Dissecting the molecular control of endothelial NO synthase by caveolin-1 using cell-permeable peptides

Abstract

In endothelia, NO is synthesized by endothelial NO synthase (eNOS), which is negatively regulated by caveolin-1 (Cav-1), the primary coat protein of caveolae. We show that delivery of Cav-1 amino acids 82-101 (Cav) fused to an internalization sequence from Antennapedia (AP) blocks NO release in vitro and inflammation and tumor angiogenesis in vivo. To characterize the molecular mechanism by which the AP-Cav peptide and Cav-1 mediate eNOS inhibition, we subdivided the Cav portion of AP-Cav into three domains (Cav-A, -B, and -C), synthesized five overlapping peptides (AP-Cav-A, -AB, -B, -BC, and -C), and tested their effects on eNOS-dependent activities. Peptides containing the Cav-B domain (amino acids 89-95) induced time- and dose-dependent inhibition of eNOS-dependent NO release in cultured endothelial cells, NO-dependent inflammation in the ear, and hydraulic conductivity in isolated venules. Alanine scanning of AP-Cav-B revealed that Thr-90 and -91 (T90,91) and Phe-92 (F92) are crucial for AP-Cav-B- and AP-Cav-mediated inhibition of eNOS. Mutation of F92 to A92 in the Cav-1 cDNA caused the loss of eNOS inhibitory activity compared with wild-type Cav-1. These data highlight the importance of amino acids 89-95 and particularly F92 in mediating eNOS inhibition by AP-Cav and Cav-1.

Fig. 1.

Truncated AP-Cav peptides attenuate endothelial cell NO release and Evans blue extravasation. (A) The caveolin scaffolding domain of AP-Cav (WT) was subdivided into three subdomains (Cav-A, -B, and -C), and five peptides containing combinations of these subdomains were synthesized (AP-Cav-AB, -BC, -A, -B, and -C). (B) Effect of truncated AP-Cav peptides on VEGF-induced NO release in BAEC. BAEC were pretreated with peptides (10 μM) for 6 h, and VEGF (1 nM)-induced NO release was accumulated for 30 min. Basal nitrite release (133 ± 21 pmol per 10⁶ cells) was subtracted from all values. (C) Dose-dependent effect of AP-Cav-B on VEGF-induced NO release. BAEC were treated with AP-Cav-B (5-50 μM) as described above. Basal nitrite release (208 ± 22 pmol per 10⁶ cells) was subtracted from all values. (D) Effect of truncated AP-Cav peptides on mustard oil-induced inflammation. Male CD-1 mice were treated for 1 h with AP-Cav (2.5 mg/kg i.p.) or truncated peptides (equivalent molar dose) and then were anesthetized (ketamine/xylazine) for administration of Evans blue (via the left jugular vein; 30 mg/kg) 1 min before painting right and left ears with 5% phenyl isothiocyanate and mineral oil (control), respectively. After 30 min, the animals were killed, and the content of Evans blue was determined spectrophotometrically at 595 nm. Respective basal Evans blue extravasation (left ear values for each group) was subtracted from stimulated (right ear) values. All graphs represent means of at least three experiments in triplicate ± SEM. *, P < 0.05 compared with vehicle-treated control (-) group.

Fig. 2.

Truncated AP-Cav peptides attenuate PAF-induced hydraulic conductivity in isolated venules. (A) Paired measurements of L_p in response to 10 nM PAF with and without peptide pretreatment. A 2-h peptide pretreatment had no effect on basal L_p (Left). The addition of PAF to control vessels (AP, 10 μM; ○) caused an increase in L_p, which was almost completely attenuated by pretreating vessels with AP-Cav (⋄) or AP-Cav-AB (•), whereas pretreatment with AP-Cav-B (▴) was not as profound. (B) PAF-induced mean peak increases in L_p with peptide pretreatment as described above. Basal L_p (0.92 ± 0.15) was subtracted from all groups. Values are means of five different experiments ± SEM. *, P < 0.05 compared with AP-treated control group.

Fig. 3.

Truncated AP-Cav peptides block recombinant eNOS activity and eNOS binding to Cav-1. (A) Purified eNOS was preincubated with either l-nitroarginine methyl ester (1 mM) or the peptides (Cav, Cav-AB, or Cav-A; 1-25 μM) for 15 min, and NOS activity was assessed as described in Methods. Data represent duplicate determinations from a single experiment that was repeated twice. (B) Confluent BAEC were treated for 6 h with AP, AP-Cav, AP-Cav-AB, AP-Cav-A, or AP-Cav-B (10 μM). Cells were lysed, Cav-1 was immunoprecipitated, and eNOS coassociation was detected by Western blot analysis. Experiments were performed in duplicate, and typical data are shown. IP, immunoprecipitation; WB, Western blot.

Fig. 4.

F92 mutants fail to block eNOS activity in vitro and in vivo.(A) Seven mutant peptides were generated, and their inhibitory potential was tested on VEGF-induced NO release by BAEC. Basal nitrite release (199 ± 20 pmol per 10⁶ cells) was subtracted for all groups. (B) BAEC were treated as described for 6 h with AP-Cav, F92A-AP-Cav, or T90/91/F92A-AP-Cav (10 μM of each) and stimulated with VEGF (1 nM) for 30 min. Basal nitrite release (180 ± 8 pmol per 10⁶ cells) was subtracted for all groups. (C) Compared with AP-Cav, AP-T90/91/F92ACav does not block mustard oil-induced plasma protein leakage in vivo. Mice were pretreated as described for Fig. 1, and plasma leakage was stimulated with mustard oil. Basal Evans blue extravasation (left ear values for each group) was subtracted from the stimulated (right ear) values. All data are means of at least three experiments in triplicate ± SEM. *, P < 0.05 compared with vehicle-treated control group.

Fig. 5.

Mutant F92A Cav-1 does not block eNOS-mediated NO release despite colocalizing with eNOS. (A) HEK cells were transfected with plasmids encoding for a control protein (β-gal; lane 1), human eNOS (lane 2), human eNOS and WT Cav-1 (lane 3), and human eNOS and F92 Cav-1 (lane 4). Basal nitrite release (24-h accumulation) was assayed in the medium with a NO chemiluminescence analyzer. Cells then were lysed, and proteins were collected for Western blot analysis: eNOS (top blot) and Cav-1 (middle blot) and β-COP protein expression as a loading control (bottom blot). (B) Cav-1 F92A mutation does not affect its subcellular targeting and colocalization with eNOS. COS cells were transfected with plasmids coding for eNOS and hemagglutinin-tagged versions of WT (Upper) and F92A Cav-1 (Lower). Cells were fixed, incubated with antibodies against hemagglutinin (HA; red) and eNOS (green), and mounted on glass slides with medium containing nuclear dye DAPI (blue). (Right) Merged images with DAPI staining are shown.