Protein Kinase C Activation Modulates Reversible Increase in Cortical Blood–Brain Barrier Permeability and Tight Junction Protein Expression during Hypoxia and Posthypoxic Reoxygenation

Abstract

Hypoxia (Hx) is a component of many disease states including stroke. Ischemic stroke occurs when there is a restriction of cerebral blood flow and oxygen to part of the brain. During the ischemic, and subsequent reperfusion phase of stroke, blood–brain barrier (BBB) integrity is lost with tight junction (TJ) protein disruption. However, the mechanisms of Hx and reoxygenation (HR)-induced loss of BBB integrity are not fully understood. We examined the role of protein kinase C (PKC) isozymes in modifying TJ protein expression in a rat model of global Hx. The Hx (6% O₂) induced increased hippocampal and cortical vascular permeability to 4 and 10 kDa dextran fluorescein isothiocyanate (FITC) and endogenous rat-IgG. Cortical microvessels revealed morphologic changes in nPKC-θ distribution, increased nPKC-θ and aPKC-ζ protein expression, and activation by phosphorylation of nPKC-θ (Thr538) and aPKC-ζ (Thr410) residues after Hx treatment. Claudin-5, occludin, and ZO-1 showed disrupted organization at endothelial cell margins, whereas Western blot analysis showed increased TJ protein expression after Hx. The PKC inhibition with chelerythrine chloride (5 mg/kg intraperitoneally) attenuated Hx-induced hippocampal vascular permeability and claudin-5, PKC (θ and ζ) expression, and phosphorylation. This study supports the hypothesis that nPKC-θ and aPKC-ζ signaling mediates TJ protein disruption resulting in increased BBB permeability.

Keywords

blood–brain barrier ischemia protein kinase C tight junction vascular permeability

Introduction

Hypoxia (Hx) is a component of many disease states including stroke (Banks, 2009). Ischemic stroke occurs when there is a permanent or temporary restriction of cerebral blood flow and oxygen to part of the brain. Studies show that specific brain regions are vulnerable to global cerebral ischemia/Hx, especially the hippocampal CA1 pyramidal neurons and cortical neurons, whereas other regions are more resistant (Pulsinelli et al, 1982; Himeda et al, 2005). During the ischemic, and subsequent reperfusion phase of stroke, the integrity of the blood–brain barrier (BBB) is lost (Mark and Davis, 2002; Witt et al, 2003, 2008). Elucidating the mechanisms of Hx and reoxygenation (HR)-induced changes in BBB integrity remains a considerable task.

The anatomical basis of the BBB is the cerebral microvascular endothelium. The low permeability of the BBB has been attributed to the lack of fenestrations, adherens, and tight junction (TJ) protein complexes located in the paracellular clefts (Furuse et al, 1993; Martin-Padura et al, 1998) and specific enzymatic and membrane transport systems (Tamai and Tsuji, 2000). The fusion of the plasma membrane of adjacent cerebral endothelial cells by TJ complexes results in high transendothelial electrical resistances (Butt et al, 1990), prevents paracellular transport of substances with a molecular weight of >180, and maintains apical/basolateral polarity (Petty and Lo, 2002). The TJ complexes are composed of transmembrane proteins: occludin (Furuse et al, 1993; Hirase et al, 1997), the claudin multigene family of proteins (i.e., claudin-1, -3, -5, -12) (Furuse et al, 1998), and junctional adhesion molecule-related proteins (Martin-Padura et al, 1998).

Despite the clinical and experimental observations that Hx alters BBB integrity, the intracellular signaling mechanisms underlying regulation of TJ proteins at the BBB remain poorly understood. Several factors have been identified, including TJ protein phosphorylation. The protein kinase C (PKC) family of serine/threonine protein kinases includes at least 12 known isozymes divided into three groups depending on the enzymes' cofactor requirements: (1) conventional (cPKC; PKC-α, PKC-βI, PKC-βII, and PKC-γ), (2) novel (nPKC; PKC-δ, PKC-ε, PKC-η, and PKC-θ), and (3) atypical (aPKC; PKC-λ and PKC-ζ) (Steinberg, 2008). These isozymes differ in their mechanism of activation, subcellular distribution, substrate type, and expression, suggesting that each of these PKC isozymes can perform unique biological tasks (Banan et al, 2003, 2004, 2007). Although several studies have shown the importance of PKC in regulating BBB function (Andreeva et al, 2001; Rao et al, 2002), the role of specific PKC isozymes in barrier regulation and their mechanisms after ischemia remain largely unknown. nPKC-δ is upregulated after transient focal ischemia in rat (Miettinen et al, 1996), whereas increased cPKC-α, -βI, II expression is seen in Hx/aglycemia in BMECs (Yang et al, 2006), and the increased cPKC-βII and -γ expression seen during Hx coincides with increased paracellular permeability in rat brain endothelial cells (Yuan, 2002; Fleegal et al, 2005). Immunofluorescence studies have shown PKC isozymes to colocalize with ZO-1 in Madin–Darby canine kidney and human colon adenocarcinoma cells (Dodane and Kachar, 1996). In other studies, activation of nPKC-θ was reported to affect intestinal epithelial barrier function through modulation of claudin isotypes (Banan et al, 2004, 2007).

In this study, we have used a nonocclusive, nonsurgical model of global Hx to investigate the role of nPKC-θ and aPKC-ζ isozymes in the underlying mechanisms of barrier regulation and changes in the dynamics of claudin-5 and occludin expression and cytoarchitecture in regulation of BBB permeability. Previous studies in our laboratory have shown increased [¹⁴C]-sucrose permeability and edema formation with changes in occludin protein expression in cerebral microvessels after Hx and HR exposure (Witt et al, 2003, 2008). Total PKC activity in isolated cerebral microvessels was also increased, whereas HR (10 minutes) attenuated the increased total PKC activity. (Fleegal et al, 2005).

Studying the role of specific PKC isozymes in the mechanism of BBB function is critical to establish the idea that specific PKC isozymes have fundamental roles in modulatory mechanisms of cellular proteins required for the maintenance of BBB function. A better understanding of effectively modulating specific PKC isozymes will lead to the development of novel therapeutic strategies for many neurologic disorders in which Hx is a central component with loss of BBB integrity.

Materials and methods

Hypoxic Treatment

Female Sprague-Dawley rats (Harlan, Indianapolis, IN, USA) weighing 200 to 250 g were group housed under standard 12/12 hours light/dark conditions and given food and water ad libitum. All treatment protocols were approved by the University of Arizona Institutional Animal Care and Use Committee, and abide by National Institutes of Health guidelines. Rats were divided into three treatment groups: normoxic (Nx), Hx, and Hx with reoxygenation (HR). The Nx rats were maintained in room air (21% O₂) for 30 or 60 minutes. For Hx treatment, conscious rats were placed in an oxygen-controlled Hx chamber (COY Laboratory Products, Grass Lake, MI, USA) as previously described (Witt et al, 2003) and maintained at 6% O₂ for 30 or 60 minutes. The unanesthetized rats showed no signs of distress in response to Hx treatment in this study or in previous studies using this model (Witt et al, 2003). Some Hx-treated rats were given an injection of chelerythrine chloride (5 mg/kg intraperitoneally; Sigma-Aldrich Inc., St Louis, MO, USA) 10 minutes before Hx treatment. For HR treatment, rats were removed from the Hx chamber and exposed to room air (21% O₂) for 15 minutes.

Brain Region Evaluation of Blood–Brain Barrier Permeability to 4 and 10 kDa Dextran Under Normoxic, Hypoxic, and Hypoxia and Reoxygenation Conditions

Quantitative evaluation of BBB disruption during Hx and HR was achieved by measuring fluorescence in defined brain areas based on a method previously described (Singhal et al, 2002). Rats were anesthetized by intraperitoneal injection (1.0 mL/kg) of ketamine (78.3 mg/mL), xylazine (3.1 mg/mL), and acepromazine (0.6 mg/mL) and unilateral cannula implanted into the common carotid artery. Rats were then divided into the three treatment groups (i.e., Nx, Hx, and HR). Ten minutes before the end of each treatment, rats were injected (0.5 mL) with fluorescent dextran (4 or 10 kDa in saline 2 mg/mL). Rats were then transcardially perfused with 0.9% saline to remove intravascular dextran. The brain was rapidly removed and dissected into four regions: hippocampus, cortex, striatum, and cerebellum. Tissue was homogenized in 50% wt/vol trichloroacetic acid (Sigma). After centrifugation (10,000 g), the supernatant was collected and fluorescence intensity (ng/mL) was measured on a microplate fluorescence reader (Tecan Trading AG, Mannedorf, Switzerland) using excitation 495 and emission 520 nm. Total fluorescence in each sample was derived from concentrations of external standards (100 to 1000 ng/mL). The difference in fluorescence between Nx, Hx, and HR tissue was calculated as tracer leakage and data are presented as percent change from Nx tissue.

Brain Slices and Endogenous Rat-IgG Protein Extravasation

Rats were exposed to Nx, Hx, HR conditions as described above. At the end of each treatment, rats were decapitated and brains were rapidly removed, snap-frozen in dry ice-cooled isopentane at −40°C and stored at −80°C. Cryostat coronal sections (30 μm) were cut and mounted on gelatin-coated glass slides and stored at −80°C until required and then probed for endogenous rat-IgG using an fluorescein isothiocyanate (FITC)-conjugated antibody.

Rat Cerebral Microvessel Isolation

Brain microvessels were isolated, based on a previously described method (Witt et al, 2005). Briefly, dissected cortical tissue was homogenized using a hand-held Teflon homogenizer in ice-cold microvessel isolation buffer (103 mmol/L NaCl, 4.7 mmol/L KCl, 2.5 mmol/L CaCl₂, 1.2 mmol/L KH₂PO₄, 1.2 mmol/L MgSO₄, 15 mmol/L HEPES, 25 mmol/L NaHCO₃, 10 mmol/L glucose, 1 mmol/L sodium pyruvate, 10 g/L 64 K dextran, 1 mmol/L Na₃VO₄, 1.0 mmol/L phenylmethylsulfonyl fluoride, 1 mmol/L NaF, pH 7.4) containing protease inhibitor cocktail (Sigma), and phosphatase inhibitor cocktail 1 and 2 (for serine/threonine and tyrosine protein phosphatases; Sigma). Ice-cold 26% dextran (m/v) was added to the homogenate. Samples were gently vortexed, centrifuged (5600 g, 10 minutes, 4°C), and supernatant aspirated. Pellets were resuspended in ice-cold microvessel isolation buffer and passed through a 70-μm filter (Becton Dickinson, Franklin Lakes, NJ, USA). Filtered homogenates were pelleted by centrifugation (3000 g, 10 minutes). The supernatant was aspirated leaving a pellet of enriched brain microvessels to be used for membrane and cytoplasmic fractions or confocal microscopy.

Enriched Membrane and Cytosolic Fractions

Isolation of membrane and cytosolic fractions was performed as previously described (Banan et al, 2004). Brain microvessel samples (isolated as described above) were homogenized using a hand-held Teflon homogenizer in ice-cold CelLytic buffer (Sigma) containing protease inhibitor cocktail (Sigma) and phosphatase inhibitor cocktail 1 and 2 (Sigma). The homogenate was cleared by centrifugation (13,000 g, 20 minutes, 4°C) and the resulting supernatant then ultra-centrifuged (100,000 g, 60 minutes, 4°C). After this high-speed centrifugation, the supernatant was removed and used as a source of the cytosolic-enriched fraction, whereas the pellet was resuspended in CelLytic (Sigma) containing protease inhibitor (Sigma) and phosphatase cocktail inhibitors 1 and 2 (Sigma) used as the source of the membrane-enriched fraction. Protein concentration of each sample was determined using the bicinchoninic acid protein assay (Pierce Biotechnology, Rockford, IL, USA) and used for Western blot analysis.

SDS-PAGE and Western Blot Analysis

Membrane- and cytosolic-enriched fractions were separated on SDS-polyacrylamide gels (10% or 12% Bis-Tris Criterion XT precast gels; Bio-Rad, Hercules, CA, USA). Separated proteins were transferred to polyvinylidene difluoride (PVDF; 0.45 mm, Perkin Elmer, Waltham, MA, USA) membranes. Nonspecific binding was blocked with Aquablock (Eastcoast Biologics Inc., North Berwick, ME, USA) for 60 minutes at room temperature. The PVDF membranes were incubated with primary antibodies diluted in Aquablock overnight at 4°C. Mouse antibodies to nPKC-θ (0.4 μg/mL; Santa Cruz Biotechnology, Inc., San Francisco, CA, USA), aPKC-ζ (0.4 μg/mL; Santa Cruz), claudin-5 (2.5 μg/mL; Zymed Laboratories, San Francisco, CA, USA), platelet endothelial cell adhesion molecule-1 (PECAM-1) (0.25 μg/mL; Serotec, Oxford, UK), and rabbit antibodies to occludin (2.5 μg/mL; Zymed), phospho-PKC-θ (Thr538) (1:1000; Cell Signaling Technology, Inc., Danvers, MA, USA), phospho-PKC-ζ (Thr410) (1:1000; Cell Signaling), and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) (0.4 μg/mL; Santa Cruz) were used. The PVDF membranes were washed with Tris-Tween buffered saline (30 mmol/L Tris, 150 mmol/L NaCl, and 0.5% (vol/vol) Tween 20 at pH 7.4) and incubated with secondary antibody, either goat anti-mouse or goat anti-rabbit conjugated to IR Dye 680 or IR Dye 800CW (0.1 μg/mL) (LICOR, Lincoln, NE, USA) in Aquablock for 120 minutes at room temperature. The PVDF membranes were washed in Tris-Tween buffered saline and phosphate-buffered saline (PBS). Protein bands were visualized by infrared laser scanning using the Odyssey Infrared Imaging system (LICOR). Molecular weights of the protein bands were calculated using Odyssey software and were quantified and corrected for background using ImageJ densitometric software (NIH, Bethesda, MD, USA). Protein levels were normalized to expression of PECAM-1 or GAPDH, which served as loading controls.

Confocal Microscopy

Microvessel-enriched preparations or coronal brain sections (+0.2 to 0.7 mm interaural line) were air-dried, fixed in 100% ethanol for 10 minutes, washed in PBS then in 1% bovine serum albumin/0.2% Tween-20 in PBS (buffer), and incubated in normal goat serum (1:50; Dako A/S, Glostrup, Denmark) in buffer for 30 minutes. Indirect immunofluorescence on isolated microvessel preparations was performed using: mouse antibodies to PECAM-1 (0.25 μg/mL; Serotec) and rabbit antibodies to claudin-5 (2.5 μg/mL; Zymed), occludin (2.5 μg/mL; Zymed), and ZO-1 (2.5 μg/mL; Zymed). For double-label immunofluorescence, polyclonal antibodies to claudin-5, occludin, or ZO-1 were coincubated with monoclonal antibody to PECAM-1. All antibodies were diluted in buffer. Primary antibodies were incubated on microvessel preparations for 120 minutes, washed in PBS/0.5% bovine serum albumin/0.2% Tween-20 and then incubated in purified goat anti-mouse-IgG and goat anti-rabbit-IgG secondary antibodies conjugated to either Alexa-Fluor-488 or Alexa-Fluor-568 (4 mg/mL; Molecular Probes, Eugene, OR, USA) for 60 minutes in the dark. Serum protein extravasation was assessed in coronal brain sections using a monoclonal antibody to rat-IgG conjugated to FITC (100 μg/mL; Dako) incubated for 120 minutes. Finally, both isolated microvessel preparations and coronal brain sections were washed in PBS/0.5% bovine serum albumin/0.2% Tween-20 then PBS and mounted in ProLong Gold antifade (Invitrogen, Carlsbad, CA, USA). All incubations were performed at room temperature. Isolated microvessels and brain sections were examined using a Zeiss LSM 510 laser scanning confocal microscope (Zeiss, Thornwood, NY, USA). Images were acquired through microvessels or 30 μm sections and maximum image projections were exported and viewed using Paint Shop Pro 7.0 (Jasc Software, Eden Prairie, MN, USA) and uniformly adjusted to optimize brightness and contrast.

Statistical Analysis

Dextran fluorescence intensity and Western blot densitometry data are reported as means ± s.e.m. from three to five separate experiments in each group. Statistical significance between treatment groups was determined using analysis of variance. A value of P<0.05 was accepted as statistically significant.

Results

Brain Regional and Temporal Changes in Blood–Brain Barrier Integrity Induced by Hypoxia and Hypoxia and Reoxygenation

We determined Hx-induced changes in BBB integrity to fluorescently tagged dextrans (4 and 10 kDa) and endogenous rat-IgG in the hippocampus, cortex, striatum, and cerebellum. After Hx (30 minutes) exposure, the hippocampus showed a significant (P<0.05) increase in permeability to both the 4- and 10-kDa fluorescent dextrans, compared with the hippocampus from Nx rats (Figures 1A and 1B). The Hx-induced hippocampal permeability was significantly (P<0.05) attenuated after HR (Hx 30 minutes + 21% O₂ 15 minutes). In cortical tissue, Hx (30 minutes) induced a significant (P<0.05) increase in cortical permeability to 4 kDa dextran, but not to 10 kDa dextran. The Hx-induced increased cortical vascular leak of 4 kDa dextran was attenuated by HR (Hx 30 minutes + 21% O₂ 15 minutes) and was comparable to cortical tissue from Nx rats (Figures 1A and 1B). Other brain regions studied (i.e., striatum and cerebellum) showed much smaller or no changes in vascular permeability after HR treatment (Figures 1A and 1B). We also studied the effect of Hx duration on vascular permeability in the hippocampus and cortex by visualizing changes in distribution of endogenous rat-IgG in coronal brain sections (Figure 1C) by use of an anti-rat-IgG-FITC-conjugated antibody. In Nx tissue, rat-IgG immunoreactivity was found within the cortical and hippocampal vasculature (Figure 1C, arrows). After Hx (30 minutes), rat-IgG immunoreactivity was greatly reduced in hippocampal vasculature (Figure 1C, *) and was seen in the surrounding brain parenchyma, suggesting loss of vascular integrity and leak of rat-IgG from vasculature into the parenchyma. Rat-IgG immunoreactivity was still seen within cortical vasculature (Figure 1C, arrows). After HR, rat-IgG immunoreactivity was once again seen within the hippocampal vasculature, suggesting some degree of vascular integrity had returned preventing further leak of rat-IgG (Figure 1C). Extending the Hx period to 60 minutes resulted in reduced rat-IgG immunoreactivity in both hippocampal and cortical vasculature and was visualized in the surrounding parenchyma. The HR (Hx 60 minutes + 21% O₂ 15 minutes) again restored some BBB integrity as endogenous rat-IgG was seen in both the vasculature again, suggesting further leak of rat-IgG had been prevented (Figure 1C).

Figure 1

Brain regional and temporal changes in vascular permeability to exogenous 4 and 10 kDa dextrans and endogenous rat-IgG induced by hypoxia (Hx) (black bars) and Hx and reoxygenation (HR) (shaded bars) compared with normoxic (Nx) conditions (white bars). (A, B) Increased hippocampal permeability to 4 and 10 kDa dextran after Hx (6% O₂ 30 minutes) was attenuated by HR. Cortical tissue showed increased 4 kDa dextran permeability, but not 10 kDa dextran during Hx. The striatum and cerebellum showed smaller or no change in vascular permeability. Values are expressed as mean ± s.e.m. (C) In coronal brain sections, intact blood–brain barrier (BBB) restricts endogenous rat-IgG extravasation, which is confined within the vasculature (arrows). Hypoxia (30 minutes) increased hippocampal permeability to rat-IgG, which was lost from the vasculature (*). Cortical tissue was unaffected. Hypoxia (60 minutes)-induced loss of rat-IgG immunoreactivity in both hippocampal and cortical vasculature, suggesting increase permeability to rat-IgG. The HR (15 minutes) saw rat-IgG immunoreactivity within both hippocampal and cortical vasculature, suggesting restoration of barrier properties preventing further leak of rat-IgG. n = 3 to 5 rats per group.

Reversible Disruption of Tight Junction Assembly and Tight Junction Protein Expression in Cortical Microvessels Induced by Hypoxia and Hypoxia and Reoxygenation

Isolated cortical microvessels labeled for TJ proteins claudin-5, occludin, and ZO-1 and vascular endothelial cells labeled with PECAM-1 (Figures 2A to 2C, insets) showed marked morphologic changes in response to Hx and HR. In Nx cortical microvessels, claudin-5 immunoreactivity showed sharply defined immunoreactivity at the margins of the endothelial cells (Figure 2A, arrows). After Hx (60 minutes), paracellular claudin-5 immunoreactivity was reduced and appeared diffuse (Figure 2A, ♦) within the endothelial cell, or was not detected (Figure 2A, *). After HR (Hx 60 minutes + 21% O₂ 15 minutes), there was reappearance of claudin-5 immunoreactivity at the endothelial cell margins (Figure 2A, arrows). Restoration of claudin-5 immunoreactivity to paracellular domains was not seen in all vessels, as some vessels still showed sparse claudin-5 immunoreactivity (*). A similar change in morphology was seen in microvessels labeled for occludin and ZO-1 (Figures 2B and 2C). In Nx endothelial cells, occludin and ZO-1 immunoreactivity was seen as a sharply defined pattern of fluorescence along cortical endothelial cell margins (arrows), which was lost after Hx (60 minutes) (Figures 2B and 2C, *). The HR (Hx 60 minutes + 21% O₂ 15 minutes)-induced partial restoration of the sharply defined pattern of immunoreactivity in some endothelial cells (arrows), but not in others (*). Western blot analysis was used to determine the effect of Hx (30 or 60 minutes) and HR (Hx 30 minutes + 21% O₂ 15 minutes) on claudin-5 and occludin protein expression in enriched microvessel membrane and cytosolic fractions (Figures 2D and 2E). A claudin-5 immunoreactive band was seen at 19.6 kDa. Hypoxia (30 minutes) induced a 26% (P<0.05) increase in claudin-5 expression increasing to 46% (P<0.05) after Hx (60 minutes), compared with Nx-membrane fraction. The HR (Hx 30 minutes + 21% O₂ 15 minutes) showed similar levels of protein expression to Hx (30 minutes) (Figure 2D). Very low levels of claudin-5 expression were seen in the cytosolic-enriched fraction (data not shown). The PVDF membranes probed for occludin revealed many bands. These have previously been described, in our laboratory, as representing monomer, dimer, and oligomer forms of occludin (McCaffrey et al, 2009). In this study, we have selected to study changes in the duel monomer 47.9 and 51.5 kDa bands, which are indicative of phosphorylated and nonphosphorylated states of occludin. Membrane-enriched fractions showed a 16% to 24% increase in expression of the occludin 51.5 kDa band, whereas the 47.9-kDa band showed little change, after Hx (30 or 60 minutes). The HR (Hx 30 minutes + 21% O₂ 15 minutes) induced 23% (P<0.05) increase in 51.5 kDa band expression, whereas the 47.9-kDa band showed no significant change in expression from Nx tissue (Figure 2E).

Figure 2

Reversible disruption of tight junction (TJ) assembly in cortical microvessels induced by hypoxia (Hx) (6% O₂ 60 minutes) and Hx and reoxygenation (HR) (15 minutes). (A) Claudin-5; sharply defined pattern of claudin-5 immunoreactivity was seen along the margins of cortical endothelial cells (arrows) under normoxic (Nx) conditions. After Hx, immunoreactivity was lost (*) or appeared as diffuse staining within the endothelial cells (♦). The HR (15 minutes), claudin-5 immunoreactivity resembled the pattern seen in Nx tissue, although immunoreactivity was not so sharply defined. (B) Occludin; a sharply defined pattern of immunoreactivity along the margins of cortical endothelial cells (arrows) was seen under Nx conditions. After Hx, occludin immunoreactivity was lost (*). After HR (15 minutes) occludin immunoreactivity was restored at paracellular domains (arrows). However, some vessels still showed an absence of occludin immunoreactivity. (C) ZO-1; similar results to those obtained with occludin were seen. Insert images show endothelial cells labeled for platelet endothelial cell adhesion molecule-1 (PECAM-1). Western blot analysis showed Hx- and HR-induced changes in TJ protein expression in cortical membrane- and cytosolic-enriched fractions. (D) Claudin-5 expression was increased in membrane fractions after Hx (30 or 60 minutes) (diagonal stripes) compared with Nx fractions (white bars). After HR (horizontal stripes), claudin-5 expression was comparable to Hx (30 minutes) levels. (E) Occludin expression in membrane-enriched fractions showed two bands at 51.5 and 47.9 kDa. Hypoxia (30 or 60 minutes) (diagonal stripes) exposure increased 51.5 kDa band expression with little change in the 47.9-kDa band compared with Nx fractions (white bars). After HR (horizontal stripes), occludin 51.5 kDa band expression was increased compared with that seen in Nx tissue. In all, 47.9 kDa band expression was comparable to that seen in Nx tissue. Protein quantification data were obtained by densitometry and were normalized using PECAM-1 and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) as loading control. Values are expressed as relative optical density and are represented as mean ± s.e.m. For each column n = 3 to 5 rats, *P<0.05.

Changes in Cortical Endothelial Cell Morphology and Expression of nPKC-θ, aPKC-ζ, and cPKC-γ Isozymes Induced by Hypoxia and Hypoxia and Reoxygenation

High levels of nPKC-θ, aPKC-ζ, and cPKC-γ isozyme expression were seen in cortical endothelial cells, which were visualized by PECAM-1 (Figures 3A to 3C). Nonendothelial cells also showed cPKC-γ immunoreactivity in our enriched microvessel preparations (Figure 3C). Hypoxia (60 minutes) and HR (Hx 60 minutes + 21% O₂ 15 minutes) induced a change in expression and cellular localization of nPKC-θ, aPKC-ζ, but not cPKC-γ (Figures 3A to 3C). Under Nx conditions, nPKC-θ demonstrated diffuse immunoreactivity throughout the vascular endothelial cell. After Hx (60 minutes) exposure, nPKC-θ expression increased, and in some regions became localized along the margins of the endothelial cell (Figure 3A, arrows) resembling the pattern seen in microvessels labeled for TJ proteins. This marginal endothelial cell immunoreactivity became more pronounced after HR (Hx 60 minutes + 21% O₂ 15 minutes) treatment (Figure 3A, arrows). Under Nx and Hx (60 minutes) conditions, aPKC-ζ immunoreactivity showed a diffuse distribution throughout the vascular endothelial cell. After HR (Hx 60 minutes + 21% O₂ 15 minutes), there was a marked increase in immunoreactive intensity within the microvessel (Figure 3B). cPKC-γ showed little change in intensity or distribution after Hx and HR in cortical endothelial cells (Figure 3C). Western blot analysis was performed on membrane- and cytosolic-enriched fractions and probed for nPKC-θ and aPKC-ζ protein expression (Figures 4A and 4B). nPKC-θ (67 kDa band) showed a marked (P<0.05) (30% to 42%) increase in protein expression in membrane-enriched fractions after Hx (30 or 60 minutes) treatment compared with that seen in Nx tissue (Figure 4A). The HR (Hx 30 minutes + 21% O₂ 15 minutes) treatment resulted in reduced nPKC-θ expression compared with that seen after Hx (30 minutes). Cytosolic-enriched fraction showed minimal changes in nPKC-θ protein expression under Hx and HR conditions (Figure 4A). Immunoblots probed for aPKC-ζ demonstrated two sets of bands, one at 66.4 kDa present in both membrane and cytosolic fractions and a band at 73.9 kDa present only in the cytosolic-enriched fraction (Figure 4B). The 66.4-kDa band showed increased (21%) (P<0.05) protein expression in the membrane-enriched fraction after Hx (30 minutes). Similar levels of expression were seen after Hx (60 minutes) treatment (Figure 4B). After HR, aPKC-ζ (66.4 kDa) expression was reduced compared with Hx treatment and was comparable to levels seen in Nx tissue. Cytosolic fractions showed little change in expression of the 66.4-kDa band compared with Nx tissue (Figure 4B). Together, these experiments provide evidence that Hx (30 or 60 minutes) and HR (Hx 30 or 60 minutes + 21% O₂ 15 minutes)-induced changes in the degree of cortical expression and cellular localization of nPKC-θ and aPKC-ζ isozymes, but not cPKC-γ.

Figure 3

Enriched cortical microvessels were immunolabeled for nPKC-θ, aPKC-ζ, and cPKC-γ isozymes. (A) nPKC-θ, normoxic (Nx) showed diffuse immunoreactivity within the isolated endothelial cells. Hypoxia (Hx) (60 minutes)-induced increased nPKC-θ immunoreactivity and expression along endothelial cell margins (arrows). Hypoxia and reoxygenation (HR) (15 minutes) treatment showed increased immunoreactivity at endothelial cells margins (arrows) (B) aPKC-ζ Nx showed diffuse immunoreactivity throughout the vascular endothelial cell. After Hx, little change in immunoreactivity was seen. After HR, there was increased diffuse fluorescence within the endothelial cells. (C) cPKC-γ Nx showed diffuse immunoreactivity within endothelial cells. Hypoxia and HR induced little change in cPKC-γ intensity and distribution. Insert images show endothelial cells labeled for platelet endothelial cell adhesion molecule-1 (PECAM-1). PKC, protein kinase C.

Figure 4

Western blot analysis of cortical microvessel membrane- and cytosolic-enriched fractions probed for nPKC-θ and aPKC-ζ isozymes and phosphorylation state, (A) nPKC-θ (67 kDa) showed increased protein expression in membrane fractions after hypoxia (Hx) treatment (30 or 60 minutes). Hypoxia and reoxygenation (HR) (15 minutes) treatment showed similar levels of expression as Hx (30 minutes). Cytosolic fractions showed little change in protein expression under Hx (30 or 60 minutes) and HR conditions. pPKC-θ (Thr538) (74 kDa) was only detected in the membrane-enriched fraction and showed marked increased expression after Hx (30 or 60 minutes). Expression was still elevated after HR. (B) aPKC-ζ showed increased protein expression during Hx treatment, which decreased to levels comparable to that detected in normoxic (Nx) tissue after HR. aPKC-ζ in cytosolic-enriched fractions showed no significant changes in expression. pPKC-ζ (Thr410) showed a slight increase in expression after Hx (30 minutes) increasing after Hx (60 minutes). The HR reduced pPKC-ζ (Thr410) expression. Cytosolic fractions showed slight increased expression during Hx and HR. Protein quantification data were obtained by densitometry and were normalized using platelet endothelial cell adhesion molecule-1 (PECAM-1) and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) as loading control. Values are expressed as relative optical density and are represented as mean ± s.e.m. For each column n = 3 to 5 rats, *P<0.05. PKC, protein kinase C.

Activation of Protein Kinase C Signaling in Rat Cortical Microvessels Induced by Hypoxia and Hypoxia and Reoxygenation

Phosphorylation of nPKC-θ on Thr538 and/or Ser676 and phosphorylation of aPKC-ζ on Thr410 is essential for nPKC-θ and aPKC-ζ activation. To determine whether the observed changes in nPKC-θ and aPKC-ζ protein expression and endothelial cellular localization represent a change in isozyme activity during Hx or HR, we probed with phospho-specific antibodies pPKC-θ (Thr538) and pPKC-ζ (Thr410). Western blot analysis revealed expression of pPKC-θ (Thr538) (74 kDa) band only in the membrane fraction and not the cytosolic fraction. Hypoxia (30 or 60 minutes) treatment induced significant (P<0.05) increased pPKC-θ (Thr538) expression when compared with Nx tissue (Figure 4A). Expression of this band remained slightly above Nx levels after HR. The PVDF membranes probed for pPKC-θ (Ser676) only showed expression in the cytosolic fraction (data not shown). The PVDF membranes probed for pPKC-ζ (Thr410) (72.8 kDa) showed a slight (4% to 16%) but not significant increase in expression in the membrane fraction after Hx (30 or 60 minutes) compared with Nx expression levels. In contrast to nPKC-θ, PKC-ζ also showed elevated (13% to 26%) phosphorylation expression in cytosolic fractions (Figure 4B).

Effect of Chelerythrine Chloride on Blood–Brain Barrier Integrity and Claudin-5, Occludin, nPKC-θ, and aPKC-ζ Expression in Rat Cortical Microvessels During Hypoxia

To test the hypothesis that PKC isozymes mediate Hx-induced changes in BBB integrity, we administered a PKC inhibitor, chelerythrine chloride. Hypoxia (30 minutes) induced a significant (P<0.05) increase in hippocampal permeability to 4 kDa fluorescent dextran (Figure 1A). Increased hippocampal permeability was significantly (P<0.05) attenuated by chelerythrine chloride (5 mg/kg intraperitoneally) (Figure 5A). However, Hx-induced cortical increase in 4 kDa dextran vascular permeability was not significantly attenuated by chelerythrine chloride (5 mg/kg intraperitoneally) (Figure 5A). Immunoblots from cortical membrane-enriched fractions showed that the Hx-induced increased claudin-5 protein expression was significantly (P<0.05) attenuated by chelerythrine chloride (5 mg/kg intraperitoneally) and showed levels comparable to that seen in Nx-membrane fractions (Figure 5B). Expression of occludin (51.5 kDa) showed no significant change, whereas the expression of the 47.9-kDa band showed a significant (P<0.05) 21% increase in expression following chelerythrine chloride (5 mg/kg intraperitoneally) administration (Figure 5C). Chelerythrine chloride (5 mg/kg intraperitoneally) also significantly (P<0.05) attenuated the Hx (30 minutes)-induced increase in nPKC-θ membrane protein expression (Figure 5D) to levels comparable to those seen in Nx conditions. A similar, but not significant, attenuation of Hx-induced elevated aPKC-ζ membrane expression was also seen (Figure 5E) following chelerythrine chloride administration.

Figure 5

Effect of chelerythrine chloride on hypoxia (Hx)-induced increased vascular permeability to 4 kDa dextran and tight junction (TJ) and protein kinase C (PKC) isozyme protein expression. (A) Hippocampal vasculature showed increased permeability to 4 kDa dextran after Hx (30 minutes) (black bars) compared with normoxic (Nx) controls (white bars). This was attenuated by chelerythrine chloride (5 mg/kg intraperitoneally) (shaded bars). Cortical tissue also showed increased permeability to 4 kDa during Hx, but was not attenuated by chelerythrine chloride. (B) Claudin-5 expression was significantly increased in the membrane-enriched fraction after Hx (6% O₂ 30 or 60 minutes), which was prevented by administration of chelerythrine chloride (5 mg/kg intraperitoneally). (C) Expression of occludin (51.5 kDa) showed no significant change, whereas 47.9 kDa band showed increased expression following administration of chelerythrine chloride (5 mg/kg intraperitoneally). (D, E) Hypoxia-induced increased expression of nPKC-θ and aPKC-ζ in membrane fractions was prevented by chelerythrine chloride (5 mg/kg intraperitoneally). Protein quantification data were obtained by densitometry and were normalized using platelet endothelial cell adhesion molecule-1 (PECAM-1) as loading control. Values are expressed as relative optical density and are represented as mean ± s.e.m. For each column n = 3 to 5 rats, *P<0.05.

Discussion

In this investigation, we have shown that activation of nPKC-θ and aPKC-ζ isozymes are required for dynamic changes in TJ protein expression and functional integrity of the BBB after Hx and HR. The mechanism underlying this biological effect of PKC isozymes appears to be reversible cortical disruption of claudin-5, occludin, and ZO-1 organization and assembly, resulting in a size and time-dependent opening of the cortical BBB. These conclusions are drawn from several lines of evidence as discussed below.

The effect of nPKC-θ and aPKC-ζ activation on TJ expression and paracellular permeability at the BBB has been studied in an in vivo in a rat model of global Hx. This model has a number of advantages as: (1) the degree of Hx stress does not induce necrotic damage of the endothelium and allows us to study a recoverable BBB, (2) oxygen availability is reduced without stopping the cerebral blood flow thus allowing nutrients still to reach the brain, and (3) the interaction of a number of cellular components (i.e., astrocytes, pericytes, neurons, microglia) may be studied.

A novel aspect of this study is assessment of vascular integrity after both Hx and HR treatments and in defined brain regions. Our results showed that Hx rapidly increased hippocampal vascular permeability to both 4 and 10 kDa dextrans and rat-IgG, whereas HR restored some degree of vascular integrity. The small increase in cortical vascular leak to 10 kDa dextran may represent a partial opening of the BBB after Hx (30 minutes) to smaller molecules such as [¹⁴C]-sucrose and 4 kDa dextran, but still restrictive to larger molecules such as 10 kDa dextran and rat-IgG. Increasing Hx exposure from 30 to 60 minutes induced loss of cortical vascular integrity to rat-IgG, which was restored by HR. Further studies are required to assess vascular permeability (4 and 10 kDa) after a longer Hx stimulus (i.e., 60 minutes). The striatum and cerebellum were less vulnerable to Hx showing smaller changes in vascular permeability. We believe that these data show, for the first time, a size selective and time-dependent Hx-induced opening of the BBB in specific brain regions. Our results are consistent with other studies, which have shown specific brain regions to be especially vulnerable, whereas other regions are more resistant to global cerebral ischemia/Hx (Pulsinelli et al, 1982; Himeda et al, 2005). Previous studies have shown that Hx-with HR-induced increased paracellular [¹⁴C]-sucrose (molecular weight ∼342) diffusion across a range of Hx conditions (6% O₂ 30 minutes to 18% O₂ 60 minutes) (Witt et al, 2003). However, vascular permeability was measured after a period of HR (> 10 minutes). As our confocal microscopy studies showed a very rapid change in TJ protein expression (claudin-5, occludin, and ZO-1) after HR (15 minutes) compared with Hx (Figure 2), it was vital to assess vascular permeability after both Hx and HR. The advantages of the present method are that it: (1) allows vascular permeability to be determined during both Hx and HR, (2) causes minimal disturbance of cerebral blood flow, (3) allows use of various sized markers (4 to 70 kDa), and (4) allows vascular permeability to be determined in specific defined brain regions. This study highlights an important difference in vulnerability in cortical and hippocampal vascular integrity after Hx (30 minutes), which may reflect differences in signaling pathways modulating TJ function in the two brain regions. The mechanism by which BBB integrity is lost during Hx is presently unknown, thus the measurement of vascular integrity in specific brain regions after Hx will enable us to identify different cellular signaling mechanisms, which infer greater or lesser vulnerability to Hx insults. We have used a nonocclusive model of global Hx to study effects on vascular permeability. It could be argued that autoregulation may reduce cerebral blood flow and blood pressure resulting in ischemic conditions. However, using this Hx model, Witt et al (2003) reported increased CBF and blood pressure during the first 10 to 20 minutes of Hx insult and the initial period of reoxygenation. Thus, ischemia is unlikely to be a component in this study.

Several factors have been identified to regulate BBB permeability including TJ protein phosphorylation status. The PKC family is recognized to affect epithelial and endothelial barriers, and have been associated with TJ protein alterations, regulating both the subcellular localization and the phosphorylation status of TJ proteins (Avila-Flores et al, 2001; Fleegal et al, 2005). In the CNS, studies have shown PKC isozymes to regulate TJ protein assembly and in maintaining paracellular integrity in endothelial cells of the BBB (Angelow et al, 2005). cPKC-α and -β are both involved in regulation of endothelial cell permeability after ischemia and inflammatory stimulation (Yuan, 2002). In the current study, we have assessed and determined changes in cortical endothelial cell distribution and expression of nPKC-θ, and aPKC-ζ isozymes in isolated microvessels and membrane- and cytosolic-enriched fractions, induced by Hx and HR. The selection of these PKC isozymes was based on reports that these two isozymes are associated with TJ proteins (Gonzalez-Mariscal et al, 2008), previous studies in our laboratory (Fleegal et al, 2005) and a recent report by Banan et al (2007) showing an association between PKC-θ and claudin-5 in the intestine. We have shown increased expression and activation of nPKC-θ and aPKC-ζ after Hx. Our biochemical results are in agreement with an earlier in vitro and in vivo study (Fleegal et al, 2005), which showed that Hx and HR increased total PKC activity through altered expression of specific PKC isozymes. In vivo, Hx induced a large increase in activity of cPKC-γ, -βII, and nPKC-ε, whereas nPKC-θ and aPKC-ζ showed modest increases, although nPKC-θ showed much greater levels of expression during HR. The relative increase in nPKC-θ and aPKC-ζ expression after Hx (60 minutes) was smaller than we report. A refinement of our current study is the use of isolated membrane- and cytosolic-enriched fractions from microvessels rather than total endothelial homogenates. Determination of nPKC-θ expression in the membrane-enriched fraction may account for the greater nPKC-θ expression reported. The increased nPKC-θ and aPKC-ζ protein expression induced by Hx or HR treatment is unlikely to be due to de novo synthesis of new proteins because of the acute nature of the study, but could be explained by a conformational change in PKC protein favoring antibody binding or by PKC movement from one cellular pool to another. The later hypothesis is supported by our immunofluorescence studies, which showed a change in endothelial expression and distribution of both nPKC-θ, aPKC-ζ, but not cPKC-γ, after Hx and HR. nPKC-θ immunoreactivity, under Nx conditions, appeared diffuse within the endothelial cell, but after Hx and HR showed increased fluorescence at the endothelial cell margins (Figure 3A, arrows). Although aPKC-ζ showed increased immunoreactivity after Hx and HR, unlike nPKC-θ it did not show endothelial cell margin expression, and remained diffusely distributed in the endothelial cell. The PKC isozymes are known to be activated by phosphorylation, and once activated some isozymes translocate from the cytoplasm to the membrane (Ron and Kazanietz, 1999; Gopalakrishna and Jaken, 2000). Phosphorylation of nPKC-θ on Thr538 or Ser676 and phosphorylation of aPKC-ζ on Thr410 is essential for nPKC-θ and aPKC-ζ activation. Using phospho-specific antibodies to nPKC-θ and aPKC-ζ, we demonstrated Hx- and HR-induced activation of both nPKC-θ and aPKC-ζ isozymes. Western blots probed for pPKC-θ (Ser676) only showed expression in the cytosolic fraction (data not shown). Further studies are in progress to determine whether other PKC isozymes show similar changes. Fleegal et al (2005) reported cPKCβII and nPKCε showed a large increase in expression after Hx stress. cPKCβII and nPKC-δ have been shown to have a role in BBB permeability in bEnd3 cells (Kim et al, 2010b) and nPKC-δ has been reported to attenuate blood–retinal barrier breakdown in diabetic retinopathy (Kim et al, 2010a).

We propose, based on the changes in morphologic nPKC-θ profiles (Figure 3A) and changes in pPKC-θ (Thr538) and nPKC-θ protein expression, that Hx and HR activate nPKC-θ by phosphorylation on PKC-θ (Thr538) inducing translocation from cytosolic pools and non-TJ membrane domains to TJ domains. In these TJ domain regions, nPKC-θ will be localized next to a TJ protein substrate (e.g., claudin, occludin, ZO-1) and thus in a position to modify TJ Ser/Thr phosphorylation status leading to altered BBB function. Supporting this hypothesis, in vitro studies have shown that aPKC-ζ colocalizes with ZO-1 in both Madin–Darby canine kidney and in human colon adenocarcinoma cell lines (Dodane and Kachar, 1996). Banan et al (2007) have shown that nPKC-θ isozymes modify claudin isotypes resulting in altered intestinal epithelium barrier properties. Schechtman and Mochly-Rosen (2001) proposed that receptors for activated C kinase (RACK) are selective anchoring proteins binding individual PKC isozymes after their activation. The PKC binding to RACK localizes each PKC isozyme next to a protein substrate and away from other protein thus giving a different functional specificity for each activated PKC isozymes (Schechtman and Mochly-Rosen, 2001). Specific RACK proteins have been identified, RACK-1 is selective for cPKC-βII and RACK-2 is selective for nPKC-ε (Schechtman and Mochly-Rosen, 2001).

Changes in nPKC-θ and aPKC-ζ expression correlated with changes in the pattern of claudin-5, occludin, or ZO-1 immunostaining of isolated cortical microvessels after Nx, Hx, and HR treatments. The pattern of immunostaining marking the margins of adjacent endothelial cells seen under Nx conditions was similar to that previously reported (Willis et al, 2004) and represents the TJ expression in an intact BBB. After Hx, as nPKC-θ and aPKC-ζ are activated, TJ protein immunoreactivity was greatly reduced or absent in cortical endothelial cells corresponding to a loss of vascular integrity to dextran markers and rat-IgG. Confocal microscopy showed claudin-5 immunoreactivity was greatly reduced at the paracellular domains and showed diffuse immunoreactivity within the endothelial cell membrane, whereas Western blot analysis showed increased membrane fraction claudin-5 expression. This picture corresponded to a loss of vascular integrity to dextran markers and endogenous rat-IgG. The increased Western blot claudin-5 expression reported here is consistent with other studies, which have shown increased claudin-5 expression is associated with a loss of barrier function (Brooks et al, 2005). Occludin and ZO-1 immunoreactivity in isolated microvessels showed similar changes in response to Hx and HR treatment as seen for claudin-5, except after Hx no diffuse cellular immunoreactivity was seen. The short-time course of the study suggests that the increased TJ protein expression after Hx treatment is due to changes in existing proteins rather than de novo synthesis. Hypoxia may induce a conformational change in TJ protein favoring antibody binding, or as shown for occludin, may induce a change in the relative amounts of oligomeric, dimeric, and monomeric occludin isoforms (McCaffrey et al, 2009).

Finally, to test the hypothesis that PKC isozymes modulate BBB permeability and TJ protein expression in vivo under Hx conditions, we inhibited PKC activity. Chelerythrine chloride is a potent and specific inhibitor of PKC (IC₅₀ = 0.7 μmol/L) compared with other kinases such as protein kinase A (IC₅₀ = 0.17 mmol/L), tyrosine protein kinase (IC₅₀ = 0.1 mmol/L), and calcium/calmodulin-dependent protein kinases (IC₅₀ > 0.1 mmol/L) (Herbert et al, 1990). Chelerythrine chloride inhibits a range both novel and atypical isozymes (Ping et al, 1999; Yuan et al, 2009). We show that chelerythrine chloride attenuated the Hx-induced hippocampal vascular permeability to 4 kDa dextran. Chelerythrine chloride had no effect on Hx (30 minutes)-induced increase in 4 kDa cortical vascular permeability, but attenuated the Hx-induced increased expression of claudin-5 and nPKC-θ and aPKC-ζ isozymes. There was also increased membrane expression of the 47.9-kDa occludin band, whereas the 51.5-kDa band showed little change. Chelerythrine chloride has been shown to attenuate Hx-induced paracellular permeability changes rat brain microvessel endothelial cell cultures (Fleegal et al, 2005). We believe this is the first in vivo demonstration that PKC isozyme inhibition attenuates Hx-induced increased vascular permeability and supports the earlier in vitro work. However, the use of chelerythrine chloride may result in inhibition of many PKC isozymes, so further studies using more isozyme-specific PKC inhibitors, such as 2,4-diamino-5-nitropyrimidines, 2,4-diamino-5-nitropyrimidines, and Go6983, are required to help us identify the roles of nPKC-θ and aPKC-ζ in modifying Hx-induced BBB integrity.

In summary, our results provide in vivo evidence that nPKC-θ and aPKC-ζ intracellular signaling pathways modulate BBB functional integrity through dynamic regulation of assembly and organization of TJ proteins in a model of global Hx and HR. A greater in vivo understanding of Hx-induced changes in BBB permeability/integrity, in specific brain regions is vital in the development of novel drugs and therapeutic regimes in the treatment of ischemic stroke and other diseases in which Hx is a feature.

Footnotes

The authors declare no conflict of interest.

References

Andreeva

Krause

Muller

Blasig

Utepbergenov

(2001) Protein kinase C regulates the phosphorylation and cellular localization of occludin. J Biol Chem 276:38480–6

Angelow

Zeni

Hohn

Galla

H-J

(2005) Phorbol ester induced short- and long-term permeabilization of blood-CSF barrier in vitro. Brain Res 1063:168–79

Avila-Flores

Rendon-Huerta

Moreno

Islas

Betanzos

Robles-Flores

Gonzalez-Mariscal

(2001) Tight-junction protein zonula occludens 2 is a target of phosphorylation by protein kinase C. Biochem J 360:295–304

Banan

Fields

Zhang

Shaikh

Farhadi

Keshavarzian

(2003) The atypical zeta (ζ) isoform of protein kinase C prevents oxidant-induced NF-κB activation in the intestinal epithelium. J Pharmacol Exp Ther 307:53–66

Banan

Zhang

Fields

Shaikh

Farhadi

Keshavarzian

(2004) Theta (θ) isoform of PKC is required for alterations in microtubule cytoskeletal dynamics in intestinal epithelium: a novel function for PKC-θ. Am J Physiol 287:C218–34

Banan

Zhang

Shaikh

Fields

Choudhary

Forsyth

Farhadi

Keshavarzian

(2007) θ Isoform of protein kinase C alters barrier function in intestinal epithelium through modulation of distinct claudin isotypes: a novel mechanism for regulation of permeability. J Pharmacol Exp Ther 313:962–82

Banks

(2009) Mouse models of neurological disorders: a view from the blood-brain barrier. Biochem Biophysia Acta (in press)

Brooks

Hawkins

Huber

Egleton

Davis

(2005) Chronic inflammatory pain leads to increased blood–brain barrier permeability and tight junction protein alterations. Am J Physiol Heart Circ Physiol 289:H738–43

Butt

Jones

Abbott

(1990) Electrical-resistance across the blood-brain-barrier in anesthetized rats—a developmental study. J Physiol 429:47–62

10.

Dodane

Kachar

(1996) Identification of isoforms of G proteins and PKC that colocalize with tight junctions. J Membr Biol 149:199–209

11.

Fleegal

Hom

Borg

Davis

(2005) Activation of PKC modulates blood-brain barrier endothelial cell permeability changes induced by hypoxia and posthypoxic reoxygenation. Am J Physiol Heart Circ Physiol 289:2012–9

12.

Furuse

Fujita

Hiiragi

Fujimoto

Tsukita

(1998) Claudin-1 and -2: novel integral membrane proteins localizing at tight junctions with no sequence similarity to occludin. J Cell Biol 141:1539–50

13.

Furuse

Hirase

Itoh

Nagafuchi

Yonemura

Tsukita

(1993) Occludin—a novel integral membrane-protein localizing at tight junctions. J Cell Biol 123:1777–88

14.

Gonzalez-Mariscal

Tapia

Chamorro

(2008) Crosstalk of tight junction components with signalling pathways. Biochimica et Biophysica Acta 1778:729–56

15.

Gopalakrishna

Jaken

(2000) Protein kinase C signaling and oxidative stress. Free Rad Biol Med 28:1349–61

16.

Herbert

Augereau

Gleye

Maffrand

(1990) Chelerythrine is a potent and specific inhibitor of protein kinase C. Biochem Biophys Res Comm 172:993–9

17.

Himeda

Hayakawa

Tounai

Sakuma

Kato

Araki

(2005) Alterations of interneurons of the gerbil hippocampus after transient cerebral ischemia: effect of pitavastatin. Neuropsychopharmacology 30:2014–25

18.

Hirase

Staddon

Saitou

Ando Akatsuka

Itoh

Furuse

Fujimoto

Tsukita

Rubin

(1997) Occludin as a possible determinant of tight junction permeability in endothelial cells. J Cell Sci 110:1603–13

19.

Kim

J-H

Kim

Jun

H-O

Kim

K-W

(2010a) Inhibition of protein kinase C δ attenuates blood-retinal barrier breakdown in diabetic retinopathy. Am J Pathol 176:1517–24

20.

Kim

Y-A

Park

Kim

M-Y

Lee

Baik

Moon

C-H

Jung

Y-S

(2010b) Role of PKCβII and PKCδ in blood-brain barrier permeability during aglycemic hypoxia. Neurosci Letts 468:254–8

21.

Mark

Davis

(2002) Cerebral microvascular changes in permeability and tight junctions induced by hypoxia-reoxygenation. Am J Physiol Heart Circ Physiol 282:H1485–94

22.

Martin-Padura

Lostaglio

Schneemann

Williams

Romano

Fruscella

Panzeri

Stoppacciaro

Ruco

Villa

Simmons

Dejana

(1998) Junctional adhesion molecule, a novel member of the immunoglobulin superfamily that distributes at intercellular junctions and modulates monocyte transmigration. J Cell Biol 142:117–27

23.

McCaffrey

Willis

Staatz

Nametz

Quigley

Hom

Lochhead

Davis

(2009) Occludin oligomeric assembly at tight junctions of the blood–brain barrier are altered by hypoxia and reoxygenation. J Neurochem 110:58–71

24.

Miettinen

Roivainen

Keinanen

Hokfelt

Koistinaho

(1996) Specific induction of protein kinase C delta subspecies after transient middle cerebral artery occlusion in the rat brain: inhibition by MK-801. J Neurosci 16:6236–45

25.

Petty

(2002) Junctional complexes of the blood-brain barrier: permeability changes in neuroinflammation. Prog Neurobiol 68:311–23

26.

Ping

Takano

Zhang

Tang

X-L

Qiu

RCX

Banerjee

Dawn

Balafonova

Bolli

(1999) Isoform-selective activation of protein kinase C by nitric oxide in the heart of conscious rabbits: a signaling mechanism for both nitric oxide–induced and ischemia-induced preconditioning. Circ Res 84:587–604

27.

Pulsinelli

Brierley

Plum

(1982) Temporal profile of neuronal damage in a model of transient forebrain ischemia. Ann Neurol 11:491–8

28.

Rao

Basuroy

Rao

Karnaky

Jr Gupta

(2002) Tyrosine phosphorylation and dissociation of occludin-ZO-1 and E-cadherin-beta-catenin complexes from the cytoskeleton by oxidative stress. Biochem J 368:471–81

29.

Ron

Kazanietz

(1999) New insights into the regulation of protein kinase C and novel phorbol ester receptors. FASEB J 13:1658–76

30.

Schechtman

Mochly-Rosen

(2001) Adaptor proteins in protein kinase C-mediated signal transduction. Oncogene 20:6339–47

31.

Singhal

Wang

Sumii

Mori

(2002) Effects of normobaric hyperoxia in a rat model of focal cerebral ischemia-reperfusion. J Cereb Blood Flow Metab 22:861–8

32.

Steinberg

(2008) Structural basis of protein kinase C isoform function. Physiol Rev 88:1341–78

33.

Tamai

Tsuji

(2000) Transporter-mediated permeation of drugs across the blood-brain barrier. J Pharm Sci 89:1371–88

34.

Willis

Leach

Clarke

Nolan

Ray

(2004) Reversible disruption of tight junction complexes in the rat blood-brain barrier, following transitory focal astrocyte loss. Glia 48:1–13

35.

Witt

Mark

Hom

Davis

(2003) Effects of hypoxia-reoxygenation on rat blood–brain barrier permeability and tight junctional protein expression. Am J Physiol Heart Circ Physiol 285:H2820–31

36.

Witt

Mark

Huber

Davis

(2005) Hypoxia-inducible factor and nuclear factor kappa-B activation in blood–brain barrier endothelium under hypoxic/reoxygenation stress. J Neurochem 92:203–14

37.

Witt

Mark

Sandoval

Davis

(2008) Reoxygenation stress on blood-brain barrier paracellular permeability and edema in the rat. Microvascular Res 75:91–6

38.

Yang

Roder

Bhat

Thekkumkara

Abbruscato

(2006) Protein kinase C family members as a target for regulation of blood-brain barrier Na,K,2Cl-cotransporter during in vitro stroke conditions and nicotine exposure. Pharm Res 23:291–302

39.

Yuan

Seo

J-S

Kang

Keinan

Steele

Michelotti

Wetsel

Beratan

Gong

Y-D

Lee

Hong

(2009) Identification of 3-hydroxy-2-(3- hydroxyphenyl)-4H-1-benzopyran-4-ones as isoform-selective PKC-ζ inhibitors and potential therapeutics for psychostimulant abuse. Mol BioSyst 5:927–30

40.

Yuan

(2002) Protein kinase signaling in the modulation of microvascular permeability. Vascul Pharmacol 39:213–23