Rapid, reversible modulation of blood-brain barrier P-glycoprotein transport activity by vascular endothelial growth factor

Abstract

Increased brain expression of vascular endothelial growth factor (VEGF) is associated with neurological disease, brain injury, and blood-brain barrier (BBB) dysfunction. However, the specific effect of VEGF on the efflux transporter P-glycoprotein, a critical component of the BBB, is not known. Using isolated rat brain capillaries and in situ rat brain perfusion, we determined the effect of VEGF exposure on P-glycoprotein activity in vitro and in vivo. In isolated capillaries, VEGF acutely and reversibly decreased P-glycoprotein transport activity without decreasing transporter protein expression or opening tight junctions. This effect was blocked by inhibitors of the VEGF receptor flk-1 and Src kinase, but not by inhibitors of phosphatidylinositol-3-kinase or protein kinase C. VEGF also increased Tyr-14 phosphorylation of caveolin-1, and this was blocked by the Src inhibitor PP2. Pharmacological activation of Src kinase activity mimicked the effects of VEGF on P-glycoprotein activity and Tyr-14 phosphorylation of caveolin-1. In vivo, intracerebroventricular injection of VEGF increased brain distribution of P-glycoprotein substrates morphine and verapamil, but not the tight junction marker, sucrose; this effect was blocked by PP2. These findings indicate that VEGF decreases P-glycoprotein activity via activation of flk-1 and Src, and suggest Src-mediated phosphorylation of caveolin-1 may play a role in downregulation of P-glycoprotein activity. These findings also imply that P-glycoprotein activity is acutely diminished in pathological conditions associated with increased brain VEGF expression and that BBB VEGF/Src signaling could be targeted to acutely modulate P-glycoprotein activity and thus improve brain drug delivery.

Figure 1.

VEGF reduces P-glycoprotein-mediated transport in isolated brain capillaries. A, Representative confocal micrographs showing steady-state (60 min) luminal fluorescence of the P-glycoprotein substrate NBD-CSA and the MRP-2 substrate Texas Red (TR). CON, Control; MAN, incubated with 100 mm mannitol to induce osmotic disruption of the tight junctions; PSC, incubated with P-glycoprotein inhibitor PSC833 (5 μm); LTC4, incubated with MRP2 inhibitor leukotriene C4 (0.3 μm); VEGF, incubated with 200 ng/ml VEGF. Scale bar, 5 μm. B, Dose–response curve for VEGF effect on P-glycoprotein-mediated transport. Data are mean luminal fluorescence of NBD-CSA ± SEM (PSC-insensitive fluorescence subtracted). VEGF significantly reduced luminal accumulation of NBD-CSA. Inset, Dose–response curve for VEGF effect on MRP2-mediated transport. Data are mean luminal fluorescence of TR ± SEM with LTC4-insensitive fluorescence subtracted. VEGF did not affect luminal accumulation of TR. Statistical comparison, **p < 0.01, ***p < 0.001 (vs NBD-CSA without VEGF), significance determined by one-way ANOVA with a Newman–Keuls multiple-comparison test. Representative experiments shown; data were collected from at least 10 capillaries per data point.

Figure 2.

VEGF effects on P-glycoprotein transport activity and protein expression. A, Time course of VEGF action. Capillaries were incubated to steady state (1 h) with NBD-CSA, and P-glycoprotein activity was measured over multiple time points following addition of VEGF (200 ng/ml). After measurements at 60 min, the media were removed and replaced with fresh media, with one of the VEGF-treated groups getting VEGF-free media. Shown are pooled data from four experiments, each involving measurement from at least 10 capillaries per treatment. Statistical comparison, ***p < 0.001 versus control at the corresponding time point, significance determined by two-way ANOVA with a Bonferroni post hoc test. B, Western blot analysis of P-glycoprotein expression. Incubation of brain capillaries with VEGF (200 ng/ml, 1 h) does not change protein expression of P-glycoprotein in membrane fractions. C, The effect of VEGF on P-glycoprotein activity is not blocked by the proteasome inhibitor lactacystin (LAC, 10 μm). D, The effect of VEGF is attenuated by the microtubule polymerization inhibitor nocodazole (NOC, 20 μm). C, D, Representative experiments shown, measurements made in at least 10 capillaries per treatment. Statistical comparison, ***p < 0.001 versus control, significance determined by one-way ANOVA with a Newman–Keuls multiple-comparison test.

Figure 3.

VEGF signaling to P-glycoprotein. Shown are representative experiments; each experiment was repeated 2–4 times. Data are mean P-glycoprotein (Pgp) activity ± SEM, collected from at least 10 capillaries per condition. In all experiments shown, capillaries were preincubated with inhibitors 30 min before addition of NBD-CSA, then incubated to steady state (1 h) before addition of VEGF, and measurements were made 1 h later. Time-matched control groups from the same preparation were exposed to PBS during inhibitor and/or VEGF treatments. A, B, The effect of VEGF on P-glycoprotein activity is blocked by flk-1 inhibitors SU5416 (500 nm) and TMPP (50 nm). C, D, The effect of VEGF on P-glycoprotein activity is not blocked by the phosphatidylinositol-3-kinase inhibitor wortmannin (WORT, 100 nm), nor by the broad-spectrum PKC inhibitor staurosporine (STR, 100 nm). E, F, The Src family kinase inhibitor PP2 (1 μm) also blocks the effect of VEGF (E), but its nonfunctional analog PP3 (1 μm) does not (F). Statistical comparisons, ***p < 0.001 versus control, significance determined by one-way ANOVA with a Newman–Keuls multiple-comparison test.

Figure 4.

Src-mediated modulation of P-glycoprotein transport activity. P-glycoprotein transport activity is decreased by the Src kinase activating peptide, YEEIP (100 μm, 1 h), an effect blocked by PP2 but not PP3. Statistical comparisons, ***p < 0.001 versus control, significance determined by one-way ANOVA with a Newman–Keuls multiple-comparison test.

Figure 5.

VEGF increases Src-mediated Tyr-14 phosphorylation of caveolin-1. A, YEEIP (100 μm, 1 h) increases Tyr-14 phosphorylation of caveolin-1. B, Incubation of brain capillaries with VEGF (200 ng/ml, 1 h) and/or the Src kinase inhibitor PP2 (1 μm) is not associated with any change in protein expression of P-glycoprotein or total caveolin-1 (cav-1). Immunoreactivity for Tyr-14 phosphorylated caveolin-1 (p14Y-cav-1) is increased by VEGF; this effect is blocked by preincubation with PP2. Representative blots from a single experiment are shown; 10 μg of protein from membrane fractions were loaded per lane. All samples shown were from a single preparation of capillaries derived from 10 rat brains.

Figure 6.

Coimmunoprecipitation of caveolin-1 and P-glycoprotein. Caveolin-1 was immunoprecipitated from 100 μg of brain capillary protein. Representative experiment shown. CON, VEGF, and VEGF + PP2 represent samples from a single capillary preparation treated as indicated. Ab, Caveolin-1 antibody carried through the IP procedure without capillary protein. +, Five micrograms of brain capillary protein (positive control). −, One hundred micrograms of brain capillary protein carried through the IP procedure without antibody (negative control). IgG heavy chain (hc) and light chains (lc) were well resolved from detected bands for P-glycoprotein and caveolin-1. No consistent changes were observed in P-glycoprotein precipitated with caveolin-1 in any condition.

Figure 7.

Effect of VEGF on in vivo brain distribution of [¹⁴C]-sucrose and [³H]-morphine. A, Sucrose and morphine distribution in control and CSA-treated rats (8 μm in perfusate). Rats were perfused with [¹⁴C]-sucrose and [³H]-morphine simultaneously, n = 4–11 per time point per condition. Sucrose distribution versus perfusion time was best fit to a unidirectional uptake model (Eq. 2). Best fit lines were not significantly different between groups (F_(2,53) = 0.2923, p = 0.7477). Morphine distribution versus perfusion time was best fit to a model with an efflux component (Eq. 3). Comparison of curve fits showed that the curves were not equivalent (F_(2,53) = 3.976, p = 0.0246), reflecting increased brain distribution of morphine in the CSA group. Curve fits and comparisons were performed with GraphPad Prism version 4.02. B, VEGF (500 ng in 2 μl of aCSF) or 2 μl of aCSF alone was injected into the lateral ventricle 30 min before in situ brain perfusion. VEGF significantly increased brain distribution of morphine (F_(2,52) = 4.935, p = 0.0109) but not sucrose (F_(2,52) = 1.792, p = 0.1768) in the cerebral hemisphere ipsilateral to the injection.

Figure 8.

Effect of VEGF and Src inhibition on in vivo brain distribution of [³H]-verapamil. Intracerebroventricular injection of VEGF significantly increases brain distribution of [³H]-verapamil in the cerebral hemisphere ipsilateral to the injection site during a 20-min brain perfusion (*p < 0.05). Intraperitoneal injection of PP2 (1 mg/kg body weight, immediately before intracerebroventricular VEGF/aCSF injection) abolishes the effect of intracerebroventricular VEGF on brain verapamil distribution (n = 4–5 per group).