Jumonji domain-containing protein 6 (Jmjd6) is required for angiogenic sprouting and regulates splicing of VEGF-receptor 1

Abstract

JmjC domain-containing proteins play a crucial role in the control of gene expression by acting as protein hydroxylases or demethylases, thereby controlling histone methylation or splicing. Here, we demonstrate that silencing of Jumonji domain-containing protein 6 (Jmjd6) impairs angiogenic functions of endothelial cells by changing the gene expression and modulating the splicing of the VEGF-receptor 1 (Flt1). Reduction of Jmjd6 expression altered splicing of Flt1 and increased the levels of the soluble form of Flt1, which binds to VEGF and placental growth factor (PlGF) and thereby inhibits angiogenesis. Saturating VEGF or PlGF or neutralizing antibodies directed against soluble Flt1 rescued the angiogenic defects induced by Jmjd6 silencing. Jmjd6 interacts with the splicing factors U2AF65 that binds to Flt1 mRNA. In conclusion, Jmjd6 regulates the splicing of Flt1, thereby controlling angiogenic sprouting.

Fig. 1.

Regulation of angiogenic sprouting by Jmjd6. (A–D) Jmjd6 was silenced in human umbilical vein endothelial cells (HUVECs) by two different siRNAs (Jmjd6-1 and -2) and the functional activity of the cells was compared to two control siRNAs (Scr1 and Scr2). (A and B) Angiogenic sprouting was compared in the spheroid model. Data are mean ± SEM for n = 3–4; ^#P < 0.05 vs. Scr1 (Mann-Whitney U test). (C and D) In addition, the Matrigel network assay was determined in vitro. Data are mean ± SEM for n = 4; *P < 0.05 vs. Scr1 (Student's t test). (Inset in A) Knock-down of Jmjd6 protein in siRNA treated cells by Western blot 48 h after transfection. (E) Invasion of HUVEC was determined in modified Boyden chambers. Data are mean ± SEM for n = 3; *P < 0.05 vs. Scr (Student's t test). (F) Cell viability was measured by MTT cell viability assay. Data are mean± SEM for n = 3. (G) Assessment of spheroid sprouting in HUVEC overexpressing Jmjd6 plasmid. Data are mean ± SEM for n = 3 (Student's t test) *P < 0.05.

Fig. 2.

Angiogenesis in Jmjd6^+/− mice. (A–C) Matrigel plugs were subcutaneously implanted in Jmjd6^+/− mice or wild-type littermates. After 7 d, FITC-lectin was injected and plaque angiogenesis was measured. (A) Quantitative analysis of lectin-perfused capillaries. Data are mean ± SEM for n = 4 animals/group. ^#P < 0.05 vs. wild-type (Mann-Whitney U test). (B and C) Representative histological images. (B) Higher magnification images in bFGF-supplemented plugs. (Scale bar, 20 μm.) (C) Lower magnification images in H&E staining (Top), FITC-lectin positive vessel invading in Matrigel plugs with (Bottom) or without (Middle) bFGF supplementation. [Scale bars, (Top) 100 μm, (Middle and Bottom) 20 μm.] (D) Jmjd6 expression in heart tissue of wild-type or Jmjd6^+/− mice. Data are mean ± SEM for n = 6–7 animals (Student's t test). (E) Lung endothelial cells (ECs) were isolated of wild-type and Jmjd6^+/− mice and were cultured in Matrigel. Quantification of the network formation is shown. (Inset) A representative result. Data are mean± SEM for n = 6–7 animals. *P < 0.05 compared with wild-type (Student's t test).

Fig. 3.

Modulation of gene expression and splicing by Jmjd6. (A and B) Jmjd6 was silenced by siRNA in HUVEC and gene expression and splicing was determined by GeneChip Human Exon 1.0 ST array 48 h after siRNA transfection. Data are mean ± SEM for n = 3 independent experiments. Selection of significantly up-regulated (A) or down-regulated genes (B). Selected genes were confirmed by qPCR, as shown in the right panel of the figures. Data are mean ± SEM for n ≥ 3. *P < 0.05 vs. Scr. (Student's t test). (C) Soluble Flt1 expression in Jmjd6 siRNA-treated HUVECs. Data are mean ± SEM for n = 3 independent experiments. *P < 0.05 vs. control siRNA (Student's t test). (D) Jmjd6 expression was silenced by siRNA in ECs. Forty-eight hours after transfection, a Northern blot was used to detect the alternative splicing forms of the Flt1 mRNA; mFlt1 and sFlt1 are indicated by arrows. A reprobe with an actin probe served as a loading control. (E) Ratio of sFlt1/mFlt1 expression in Jmjd6 siRNA-treated HUVECs. Data are mean ± SEM for n = 3 independent experiments. *P < 0.05 vs. normoxia (Student's t test). (F) Ratio of sFlt1/mFlt1 expression after incubation of HUVECs with 50 μM DFO for 24 h was detected by qPCR. Data are mean ± SEM for n = 4; *P < 0.05 vs. control. (Student's t test). (G) Ratio of sFlt1/mFlt1 expression after incubation of HUVECs under hypoxia (0.1% O₂) for 24 h was detected by qPCR. Data are mean ± SEM for n = 3. *P < 0.05 vs. normoxia (Student's t test).

Fig. 4.

Rescue of Jmjd6 silencing mediated impaired sprouting angiogenesis. (A and B) Jmjd6 was silenced by siRNA and angiogenic sprouting was determined using the spheroid assay in the presence or absence of recombinant VEGF, PlGF, or VEGF/PlGF heterodimer (A) or neutralizing anti-sFlt1 antibody (B). Data are mean ± SEM for n = 3–4. Five or more spheroids were counted per experiment. *P < 0.05 (ANOVA).

Fig. 5.

Jmjd6 mediates splicing of Flt1 by interacting with U2AF65. (A) U2AF65 was immunoprecipitated and bound RNA was quantified by qPCR for sFlt1 or the control RNA U1. Data are mean ± SEM for n = 3 independent experiments. (B) Jmjd6-Flag was immunoprecipitated and bound U2AF65 was detected by Western blot. Three percent of the Lysates were loaded as input control.