The Mechanisms of Cerebral Vascular Dysfunction and Neuroinflammation by MMP-Mediated Degradation of VEGFR-2 in Alcohol Ingestion

Activation of MMPs Diminishes TJ Protein Expression in Brain Endothelium and Microvessels

Degradation of extracellular matrix and basement membrane proteins is the key function of MMPs. We have shown that VEGFR is involved in activation of MMPs and BBB damage in acute alcohol exposure.⁹ To identify the specific types of MMPs, we first examined the changes in the expression of 10 different types of MMPs by proteome array in primary hBECs culture after exposure to 50 mmol/L ethanol for 24 hours. Out of the 10 MMPs, ethanol exposure significantly elevated the levels of MMP-3 and MMP-9 compared with untreated hBECs in culture (Figure IA–IC in the online-only Data Supplement), and these observations were confirmed by immunofluorescence staining and Western blotting. Further, detection of MMPs activity by zymography validated the changes in expression and protein levels of MMPs in brain endothelial cells and rat brain tissues and microvessels. Thus, we found that alcohol increased the gelatinolytic (for MMP9) or stromelysin (for MMP3) activity with/without VEGF treatment in cell culture or in brain tissue and microvessels (Figure II in the online-only Data Supplement). Thus, in this study, MMP-3 and MMP-9 were targeted as the causative factors for disruption and leakiness of neurovascular function (the BBB). As expected, colocalization of MMPs and TJ proteins revealed that increase in the expression of MMP-3 and MMP-9 reciprocally diminished the expression of occludin and ZO-1 by alcohol treatment compared with untreated control cells (Figure 1A through 1C and Figure III in the online-only Data Supplement). In parallel, we also analyzed the colocalization of MMP-3 or MMP-9 and occludin in intact rat brain microvessels from control and chronic alcohol intake. In agreement with the findings in primary hBEC culture, we observed that increase in MMP-3 or MMP-9 expression correlated with a decrease in occludin expression in intact capillary vessels (Figure 1D) and in microvessels of brain tissue section (Figure IV in the online-only Data Supplement) from alcohol ingested animal compared with controls.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Activation of MMPs by ethanol degrades TJ proteins. A through D, Immunofluorescent staining of A, MMP-9 (red), occludin (green), and DAPI (blue); B, MMP-3 (red), ZO-1 (green), and DAPI (blue); C, MMP-9 (red), ZO-1 (green), and DAPI (blue) in hBECs; and D, MMP-3 (red), occludin (green), and DAPI (blue) in intact brain microvessels of rat. E and F, Western blot analyses of MMP-3, MMP-9, and actin (E); occludin, claudin-5, ZO-1 and actin (F) in hBEC lysate protein, whole brain tissue homogenates, and in protein extract from isolated microvessels. For all hBECs culture work, cells were treated with or without EtOH (50 mmol/L) in the presence or absence of MMP-3 or MMP-9 (100 ng/mL each) or TIMP1 (100 ng/mL) for 24 hours. Bar graphs show the results that are expressed as ratio of MMP-3/9, occludin, claudin-5, or ZO-1 to that of β-actin bands. Values are mean±SEM; n=4. *P<0.05; **P<0.01 versus control in E and F; ##P<0.01 versus EtOH in E or MMP-3/9 in F. Scale bar=40 µm in A, B, and C; and 5 µm in D.

Reduction in TJ Proteins by MMPs Disrupts the Integrity of BBB Interface

Changes in the expression of MMP-3, MMP-9, and TJ proteins were validated by Western blot analyses in lysate protein from hBEC culture, intact brain microvessels, and brain tissue homogenates. Similar to microscopy imaging data, Western blot analyses confirmed that alcohol in deed enhanced the levels of MMP-3 and MMP-9 proteins (Figure 1E) as well as decreased the levels of occludin, claudin-5 and ZO-1 proteins (Figure 1F) in hBEC culture, intact capillary vessels, and in brain tissue. Degradation of occludin, claudin-5 and ZO-1 by addition of exogenous MMP-3 or MMP-9 proved the point that BBB is targeted by MMPs activation. The attenuating effect of TIMP1 on TJ proteins in the absence or presence of ethanol further validated the notion that activation of MMPs lead to disruption and leakiness of the BBB integrity.

Loss of TJ Integrity Enhances Permeability and Leakiness of BBB

As functional assays, we assessed the integrity of BBB by TEER, permeability of BBB by sodium fluorescein/Evan blue (NaFl/EB) tracers, and infiltration of immune cells across the BBB and into the brain by calcein-AM or Fluo-3–labeled monocytes. Assessing the changes in real-time TEER by a highly sensitive ECIS system showed a significant reduction in BBB electric resistance by EtOH or in combination with MMP-3 or MMP-9, which exacerbated the effect of EtOH in a time-dependent manner (Figure 2A). The fact that TIMP1 prevented the effect of EtOH on TEER reduction reinforced the direct role MMPs on BBB dysfunction. The role of VEGFR-2 in the integrity of BBB was come from the treatment of Ki8751 (VEGFR-2 kinase inhibitor), where TEER value was significantly reduced and with EtOH the effect was exacerbated (Figure 2A). Disruption of this BBB integrity was further supported by increased permeability of small molecular weight tracer NaFl (MW=376) and large molecular weight tracer EB (MW=961) into the brain of chronic alcohol intake rats compared with respective controls (Figure 2B).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

Activation of MMP-3/9 causes BBB leakiness. A, Kinetic profile of the transendothelial electric resistance (TEER) across the BBB after EtOH, MMP-3/9, TIMP1, VEGF-A, or Ki8751 treatment. After maintaining a stable resistance for 2 hours, TEER was monitored at 400 Hz with 10-minute intervals. Reproducible results were obtained in 3 individual experiments. B, In vivo permeability of Evans blue (EB, 5 µmol/L) and sodium fluorescein (Na-Fl, 5 µmol/L) across the BBB (n=5). C, Migration of monocytes across the in vitro model of BBB after treatment of various test compounds as shown in figure. Data are expressed as fold difference from untreated cells. D, Effects of EtOH (50 mmol/L), VEGF-A (100 ng/mL), Ki8751 (10 µmol/L), EtOH+VEGF-A, or EtOH+Ki8751 on occludin, claudin-5, and ZO-1 levels in hBECs after 24 hours of exposure (Western blotting). E, Fluo-3–labeled macrophage adhesion/migration in brain capillary after infusion of cells into the common carotid artery. F, Immunofluorescent staining of CD68 (green) and von Willebrand factor (green) merged with DAPI (blue) in microvessel of brain tissue section. Scale bar indicates 5 µm in all panels. *P<0.05; **P<0.01 versus untreated or control in A, B, and C; ^#P<0.05; ^##P<0.01 versus EtOH (second bar) in C.

Leakiness of BBB Enhances Immune Cell Infiltration Into the Brain

Migration of calcein-AM–labeled monocytes across the endothelium showed a significant 2- to 2.5-fold increments of immune cell migration in EtOH or EtOH in combination with MMP-3 or MMP-9 compared with controls (Figure 2C). The effect of EtOH on cell migration was comparable to that of lipopolysaccharide (LPS), which was used here as positive control. As expected, TIMP1, an inhibitor of MMPs attenuated the effect of EtOH on monocyte migration. VEGF-A alone did not cause robust cell migration. However, inhibition of VEGFR-2 by Ki8751 exacerbated the effect of EtOH on cell adhesion and migration across the BBB. Thus, we further investigated the involvement of VEGFR-2 on BBB integrity in hBEC culture. Western blot analyses revealed that treatment of hBECs with Ki8751 selectively downregulated the levels of occludin, claudin-5, and ZO-1 proteins either in the presence or absence of alcohol exposure, suggesting the synergistic association between VEGFR-2 and brain vascular function (Figure 2D). Moreover, addition of VEGF-A (100 ng/mL) exogenously decreased the level of occludin, claudin-5 and ZO-1 proteins (Figure 2D), suggesting that antagonizing the VEGFR-2 function by a high concentration of VEGF-A can impair the integrity of BBB.

Finally, utilization of our animal model of chronic alcohol intake allowed us to validate the notion that neuroinflammatory process was initiated by BBB damage. We found that infusion of bone marrow derived rat monocytes and macrophages labeled with Fluo-3 into the right common carotid artery demonstrated an increase in cell adhesion and infiltration at the damage sites of the brain capillary alcohol intake animal compared with pair-fed controls (Figure 2E). Further, infiltration of immune cells into brain was analyzed by staining with antibody to CD68 in brain tissue sections from rats that were not subjected to cell infusion. Our data confirmed that there was more number of infiltrated immune cells in the brain of alcohol intake than the pair-fed controls (Figure 2F). Taken together, these results indicate that activation of MMPs and diminished repair function of VEGFR-2 may lead to vascular leakiness and enhanced cell adhesion/infiltration for initiation of cerebral inflammatory disease in alcohol abuse.

Degradation of VEGFR-2 Elevates Cellular and Serum VEGF-A Levels

The decrease in VEGFR-2 protein that we observed is expected to elevate the cellular and circulatory VEGF-A levels because the platelets and endothelial cells will continue to secrete VEGF-A, whereas VEGFR-2/VEGFR-1 is unable to bind to VEGF-A inefficiently. Thus, we quantified the levels of VEGF-A in blood serum samples and supernatants of hBECs cultured media by ELISA. Our results show that alcohol elevated the levels of VEGF-A in serum and cellular culture media compared with respective controls (Figure 4A and 4B). Interestingly, MMP-3 and MMP-9 also increased the levels of VEGF-A either in the presence or absence of alcohol, whereas TIMP1 (inhibitor of MMPs) decreased the effect of alcohol on VEGF-A regulation. These data suggest that MMP-3/9 was involved in the malfunctioning of VEGFR-2 causing increase extracellular VEGF-A levels. To address the idea that this excess level of VEGF-A in turn triggers the activation of MMP-3/9, we then determined the changes in protein levels of MMP-3, MMP-9, and VEGFR-2 after treatment of VEGF-A with hBECs in culture. The kinetic profile data showed that an increase in MMP-3 or MMP-9 protein level was inversely proportional to a decrease in VEGFR-2 protein (Figure 4C) after VEGF-A treatment. The subtle difference between MMP-3, MMP-9, and VEGFR-2 alterations appeared to occur at 12–24 hours after VEGF-A exposure. Thus we treated hBEC culture with VEGF-A for 24 hours with or without EtOH and analyzed the alterations of MMP-3, MMP-9, and VEGFR-2 proteins level by Western blotting. Indeed, we confirmed that activation of MMP-3 or MMP-9 by physiological excess VEGF-A level led to the degradation of VEGFR-2 protein, which was further aggravated by the presence of alcohol (Figure 4D). We then examined the phosphorylation of VEGFR-2 at 2 sites of phospho-VEGFR-2 tyr1054 and phospho-VEGFR-2 tyr1175 in cell culture and animal model of chronic alcohol intake. Using specific antibody to phospho-VEGFR-2 tyr1054 and phospho-VEGFR-2 tyr1175, we detected the phosphorylation of VEGFR-2 by Western blotting. EtOH or VEGF significantly increased the phosphorylation of VEGFR-2 at these 2 tyrosine residues in hBECs culture similar to alcohol-induced increase in whole brain or microvessel tissues compared with respective controls (Figure 4E). As expected, TIMP1 or Ki8751 significantly inhibited the phosphorylation of VEGFR-2.

• Download figure
• Open in new tab
• Download powerpoint

Figure 4.

Degradation of VEGFR-2 causes elevation of VEGF-A levels. A and B, ELISA shows the levels of VEGF-A in A, hBEC cell culture supernatants in different treatment conditions; B, blood serum; and C, Western blot analysis of alterations in VEGFR-2, MMP-3, and MMP-9 proteins after treatment of hBECs with VEGF-A (100 ng/mL) at different time points. D, Changes in protein levels of VEGFR-2, MMP-3, and MMP-9 proteins after treatment of hBECs with EtOH (50 mmol/L) or VEGF-A or in combination for 24 hours. E, Western blot analysis for the changes in protein levels of phosphorylated VEGFR-2 (at tyr1054 and tyr1175) after treatment of hBECs in different treatment conditions for 24 hours as shown and in EtOH treated rat brain tissue and microvessels lysates. Results are expressed as ratio of VEGFR-2 or MMP-3/9 to that of β-actin in D and p-VEGFR-2tyr1054 or p-VEGFR-2tyr1175 to that of β-actin bands in E. Values are mean±SEM, n=3–5 in all experiments. *P<0.05; **P<0.01 versus untreated or control in A and B; ^##P<0.05; ^##P<0.01; ^###P<0.001 versus EtOH (second bar) in A and E.

These findings suggest that reduction of VEGFR-2 by MMPs may be further exacerbated by VEGF signaling loop in alcohol-induced neurovascular disease. These data suggest that alcohol-induced degradation of VEGFR-2 causes increase in phosphorylation of VEGFR-2 protein as result of VEGF elevation and translocation of VEGFR-2 to the nuclear membrane, which has been observed in endothelial cells.^31,32

Activation of Caspase-1 Matures IL-1β

Because activation of caspase-1 leads to production of interleukin 1 beta (IL-1β), we then analyzed the level of IL-1β in cell-cultured media and in blood serum from chronic animal studies by ELISA. We observed that treatment of hBECs with exogenous MMP-3, MMP-9, VEGF-A, or Ki8751 significantly increased the levels of IL-1β in the presence or absence of EtOH; however, TIMP1 and z-YVAD-fmk prevented the effects of EtOH on IL-1β elevation (Figure 6A). These in vitro studies findings were further validated by our in vivo studies, where the levels of IL-1β in the blood serum of EtOH-diet consumed animal had about 7-fold higher than the pair-fed control (Figure 6B). Note that the levels of IL-1β in the blood serum were much higher than that present in cell cultured supernatants. Further, to validate the results of these extracellular and circulating IL-1β levels, we also analyzed the actual protein contents of IL-1β in cell-cultured supernatants, cell lysates, and whole brain tissue/microvessel homogenates by Western blotting. Similar to ELISA assayed results, inclusion of MMP-3, MMP-9, VEGF-A, or Ki8751 in alcohol condition significantly elevated the levels of IL-1β in cell-cultured supernatants (Figure 6C), in cell lysates protein (Figure 6D), and in whole brain tissue or microvessel homogenates (Figure 6E). Taken together, these data indicate that MMP-mediated loss of VEGFR-2 function and subsequent elevation of VEGF-A levels in alcohol intake lead to activation of caspase-1 and release of proinflammatory IL-1β. Thus, release of this cytokine via the VEGF signaling pathway may be the reversible loop mechanism for the activation of MMP-3/9 by VEGF-A, which was observed in data presented in Figure 4A and 4B.

• Download figure
• Open in new tab
• Download powerpoint

Figure 6.

Activation of caspase-1 matures IL-1β and causes neuroinflammation and cell apoptosis. A and B, ELISA of IL-1β levels in A, hBEC culture supernatants of different treatment conditions, and B, blood serum (100 µL of serum was used per well) of control and chronic alcohol intake rats. C and E, Western blot analyses of IL-1β protein in C, hBEC culture supernatants expressed as fold induction of IL-1β protein levels; D, hBEC lysate protein; and E, whole brain tissue and microvessel homogenates. Bar graphs show the fold induction of expression of IL-1β in different treatments in C; ratio of expression of IL-1β to that of β-actin bands in D and E. F, TUNEL staining in rat brain cortical tissue section. Arrow indicates the TUNEL-positive cells. Scale bar=40 µm. G and H, Changes in PARP protein levels by Western blotting using anti-PARP antibody in G, hBEC lysate, and H, whole brain cortical tissue or microvessel tissue homogenates. Bar graphs data show the changes in arbitrary relative intensity of 113-kDa and 89-kDa fragments of PARP in cell culture and brain tissue samples. Values are mean±SEM; n=3–4. **P<0.01; ***P<0.001 versus untreated or control in C, D, and E; ^##P<0.01 versus EtOH (second bar) in C and D. ***P<0.001 versus respective fragment in G and H.

Caspase-1 and IL-1β Causes Cell Apoptosis

As a functional read out assay for caspase-1 activation and IL-1β release, we next evaluated the extent of cell apoptosis by TUNEL and Poly-ADP-Ribose-Polymerase (PARP) induction Western blot in hBECs culture and in animal studies. It was obvious that there was much higher positive staining of TUNEL in brain tissue from alcohol-diet ingested animal than the pair-fed control brain tissue section (Figure 6F). These findings were further substantiated by the induction of 89 kDa PARP in hBECs (Figure 6G) and in rat brain tissue or microvessel homogenates (Figure 6H) after alcohol exposure. Similar to IL-1β release data, Ki8751 with or without EtOH upregulated the level of 89 kDa PARP, whereas TIMP1 or z-YVAD-fmk decreased the levels of 89 kDa PARP induction. These results suggest that downregulation of VEGFR-2 mediated caspase-1 activation and that IL-1β release causes cell apoptosis.