Molecular and Cellular Neuroscience 22 (2003) 353–364
www.elsevier.com/locate/ymcne
Temporal and spatial gene expression patterns after experimental stroke in a rat model and characterization of PC4, a potential regulator of transcription Adrian Roth,a,1 Ramanjit Gill,b and Ulrich Certaa,* a
Roche Center for Medical Genomics, F. Hoffmann–La Roche Ltd., Building 93/610, CH-4070 Basel, Switzerland b Preclinical Research (PRBD-N), F. Hoffmann–La Roche Ltd., CH-4070 Basel, Switzerland Received 10 July 2002; revised 5 October 2002; accepted 19 November 2002
Abstract We have used the middle cerebral artery occlusion model in the rat in combination with microarray transcript imaging to study changes in gene activity after ischemic stroke. We analyze transcriptional changes in three regions of the affected, ipsilateral brain sphere using contralateral tissues from the same animal as a control over several time points in 180 individual RNA samples. After 1 h transcription factors and signaling molecules are expressed in all tissues followed by the induction of tissue repair-related genes in the cortices which undergo regeneration. Some of these genes are turned on by PC4, which is upregulated in tissues surrounding the infarct core. Interestingly, PC4 is a nerve growth factor (NGF)-inducible gene and has been associated in earlier studies with neuronal growth processes. The expression mode of PC4, the cellular localization of the gene product, and the functional properties of downstream genes induced in vivo and in vitro using transgenic cell lines suggest that PC4 is a regulator of transcription involved in tissue regeneration after ischemic stroke. The novel experimental strategy applied here is suited to provide insight into the molecular mechanisms underlying stroke and tissue regeneration and may enable the discovery of preventive medicines. © 2003 Elsevier Science (USA). All rights reserved.
Introduction The brain is unable to store energy and any interruption in the blood supply rapidly causes neuronal dysfunction and damage. Ischemic damage can be caused by cardiac arrest or by a clogged artery in the brain that ultimately results in stroke, one of the major causes of death in industrial countries. Interruption of the blood flow results in severe tissue damage in the so-called ischemic core, whereas the tissue surrounding this area is able to maintain certain functions such as ionic homeostasis. This tissue referred, to as the “penumbra,” can be salvaged by recirculation of blood or pharmacological intervention (Gill et al., 1995). The success of treatment in humans is poor so far due to dose-
* Corresponding author. Fax: ⫹41-61-688-1448. E-mail address:
[email protected] (U. Certa). 1 Present address: Biozentrum, Division of Pharmacology/Neurobiology, University of Basel, CH-4056 Basel, Switzerland.
limiting side effects and difficult, long-lasting clinical trials. Some limited knowledge about the molecular mechanism involved in tissue regeneration has been gained from animal model experiments. In this study we used the permanent rat middle artery occlusion (MCAO) model, which is well characterized in terms of the pathological changes, development of infarct assessed using magnetic resonance imaging, changes in blood flow, and neuroprotective strategies. The model replicates, in many aspects, the neuropathological changes seen following stroke in humans. In addition, only the ipsilateral side of the brain is affected by the ischemic damage; the contralateral side remains mostly unaffected, allowing collection of experimental and reference tissue from the same animal which eliminates potential artifacts caused, for instance, by animal handling. The striatum and the parietal cortices are tissues most heavily affected by ischemia, whereas the frontal cortex shows virtually no signs of damage after 3 h (Gill et al., 1995). This model was recently
1044-7431/03/$ – see front matter © 2003 Elsevier Science (USA). All rights reserved. doi:10.1016/S1044-7431(02)00039-8
354
A. Roth et al. / Molecular and Cellular Neuroscience 22 (2003) 353–364
used in our laboratory to assess global gene expression after ischemia using oligonucleotide DNA microarrays with the capacity to detect transcript levels of an unfocused, random set of about 750 genes derived from a variety of rat tissues (Soriano et al., 2000). This technology replicated most results on differential gene expression after focal ischemia obtained with conventional, nonparallel experimental approaches (Gill et al., 1991, 1992, 1995; Kogure and Kato, 1993; Macrae, 1992). A number of genes induced after an ischemic insult are involved in neuroprotection, remodeling, and neuron regeneration, which led to the currently widely accepted view of an active repair mechanism of damaged tissue. (Kogure and Kato, 1993; Koistinaho and Hokfelt, 1997; Soriano et al., 2000). Characterization and identification of the molecular signal that triggers this cascade of neuron regeneration and repair events may open the door to the development of a novel generation of drugs with high selectivity for cerebral ischemia. The continuous improvement of microarray technologies allowed us to measure the expression of about 13,000 genes in parallel as compared with 750 in our previous study (Soriano et al., 2000). The DNA array used here represents about 13,000 genes with 40 oligonucleotide probe pairs distributed on two chips. We selected all known genes and ESTs from the 1998 version of the EMBL database for the chip design. Tissue repair after ischemia becomes evident about 24 h after the initial damage in the MCAO model (Gill et al., 1995). The expression of potential activator and regulatory molecules that induce transcription of genes involved in repair must occur shortly after damage. To identify this class of activators we performed DNA microarray-based gene expression analysis at several time points, namely 1, 3, 6, and 24 h after artery occlusion. We focused on genes that are induced early and display properties of molecules that induce downstream events possibly important for repair and regeneration. This strategy led to the identification of known transcription factors but also to a polypeptide termed PC4. The gene has a low expression level typical of transcription factors and has the highest gene activity 3 to 6 h after ischemia. We used stable cell lines in which a tetracycline-regulatable promoter controls PC4 expression in combination with microarray analysis to identify genes downstream of PC4. This novel strategy identified about 20 genes expressed in the MCAO model and the transgenic cell lines.
Results Global transcriptional activity after MCAO Focal cerebral ischemia was induced in rats by occlusion of the middle cerebral artery resulting in apparent tissue damage of the striatum and the parietal cortex 3 h after the insult. Consistent with previous studies the frontal cortex area did not show any significant damage (Gill et al., 1995).
We established mRNA expression profiles of the damaged brain tissues (striatum, frontal cortex, parietal cortex) 1, 3, 6, and 24 h after the insult and used the corresponding samples from the nonischemic control hemisphere of the brain to measure the normal mRNA abundance of the modulated genes in each tissue at each time point. For each time point five animals were used and samples processed individually. In general, transcripts of about 40% of the genes represented on the arrays were detected independent of the tissue source or of time elapsed after damage (data not shown). A total of 159 genes are induced and 118 repressed when all time points and tissues are included in the analysis (Fig. 1A). Only seven genes are concurrently upregulated in all three areas: Three hours after MCAO the immediate early genes pJunB, krox24, and ngfi-b and the stress and inflammation responding genes pc3, cl100 (a MAP kinase phosphatase), and gadd45 gamma (growth arrest and DNA damage-inducible gene 45) are simultaneously induced, followed by the insulin-induced growth-response protein cl-6 soon after after the insult. Probably due to severe tissue damage, the number of differentially expressed genes is lowest in the striatum (Fig. 1A) because the mRNA is degraded in the core zone of the lesion. The number of differentially expressed genes steadily increases with time after MCAO, supporting the concept of an active tissue repair mechanism after experimental stroke. In addition, in the cortices about 25% of the induced genes encode either transcription or growth factors (data not shown). Identification of candidate genes mediating neuroprotection The transcriptional changes observed and described above are consistent with a model in which a small number of early regulator molecules trigger execution of a repair mechanism active 24 h after damage. The main goal was therefore to identify such genes based on their kinetics of expression, the expression level itself, and also functional properties. Pharmacological activation of such genes or their products may thus activate the repair mechanism without ischemic insult. The frontal cortex appears to be the appropriate source for such a gene because this tissue shows almost no direct damage 24 h after MCAO and seems to be most resistant to the late-stage damage induced by spreading depression and delayed neuronal cell death. Even though almost no obvious signs of damage can be observed at the time points considered (0 –24 h after insult) frontal cortex clearly seems to be affected by the insult given the large number of differentially regulated genes in this area after induction of MCAO (Fig. 1). But whereas the growing infarct induces strong damage in the striatum and the parietal cortex, in the frontal cortex this could not be observed (Gill et al., 1995). We therefore suggest that this area succeeds best in resisting and/or recovering from the insult and does so by inducing specific programs. For the identification
A. Roth et al. / Molecular and Cellular Neuroscience 22 (2003) 353–364 355
Fig. 1. Gene cluster analysis after ischemia. (A) Differentially expressed genes were clustered according to fold change across the four time points in the three affected tissues. The numbers of induced and repressed genes are balanced and the most significant changes occur 24 h after damage, while only minor changes in expression occur 1 h after the insult. Shown is one gene per line; red indicates upregulation, and blue, downregulation. Changes are shown in color increments ranging from 2-fold (lightest) to 5-fold or greater (most intense) change. RNA degradation in the 24-h striatum sample disabled mRNA profiling. (B) Chart representation of typical gene clusters (A–G). Genes in clusters A and B show the highest induction at around 3 h after ischemic damage and were therefore considered the most promising sets for the identification of key regulator molecules. PC4 was selected for further characterization based on its expression profile, the expression level (lower left corner), and the known biological function.
356
A. Roth et al. / Molecular and Cellular Neuroscience 22 (2003) 353–364
of relevant candidate genes in this tissue the upregulated genes were hierarchically clustered yielding seven groups with distinct temporal mRNA expression profiles (Fig. 1C). Genes in clusters A and B contain potential candidates because their peak of expression is 3 h after MCAO, with change factors up to 60 in A and 10 in B (Fig. 1C). Genes in clusters C, D, E, F, and G are activated at later time points and therefore not considered as candidates encoding regulator proteins. Genes in clusters A and B encode, consistent with our working hypothesis, transcriptional activators (krox24, krox20, ngfi-b, c-fos), early inflammatory markers (hsp40, pc3, MAP kinase phosphatase cl100), and the seizure- and neuronal plasticity-related protein inhibin beta-A or the nuclear membrane protein emerin. The expression kinetics of genes in cluster B is similar to that of genes in A except that the induction and expression levels are significantly lower. The genes in clusters C–G are functionally related to structure and adhesion (neuron-specific gene family member p21, tropomyosin tm-4, the cytoplasmic adapter protein pacsin2) or protein or peptide metabolism (␣ subunit of tcp-1, release factor eRF3, the tyrosine phosphatase prl-1, tissue plasminogen activator, the pseudogene b of the potential radical scavenger metallothionein (mt-1 b)). One gene in cluster B termed PC4 has the expected properties of a potential control protein: it shares sequence homology with interferons and is inducible by NGF, a molecule induced early after MCAO and known to be involved in regeneration and repair processes after damage (Sofroniew et al., 2001). Moreover, PC4 was associated with nerve growth and might play a role after tissue injury (Guardavaccaro et al., 1995; Iacopetti et al., 1996; Nelson et al., 2002; Rubin et al., 1998; Tirone and Shooter, 1989). We thus decided to focus on PC4 to elucidate its role in ischemic stroke. Induction of PC4 in the MCAO model We first expressed the central portion of PC4 as a tagged fusion protein in Escherichia coli and used the gel-purified protein to raise polyclonal antibodies in rabbits. The immune serum detected an NGF-inducible protein band of the expected size (⬃54 kDa) in PC12 cell extracts which was not recognized by preimmune serum (data not shown). This allowed us to assess PC4 protein levels by immunoblotting in the experimental animals. Consistent with microarray data, low levels of PC4 are detectable in the undamaged brain halves of two experimental rats and in the shamoperated control (6-h time point) (Fig. 2A, panels 1 and 3). An estimated four- to fivefold induction of PC4 is evident after MCAO in the affected frontal cortex. In one animal (lane 2), elevated levels of PC4 are detected in the contralateral brain half probably due to earlier induction or expression signal spreading. Nevertheless, PC4 is strongly upregulated in this MCAO-treated animal compared with the sham-operated animal.
Fig. 2. Detection of PC4 protein and mRNA in rat brain samples. (A) Western blot analysis of brain sample lysates using polyclonal rabbit serum against recombinant PC4. “Sham” refers to a sham-operated animal on which artery occlusion was not performed. 1, 2, and 3 are frontal cortex samples from three animals taken 6 h after MCAO. Low levels of PC4 are detectable in the control and contralateral brain hemisphere (c) and a marked induction is evident in the affected, ipsilateral brain hemisphere (i). Elevated levels of PC4 are detected in the contralateral site of animal 2 probably due to signal spreading. Also, this result highlights the importance of experimental repeats. (B) Detection of PC4 mRNA by in situ hybridization. Brains of MCAO-treated (right) and sham-operated animals were microdissected and the sample immobilized on glass slides followed by hybridization using radiolabeled 60-mer oligonucleotides. Signals were revealed by autoradiography after 4 weeks of exposure.
Spatial induction of PC4 mRNA in the rat brain after MCAO The sensitivity of the PC4 rabbit serum was insufficient for reliable detection of the protein by immunofluorescence in brain sections (data not shown). This is not surprising based on the low level of mRNA detected by the microar-
A. Roth et al. / Molecular and Cellular Neuroscience 22 (2003) 353–364
rays. For this we revealed the spatial distribution of PC4 expression by the more sensitive in situ hybridization technique using radiolabeled 60-mer oligonucleotides (Fig. 2B). Consistent with Western blot data elevated levels of mRNA are specifically detected in the affected brain half, using coronal sections. The strongest signals come from the frontal and parietal cortex and the nucleus accumbens, which surround the infarct zone. A moderate signal reduction is evident 24 h after the insult, which is in agreement with the microarray quantification data for PC4 mRNA (Fig. 1B). Cellular localization of PC4 It is known that PC4 is inducible in neuronal cell lines, leading to changes in neurite outgrowth and morphology (Guardavaccaro et al., 1995; Iacopetti et al., 1996; Tirone and Shooter, 1989). We thus compared the reactivity of PC4 rabbit serum with that of either NGF-treated or nontreated neuronal PC12 cells (Fig. 3A, IM). Preimmune serum served as background control (Fig. 3A, PI). In nontreated cells, only low levels are detectable but a marked signal increase occurs in treated cells together with the expected change in morphology. Notably, a significant portion of PC4 is localized in the newly outgrowing neurites. As a control, we treated rat kidney NRK cells in which the pathway leading to the activation of PC4 expression is obviously missing. Independent of NGF treatment, relatively low levels of PC4 are detectable in the cytoplasm and no apparent change in shape is evident in this nonneuronal cell type. The mainly cytoplasmic localization of PC4 was confirmed by probing subcellular fractions of NGF-induced PC12 cells that were generated by differential centrifugation. Consistent with immune fluorescence, the antigen is detected predominantly in the cytosolic fraction (Fig. 3C, lane 4). Despite significant sequence homology, the cytosolic localization of PC4 diminishes the possibility that it is functionally related to the secreted interferons. The cellular localization of PC4 makes it unlikely that this polypeptide is a classic transcriptional factor and favors the view that it might participate in a cytoplasmic signaling cascade regulating downstream gene activity. Microarray-based identification of PC4 target genes in transgenic cell lines To identify PC4-inducible genes, we expressed the coding region in a kidney cell line and a neuronal cell line under control of a regulatable promoter followed by differential microarray analysis. This combinatorial approach should identify PC4 target genes, which are modulated independent of the cell type and the microenvironment. Comparing the MCAO rat brain and cell line expression data then identifies genes directly regulated by PC4. The cDNA was cloned behind a tetracycline-responsive promoter using a
357
variant of the original system (Gossen et al., 1993; Gossen and Bujard 1992; Schultze et al., 1996) where all genetic components required for regulated gene expression are combined on a single plasmid. We transfected neuronal PC12 and NRK kidney cells followed by zeomycin selection in the presence of tetracycline, which represses PC4 expression and potential toxic effects. About 20 clones were analyzed by Western blots and the most tetracycline-responsive clone from each line was selected for the subsequent microarray analysis (Fig. 4). We compared induced cells (no tetracycline) with repressed cells and selected genes with a onefold change and an average difference expression level greater than 100 units. This low-stringency filter criteria selected from each line more than 1000 inducible genes. PC4 has 97% identity in 64 overlaps on the level of the amino acid sequence to mouse -interferons and rat ␥-interferon (Dijkema et al., 1985; Tirone and Shooter, 1989). It is therefore surprising that none of the known interferon-inducible genes present on the chips was induced by PC4 overexpression (data not shown). This outcome is consistent with the cellular localization of PC4 (Fig. 3) and lack of a secretion signal in the primary sequence. Furthermore all genes modulated in both induced cell lines are upregulated which supports the view that PC4 is a transcriptional regulator. The expression of 1060 genes in NRK cells and 1348 genes in PC12 is cell line specific and activation may depend on certain cell type-specific cofactors, whereas a total of 399 genes are induced in both cell lines. Notably, overexpression of PC4 in the NRK rat kidney cell line induces six genes, which are strictly neurospecific under natural conditions (data not shown). Examples include Munc13-1, a presynaptic phorbol ester receptor and essential protein for synaptic vesicle maturation (Augustin et al., 1999; Betz et al., 1998), and RHO-1, a subunit of the GABA-A receptor (Cutting et al., 1991). To identify genes modulated in both experimental setups, we selected the set of genes induced in the frontal cortex of the MCAO model that were silent at early time points followed by marginal activation after 6 h and maximum activation 24 h after ischemia. The number of differentially expressed genes reaches a maximum in the damaged brain hemispheres at this late time point (data not shown). This gene selection criterion yielded a total of 578 genes from the MCAO experiment. The majority encode proteins involved in repair processes such as restructuring of the cytoskeleton or the extracellular matrix (data not shown). When this gene collection was merged with the 399 induced transcripts upregulated in both transgenic cell lines, 20 induced genes were identified that are probably the direct downstream targets activated by PC4 (Table 1). They belong to two major functional categories: One group contains genes involved in transcription and translation (carm1, p34, chd4, eif2-␣, eif3-p110, eif6) and in protein processing or trafficking (e.g., calm, ribophorin). The other group consists of genes encoding cytoskeleton components or cell adhesion molecules such as abp-280, transgelin-2, and beta cat-like.
358
A. Roth et al. / Molecular and Cellular Neuroscience 22 (2003) 353–364
Fig. 3. Detection and cellular localization of PC4 by immunofluorescence. (A) Induction of PC4 by NGF in rat PC12 cells. Traces of PC4 are detected before NGF stimulation using rabbit serum against recombinant PC4 (see Experimental Methods). After incubation the signal intensity increases and the cells undergo morphological changes like the outgrowth of PC4-positive neurites. The left panel shows reactivity of rabbit preimmune serum. (B) Control experiment using rat kidney NRK cells in which no obvious induction of PC4 is evident. (C) Cellular localization of PC4. NGF-stimulated PC12 cells were lysed and fractionated by standard differential centrifugation. Lane 1: nuclei and intact cells. Lane 2: heavy membrane and vesicle fraction. Lane 3: light membrane fraction. Lane 4: cytosolic fraction. Lane 5: whole-cell lysate.
A. Roth et al. / Molecular and Cellular Neuroscience 22 (2003) 353–364
359
Discussion
Fig. 4. Regulated expression of PC4 in PC12 (top) and NRK (bottom) cells. Expression of PC4 in the cell culture was repressed by tetracycline or induced by removal of tetracycline from the medium. After 24 h cells were harvested and protein lysates analyzed on Western blots. For PC12 cells clone 2 and for NRK cells clone 4 were selected for further experiments because these clones displayed the best repression and induction levels.
In summary, regulated overexpression of PC4 in two cell lines derived from kidney or neuronal tissue results in the induction of genes that are also induced after ischemia.
Focal cerebral ischemia was induced in rats by occlusion of the middle cerebral artery using the well-established MCAO model, which is in many aspects similar to stroke in humans (Dietrich, 1998; Gill et al., 1991, 1992, 1995; Tamura et al., 1981). In this model, differential expression of multiple genes after ischemia has been shown by us and others using independent experimental strategies, like realtime PCR, low-density microarrays, and more recently SAGE (Bates et al., 2001; Koistinaho and Hokfelt, 1997; Soriano et al., 2000). The detection of known stroke-inducible genes such as the NGF-inducible proteins (NGFI-A, -B, -C), heat shock proteins, or c-fos has validated recent technologies such as DNA microarrays (Soriano et al., 2000). An inherent limit of these multiparallel approaches is the fact that only a relative small proportion of the results can be translated into gene function. Here we apply the DNA microarray technology, established in a pilot study, to analyze the spatial and temporal expression patterns of genes modulated as a result of ischemic insult. The genes identified in the pilot study of Soriano et al. (2000) compare well with those in the presented study in the case of genes that are on both arrays. For example, the eight genes described in that report in more detail were identified in this study again (c-fos, NGFI-A, NGFI-C, Krox-20, Inhibin-beta-A), whereas Arc, MKP-1,
Table 1 PC4 inducible genes in the MCAO rat model and PC4 transgenic neuronal cell lines Gene product
Accession no.a
Descriptionb
Functionc
Carm1 Chd4 Nucleolin eif2-␣ eif3-p110 eif6 4F2hc e-septin calm srpr Ribophorin II b-cat like coronin-like p-41 Arc ab280 transgelin 2 enpl nmmhc-a unknown
AF117887 Q14839 M55022 J02646 U46025 AF047046 AB015433 AF180526 AF041373 BC001162 X55298 AJ301634 AJ006064 AF083269 X53416 BC002616 S69315 U31463
Protein arginine methyltransferase Chromodomain helicase DNA-Binding protein 4 Nucleolin protein c23 Translation initiation factor ␣ subunit Translation initiation factor p110 subunit Translation initiation factor 6 4f2 cell surface antigen heavy chain Cytoskeletal GTPase protein E-septin Clathrin assembly protein (short form) Signal recognition particle receptor ␣ subunit Transmembrane Glycoprotein 65-kDa subunit precursor -Catenin-like protein Homolog of actin-binding protein coronin 1b Arp2/3 family member Endothelial actin-binding protein sm22-␣ homolog, actin-binding protein Endoplasmin precursor Nonmuscle myosin heavy chain A [No significant homology in Swissprot/EMBL]
Transcription coactivation Transcription activation Transcription Protein synthesis Protein synthesis Protein metabolism Amino acid transport Cytokinesis, trafficking Membrane trafficking ER membrane protein ER membrane protein wnt signaling Cytokinesis, signaling Actin polymerization Branching of actin Development Stress response Stress response, structure
Note. PC4 was stably expressed in rat cell lines NRK and PC12 using a tetracycline-regulatable promoter and G418 as a selectable marker. Cultures were grown in the presence (uninduced) or absence (induced) of tetracycline for 48 h. The cells were harvested and lysed, and isolated total RNA was processed for microarray analysis using Affymetrix-EMBL 11a/12a arrays. Compared with the uninduced cultures, 1368 genes were regulated in PC12 cells and 1006 in NRK cells. Of the 399 genes modulated in both transgenic lines, 20 were also detected in the frontal cortex of the MCAO animals. a GenBank sequence accession number. b Gene names and descriptions due to database search (Swissprot and EMBL). Sequences on the array with homology ⬍80% to any known sequence of the two databases are indicated as unknown. c Functions of the genes were assessed based on information supplied in the databases and from the literature.
360
A. Roth et al. / Molecular and Cellular Neuroscience 22 (2003) 353–364
Fig. 5. Summary of the transcriptional responses in the rat brain after experimental ischemia. The model summarizes experimental results from the current study and the literature. The insult induces expression of immediate early genes (IEGs) in damaged areas of the damaged brain sphere. In the heavily damaged striatum, genes related to proteolysis and necrosis are expressed, followed by the breakdown of metabolism and ultimately cell death (see also Fig. 1A, striatum panel, 24 h). In the cortices genes typical of growth promotion and inflammation are induced, leading 24 h after damage to the induction of genes that are functionally related to tissue remodeling.
and MKP-3 are not present on the array used here and therefore could not be identified again. As outlined above and due to limited throughput, it is not possible to confirm the differential gene expression results detected on chips with more established techniques such as Northern blotting or RT-PCR. We thus reduced the problem of false positive or negative results by including five experimental replicates per time point and brain area, which gave rise to a total of 180 individual samples to be processed. Only genes with statistically significant average difference values (P ⬍ 0.03) were chosen for further selection based on kinetics, change factor, and brain area. The values for average differences for each brain region between the contralateral and ipsilateral sides were compared using an analysis of variance (ANOVA) test with repeated measures followed by a post hoc contrast test that corrects for multiple comparisons (BMDP Statistical Software, Cork, Ireland). Application of these stringent filter criteria to the entire data set resulted in 277 differentially genes (159 induced and 118 repressed) modulated after the artery occlusion stimulus. This subset included a number of established stroke genes which validates our statistical approach and filter criteria. The total number of transcripts detected by the probe sets on the microarrays was 40% independent of the tissue source and time point with the exception of the striatum, where tissue damage is apparent 24 h after the insult. This eliminates major RNA degradation in the samples and shows that only a relatively small percentage of genes are differentially expressed in the MCAO model using these statistical filters. The 277 differentially expressed genes fall into 11 major
functional categories together with uncharacterized ESTs based on database and literature surveys. For example, genes encoding structural and inflammatory proteins or adhesion molecules are induced virtually only in the cortices, whereas genes that have been linked to cell death or proteolysis are encountered in the more severely damaged striatum. Early after induction of damage, transcription factors like the NGF-inducible genes and heat shock proteins are induced in all areas. These factors then seem to activate specific downstream programs in each tissue depending on the level of damage. A purely random activation or repression of gene expression is therefore unlikely based on these fundamental observations. The gene expression data rather suggest execution of a controlled repair program induced by MCAO (Fig. 5). PC4 selection One goal of our study was the identification of regulatory molecules that induce the genes involved in neuroprotection or repair. We selected candidate genes based on their spatial and temporal expression after ischemic stroke in the rat brain together with a functional approach using transgenic, regulatable cell lines expressing PC4. This candidate gene is coexpressed in affected brain tissues with other transcription factors and regulators 3 and 6 h after the insult. The relatively low transcript levels and the cytoplasmic localization are typical of regulatory molecules of signaling pathways such as the JAK/STAT cascade activated by interferons (Imada and Leonard, 2000; Leonard, 2001). In this connection it is important to note that PC4 does not activate
A. Roth et al. / Molecular and Cellular Neuroscience 22 (2003) 353–364
any known interferon-inducible genes despite significant sequence homology (Tirone and Shooter, 1989). After NGF stimulation of PC12 cells neurite outgrowth occurs and the protein is detected by immunofluorescence abundantly in the axons (Fig. 3). Under natural conditions, this might facilitate cell-to-cell communication and spread of a repair signal throughout the damaged tissue (Ebadi et al., 1997; Raivich et al., 1999; Sofroniew et al., 2001). The output of signaling pathways such as JAK/STAT and ERK (reviewed by Imada and Leonard, 2000; Leonard, 2001) is the activation of downstream genes, which execute the signal triggered by various stimuli. It was therefore important to show that PC4 expression results in gene activation. Due to experimental limitations this cannot been shown in the rat in vivo using genetic approaches. We thus applied a novel microarray-based approach in which PC4 is overexpressed in a cell line using a tetracycline-regulatable promoter. Comparing the global gene transcription patterns of uninduced and induced cultures identifies potential downstream genes. The use of two independent lines (NRK and PC12) allows elimination of cell type-specific genes or genes modulated by unspecific transcriptional activation of the transactivator used for regulated gene expression. This approach demonstrated clearly that PC4 indeed seems to activate transcription of about 300 genes induced in both lines. Interestingly, only a handful of genes were downregulated. Twenty genes identified in both cell lines were also expressed in the frontal cortex 24 h after ischemia. The PC4-inducible genes fall into two major functional categories: The first group comprises genes involved mainly in transcription and translation (Table 1). carm1 (coactivator-associated arginine methyltransferase 1) operates as a secondary coactivator through its association with p160 family members of coactivators which mediate transcriptional activation by nuclear hormone receptors by methylation of histone h3 (Chen et al., 1999). Also, chd4 has properties of a transcriptional activator (Seelig et al., 1995). Several studies documented an inhibition of protein synthesis after cerebral ischemia as a result of phosphorylation of eif2-␣ by the kinase perk (Alirezaei et al., 2001; Bergstedt et al., 1993; Hu and Wieloch, 1993; Kumar and Wu, 1995). The co-induction of three members of a family of eukaryotic protein synthesis initiation factors, eif2-␣, eif6, and the p110 subunit of eif3, would point to a reversal of this inhibition resulting in de novo protein synthesis after ischemia. eif2-␣, for example, participates in the very early steps of protein synthesis by forming a ternary complex with GTP and initiator tRNA necessary for binding to the 40S ribosomal subunit (Kimball, 1999). Upregulation of the protein synthesis marker nucleolin, essential for pre-RNA transcription and ribosome assembly, is consistent with this view (Bourbon and Amalric, 1990). The second group of inducible genes is involved in protein metabolism and vesicle trafficking. calm (or ap180), for instance, is a clathrin assembly protein expressed in synapses that plays a role in intracellular trafficking of protein and lipid (Kim and Kim,
361
2000, 2001; Yao et al., 2000). 42fhc (also termed cd98) activates amino acid transporters (Kanai et al., 1998) and interacts with integrins (Fenczik et al., 2001). Another example of a molecule involved in cytogenesis is E-septin, which targets proteins to specific intracellular sites and participates in vesicle trafficking and morphogenesis (Fung and Scheller, 1999; Kartmann and Roth, 2001). Among the remaining genes Beta-catenin-like is worth mentioning, because it encodes a calcium-dependent cell adhesion molecule that operates at the subcellular level by organizing multicellular structures as part of the wnt-signaling pathway involved in cell proliferation, differentiation, and survival (Blankesteijn et al., 2000; Byers et al., 1994; Hirano et al., 1992; Satoh and Kuroda, 2000). The functional properties of the PC4-inducible genes reviewed above and those of the early response genes allow suggestion of a broad model describing the transcriptional events after MCAO in the three areas of the rat brain examined here (Fig. 5). The first event triggered by blood flow interruption is a global signal that induces the activation of IEGs. This is followed after 6 h by tissue-specific induction of other transcription factors and signaling molecules and is likely a response to IEG induction. The final outcome of this transcriptional program is evident after 24 h: tissue structures including RNA are degraded in the most severely damaged tissue, the striatum (see Fig. 1). In parietal cortex and especially in the frontal cortex, transcriptional events pointing to tissue remodeling occur, resulting in reduced damage as well as partial reversal of the damage, which is also evident at the macroscopic level (Gill et al., 1995). A role for PC4 as a generic regulator of injury processes has been suggested in a model of cardiac ischemia–reperfusion (Nelson et al., 2002) where the gene is expressed together with the immediate early genes c-fos and c-jun, supporting our conclusions. Furthermore Rubin et al. (1998) suggest a role for PC4 in augmenting the adaptive response of the tissue to damage by stimulating differentiation of cells. One aim of this study was to identify potential targets for pharmacological intervention of stroke. Drug-mediated activation of the repair mechanism described would probably reduce the damage caused by stroke but it seems unlikely that such molecules would also fully protect the brain from damage since the strong physiological stress induced by interrupted blood flow causes rapid and severe damage and subsequent breakdown of cell metabolism. Nevertheless, our data demonstrate with an integrated approach the existence of an active and complex neuroprotection and repair program executed after ischemic damage.
Experimental methods Surgical procedures and animal handling All experiments were performed in accordance with the Swiss Federal Act on Animal Protection. Male Sprague–
362
A. Roth et al. / Molecular and Cellular Neuroscience 22 (2003) 353–364
Dawley (320 to 350 g) rats were maintained on a 12-h light– dark cycle and allowed food and water ad libitum. The animals were anesthetized with a mixture of 3% isoflurane, 70% air, and 30% oxygen. They were then intubated and ventilated. Anesthesia was maintained using 1 to 2% isoflurane, 70% air, and 30% oxygen. The femoral artery was cannulated to enable monitoring of physiological variables such as mean arterial blood pressure and blood gases. The temperature of the animals was maintained strictly at 37 ⫾ 0.1°C. The animals were subjected to permanent occlusion of the left middle cerebral artery (MCA) as described previously (Gill et al., 1992; Tamura et al., 1981). The left MCA was occluded permanently using bipolar coagulation; afterward the skin incision was sutured and the femoral catheter removed. Then the anesthetic was turned off, the animals were allowed to breathe air, and once consciousness was regained they were extubated and allowed to breathe spontaneously. The animals (n ⫽ 5) were decapitated 1, 3, 6, and 24 h after permanent MCA occlusion. The brains were rapidly removed; the frontal and parietal cortices and striatum were dissected, snap-frozen in liquid nitrogen, and stored at ⫺80°C until RNA isolation. Cell culture The rat PC12 pheochromocytoma cell line was a gift from E. Shooter (Stanford, CA, USA). Normal rat kidney (NRK) cells were obtained from the American Type Culture Collection (Rockville, MD, USA). Both cell lines were grown in Dulbecco’s modified Eagle’s medium, 1% penicillin–streptomycin, 5% horse serum, and 10% fetal bovine serum (all from Life Technologies, Inchinnan Business Park, UK) in a humidified incubator with 8% CO2 at 37°C. Cloning of PC4 into pUC-combicmv/zeo The PC4 gene was amplified by 35 cycles of (RT) PCR using rat cDNA as template (6-h time point). The primers were selected from the database sequence J04511: 5⬘ TCCATCGGCTGATCCTTGCTGAGCTCCAAG 3⬘ (nt 91–120), 5⬘ AAAGTCTCTAAAAGCTTATFGTACATTAGAA 3⬘ (nt 1583–1553). The amplicon had the expected size and 9 of 12 pGEM-Teasy (Promega, Madison, WI, USA) subclones the expected sequence (data not shown). The fragment was released and subcloned into pUC-combicmv/zeo (Schultze et al., 1996). Two micrograms of plasmid DNA was used to transfect PC12 and NRK cells using the LipofectAMINE reagent (Life Technologies) according to the manufacturer’s instructions. Stable clones were selected using 400 g/ml zeocin (Invitrogen, Carlsbad, CA, USA) and 0.2 g/ml tetracycline to repress the promoter. For each line 12 clones were picked and extended followed by Western blot analysis of induced and repressed cultures. Two clones (PC12-3,
NRK-9) responded well to tetracycline and were selected for further experiments. Isolation of total RNA The brain tissue was homogenized in FastPrep tubes for 20 in a Savant homogenizer (Bio101, Buena Vista, CA, USA) containing Teflon beads and 1 ml RNAzol B reagent (Biotecx, Friendswood, TX, USA). Cell cultures were harvested directly with RNAzol. Total RNA was isolated according to the manufacturer’s instructions and quality was assessed using gel electrophoresis. Microarray experiments Biotin-labeled target RNA was transcribed from total RNA as described (Certa et al., 2001). Hybridization was carried out overnight at 42°C using proprietary chips obtained from Affymetrix Inc. (Santa Clara, CA USA; EMBL12a and EMBL13a) containing about 11,000 annotated genes and sequence tags from public and private databases. Forty pairs of oligonucleotides, which are synthesized in a proprietary process using light-directed synthesis, represent each gene. Washing and scanning of the arrays were performed as recommended by the supplier. Data analysis Raw expression data were collected using commercial GeneChip software (Version 2.0) from Affymetrix. The values for average differences for each brain region between the contralateral and ipsilateral sides were compared with the unpaired t test and an outlier-removal test (Nalimov test). Differential expression was calculated as the increase between two conditions, e.g., between the ischemia-induced ipsilateral brain side and the unaffected contralateral half. A gene was considered differentially expressed when the signal increase or decrease was at least threefold at one of the time points. In addition, the standard deviation had to be significantly smaller than the absolute change in average difference and the calculated confidence level of a gene was set greater than 97% (P value ⬍ 0.03 based on unpaired t test). In situ hybridization A synthetic, [35S]dATP-labeled 60-mer antisense oligonucleotide PC4 probe (nt residues 932–992; Genosys Biotechnology’s, Woodlands, TX, USA) was used as described (van Lookeren Campagne and Gill, 1998). The probe was hybridized to coronal 20-m sections of brain regions from the experimental animals. Sections were exposed for 4 weeks at 4°C to -Max Hyperfilms (Amersham, Buckinghamshire, U.K.).
A. Roth et al. / Molecular and Cellular Neuroscience 22 (2003) 353–364
Production of polyclonal rabbit sera against PC4 A 191-bp fragment of PC4 (position 191–395) was PCRamplified using the verified pUC-combi clones as template and PCR primers with artificial BamHI cloning sites (boldface): 5⬘ TCCCCACCGCGGATCCGGGGGGTGGCGGCG 3⬘, 5⬘ TCATGTTCAACTTCCCTAATTAACTGGATCC 3⬘. The amplicon was digested with BamHI and subcloned into the commercial E. coli expression vector pQE40 (Qiagen, Hilden, Germany). Cultures of induced colonies were screened for expression of a fusion protein on Coomassiestained gels. The histidine-tagged fusion protein of clone AW-23 was purified in 8 M urea, 0.1 M sodium phosphate, 10 mM Tris, pH 8, using commercial Ni-NTA spin columns with decreasing pH according to the manufacturer’s instructions (QIA Express Kit, Qiagen). The pH 2.7 fraction contained the recombinant protein. Two hundred micrograms of PC4-6H was isolated from a gel slice of a preparative SDS– gel and the identity of the protein was confirmed by mass spectrometry (data not shown). After overnight dialysis against PBS the protein was concentrated and used to immunize rabbits using standard protocols. The final serum was injected into two rabbits. After three immunizations both sera (1:500 dilution) recognized a protein of the expected size in NGF-induced PC12 cells. Western blot analysis Ten micrograms of total protein lysate was run on a 10% SDS–polyacrylamide gel under reducing conditions. The protein was blotted on a Hybond ECL membrane (Amersham Pharmacia Biotech) with a semidry blotter at 4.5 V for 30 min (Bio-Rad, Hercules, CA, USA). The membrane was blocked with the SuperBlock reagent to quench background reactivity (Pierce, Rockford, IL, USA) and incubated with PC4 polyclonal antibodies and an anti-rabbit IgG horseradish peroxidase (Santa Cruz Biotechnology, Santa Cruz, CA, USA). Protein bands were visualized with the SuperSignal Detection Kit (Pierce) and exposed to Biomax MR films (Eastman Kodak Company, Rochester, NY, USA). Immunofluorescence Cell cultures were washed twice with PBS and subsequently fixed with cold (⫺20°C) methanol. Cells were rehydrated with PBS and incubated with antibody (1:500 dilution) for 20 min at 37°C followed by three washes with PBS. Antigen/antibody complexes were detected with goatanti rabbit FITC conjugate (1:80 in PBS; Nordic Immunology, Tilburg, The Netherlands) for 20 min at 37°C followed by three PBS washes. Before microscopic inspection the cells were suspended in 50% glycerol in PBS. Pictures were recorded at a 100-fold magnification and appropriate dyespecific excitation and emission settings.
363
Subcellular fractionation of PC12 cells by differential centrifugation NGF-induced cells were harvested and pelleted for 3 min at 1000g and 4°C and washed twice with PBS. The cell pellet was homogenized in 10 mM Hepes, pH 7.3, containing 320 mM sucrose and 1 mM K2-EDTA in a glass douncer with a tiget pestle. The resulting lysate was centrifuged for 5 min at 1000g and 4°C. This low-speed fraction containing essentially intact cells and nuclei was stored frozen and the supernatant was centrifuged for 10 min at 10,000g and 4°C. The resulting pellet containing the heavy membrane fraction was stored frozen and the supernatant was centrifuged for 20 min at 100,000g and 4°C (Beckman TL-100, Rotor TLA 45; Beckman Instruments, Palo Alto, CA, USA). This final pellet contains mainly light membranes and the supernatant represents the soluble, cytosolic fraction. Each fraction containing equivalent amounts of total protein was analyzed by Western blotting using PC4 polyclonal rabbit serum at a 1:1000 dilution.
Acknowledgments We thank Dr. Martin Neeb (Hoffmann–La Roche Bioinformatics) for microarray analysis software and assistance during data analysis. We express thanks to Dr. Marc Soriano and Michel Tessier (Hoffmann–La Roche RCMG) for help in the initial phase of the experiments. We are grateful to Professors Willi Schaffner and Heinrich Reichert (Pharmazentrum University of Basel) for continuous interest in the experiments, and Peter Jakob-Cristal (Hoffmann–La Roche RCMG) for mass spectrometric analysis of recombinant PC4. Finally, we acknowledge Professor Markus Ru¨ egg (Biozentrum University of Basel) for comments on the manuscript.
References Alirezaei, M., Marin, P., Nairn, A.C., Glowinski, J., Premont, J., 2001. Inhibition of protein synthesis in cortical neurons during exposure to hydrogen peroxide. J. Neurochem. 76, 1080 –1088. Augustin, I., Rosenmund, C., Sudhof, T.C., Brose, N., 1999. Munc13-1 is essential for fusion competence of glutamatergic synaptic vesicles. Nature 400, 457– 461. Bates, S., Read, S.J., Harrison, D.C., Topp, S., Morrow, R., Gale, D., Murdock, P., Barone, F.C., Parsons, A.A., Gloger, I.S., 2001. Characterisation of gene expression changes following permanent MCAO in the rat using subtractive hybridisation. Brain Res. Mol. Brain Res. 93, 70 – 80. Bergstedt, K., Hu, B.R., Wieloch, T., 1993. Postischaemic changes in protein synthesis in the rat brain: effects of hypothermia. Exp. Brain Res. 95, 91–99. Betz, A., Ashery, U., Rickmann, M., Augustin, I., Neher, E., Sudhof, T.C., Rettig, J., Brose, N., 1998. Munc13-1 is a presynaptic phorbol ester receptor that enhances neurotransmitter release. Neuron 21, 123–136. Blankesteijn, W.M., van Gijn, M.E., Essers-Janssen, Y.P., Daemen, M.J., Smits, J.F., 2000. Beta-catenin, an inducer of uncontrolled cell prolif-
364
A. Roth et al. / Molecular and Cellular Neuroscience 22 (2003) 353–364
eration and migration in malignancies, is localized in the cytoplasm of vascular endothelium during neovascularization after myocardial infarction. Am. J. Pathol. 157, 877– 883. Bourbon, H.M., Amalric, F., 1990. Nucleolin gene organization in rodents: highly conserved sequences within three of the 13 introns. Gene 88, 187–196. Byers, S.W., Sujarit, S., Jegou, B., Butz, S., Hoschutzky, H., Herrenknecht, K., MacCalman, C., Blaschuk, O.W., 1994. Cadherins and cadherinassociated molecules in the developing and maturing rat testis. Endocrinology 134, 630 – 639. Certa, U., de Saizieu, A., Mous, J., 2001. Hybridization analysis of labeled RNA by oligonucleotide arrays. Methods Mol. Biol. 170, 141–156. Chen, D., Ma, H., Hong, H., Koh, S.S., Huang, S.M., Schurter, B.T., Aswad, D.W., Stallcup, M.R., 1999. Regulation of transcription by a protein methyltransferase. Science 284, 2174 –2177. Cutting, G.R., Lu, L., O’Hara, B.F., Kasch, L.M., Montrose-Rafizadeh, C., Donovan, D.M., Shimada, S., Antonarakis, S.E., Guggino, W.B., Uhl, G.R., et al., 1991. Cloning of the gamma-aminobutyric acid (GABA) rho 1 cDNA: a GABA receptor subunit highly expressed in the retina. Proc. Natl. Acad. Sci. USA 88, 2673–2677. Dietrich, W.D., 1998. Neurobiology of stroke. Int. Rev. Neurobiol. 42, 55–101. Dijkema, R., van der Meide, P.H., Pouwels, P.H., Caspers, M., Dubbeld, M., Schellekens, H., 1985. Cloning and expression of the chromosomal immune interferon gene of the rat. EMBO J. 4, 761–767. Ebadi, M., Bashir, R.M., Heidrick, M.L., Hamada, F.M., Refaey, H.E., Hamed, A., Helal, G., Baxi, M.D., Cerutis, D.R., Lassi, N.K., 1997. Neurotrophins and their receptors in nerve injury and repair. Neurochem. Int. 30, 347–374. Fenczik, C.A., Zent, R., Dellos, M., Calderwood, D.A., Satriano, J., Kelly, C., Ginsberg, M.H., 2001. Distinct domains of CD98hc regulate integrins and amino acid transport. J. Biol. Chem. 276, 8746 – 8752. Fung, E.T., Scheller, R.H., 1999. Identification of a novel alternatively spliced septin. FEBS Lett. 451, 203–208. Gill, R., Andine, P., Hillered, L., Persson, L., Hagberg, H., 1992. The effect of MK-801 on cortical spreading depression in the penumbral zone following focal ischaemia in the rat. J. Cereb. Blood Flow Metab. 12, 371–379. Gill, R., Brazell, C., Woodruff, G.N., Kemp, J.A., 1991. The neuroprotective action of dizocilpine (MK-801) in the rat middle cerebral artery occlusion model of focal ischaemia. Br. J. Pharmacol. 103, 2030 –2036. Gill, R., Sibson, N.R., Hatfield, R.H., Burdett, N.G., Carpenter, T.A., Hall, L.D., Pickard, J.D., 1995. A comparison of the early development of ischaemic damage following permanent middle cerebral artery occlusion in rats as assessed using magnetic resonance imaging and histology. J. Cereb. Blood Flow Metab. 15, 1–11. Gossen, M., Bonin, A.L., Bujard, H., 1993. Control of gene activity in higher eukaryotic cells by prokaryotic regulatory elements. Trends Biochem. Sci. 18, 471– 475. Gossen, M., Bujard, H., 1992. Tight control of gene expression in mammalian cells by tetracycline-responsive promoters. Proc. Natl. Acad. Sci. USA 89, 5547–5551. Guardavaccaro, D., Ciotti, M.T., Schafer, B.W., Montagnoli, A., Tirone, F., 1995. Inhibition of differentiation in myoblasts deprived of the interferon-related protein PC4. Cell Growth Differ. 6, 159 –169. Hirano, S., Kimoto, N., Shimoyama, Y., Hirohashi, S., Takeichi, M., 1992. Identification of a neural alpha-catenin as a key regulator of cadherin function and multicellular organization. Cell 70, 293–301. Hu, B.R., Wieloch, T., 1993. Stress-induced inhibition of protein synthesis initiation: modulation of initiation factor 2 and guanine nucleotide exchange factor activities following transient cerebral ischemia in the rat. J. Neurosci. 13, 1830 –1838. Iacopetti, P., Barsacchi, G., Tirone, F., Cremisi, F., 1996. Expression of the PC4 gene in the developing rat nervous system. Brain Res. 707, 293–297. Imada, K., Leonard, W.J., 2000. The Jak-STAT pathway. Mol. Immunol. 37, 1–11.
Kanai, Y., Segawa, H., Miyamoto, K., Uchino, H., Takeda, E., Endou, H., 1998. Expression cloning and characterization of a transporter for large neutral amino acids activated by the heavy chain of 4F2 antigen (CD98). J. Biol. Chem. 273, 23629 –23632. Kartmann, B., Roth, D., 2001. Novel roles for mammalian septins: from vesicle trafficking to oncogenesis. J. Cell Sci. 114, 839 – 844. Kim, H.L., Kim, J.A., 2000. Purification of clathrin assembly protein from rat liver. Exp. Mol. Med. 32, 222–226. Kim, J.A., Kim, H.L., 2001. Cell-free expression and functional reconstitution of CALM in clathrin assembly. Exp. Mol. Med. 33, 89 –94. Kimball, S.R., 1999. Eukaryotic initiation factor elF2. Int. J. Biochem. Cell Biol. 31, 25–29. Kogure, K., Kato, H., 1993. Altered gene expression in cerebral ischemia. Stroke 24, 2121–2127. Koistinaho, J., Hokfelt, T., 1997. Altered gene expression in brain ischemia. NeuroReport 8, i–viii. Kumar, K., Wu, X.L., 1995. Expression of beta-actin and alpha-tubulin mRNA in gerbil brain following transient ischemia and reperfusion up to 1 month. Brain Res. Mol. Brain Res. 30, 149 –157. Leonard, W.J., 2001. Role of Jak kinases and STATs in cytokine signal transduction. Int. J. Hematol. 73, 271–277. Macrae, M., 1992. New models of focal cerebral ischemia. Br. J. Clin. Pharmacol. 34, 302–308. Nelson, D.P., Wechsler, S.B., Miura, T., Stagg, A., Newburger, J.W., Mayer Jr., J.E., Neufeld, E.J., 2002. Myocardial immediate early gene activation after cardiopulmonary bypass with cardiac ischemia–reperfusion. Ann. Thorac. Surg. 73, 156 –162. Raivich, G., Jones, L.L., Werner, A., Bluthmann, H., Doetschmann, T., Kreutzberg, G.W., 1999. Molecular signals for glial activation: proand anti-inflammatory cytokines in the injured brain. Acta. Neurochir. Suppl. 73, 21–30. Rubin, D.C., Swietlicki, E.A., Wang, J.L., Levin, M.S., 1998. Regulation of PC4/TIS7 expression in adapting remnant intestine after resection. Am. J. Physiol. 275, G506 –G513. Satoh, J., Kuroda, Y., 2000. Beta-catenin expression in human neural cell lines following exposure to cytokines and growth factors. Neuropathology 20, 113–123. Schultze, N., Burki, Y., Lang, Y., Certa, U., Bluethmann, H., 1996. Efficient control of gene expression by single step integration of the tetracycline system in transgenic mice. Nat. Biotechnol. 14, 499 –503. Seelig, H.P., Moosbrugger, I., Ehrfeld, H., Fink, T., Renz, M., Genth, E., 1995. The major dermatomyositis-specific Mi-2 autoantigen is a presumed helicase involved in transcriptional activation. Arthritis Rheum. 38, 1389 –1399. Sofroniew, M.V., Howe, C.L., Mobley, W.C., 2001. Nerve growth factor signaling, neuroprotection, and neural repair. Annu. Rev. Neurosci. 24, 1217–1281. Soriano, M.A., Tessier, M., Certa, U., Gill, R., 2000. Parallel gene expression monitoring using oligonucleotide probe arrays of multiple transcripts with an animal model of focal ischemia. J. Cereb. Blood Flow Metab. 20, 1045–1055. Tamura, A., Graham, D.I., McCulloch, J., Teasdale, G.M., 1981. Focal cerebral ischaemia in the rat: 1. Description of technique and early neuropathological consequences following middle cerebral artery occlusion. J. Cereb. Blood Flow Metab. 1, 53– 60. Tirone, F., Shooter, E.M., 1989. Early gene regulation by nerve growth factor in PC12 cells: induction of an interferon-related gene. Proc. Natl. Acad. Sci. USA 86, 2088 –2092. van Lookeren Campagne, M., Gill, R., 1998. Increased expression of cyclin G1 and p21WAF1/CIP1 in neurons following transient forebrain ischemia: comparison with early DNA damage. J. Neurosci. Res. 53, 279 –296. Yao, P.J., Weimer, J.M., O’Herron, T.M., Coleman, P.D., 2000. Clathrin assembly protein AP-2 is detected in both neurons and glia, and its reduction is prominent in layer II of frontal cortex in Alzheimer’s disease. Neurobiol. Aging 21, 921–929.
