- Top of page
- Materials and methods
Microsphere embolism (ME)-induced up-regulation of endothelial nitric oxide synthase (eNOS) in endothelial cells of brain microvessels was observed 2–48 h after ischemia. eNOS induction preceded disruption of the blood–brain barrier (BBB) observed 6–72 h after ischemia. In vascular endothelial cells, ME-induced eNOS expression was closely associated with protein tyrosine nitration, which is a marker of generation of peroxynitrite. Leakage of rabbit IgG from microvessels was also evident around protein tyrosine nitration-immunoreactive microvessels. To determine whether eNOS expression and protein tyrosine nitration in vascular endothelial cells mediates BBB disruption in the ME brain, we tested the effect of a novel calmodulin-dependent NOS inhibitor, 3-[2-[4-(3-chloro-2-methylphenyl)-1-piperazinyl]ethyl]-5,6-dimethoxy-1-(4-imidazolylmethyl)-1H-indazole dihydrochloride 3.5 hydrate (DY-9760e), which inhibits eNOS activity and, in turn, protein tyrosine nitration. Concomitant with inhibition of protein tyrosine nitration in vascular endothelial cells, DY-9760e significantly inhibited BBB disruption as assessed by Evans blue (EB) excretion. DY-9760e also inhibited cleavage of poly (ADP-ribose) polymerase as a marker of the apoptotic pathway in vascular endothelial cells. Taken together with previous evidence in which DY-9760e inhibited brain edema, ME-induced eNOS expression in vascular endothelial cells likely mediates BBB disruption and, in turn, brain edema.
Production of nitric oxide (NO) by nitric oxide synthase (NOS) has important roles in physiological and pathological events in the central nervous system. Accumulating evidence suggests that both neuronal NOS (nNOS) and inducible NOS (iNOS) have detrimental effects on neurons in the ischemic brain, whereas endothelial NOS (eNOS) activity has protective effects (Iadecola 1997). Endothelium-derived NO regulates blood pressure, augments regional blood flow, improves cerebral circulation and inhibits platelet aggregation (Radomski et al. 1990; Morikawa et al. 1994; Huang et al. 1995; Iadecola 1997). As several studies suggest that treatment with 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase inhibitors (statins) up-regulates eNOS expression, thereby improving endothelial function (Endres et al. 1998; Laufs et al. 1998, 2000, 2002), statins are believed to reduce the risk of myocardial infarction and stroke (Ganz et al. 2000). Beasley et al. (1998) showed eNOS up-regulation via an indomethacin-sensitive mechanism in a global ischemia model induced by increasing intracranial pressure, whereas it is still controversial whether cerebral ischemia itself causes eNOS induction in vessels (Zhang et al. 1993; Limbourg et al. 2002; Veltkamp et al. 2002). We recently showed that eNOS up-regulation in vascular endothelial cells is elicited by sublethal ischemia, inducing preconditioning and lethal ischemia in the gerbil forebrain, and that the phosphatidylinositol 3-kinase (PI3-K)/Akt pathway mediates eNOS up-regulation (Hashiguchi et al. 2004). eNOS expression following sublethal ischemia likely mediates preconditioning-induced neuroprotection.
Notably, in lethal ischemia, we found a marked increase in NO production 24–48 h after transient forebrain ischemia in the gerbil (Hashiguchi et al. 2003) and microsphere embolism (ME) in the rat (Shirakura et al. 2005). In both cases, increased NO production coincided with increased eNOS expression without changes in nNOS and iNOS. Thus, the pathological relevance of eNOS up-regulation and prolonged NO production in the late phase by lethal ischemia remains unclear. Interestingly, aberrant expression of nNOS, iNOS and eNOS has also been reported in the Alzheimer's (AD) brain (Luth et al. 2002). In that case, nNOS is mainly expressed in cortical pyramidal neurons, whereas iNOS and eNOS are highly expressed in astrocytes. It is also notable that aberrant NOS expression in the AD brain co-localized with immunoreactivity against nitrotyrosine, a marker of formation of peroxynitrite (ONOO–). Peroxynitrite is produced by generation of both peroxide () and NO, thereby leading to oxidization of proteins, membrane lipid and DNA, or nitration of proteins (Ischiropoulos and al-Mehdi 1995). Peroxynitrite also modifies free tyrosine and tyrosine residues in proteins, resulting in their loss of function. For example, inactivation of manganese superoxide dismutase by tyrosine nitration occurs in rat renal ischemia/reperfusion injury (Cruthirds et al. 2003).
The blood–brain barrier (BBB) in brain microvessels maintains homeostasis of the brain microenvironment mostly through maintenance of tight junctions between brain vascular endothelial cells, thereby preventing passage of hydrophilic molecules or toxic substances from blood to brain. NO and peroxynitrite are known to elicit cerebral microvascular injury, resulting in BBB disruption, following cerebral ischemia (Janigro et al. 1994; Mayhan and Didion 1996; Greenacre et al. 1997; Tan et al. 2004). Of note, inhibition of NOS attenuates BBB disruption during middle cerebral artery occlusion (MCAO) (Nagafuji et al. 1995) and experimental meningitis (Boje 1996). However, the precise molecular mechanisms underlying NO/peroxynitrite-induced BBB disruption are not fully understood.
Here, we report that up-regulation of eNOS in ME-induced cerebral ischemia predominantly occurs in vascular endothelial cells, thereby eliciting protein tyrosine nitration in the same cells. Increased protein tyrosine nitration in microvascular endothelial cells leads to cell damage, thereby inducing BBB disruption. To confirm this hypothesis, we tested the effect of a novel calmodulin-dependent NOS inhibitor, 3-[2-[4-(3-chloro-2-methylphenyl)-1-piperazinyl]ethyl]-5,6-dimethoxy-1-(4-imidazolylmethyl)-1H-indazole dihydrochloride 3.5 hydrate (DY-9760e) (Fukunaga et al. 2000), on ME-induced BBB disruption. ME induced multi embolic infarcts mainly in the striatum, cortex, thalamus and hippocampus, and DY-9760e treatment (50 mg/kg, i.p.) elicited a significant reduction (about 60% of reduction compared with vehicle-treated animals) in infarct area after ME (Shirakura et al. 2005). Using DY-9760e, we confirm that inhibition of protein tyrosine nitration protects endothelial cells, thereby attenuating BBB disruption. The present study also supports the idea that DY-9760e is a novel therapeutic agent to treat brain edema following multiple cerebral infarction.
- Top of page
- Materials and methods
The role of NO in the development of post-ischemic cerebral infarction has been extensively studied (Kumura et al. 1996; Coeroli et al. 1998; Onitsuka et al. 1998; Lei et al. 1999), but few studies have investigated its role in microvascular damage occurring after cerebral ischemia. Using a BBB disruption model in ME-induced ischemia, we report a relationship between BBB leakage and eNOS induction after ME-induced ischemia. The critical observation in this study is that induction of eNOS and, in turn, peroxynitrite formation in the microvascular endothelium precedes events of microvessel disruption in an ME model. Expression of eNOS in the microvascular endothelium is closely associated with increased protein tyrosine nitration and concomitant increased cleaved-PARP immunoreactivity, which implies induction of apoptosis in microvascular cells.
Previously, we found that eNOS expression was induced by forebrain ischemia in the gerbil (Hashiguchi et al. 2003). Like the ME rats analyzed in this study, increased eNOS expression in gerbils peaked 24–48 h after transient ischemia. However, increased eNOS expression in the gerbil hippocampus was not associated with increased protein tyrosine nitration, which was transient and returned to basal levels within 24 h. In contrast with the transient forebrain ischemia, a more pronounced increase in protein tyrosine nitration was observed at 24–48 h with concomitant eNOS overexpression in ME. Notably, under a lethal and permanent ischemic condition such as ME, eNOS overexpressed in the endothelial cells results in protein tyrosine nitration, thereby eliciting injury of microvascular endothelium. ME-induced permanent cerebral ischemia is known to cause more severe injury than ischemia-reperfusion models in brain (Miyake et al. 1993). Indeed, ME-induced cerebral ischemia showed a more pronounced and prolonged increase in protein tyrosine nitration (Shirakura et al. 2005). Co-localization of eNOS with anti-nitrotyrosine immunoreactivity in microvessels further confirms that eNOS overexpression in the endothelial cells is causative for injury of microvascular endothelium. In support of this idea, protein tyrosine nitration was predominant in microvascular endothelium immunoreactive with anti-PARP-p85. Thus, eNOS expression in the endothelium in permanent ischemia leads to aberrant protein tyrosine nitration and PARP activation. As peroxynitrite is formed through generation of both NO and superoxide, superoxide is likely generated by eNOS in endothelial cells. Notably, overexpression of eNOS accounts for superoxide generation as well as NO in the ME setting, because eNOS is known to produce superoxide in the absence of substrates such as tetrahydrobiopterin and arginine. Taken together, under lethal ischemia conditions, eNOS expression in endothelial cells accounts for increased protein tyrosine nitration, thereby leading to cell injury and PARP cleavage.
In addition to increased eNOS expression following ME, small but significant increases in nNOS expression were observed in the ipsilateral hemisphere without changes in iNOS expression. A slight increase in nNOS was observed within 2 h, similar to eNOS expression after ME. The change in nNOS was not significant in immunohistochemical studies (data not shown). Although a slight increase in immunoreactivity with an anti-nitrotyrosine antibody was also observed in cortical neurons, we could not conclude whether the increased nNOS had a detrimental effect on neurons or on the BBB in the present study. In transient ischemia in gerbils, we previously demonstrated increased protein tyrosine nitration in hippocampal pyramidal neurons, which precedes delayed neuronal death in the hippocampus (Hashiguchi et al. 2003). Similarly, inhibition of protein tyrosine nitration in pyramidal neurons by DY-9760e treatment largely attenuated delayed neuronal death in the hippocampus. Taken together, protein tyrosine nitration in cortical neurons likely has a detrimental effect on neurons. Up-regulation of both nNOS and eNOS with concomitant increases in protein tyrosine nitration was also reported in neonatal rat brain subjected to unilateral carotid artery occlusion (Van den Tweel et al. 2005). In neonatal brain ischemia, increased expression of both nNOS and eNOS was transient and returned to basal levels within 24 h of ischemia. Increased nitrotyrosine formation was also transient and seen only 30 min after ischemia. However, the authors of that study did not show that nitrotyrosine production was associated with neuronal damage. Similarly, Ochiai-Kanai et al. (1999) reported that increased immunoreactivity against nitrotyrosine was induced following hypoxia or NMDA treatment in neonatal rat cerebrocortical slices. The clinical significance of induction of NOSs and nitrotyrosine formation was also demonstrated in Alzheimer's brains using immunohistochemical methods (Luth et al. 2002). In that study, aberrant expression of nNOS, eNOS and iNOS in the Alzheimer's brain was correlated with increased protein nitration. Specifically, aberrant expression of nNOS co-localized with nitrotyrosine immunoreactivity in cortical pyramidal cells. On the other hand, both iNOS and eNOS were highly expressed in astrocytes with concomitant increased nitrotyrosine in the Alzheimer's brain. The authors concluded that increased expression of all NOS isoforms in astrocytes and neurons contributes to synthesis of peroxynitrite, thereby leading to nitrotyrosine generation. Similarly, calcium-dependent NOS activity and iNOS immunoreactivity increased in the cerebral cortex of amyloid-precursor protein Tg2576 transgenic mice (Rodrigo et al. 2004). Taken together, up-regulation of NOS isoforms expressed in the brain could trigger peroxynitrite formation in various cell types, thereby leading to cellular damage through protein tyrosine nitration. Free 3-nitrotyrosine released from nitrated proteins also accounts for vascular endothelial dysfunction and neurotoxicity through promotion of DNA damage (Mihm et al. 2000). That study showed that protein nitration also caused dysfunction of enzymes, including manganese superoxide dismutase, an enzyme that scavenges toxic superoxide (Mn-SOD). Other proteins critical for respiratory activity in mitochondria are nitrated during inflammatory events in the liver of lipopolysacchride-treated rats (Aulak et al. 2001). Although we cannot define the substance that stimulates increased NO production and protein tyrosine nitration in endothelial cells, several cytokines and neurotransmitters could trigger NO and peroxynitrite production following brain ischemia (Bredt 1999; Bachschmid et al. 2003; Ohtaki et al. 2003; Walford et al. 2004).
We showed here that PARP cleavage, an event triggering apoptosis, occurred in tyrosine nitration-positive endothelial cells following ME ischemia. Persistent cleavage of PARP after ME suggests that it is caused by caspases mediating apoptotic cell death in the microvasuclar endothelium. Anti-PARP-p85 immunoreactivity was predominantly co-localized with anti-nitrotyrosine immunoreactivity in endothelial cells. This result in consistent with a previous study indicating that apoptosis in brain vascular endothelial cells is associated with caspase activation and PARP cleavage (Meguro et al. 2001; Akin et al. 2002). Furthermore, co-localization of tyrosine nitration immunoreactivity with leaked rabbit IgG immunoreactivity also confirmed that protein tyrosine nitration in endothelial cells led to leakage of serum protein. However, it is unclear whether PARP cleavage in endothelial cells causes BBB disruption. Further study is required to define key molecules involved in BBB disruption.
We recently described a neuroprotective agent, DY-9760e, that inhibits NO generation and protein tyrosine nitration formation (Fukunaga et al. 2000; Hashiguchi et al. 2003; Shirakura et al. 2005). DY-9760e treatment inhibits BBB disruption in a rat middle cerebral occlusion model (Sato et al. 2003). Here, we found that DY-9760e preferentially inhibits protein tyrosine nitration induced by eNOS in endothelial cells. Its inhibition of BBB disruption was closely associated with inhibition of protein tyrosine nitration and PARP cleavage in the microvascular endothelium. DY-9760e did not, however, affect ME-induced eNOS expression in endothelial cells. Taken together, inhibition of protein tyrosine nitration predominantly contributes to the BBB protective action of DY-9760e. In addition, we recently demonstrated that DY-9760e inhibits calpain-induced proteolysis of structural proteins, such as fodrin/spectrin. Fodrin co-localizes with tight junction proteins including ZO-1 and connexin-43 in cardiomyocytes (Toyofuku et al. 1998). As the tight junction proteins in vascular endothelial cells are important for BBB integrity (Ballabh et al. 2004), DY-9760e-induced inhibition of fodrin breakdown likely mediates, in part, BBB integrity.
In conclusion, we have shown that induction of eNOS following ME-induced prolonged ischemia produced a marked increase in protein tyrosine nitration, predominantly in vascularendothelial cells, thereby triggering apoptotic pathways. Protein tyrosine nitration and/or calpain-induced fodrin breakdown likely accounts for ME-induced BBB disruption, thereby causing brain edema. A neuroprotective agent, DY-9760e, is an attractive therapeutic drug exhibiting a potent protective action on the BBB in brain ischemia.