retinal and choroidal neovascularization. JCI Insight 4: e122442 - PMC - PubMed
    1. Banumathi E, O'Connor A, Gurunathan S, Simpson DA, McGeown JG, Curtis TM (2011) VEGF‐induced retinal angiogenic signaling is critically dependent on Ca2+ signaling by Ca2+/calmodulin‐dependent protein kinase II. Invest Ophthalmol Vis Sci 52: 3103–3111 - PubMed

Publication types

Associated data

-