Abstract
Protein kinase C (PKC) consists of a family of closely related Ca2+/phospholipid-dependent phosphotransferase isozymes, most of which are present in the brain and are differentially activated by second messengers. Calcium-dependent PKC activity may cause neuronal degeneration after ischemic insult. PKC is also involved in trophic-factor signaling, indicating that activity of some PKC subspecies may be beneficial to the injured brain. Therefore, we screened long-term changes in the expression of multiple PKC subspecies after focal brain ischemia. Middle cerebral artery occlusion was produced by using an intraluminal suture for 30 min or 90 min. Inin situ hybridization experiments, mRNA levels of PKCα, -β, -γ, -δ, -ε and -ζ were decreased in the infarct core 4 hr after ischemia and were lost completely 12 hr after ischemia. In areas surrounding the core, PKCδ mRNA was specifically induced 4, 12, and 24 hr after ischemia in the cortex. At 3 and 7 d, the core and a rim around it showed increased mRNA levels of PKCδ. No other subspecies were induced. At 2 d, immunoblotting demonstrated increased levels of PKCδ protein in the perifocal tissue, and immunocytochemistry revealed an increased number of PKCδ-positive neurons in the perifocal cortex. In the core, PKCδ-positive macrophages and endothelial cells were seen. Pretreatment with MK-801, an NMDA antagonist, inhibited cortical PKCδ mRNA induction. The data show that focal brain ischemia induces PKCδ mRNA and protein but not other PKC subspecies through the activation of NMDA receptors and that the upregulation lasts for several days in neurons of the perifocal zone.
In most rat models, focal brain ischemia lesions introduced by middle cerebral artery (MCA) occlusion consist of the focus, comprising the lateral striatum and the overlying neocortex, and the perifocal “penumbra,” which consists of the surrounding zone of tissue insufficiently supplied by collateral flow from the leptomeningeal and the anterior and posterior cerebral arteries (Koizumi et al., 1986; Longa et al., 1989; Memezawa et al., 1992). The densely ischemic focal tissue is destined to die, but the penumbra is salvageable by recirculation or pharmacological intervention within 2–3 hr after MCA occlusion. Within a week or less after MCA occlusion, retrograde neuronal degeneration takes place in the ipsilateral thalamus (Iizuka et al., 1990), and disinhibitory overexcitation is thought to cause neuronal death in the ipsilateral substantia nigra (Saji and Reis, 1987; Nagasawa and Kogure, 1990; Tamura et al., 1990) .
Calcium overload resulting from its influx through glutamate receptors and calcium channels and its release from endoplasmic reticulum is thought to be a key mediator of neuronal death in ischemic injury (Siesjö et al., 1995). High levels of intracellular Ca2+ trigger activation of various enzymes, resulting in altered protein synthesis and phosphorylation, increased proteolysis, DNA fragmentation, lipolysis, and production of free radicals (Siesjö et al., 1995). Phosphorylation of target proteins by protein kinases is a major event in the mediation of cellular responses to neuronal injury. Several protein kinases, including protein kinase C (PKC) and Ca/calmodulin-dependent protein kinase are activated by calcium. Both of these kinases have been implicated in ischemic neuronal death (Hara et al., 1990; Madden et al., 1990; Aranowski et al., 1992; Maiese et al., 1993; Hanson et al., 1994; Waxham et al., 1996).
PKC consists of a family of >10 closely related Ca2+/phospholipid-dependent phosphotransferase isozymes, most of which are present in the brain and are differentially activated by second messengers (Tanaka and Nishizuka, 1994). The enzyme family is subdivided into three major classes: the classical PKCs comprising α, β, and γ isoforms, all of which are calcium-dependent; the novel calcium-independent PKCs (δ, ε, η, and θ isoforms); and the atypical group whose members (ζ, λ, and μ) are both calcium- and diacylglycerol-independent enzymes (Tanaka and Nishizuka, 1994). In neurons, PKC has been implicated in the regulation of cell growth and differentiation (Dekker et al., 1989, 1990; Tanaka and Nishizuka, 1994;Roivainen et al., 1995), release of GABA, glutamate, and other neurotransmitters (Tanaka and Nishizuka, 1994; Basudev et al., 1995), apoptosis (Mailhos et al., 1994; Zhang et al., 1995a), gene expression (Angel et al., 1987; Nishizuka, 1992; Hirai et al., 1994; Tanaka and Nishizuka, 1994), synaptic plasticity (Dekker et al., 1989, 1990;Nishizuka, 1992; Tanaka and Nishizuka, 1994), ion channels (Chen and Huang, 1991; Tanaka and Nishizuka, 1994), and activity of neuronal nitric oxide synthase (nNOS) (Maiese et al., 1993; Tanaka and Nishizuka, 1994). In vitro, downregulation or inhibition of PKC increases neuronal protection against excitotoxicity (Favaron et al., 1990; Mattson, 1991), and in vivo some PKC inhibiting agents reduce ischemic injury (Kharlamov et al., 1993), suggesting that PKC is involved in ischemic neuronal death. Even though some studies indicate transient activation of Ca2+-dependent PKCs after global brain ischemia (Cardell et al., 1991; Wieloch et al., 1991;Cardell and Wieloch, 1993), several groups have reported decreased PKC activities after global and focal brain ischemia (Crumrine et al., 1990, 1992; Louis et al., 1991; Domanska-Janik and Zalewska, 1992;Busto et al., 1994). In addition, membrane translocation of PKC in cardiac muscle cells provides transient protection against ischemia (Speechly-Dick et al., 1994; Liu et al., 1995; Mitchell et al., 1995), suggesting that immediate PKC activation rather than inhibition may be beneficial for cells with compromised energy sources.
Under certain physiological circumstances, a long-lasting and persistent Ca2+-independent activation of PKC involving increased gene expression can take place in the mammalian brain (Young, 1989; Thomas et al., 1994b). The maintenance of long-term potentiation in the hippocampus, for example, was recently reported to be associated with such PKC activation (Klann et al., 1991; Thomas et al., 1994b) and to be dependent on NMDA receptors (Thomas et al., 1994a). Other reports have shown that free radicals are able to induce persistent activation of PKC in cultured astrocytoma cells (Brawn et al., 1995) and hippocampal homogenates (Palumbo et al., 1992). Because both free radicals and NMDA glutamate receptors are likely to be involved in neuronal death in focal brain ischemia, we decided to examine whether long-term alterations in mRNA or protein expression of PKC subspecies representing different subclasses takes place after transient unilateral MCA occlusion.
MATERIALS AND METHODS
Ischemia induction and tissue dissection. Focal cerebral ischemia was produced by intraluminal nylon thread introduction. Male Wistar rats (250–300 gm) were anesthetized with 4% halothane (70% N20/30% O2); during the operation, halothane concentration was reduced to 0.5%. The rectal temperature of the animal was maintained between 37.0 and 37.5°C with a heating pad. The left common artery was exposed, and the external carotid artery was ligated. A 0.25 mm nylon thread was inserted into the internal carotid artery up to the anterior cerebral artery. After 30 or 90 min of ischemia, restoration of the MCA blood flow was performed by removing the suture. At intervals of 0, 1, 4, or 12 hr, or 3 or 7 d after ischemia, the animals (n = 4 in each group) were anesthetized with pentobarbital (Mebunat; 40 mg/kg, i.p.) and decapitated for in situ hybridization. For Western and Northern blotting, two control and four ischemic animals were decapitated 2 d after 90 min of ischemia, and tissues representing the striatal infarct core and cortical perifocal area were dissected out from both the ipsilateral and contralateral sides. For immunocytochemistry, four rats were perfusion-fixed with 4% paraformaldehyde (Pease, 1962) 2 d after 90 min of ischemia, and the brains were post-fixed for 3 hr in the same fixative. Four additional animals were used to study the effect of pretreatment with MK-801, an NMDA antagonist. Three milligrams/kilogram M-K801 (Research Biochemical International, Natick, MA) were injected i.p. 30 min before 90 min of ischemia, and after 12 hr of reperfusion the animals were anesthetized with pentobarbital and processed for in situhybridization. Control animals underwent identical surgery except that the thread inserted into the internal carotid artery did not reach the cerebral arteries.
In situ hybridization. Ten micrometer sections were cut on a cryostat at −20°C, collected on SuperFrost Plus slides (Fisher Scientific, Pittsburgh, PA), and stored at −20°C until used. The synthetic oligonucleotides used were as follows: 5′-CGGGGCCCAGCTTGGCTTTCTCGAACTTCTGCCTG-3′ (PKCα); 5′-CCTTGGTACCTTGGCCAATCTTGGCTCTCT-3′ (PKCβ); 5′-GAATGGGAGAGGAAGAGGGGCCCATCCGCACTCTC-3′ (PKCγ); 5′-AGACAGCTGTCTTCTCTCGAATCCCTGGTATATT-3′ (PKCδ); 5′-TAGACGACGAGGCTCGGTGCTCCTCTCCTCGGTTG-3′ (PKCε); 5′-GTCTGG-GTGGCCAGCATCCCTCTCTGGCTGCTTGG-3′ (PKCζ). Oligonucleotides with the length and GC-ratio similar to corresponding antisense oligonucleotides but without homology to any known gene sequences were used as controls. The probes were end-labeled with 35S-ATP using terminal deoxynucleotidyl transferase (New England Nuclear, Boston, MA) and purified over Nuctrap Push Columns (Stratagene, La Jolla, CA). The slides were hybridized overnight in the hybridization solution containing 10 × 106 cpm/ml probe, 40 μl of 5 m dithiothreitol, 50 μl of salmon sperm DNA (10 mg/ml), and 900 μl of hybridization cocktail (50 ml of formamide, 20 ml of 20× SSC, 2 ml of 50 × Denhardt’s reagent, 10 ml of 0.2 m sodium phosphate buffer, pH 7.4, 10 gm of dextran, 4 ml of 25% sarcosyl). The sections were washed in 1× SSC at 55°C for 2 hr, rinsed 2 × 5 min in deionized water at room temperature, dehydrated for 30 sec in 60% and 90% ethanol, air-dried, and covered with Kodak XAR-5 film for 16 d.
Northern blotting. For Northern blotting, total RNA was isolated from frozen rat brain using the TRIzol TM reagent (Life Technologies, Gaithersburg, MD) according to the instructions of the manufacturer. Samples of 30 μg/lane were electrophoresed through a formaldehyde/1.2% agarose gel and transferred to a Hybond N (Amersham, Berkshire, England) nylon membrane by capillary blotting. The membrane was hybridized with the 32P-labeled oligonucleotide probe (the same as in in situ hybridization experiments; specific activity, 3.8 × 108 cpm/pmol) in 5× SSC/5× Denhardt’s/50% formamide/1% SDS at 42°C overnight, and washed twice at room temperature for 5 min each in 2× SSC/0.1% SDS and in 0.2× SSC/0.1% SDS for 5 min at room temperature, and for 15 min at 42°C. The blot was kept wet and exposed for 8 d to Fuji x-ray film at −80°C using intensifying screens.
Immunoblotting. Tissues of the striatal core and perifocal cortical regions from the ischemic hemisphere, and of corresponding regions from the contralateral side, were homogenized separately in a buffer containing 20 mm Tris, pH 7.5, 2 mmEDTA, 5 mm EGTA, and 10 μg/ml aprotinin. Concentrated Laemmli sample buffer was added to a final concentration of 62.5 mm Tris-HCl, pH 6.8, 2% SDS, 10% glycerol, and 5% 2-mercaptoethanol, and samples were heated to 95°C for 5 min. Ten to twenty micrograms of each sample were electrophoresed in 10% polyacrylamide gels. Proteins were electrophoretically transferred to nitrocellulose membranes (Hybond-C extra, Amersham), which were blocked for 1 hr at 25°C with 2% bovine serum albumin diluted in 0.02 m PBS, pH 7.2, containing 0.2% Tween. Blots were incubated with rabbit PKCα, -γ, and -δ antibodies (Life Technologies) (1:300–600 in blocking solution), mouse PKCε and -ζ antibodies (Boehringer Mannheim, Mannheim, Germany) (1:1000), and mouse PKCβ antibody (Seikagaku America, Rockville, MD) (1:1000) overnight at 4°C. After incubation for 1 hr at 25°C with anti-rabbit or anti-mouse peroxidase-conjugated antibody (Amersham) (1:1000), blots were washed four times in PBS-Tween, incubated with ECL detection reagent (Amersham), and exposed to Kodak XAR-5 film.
Immunocytochemistry. The same polyclonal PKCδ antibody used in immunoblotting was diluted 1:1000–2000 in 0.1 msodium phosphate buffer, pH 7.4, containing 0.3% Triton X-100 and 1% bovine serum albumin. Fifty-micrometer-thick vibratome sections were incubated in the primary antibody for 36–72 hr at 4°C, and the bound antibody was visualized with the avidin–biotin–peroxidase method (Vectastain Kit, Vector Labs, Burlingame, CA) using 3,3′-diaminobenzidine as the peroxidase substrate. Control staining included incubations with the primary antibody preabsorbed with the antigen peptide and incubations without the primary antibody.
For double-staining studies, the fixed brains were immersed in 20–30% sucrose buffer for 48 hr, snap-frozen in liquid nitrogen, and cut at 10 μm thickness in a cryostat. The sections were incubated for 48 hr in the mixture of PKC-δ (1:250) and mouse monoclonal glial fibrillary acidic protein (GFAP) (Sigma) (1:300) antibodies or PKC-δ (1:250) and mouse monoclonal OX-42 (Serotec, Oxford, UK) (1:200) antibodies. The OX-42 antibody is against complement C3 receptor and detects activated microglia and macrophages. Fluorescein isothiocyanate (FITC)-conjugated anti-rabbit (Jackson ImmunoResearch, West Grove, PA) and lissamon rhodamine-conjugated anti-mouse (Jackson ImmunoResearch) IgGs were used as secondary antibodies. The sections were coverslipped in glycerol/PBS (3:1) and examined in a Leica DMRB microscope equipped with FITC and rhodamine filter sets.
RESULTS
Expression of PKC mRNAs
In situ hybridization experiments showed distribution of the PKC subspecies similar to that reported previously (Young, 1989) (Figs. 1 and 2). Four hours after 30 or 90 min of ischemia, the mRNA levels of all PKC subspecies showed significant decrease or loss in the infarct core (Fig. 1–3). In the perifocal striatum and the cortex, mRNA levels of PKCδ were increased slightly, especially after 90 min of ischemia (Fig.3A). Although other PKC subspecies showed loss of mRNA expression in the infarcted area and no changes in perifocal regions, PKCδ mRNA was increased further in the perifocal area and in the neocortex, cingulate, and retrosplenial cortex at 12 and 24 hr (Fig. 3B). Occasionally, when the infarct area was very large, a slight PKCδ mRNA induction was detected at 12 hr in the hippocampus (Fig. 4A). Three days after the ischemia, expression of PKCδ mRNA tended to be increased in the core and was still high in a rim around the core. Seven days after the ischemia, PKCδ mRNA expression was increased further throughout the infarcted area (Fig. 3C). The lateral section of the ipsilateral thalamus showed a decreased signal of PKCδ both 3 and 7 d after 90 min of ischemia (Fig. 3C).
In situ hybridization autoradiographs showing the distribution of PKCα, PKCβ, and PKCγ mRNAs in the rat brain at posterior hippocampal (top), dorsal hippocampal (middle), and caudate (bottom) levels 24 hr after 90 min of MCA occlusion. The loss of the signal is seen in the infarct core (left hemisphere). In this particular brain, the expression PKC-β mRNA has not changed in the cortex. No changes in the mRNA expression of PKCα, PKCβ, and PKCγ subspecies are seen in the perifocal region.Asterisks indicate infarcted areas. Magnification, 3.5×.
In situ hybridization autoradiographs showing the distribution of PKCε and PKCζ mRNAs in the rat brain at posterior hippocampal (top), dorsal hippocampal (middle), and caudate (bottom) levels 24 hr after 90 min of MCA occlusion. The loss of the signal is seen in the infarct core (left hemisphere). No changes in the mRNA expression of PKCε and PKCζ subspecies are seen in the perifocal region. Asterisks indicate infarcted areas. Magnification, 5×.
In situ hybridization autoradiographs showing the distribution and induction of PKCδ mRNA in the rat brain at posterior hippocampal (top), dorsal hippocampal (middle), and caudate (bottom) levels 4 hr (A), 24 hr (B), and 7 d (C) after 90 min of MCA occlusion. In the infarct core (right), the expression of PKCδ is decreased or lost at 4 hr (A), but is increased substantially 7 d (C) after 90 min of ischemia. Concomitantly, the perifocal and cortical expression is increased 4 hr (A) and 24 hr (B) (arrowheads) after the insult, but it is back to control levels 7 d (C) after the insult. The signal is decreased in the lateral section of the thalamus 7 d (C) after ischemia. Arrows(A–C) point to the upper margin of the infarcted areas, and arrowheads (A, B) point to the perifocal area with high expression of PKCδ mRNA.Stars (A–C) show the thalamic nuclei with high expression of PKCδ mRNA. Magnification, 8×.
In situ hybridization autoradiographs showing inhibition of ischemia-induced perifocal PKCδ mRNA by administration of MK-801 (3 mg/kg) 30 min before 90 min of ischemia and followed by 12 hr of reperfusion. Sections at dorsal hippocampal (top) and caudate levels are shown. The ischemia-induced expression of PKCδ mRNA (arrowheads) is reduced significantly by MK-801 in perifocal regions, especially in the cortical area including the cingulate cortex. Ashows an ischemic animal pretreated with 0.9% NaCl; Bshows an ischemic animal pretreated with MK-801. In the brain shown inA, the lesion-induced PKCδ mRNA also encompasses the hippocampus. Arrows point to the infarct margins, andasterisks point to the thalamus that shows a high basal level of PKCδ mRNA expression. Magnification, 5×.
In brains treated with 3 mg/kg MK801 30 min before 90 min of ischemia, the PKCδ mRNA induction was inhibited in the perifocal cortex and also to a lesser extent in the perifocal striatum (Fig. 4).
To confirm the specificity of the RNA induced under ischemic conditions and recognized with the PKCδ oligonucleotide, hybridizations were performed in a 100-fold excess of unlabeled probe or with labeled oligonucleotides of the same length and GC-ratio but with the homology <80%. No PKCδ hybridization signal was detected in these experiments (not shown). In addition, Northern blotting with the PKCδ oligonucleotide revealed a single band of ∼3.1 kb in the blot (Fig.5).
Northern blotting analysis of total RNA from the rat brain 2 d after ischemia. A 32P-dATP-labeled PKCδ-oligonucleotide probe was used. The blot shows a band of ∼3.1 kb. C, Contralateral cortex; I, ischemic perifocal cortex; T, thalamus. The signal is weak in the cortical tissues but is increased in the ischemic cortex.
PKC-immunoreactive proteins
Immunoblotting demonstrated protein expression of all the PKC subspecies in both the cortical and striatal regions supplied by the middle cerebral artery (Fig. 6). In samples representing the perifocal cortical area, the PKCδ-immunoreactive band was more prominent in the ischemic than in the contralateral side in three separate experiments. No such difference was seen in tissues representing the infarct core area. Similarly, no consistent differences were seen in other PKC subspecies in tissues representing either perifocal or infarct core areas.
Western blots of PKC subspecies in homogenates of tissues from the striatal core (A) and perifocal cortical (B) regions from the ischemic hemisphere (I) and of the corresponding region on the contralateral side (C). Arrowheadsindicate positions of PKCα, -β, -γ, -δ, and -ζ subspecies, and a small arrow points to PKCε, visible in the blots. The only clear change seen consistently in three separate experiments is the increase of PKCδ in the perifocal tissue from ischemic hemispheres (B, I, andC).
In control brains, immunocytochemistry showed a high density of strongly PKCδ-positive cell bodies in the thalamus and septum (not shown). In the frontoparietal and cingulate cortex of control animals and in the contralateral side of ischemic brains (Fig.7A), few immunolabeled neurons were observed. In addition, immunoreactive nerve fibers were seen throughout the brain and were stained most intensely in the striatum (Fig. 7E). Two days after 90 min of ischemia, an increased number of immunostained neurons were seen in the ipsilateral cortex around the infarcted core (Fig. 7B–D). In the perifocal striatum, the PKCδ-immunoreactive nerve fibers were stained more intensely than the fibers on the contralateral side (Fig. 7E,F). In the core, several immunoreactive cells were seen, especially around blood vessels, very likely representing endothelial cells (Fig.7G). Many macrophage-like cells stained with the PKCδ antibody were also seen in the infarcted core (Fig.7H).
PKCδ-immunoreactive cells in the rat brain 2 d after 90 min of ischemia. On the contralateral side of the frontoparietal (A) and cingulate (B,left) cortex, only a few immunoreactive neurons are seen, whereas in corresponding areas on the ipsilateral side (B, right; C), a large population of cortical neurons are immunoreactive. In cortical neurons, immunolabeling is extranuclear (D). In the contralateral striatum, numerous immunoreactive bundles of nerve fibers are seen (E). In the ipsilateral striatum (F), these nerve fibers have disappeared in the ischemic core (c), but have become strongly immunoreactive in the perifocal zone (p). In the infarct core, immunoreactive material is seen around small blood vessels, presumably in endothelial cells (G, curved arrows) and in glial-like cells (H). Scale bars: A–C, E, F, 250 μm; D, 25 μm; G, 125 μm; H, 50 μm.
At 3 d, double-staining experiments showed that several PKCδ-immunopositive cells in the infarcted area (Fig.8A,B) and in the thin perifocal rim were OX-42-positive (Fig. 8C,D). These cells were usually round and had only few if any processes (Fig. 8A,B), indicating that they represented invaded macrophages rather than activated microglia. Some PKCδ-immunoreactive structures seen around blood vessels were stained only faintly or not at all with the OX-42 antibody, supporting the conclusion that PKCδ is also expressed by endothelial cells in the infarct core. GFAP-immunoreactive cells were not seen in the infarct core. In the perifocal area, GFAP was not colocalized with PKCδ (Fig. 8E,F).
Immunofluorescence micrographs of focal (A, B) and perifocal (C–F) PKCδ immunoreactive cells double-stained with antiserum to complement C3 receptor, a marker for activated microglia and macrophages (A–D), and with an antiserum to GFAP (E, F), a marker for activated astrocytes, 3 d after 90 min of ischemia. Most of the PKCδ immunoreactive cells are macrophage-like cells with OX-42 immunoreactivity, round cell body, and few processes. GFAP-immunoreactive structures are not labeled with PKCδ antibody. Large arrows andarrowheads point to double-labeled macrophage-like cells. Small arrows show a small blood vessel surrounded by PKCδ-immmunoreactive cells and processes, some of which are also OX-42-immunoreactive. A field shown in E andF includes PKCδ-immunoreactive neuronal cell bodies (white stars) that are surrounded by GFAP-positive astrocyte processes (F). Scale bar (shown inA): 40 μm (A, B, E, F); 80 μm (C, D).
DISCUSSION
This study shows that from the multiple PKC subspecies expressed in the brain, transient focal brain ischemia specifically induces PKCδ mRNA and protein in neurons in the perifocal cortex and striatum, followed by long-lasting expression in macrophage-like cells and possibly in endothelial cells in the ischemic region. The early induction takes place in surviving neurons and involves the activation of NMDA receptors, suggesting that spreading depression (SD) accounts for the upregulation. Even though cortical SD is believed to exacerbate ischemic injury (Nedergaard and Astrup, 1986; Hansen and Nedergaard, 1988; Siesjö, 1991; Siesjö et al., 1995), recent studies have demonstrated that brief focal ischemia (Glazier et al., 1994) and SD cause cortical and hippocampal neurons to acquire tolerance to ischemia (Kawahara et al., 1993; Kobayashi et al., 1995). Altogether, the regulation of PKCδ subspecies differs remarkably from the other PKC subspecies examined and serves to adapt the energy-compromised perifocal tissue to altered requirements of protein phosphorylation during recovery from ischemic insult.
SD consists of repeated ionic transients involving release of K+ and uptake of Ca2+, Na+, and Cl−, and it is most likely initiated and propagated by massive presynaptic release of glutamate and activation of NMDA receptors subsequent to local brain injury, including focal brain ischemia (Kraig and Nicholson, 1978; Hansen, 1985; Nedergaard and Astrup, 1986; Hansen and Nedergaard, 1988; Marrannes et al., 1988). Because restoration of the ionic gradients consumes high-energy phosphates, SD may lower the threshold of neuronal death during and immediately after ischemia. The induction of PKCδ occurred several hours after the insult, indicating that PKCδ is not likely to have a causal role in immediate ischemic cell death. Li et al. (1995) reported apoptotic cell death to be an ongoing process from 30 min through 4 weeks after focal brain ischemia, peaking at 1–2 d, with a scattered distribution in the ischemic region. The temporal profile of apoptotic cell death covers somewhat the period of upregulated expression of PKCδ described in the present study; however, PKCδ expression peaked at 1 d in surviving perifocal neurons, stayed moderately high in a thin perifocal rim through 7 d, and was further increased in non-neuronal cells in the infarct core at 7 d, suggesting that cells expressing PKCδ included cells not going through apoptosis, or did not overlap with the dying cells at all. The immediate loss of PKCδ mRNA in the infarct core supports the hypothesis that induction of PKCδ gene does not contribute to the immediate ischemic neuronal death in focal brain ischemia.
At later time points, the neuronal PKCδ expression was restricted to the infarct core and a narrow zone surrounding it. This penumbral zone also shows upregulation of a number of other inducible genes, such as heat-shock proteins (Welsh et al., 1992; Kinouchi et al., 1993), amyloid precursor proteins (Pyykönen et al., 1995), cyclo-oxygenase-2 (our unpublished observations), heme oxygenase-1 (Koistinaho et al., 1996), many immediate early genes (An et al., 1993;Kinouchi et al., 1994a,b), and growth factors (Hsu et al., 1993; Iihara et al., 1994), indicating that cells at most immediate risk to die induce expression of a specific set of genes and their protein products. It is not known which proteins are involved in processes causing neuronal death and which proteins represent survival attempts by the neurons. Many of these cells are also expressed by non-neuronal cells. The expression of certain immediate early genes such as c-fos sometimes precedes programmed cell death (Schilling et al., 1993; Kasof et al., 1995), whereas it has been suggested that c-jun is a marker for surviving neurons (Sommer et al., 1995) but also necessary for apoptotic neuronal death (Estus et al., 1994; Schlingensiepen et al., 1994; Andersson et al., 1995; Ham et al., 1995). In certain non-neuronal cells, proteolytic activation of PKCδ has been associated with delayed cell death (Emoto et al., 1995). Because no proteolytic PKCδ fragments were detected 2 d after ischemia, the possibility that PKCδ induction is involved in delayed neuronal death after the insult is unlikely, but it cannot be excluded.
In the lateral part of the ipsilateral thalamus, which is not supplied by MCA, the expression of PKCδ mRNA declined 3 d after 90 min of ischemia, and 4 d later the expression was clearly below the control levels. The time course corresponds to the delayed degeneration of ipsilateral thalamic neurons, which is thought to be secondary to axonal damage in the cortical infarct (Iizuka et al., 1990). The reduction of PKCδ mRNA in the ipsilateral thalamus therefore is likely attributable to the degeneration of neurons expressing the gene.
Crumrine et al. (1992) reported a 50% reduction in PKC activity 6 hr after permanent focal brain ischemia, which is in agreement with many other studies measuring PKC activity after spinal cord (Kochar et al., 1989) or global brain ischemia (Crumrine et al., 1990; Louis et al., 1991; Domanska-Janik and Zalewska, 1992; Busto et al., 1994). Some studies have shown that after global ischemia there is a rapid translocation of cPKC subspecies, which is not associated with increased PKC activity (Cardell et al., 1991; Wieloch et al., 1991;Cardell and Wieloch, 1993). It has been suggested that the discrepancy can be explained by the activation of a PKC inhibitor or alternatively by a rapid downregulation of PKC during ischemia (Kochar et al., 1989;Cardell and Wieloch, 1993). It is also possible that total PKC activity decreases when specific subspecies, such as PKCδ in the present study, are increased concomitantly with a decrease in other subspecies. Interestingly, decreased expression of any PKC subspecies mRNA was not followed by an immediate reduction in the protein levels. The discrepancy may be attributable to the increased expression of PKC subspecies in endothelial (Krizbai et al., 1995) and other non-neuronal cells (Nishizuka, 1992; Gott et al., 1994) and to the fact that neuronal PKC proteins are not completely degraded within 2 d after transient focal ischemia.
There are several intracellular functions that could be influenced by the observed long-term induction of PKCδ in focal ischemia. In macrophages, PKCδ is the most abundant subspecies (Jun et al., 1994), and its expression may be relevant, for example, in the regulation of inducible nitric oxide synthase after ischemia-reperfusion injury. PKC also regulates neuron-specific nNOS (Maiese et al., 1993), but the wide distribution of the ischemia-induced expression of PKCδ seems not to be colocalized with scattered nNOS-expressing neurons (Dawson et al., 1991). In endothelial cells, PKC phosphorylation causes a 30% reduction in the activity of endothelial NOS (Hirata et al., 1995), which is localized also in hippocampal CA1 neurons (Dinerman et al., 1994). In addition to the functions in gene regulation, PKCδ expression may regulate specific glutamate receptor subtypes, release of excitatory or inhibitory neurotransmitters, or membrane-bound ionic pumps (Chen and Huang, 1991; Nishizuka and Tanaka, 1994; Basudev et al., 1995), all of which are evidently relevant for plastic changes after focal brain ischemia.
Footnotes
This study was supported by the Academy of Finland and the Swedish Medical Research Council (Grant 14V-11189). We thank Hannele Ylitie for expert technical assistance and Dr. Jarmo Laitinen for constructive criticism on this manuscript.
Correspondence should be addressed to Jari Koistinaho, A. I. Virtanen Institute, University of Kuopio, P.O. Box 1627, FIN-70211 Kuopio, Finland.
