INDEX
Expression of p 21 mRNA after transient focal cerebral ischemia

Hideaki Taki1, Masao Nakatsuka1, Atsuo Masago1, Akira Iwata1, Takahisa Fuse1, Makoto Nakanishi2, Kazuo Yamada1

1Department of Neurosurgery and 2Biochemistry, Nagoya City University Medical School, 1 Kawasumi, Mizuho-cho, Mizuho-ku, Nagoya 467-8601, Japan

Address for correspondence:
Hideaki Taki,
Department of Neurosurgery, Nagoya City University, Medical School,
1 Kawasumi, Mizuho-cho, Mizuho-ku, Nagoya 467-8601, Japan.
Phone (81) 52-853-8286.
Fax (81) 52-851-5541.

Summary
The p21 is a ubiquitous inhibitor of cyclin-dependent kinase. We examined p21 mRNA expression with in situ hybridization in a rat transient focal ischemia for 2h. Upregulation of p21mRNA expression started at 3h after recirculation, when hsp70 mRNA expression was upregulated in high level. By 12h, p21 mRNA expression reached to the maximum level at the cortex, caudoputamen, striatum and thalamus, where glyceroaldehyde 3-phosphate dehydrogenase (GAPDH) mRNA expression was suppressed because of ischemic stress. At 24h and 48h, expression of p21 mRNA persisted in some parts of these regions. The p21 protein was detected with Western blotting and found upregulation with peak at 24h. Expression of intrinsic p21 mRNA and protein in the ischemic tissue might affect cell cycle arrest and influence cell death or tissue recovery from ischemia.

Key words: p21, hsp70, GAPDH, transient focal ischemia, in situ hybridization

Introduction
The p21 is an integral component of cell cycle control1). The p21 inhibits cyclin-dependent kinase and stops cell cycle at G1 phase2). Induction of p21 is transcriptionally affected by wild-type p532-4). Forced expression of p21 gene suppressed growth of human brain, lung, and colon tumor cells in culture3). This growth inhibition occurs with phosphorylation of growth-regulatory proteins such as retinoblastoma tumor suppressor protein5). We have, however, recently shown that p21 is upregulated without expression of p53 during heat shock response of glioma cell line6).
In the cerebral ischemia, there remains controversial point whether p53 is upregulated 7,8), unchanged 9) or reduced10). We have previosly shown that p53 mRNA is not upregulated in this model (unpublished observation). Therefore, we planed present study to detect p21 mRNA and protein expression in a transient middle cerebral artery (MCA) occlusion model with in situ hybridization technique. We also studied extent of cerebral ischemia by measuring mRNA expression of heat shock protein 70 (hsp70) and glyceroaldehyde 3-phosphate dehydrogenase (GAPDH), and compared those result with expression pattern of p21 mRNA.

Materials and methods
Ischemia model
Focal ischemia was induced in the Wistar rats weighing 300-350gr by MCA occlusion using an intraluminal suture. Fifty-four rats were anesthetized with 0.5-2 % halothane and nitrous oxide/oxygen mixture (70:30). The brain temperature was monitored at the temporal muscle and maintained at 37¡î with a heating pad. The right common carotid artery was exposed, and the external carotid and pterygopalatine arteries were ligated. A 4-0 monofilament nylon suture coated with silicone rubber on the distal end was threaded into the internal carotid artery through the stump of the external carotid artery and brought up to the intracranial carotid bifurcation to occlude the right MCA.
After occlusion, anesthesia was discontinued and the rats were allowed to move freely. During ischemia, rats showed left hemiparesis with rotation to the left side and they were used for the experiment. Pulling out the suture after 2h of MCA occlusion made recirculation of the blood flow. The rats were randomly divided into 5 groups, and they were sacrificed at 3h, 6h, 12h, 24h and 48h after the occlusion (n=5 for each period). Another group of rats (n=5) had exposure of neck vessels but thread was not inserted, and they were used as a sham control. The animals were sacrificed by decapitation under deep anesthesia with halothane. The brains were removed, frozen, embedded in Tissue-tek O.C.T. compound (Miles Inc., U.S.A.), and stored at -80¡î. Twenty-micron-thick coronal sections were cut on a cryostat at -20¡î, thaw-mounted onto 3-aminopropyl triethoxy silan-coated slides, and processed for in situ hybridization. Another 14 rats were used for sampling of Western blots including sham control and 6h, 24h and 48h after recirculation (n=3-4 for each group).

Preparation of RNA probe
The partial cDNA fragment (500 base pairs)11) of mouse p21 was used for this study. The cDNA was linearized by digestion with restriction endonucleases of Bam H1 (antisense) or Eco R1 (sense). The linearized cDNA was incubated with a mixture of reagents (2 ml of transcription buffer (x5), 0.5 ml of 100 mM dithiothreitol, 0.5 ml of ribonuclease (RNase) inhibitor, 0.5 ml of 10 mM ATP, CTP, and GTP, 5 ml of [35S]-labeled UTP (Takara , Japan), 0.5 ml of DNA template (1 mg / ml) with 1ml of appropriate RNA polymerase (T3 RNA polymerase for antisense ; T7 RNA polymerase for sense) at 37¡î for 30 min. Then deoxyribonuclease (DNase) digestion was performed by adding 2 ml of DNase and incubating at 37¡î for 10 min. The incorporation rate of radioactivity was then estimated by counting radioactivity.

In situ hybridization using ribonucleotides
In situ hybridization for p21 and GAPDH was performed using ribonucleotide probes. Each ribonucleotide probe was labeled for in situ hybridization with [a-35S] dUTP. The specific activity of the labeled probes was 0.5-1.0 dpm/mg. In situ hybridization techniques for ribonucleotide probe were based on those of Wilkinson et al12) with some modification. Briefly, after being warmed to room temperature, slide-mounted sections were re-fixed with 4 % formaldehyde in 0.1 M sodium phosphate buffer (pH 7.2) for 30 min. After washing with 0.1 M phosphate buffer saline (PBS), the sections were treated with 10 mg / ml of proteinase K in 50 mM Tris-HCl and 5 mM ethylenediamine tetraacetic acid (EDTA) (pH 8.0) for 5 min at room temperature. They were postfixed in the same fixative as described above, acetylated with acetic anhydride in 0.1 M triethanolamine, rinsed with PBS, and dehydrated in an ascending alcohol series (70-100 %). The sections were subsequently defatted with chloroform for 6 min and immersed twice in 100 % ethanol for 3 min before being subjected to hydration.
The [35S]-labeled RNA probes (sense and antisense) in hybridization buffer were placed on sections and covered with siliconized coverslips. The hybridization was performed in a humid chamber overnight at 55¡î. The hybridization buffer consisted of 50% deionized formamide, 0.3 M NaCl, 20mM Tris-HCl, 5 mM EDTA, 10 mM phosphate buffer, 10% dextran sulfate, 1 x Denhardt's solution, 0.2% Sarcosyl, 500 mg/ml yeast t-RNA and 200mg/ml salmon sperm DNA (pH 8.0). The probe concentration was 5x105 cpm / 100ml per slide. After hybridization, slides were immersed in 5x sodium saline citrate solution (SSC) (pH 7.2; NaCl 35 g/l, sodium citrate 17.6 g/l) at 55¡î, and coverslips were allowed to fall off. The sections were incubated at 56¡î in 50% deionized formamide, 2x SSC for 30 min. After rinsing with RNase buffer [0.5 M NaCl , 10 mM Tris-HCl, 5 mM EDTA (pH8.0)] three times for 10 min each at 37¡î, the sections were treated with 1 mg/ml RAase A in RNase buffer for 10min at 37¡î. After an additional wash in RNase buffer, the slides were incubated at 56¡î in 50% formamide, 2x SSC for 30 min , rinsed 2x SSC and 0.1x SSC for 10 min each at room temperature, dehydrated in an ascending alcohol series, and air-dried. To obtain macroautoradiograms, X-ray films were exposed to uncoated sections for 4 or 5 days and films were developed.

In situ hybridization using oligonucleotides
In situ hybridization for hsp70 was performed using oligonucleotide. The probe sequence of hsp70 was as follows: rat hsp70 (30 mer) complementary to base numbers 539-568 of the cloned rat hsp70 cDNA13); 5'-CGATCTCCTTCATCTTGGT CAGCACCATGG-3'. The oligonucleotide probe of hsp70 was synthesized and purified by Rikaken Inc (Nagoya, Japan). The oligonucleotide probe was labeled for in situ hybridization with [a-35S] dATP (Amersham, UK) at 3' end using terminal deoxynucleotidyl transferase (TdT) (Takara, Japan). The specific activity of the labeled probe was 0.5-1.0 dpm/mg.
In situ hybridization techniques for oligonucleotide probe have been described elsewhere14). In brief, after being warmed to room temperature, slide-mounted sections were fixed with 4% formaldehyde in 0.1 M phosphate buffer (pH 7.2) for 30 min, rinsed three times (5 min each) in 4x SSC, and dehydrated through an graded ethanol series (70-100 %). The sections were subsequently defatted with chloroform for 6 min and immersed twice in 100 % ethanol for 3 min before hybridization. Hybridization was performed by incubating the sections with a buffer containing 4x SSC, 50 % deionized formamide, 0.12 M phosphate buffer (pH 7.2), 1x Denhardt's solution, 2.5 % tRNA, 10 % Sarkosyl, and [a-35S] dATP-labeled probes (0.5-1.0 x 106 dpm/slide, 200 ml/slide) for 16 h at 42¡î. After hybridization, the sections were rinsed in 1x SSC (pH 7.2) for 10 min at room temperature, followed by 4 rinses in 1x SSC at 60¡î for 15 min. The sections were dehydrated through a graded ethanol series (70-100 %).

Western blotting
Western blotting for p21 was performed with the same method as reported from our laboratory6). Protein concentration was measured using bicinchoninic acid protein assay regents (Pierce, IL). Ten-microgram protein which was extracted from the ischemic corticies and corresponding area of sham control was boiled in SDS sample buffer for 5min and loaded onto an SDS-PAGE (10% acrylamide). Following transfer to PVDF membranes (Millipore, Bedford, MA), p21protein was detected by incubation with anti-p21 antibody (Santa Cruz Biotechnology, CA) and enhanced with chemiluminescence methods (ECL, Amersham, IL).

Results
GAPDH mRNA expression
In sham operated animals, GAPDH mRNA was expressed ubiquitously in the cortex, caudoputamen, hippocampus, and thalamus (Figure 1A). The pattern of GAPDH mRNA depended on neuronal density. The pattern remained same at 3 h after reperfusion as compared to sham control (Figure 1B). At 6 h after reperfusion, GAPDH mRNA expression was reduced in the lesion-side cortex and lateral caudoputamen where ischemia developed (Figure 1C). Reduction of GAPDH mRNA expression in those areas lasted till 48 h after reperfusion (Figure1C-F). Furthermore, GAPDH mRNA expression reduced in the pyramidal layer of the hippocampus and ipsilateral thalamus 24h to 48h after reperfusion (Figure 1 E, F).

hsp70 mRNA expression
In sham-operated rats, in situ hybridization autoradiograms demonstrated low levels of hsp70 mRNA (Figure 2A). It began to increase at 3 h after reperfusion and increased markedly till 12h in the ipsilateral cortex and lateral caudoputamen where ischemia developed (Figure 2B-D). The expression also developed in the pyramidal layer and dentate gyrus of the hippocampus and in the ipsilateral thalamus (Figure 2B-D).
The core of ischemia became infarcted at 24h to 48h, and hsp70 mRNA expression disappeared (Figure 2E-F). In the periinfarcted zone, the gene expression lasted till 24h (Figure 2E). By 48h the gene expression disappeared completely in the infarcted side hemisphere, indicating cessation of stress response (Figure 2F).

p21 mRNA expression
Hybridization with antisense p21 probe only showed signals (Figure3), and no hybridization signals were detected with sense p21 probe (data not shown) indicating good sensitivity of the antisense probe. In sham controls, p21 mRNA expression was detected in low level (Figure 3A). Upregulation of p21 mRNA expression began at 3h after reperfusion in the ipsilateral cortex, lateral caudoputamen and CA1 to CA3 pyramidal layer of the hippocampus (Figure 3B). The p21mRNA expression became high level at 6h in the cerebral cortex, caudoputamen, and CA1 to CA3 pyramidal layers and dentate gyrus of the hippocampus (Figure 3C). By 12h, p21 mRNA expression became maximum level in the ischemic cortex, caudoputamen, striatum and thalamus (Figure 3D). The expression lasted in the same areas at 24h (Figure 3E) and it tended to disappear at 48h (Figure 3F)

p21 protein expression
Expression of p21 protein was studied in the same model with Western blotting. In sham-operated rats, expression of p21 protein was not detected (Figure 4B). Increase of p21 protein in the ischemic cortex was noted at 6h after occlusion (Figure 4C), and the peak level of p21 protein was detected at 24h (Fig. 4D). The p21 protein reduced again at 48h (Figure 4E).

Discussion
Mammalian cells respond to environmental stresses such as temperature, ultra violet, anoxia or ischemia. This response is characterized by transient expression of stress or heat shock protein15). In relation to the stress response, cell cycle arrest occurs as a function of cyclin-dependent kinase inhibitors. The p21 is one of the upregulated gene cascades originating from p53 and acts as a cell-cycle regulator. The p21 inhibits cyclin-dependent protein kinase causing G1 arrest of cell cycle16). Furthermore, p21 was directly upregulated without expression of p53 in the heat treatment of glioma cell line, suggesting different signal cascade other than p536).
In the preliminary experiment, we could not detect induction of p53 mRNA in the same ischemia model as was used in this study (unpublished observation). Therefore, we tested whether p21 is upregulated in the mammalian brain tissue by ischemic stress. Indeed, we clearly demonstrated that brain ischemia induced p21 mRNA expression in the ischemic and periischemic tissue. The expression reached maximum level at 12 hours after ischemia in the cerebral cortex, caudoputamen and hippocampus of the ischemic side. The p21 protein expression was delayed for another 12 hours making peak level at 24h after recirculation. This might be the result of delayed translational process due to ischemia. The fact that p21 mRNA was expressed without p53 mRNA expression further support our previous view6) that p53 independent pathway is present in the cerebral ischemia.
Cerebral ischemia is a major stimulant for upregulation of heat shock genes, among which hsp70 is a major responsive gene and stress answering protein17). In a model of transient focal ischemia, Kinouchi et al.18) have reported that hsp70 mRNA was induced in the ischemic tissue and regions remote from infarction such as thalamus, hippocampus and subatantia nigra. This is confirmed by our previous report19) and reconfirmed in the present report. Our study suggests an interesting temporal relationship between expression of hsp70 mRNA and p21 mRNA. Increase in hsp70 mRNA appeared at 3h and reached maximum level at 6h post-reperfusion in the ipsilateral cortex, lateral caudoputamen, hippocampus and ipsilateral thalamus, where we found maximal p21 mRNA expression at 12h post-reperfusion. The p21 mRNA expression remained apparent in these regions 24h and 48h post-reperfusion, when hsp70 expression disappeared. These relationship clearly demonstrated that p21 is upregulated in the same tissue as ischemic stress caused upregulation of hsp70 mRNA. Though we could not show direct evidence that hsp70 triggers p21 mRNA expression, there may be close relationship between hsp70 expression and p21 expression.

Conclusion
We demonstrated expression of p21 mRNA and protein in the focal brain ischemia model with in situ hybridization and Western blot analysis. By 12h after focal ischemia and reperfusion, p21 mRNA was markedly induced in the ischemic cortex, lateral striatum, hippocampus and thalamus, and it lasted till 24h. This p21 expression occurred in a p53-independent manner. The hsp70 mRNA expression occurred in close relation to p21 expression though no direct evidence was identified.

Acknowledgement
This work is supported by Health Sciences Research Grants directed to Research on Brain Science from Ministry of Health and Welfare.



Nagoya City University
Medical School