Endothelial Krüppel-like factor 4 modulates pulmonary arterial hypertension

Abstract

Krüppel-like factor 4 (KLF4) is a transcription factor expressed in the vascular endothelium, where it promotes anti-inflammatory and anticoagulant states, and increases endothelial nitric oxide synthase expression. We examined the role of endothelial KLF4 in pulmonary arterial (PA) hypertension (PAH). Mice with endothelial KLF4 knockdown were exposed to hypoxia for 3 weeks, followed by measurement of right ventricular and PA pressures, pulmonary vascular muscularization, and right ventricular hypertrophy. The effect of KLF4 on target gene expression was assessed in lungs from these mice, verified in vitro by small interfering RNA (siRNA) knockdown of KLF4, and further studied at the promoter level with cotransfection experiments. KLF4 expression was measured in lung tissue from patients with PAH and normal control subjects. We found that, after hypoxia, right ventricular and PA pressures were significantly higher in KLF4 knockdown animals than controls. Knockdown animals also had more severe pulmonary vascular muscularization and right ventricular hypertrophy. KLF4 knockdown resulted in increased pulmonary expression of endothelin-1 and decreased expression of endothelial nitric oxide synthase, endothelin receptor subtype B, and prostacyclin synthase. Concordant findings were observed in vitro, both with siRNA knockdown of KLF4 and promoter activity assays. Finally, KLF4 expression was reduced in lungs from patients with PAH. In conclusion, endothelial KLF4 regulates the transcription of genes involved in key pathways implicated in PAH, and its loss exacerbates pulmonary hypertension in response to chronic hypoxia in mice. These results introduce a novel transcriptional modulator of PAH, with the potential of becoming a new therapeutic target.

Figure 1.

Pressure measurements and pulmonary vascular muscularization after 3 weeks of hypoxia. (A) Right ventricular (RV) systolic pressure (RVSP) in VE-cadherin Cre mice with either Krüppel-like factor (KLF) 4 wild-type (Cre/WT) or KLF4 floxed (Cre/Flox) with and without hypoxia (five animals per group). (B) Mean pulmonary arterial (PA) pressure (mPAP) in the same mice. (C) Lung immunofluorescent staining for von Willebrand factor (vWF; green) and α-smooth muscle actin (α-SMA; red), showing examples of vessels that are fully muscularized (yellow arrow), nonmuscularized (blue arrow), or partially muscularized (white arrow). (D) Percentages of vessels with different degrees of muscularization in Cre/WT and Cre/Flox animals after hypoxia exposure. Vessels were counted from five animals per group, with at least three sections per animal. Error bars represent SEM. *P ≤ 0.05, **P ≤ 0.01.

Figure 2.

RV hypertrophy after hypoxia exposure. (A) Heart cross-sections from Cre/WT and Cre/Flox mice after hypoxia, showing greater thickening of the RV wall in Cre/Flox mice. (B) Weight ratio of the RV to left ventricle plus septum (RV/[LV + S]) in the two mouse groups with and without hypoxia. (C) Fluorescent staining with rhodamine-labeled wheat germ agglutinin for RV myocytes in sections of the RV from a Cre/WT mouse and a Cre/Flox mouse after hypoxia, showing larger cross-sectional areas of RV myocytes from the Cre/Flox mouse. (D) Average cross-sectional area of RV myocytes from Cre/WT and Cre/Flox animals with and without hypoxia (three to four animals per group, with at least three sections per animal). Error bars represent SEM. **P ≤ 0.01.

Figure 3.

Expression of genes altered with KLF4 knockdown. (A) Relative mRNA expression in lung tissue homogenates from Cre/WT and Cre/Flox animals after hypoxia for the genes of KLF4, endothelial nitric oxide synthase (eNOS), endothelin (EDN) 1, EDN receptor subtype B (EDNRB), and prostacyclin synthase (PTGIS) (n = 15 animals per group). (B) Western blot analysis for protein levels of eNOS and EDN1 (ET-1) in lung homogenates. (C) small interfering RNA (siRNA) knockdown was performed in cultured human umbilical vein endothelial cells (HUVECs) and the same targets as in (A) were assessed. mRNA levels are expressed for cells treated with KLF4-specific siRNA (siKLF4) relative to those treated with nonspecific siRNA (siNS; n = 10 experiments). (D) Representative Western blot analysis for selected gene targets in HUVECs treated similar to those in (C) (n = 3 experiments). Error bars represent SEM. *P ≤ 0.05, **P ≤ 0.01.

Figure 4.

Promoter activity assays. Transient transfection was performed on bovine PA endothelial cells (BPAECs) with reporter plasmids of respective targets together with expression plasmids for either full-length KLF4 (KLF4), KLF4 with deleted zinc finger DNA-binding domain (KLF4∆ZnF), or empty plasmid vector (Vector). (A) Full-length KLF4, but not KLF4∆ZnF, induces eNOS promoter activity. (B) Full-length KLF4 represses EDN1 promoter activity, whereas KLF4∆ZnF induces it. (C) Full-length KLF4 induces EDNRB promoter activity, but KLF4∆ZnF causes less induction. (D) Full-length KLF4, but not KLF4∆ZnF, induces PTGIS promoter activity (n = 3–5 experiments for each promoter, done in triplicates). Error bars represent SEM. **P ≤ 0.01 in comparison to empty vector cotransfection.

Figure 5.

(A) KLF4 expression is decreased at both the mRNA and protein levels in lungs from patients with idiopathic PA hypertension (IPAH) compared with control subjects (n = 4 per group). Error bars represent SEM. *P ≤ 0.05. (B) A proposed mechanism by which KLF4 in the endothelial cell (EC) modulates PAH by regulating the expression of the genes for eNOS, EDN1, EDNRB, and PTGIS.