Available online at www.sciencedirect.com

7 (2007) 319–327www.elsevier.com/locate/vph

Vascular Pharmacology 4

Large scale analysis of genes contributing to the herbal preparationdependent hippocampal plasticity in postischemic rehabilitation

Zhong Wang a,⁎, Qingyou Du b, Fusheng Wang b, Qiuping Xu c, Zhongrong Liu d,Baigang Li d, Anmin Wang e, Yongyan Wang a

a Institute of Basic Research in Clinical Medicine, China Academy of Chinese Medical Sciences, 18 Baixincang, Dongzhimennei, Beijing 100700, Chinab Bioengineering Laboratory, 302 Hospital of PLA, Beijing, China

c Basic Medical Institute, Beijing University of Traditional Chinese Medicine, Chinad Pharmaceutical Research Institute, Chengdu Di'ao Pharmaceutical Group Co, Ltd, China

e Institute for Basic Medicine of TCM, China Academy of Traditional Chinese Medicine, 18 Baixincang, Dongzhimennei, Beijing 100700, China

Received 8 April 2007; accepted 7 September 2007

Abstract

Herbal preparations can affect the expression of many genes involved in the ischemic process. These genes have been providing insights intothe molecular basis of brain plasticity in stroke rehabilitation. However, the extent of plasticity has not been investigated using a chemogenomicapproach. A herbal preparation (270 mg/kg) used to treat ischemic mice for 45 days after global ischemia resulted in a significant decrease ininfarct volume and neurological score compared with that of vehicle. This effect was characterized by investigating chemical genomic profiles ofthe mouse hippocampus with a cDNA microarray containing 1176 known genes. Treatment with the herbal preparation reversed the expression of46 genes out of 100 genes altered in untreated ischemic mouse hippocampus. These data indicated that more genes were upregulated (60.78%)than downregulated (30.61%), and only 46 genes (46%) appear to be prime targets for therapeutic intervention in ischemia. The altered genes canbe classified into seven groups, including signal transduction (12 genes, 27%), oncogene (8 genes, 17%), and transcriptional regulation (7 genes,15%). Such multiple plasticity of expression could be considered as the beneficial role of this herbal preparation in stroke rehabilitation. Changesin gene expression of nuclear factor of activated T cells, 14-3-3eta, and β-arrestin suggest a potential role for the immune system in this plasticity.Brain plasticity originates from a balance of up and downregulated genes (Yin and Yang), and reversal of gene expression in multiple pathwaysindicates that a complex signaling network may be constructed and investigated further.© 2007 Elsevier Inc. All rights reserved.

Keywords: Brain plasticity; Stroke rehabilitation; Reorganization; Cerebral ischemia/reperfusion; Herbal therapy; cDNA microarray

Abbreviations: MCAO, Middle cerebral artery occlusion; GPCRs, G-protein-coupled receptors; GRKs, G-protein-coupled receptor kinases; NFAT,Nuclear factor of activated T cells; NFATc4, nuclear factor of activated T-cellsisoform 4; DAG, diacylglycerol; LIM domain, Lin-11/Isl-1/Mec-3 domain;LIMK, LIM-domain-containing protein kinase; CsA, cyclosporin A; BDNF,brain-derived neurotrophic factor; IP3, inositol 1,4,5-trisphosphate; PKC,protein kinase C; PI3K, PI-3 kinase; SAGE, serial analysis of gene expression;ERK1/2, extracellular signal-regulated kinases 1 and 2; MEK, mitogen-activated protein kinase; MAPK, MAP/ERK kinase; FKHR/DAF16, Forkheadrelated transcription factor; Akt, Akt protein kinase; BAD, Bcl-2/Bcl-XL-associated death promoter; CNS, central nervous system; 5-HT, 5-hydroxytryptamine; PKA, protein kinase A; ER, endoplasmic reticulum;CaM, calmodulin; TTC, 2,3,5-triphenyltetrazolium chloride; MMLV,moloney murine leukemia virus; NMDA, N-methyl-D-aspartate; AMPA,α-amino-3-hydroxy-5-mythyl-4-isoxazolepropionic acid.⁎ Corresponding author. Fax: +86 10 62874049.E-mail address: zhonw@sina.com (Z. Wang).

1537-1891/$ - see front matter © 2007 Elsevier Inc. All rights reserved.doi:10.1016/j.vph.2007.09.002

1. Introduction

Stroke ranks third in frequency among human diseasesworldwide. Rehabilitation of function is commonly believedto be associated with brain plasticity, which is the ability ofthe brain to change and repair from functional modificationsof existing structures to the formation, by growth andproliferation, of new structures and neurons (Bergado-Rosadoand Almaguer-Melian, 2000). The possible role of pharma-cological intervention in the postischemic rehabilitation phasehas been extensively reviewed (Goldstein, 1993, 1998) andseveral mechanisms are likely to be involved (Shaw et al.,1994). However, a comprehensive understanding of themechanisms involved depends on the characterization of

320 Z. Wang et al. / Vascular Pharmacology 47 (2007) 319–327

gene expression profiles that contribute to plasticity, which isdependent on high-throughput screening technology, such asmicroarray that have the capacity to identify thousands ofgenes simultaneously. This approach has been applied toinvestigate aging (Lee et al., 2000), multiple sclerosis(Whitney et al., 1999), and Alzheimer's disease andschizophrenia (Mirnics et al., 2001). It has also beenemployed to study gene expression patterns 3 h afterpermanent middle cerebral artery occlusion (MCAO) inSprague Dawley rats (Soriano et al., 2000) and after 14 hof reperfusion following 2 h of MCAO in mice (Trendelen-burg et al., 2002). However, to the best of our knowledgethere have been no reports that describe chemogenomicprofiling (Giaever et al., 2004) of brain plasticity in responseto Chinese medical herbs. Baicalein was suggested toselectively inhibits the NO-dependent apoptotic pathway ofactivated microglia by suppressing cytotoxic NO production(Kyoungho et al., 2003). Dioscin exerted significant inhibi-tory effects on the growth of the human leukemia cell HL-60,inducing differentiation and apoptosis and affects manycancer cells (Wang et al., 2001). However, the mechanismof the pharmacological effect on gene expression has not yetbeen elucidated. A herbal preparation composed of baicaleinand dioscin improved deficiencies in spatial learning memoryin a dose-dependent manner in our previous study (Wang etal., 2004). Here, we explore the chemical genomic pattern ofstroke rehabilitation except spatial learning memory bymicroarray analysis of mRNA expression in the postischemicmouse brain treated with the same herbal preparation.

2. Materials and methods

2.1. Animal model

Animal experiments were carried out in accordance with thePrevention of Cruelty to Animals Act 1986 and the NationalInstitute of Health guidelines for the care and use of laboratory

Fig. 1. Neuroprotection of global cerebral ischemia in mice treated with the herbaltreated with the herbal preparation were significantly reduced (vs. vehicle) 45 days aas means±SD.

animals for experimental procedures, and were approved afterreview by a local committee. Sixty adult male mice (threemonths old, 38–48 g Kunming strain, China) were divided intothree groups each consisting of 20 subjects. Global ischemiawas induced by bilateral common carotid artery (CCA)occlusion under controlled ventilation as described previously(Kawase et al., 1999). Briefly, The bilateral CCA was exposedand ligated twice for 15 min using microvascular clips, andreperfused for 10 min between those two occlusions. Sham-operated mice underwent identical procedures without theartery occlusion. Blood pressure, blood gas, and glucose weremonitored, and rectal temperature was maintained at 37.0–37.5 °C with a heating pad. Brain temperature was monitoredwith a 29-gauge thermocouple in the right corpus striatum andmaintained at 36–37 °C with a temperature-regulating lamp.The EEG was monitored to ensure isoelectricity duringischemia.

2.2. Herbal therapy

Experimental animals were divided into three groups: sham-operated mice, and ischemic mice receiving either the herbalpreparation (270 mg/kg) or vehicle (100% DMSO; 2 mL/kg)intragastrically twice a day. The herbal preparation was achemically standardized mixture of baicalein and dioscin (1:1)dissolved in DMSO just before use (China Natural Institute forthe Control of Pharmaceutical and Biological Products).Treatment commenced five days before surgery, and wasmaintained for 45 days after surgery. Sham-operated mice weretreated with vehicle.

2.3. Neurological score and infarct volume

Forty-five days after surgery, animals were neurologicallyexamined using a set of modified Neurological Severity Scores(NSS) (Chen et al., 2001) by an investigator blinded to eachgroup. Neurological function was graded on a scale of 0 to 18

preparation. Infarct volume (A) and composite neurological score (B) in micefter global cerebral ischemia (⁎⁎Pb0.01, n=10, ANOVA). Data are expressed

Fig. 2. Neuroplasticity of CA1 pyramidal neurons treated with the herbalpreparation postischemia. Toluidine blue stained coronal brain sections showingdorsal hippocampus from sham-operated mice (A, B; n=6), vehicle treatedischemic mice (C, D; n=5), and herbal preparation treated ischemic mice (E, F;n=7). Quantitation of cells per unit length of CA1 (G; F=15.34, ⁎Pb0.05,⁎⁎Pb0.01). A, C, F, Scale bar: 400 μm; B, D, F, Scale bar: 40 μm.

321Z. Wang et al. / Vascular Pharmacology 47 (2007) 319–327

(normal score, 0; maximal deficit score, 18). NSS is a compositeof motor, sensory, reflex, and balance tests. One point isawarded for the inability to perform the test or for the lack of atested reflex; thus, the higher the score, the more severe is theinjury.

Eight mice from each group were used to calculate theinfarction ratio. In brief, the cerebrum was removed and cut intofive slices in the coronal plane 1, 3, 5, and 7 mm from theprefrontal cortex. The slices were transferred to 4% 2,3,5-triphenyltetrazolium chloride (TTC) solution and incubated for30 min at 37 °C in darkness and then transferred into 10%formalin. Images of the slices were captured using a digitalcamera (Color CCD camera TP-6001A, Topica Inc., Japan).The areas of the infarction region were calculated using aPathology Image Analysis System (Topica Inc., Japan) and theratio of the infarction area to the total slice area was determined.

2.4. Quantitative cell count

Cell loss was evaluated by histological examination of thedorsal hippocampus in brain sections 45 days after surgery. Sixmice from each group were placed under deep anesthesia andfixed by transcardiac perfusion with ice-cold 4% paraformal-dehyde. Brains were removed and immersed in fixative (4 °C,overnight). Brains were sectioned (15 μm) and stained withtoluidine blue in a serial manner, and a total of approximately 20slices per animal were saved, taking care to avoid adjacentslices. From this initial set, random sections were selected witha random number table, and a total of 10 photomicrographsfrom the dorsal CA-1 per animal were taken at 100× for the cellcounts. Eight photomicrographs of CA-1 were taken from eachanimal and analyzed by a blinded investigator.

2.5. Microarray

The Atlas™ Mouse 1.2 Brain Array (Clontech, Palo Alto,CA, USA) containing 1176 genes from brain tissue was used toconduct gene expression profiling. Similar studies with theseAtlas arrays were found in different domain (Michal et al.,2004; Kumar et al., 2001). Genes were chosen for their possiblerole in mouse brain and included those encoding transporters,signal transduction molecules, nervous system transcriptionfactors, general DNA binding proteins, cell surface antigens,cell adhesion molecules, cell cycle regulators, and ion channels(www.clontech.com). Gene classification made by GeneOntology (Ashburner et al., 2000).

2.6. Probe synthesis and hybridization

The hippocampus (subfields CA1-4 of Ammon's horn) wasdissected from three mice in each experimental group, flash-frozen and stored at −75 °C. RNA was extracted using theAtlas™ Pure Total RNA Labeling System (Clontech, Palo Alto,CA, USA) according to the manufacturer's recommendations.For each group, three mice hippocampus were homogenized indenaturing solution with a Polytron, respectively. Total RNAwas isolated from tissue homogenates with phenol–chloroform

extraction and dissolved in RNase-free water. DNase I-treatedtotal RNA was precipitated by a second round of phenol-denaturing gel analysis and absorbance measurements at A260/A280 assessed chloroform extraction and the quality of RNA.Total RNA samples exhibited A260/A280 ratios of 1.8 or higherand degradation was not visible by denaturing gel analysis.After enrichment of poly A+RNA by oligo (dT) separation,cDNA probe synthesis was performed using the gene-specificcDNA Synthesis Primer Mix provided and incorporation of[alpha-33P] dATP (YAFEI Life Science Products, Inc., Beijing)by MMLV reverse transcriptase. Unincorporated nucleotideswere removed by column chromatography according to themanufacturer's protocol. cDNA probes prepared from each ofthe three experimental groups were hybridized to separatemembranes with continuous agitation at 68 °C using a kitaccording to the manufacturer's recommendations. After 24 h,200 mL of solution 1 (0.1×SSC, 1% SDS) and solution 2(0.1×SSC, 0.5% SDS) prewarmed to 68 °C were usedconsecutively to wash the membrane with agitation at roomtemperature for 5 min.

322 Z. Wang et al. / Vascular Pharmacology 47 (2007) 319–327

2.7. Microarray scanning, quantitation, and statistical analysis

Filters were exposed to a phosphor screen and scanned witha STORM 860 PhosphorImager. Radioactive intensity of eachspot was linearly scanned to 65,536 gray-grade in a pixel sizeof 50 μm. After subtracting the background chosen from an areawhere no PCR product was located, genes with intensity greaterthan 10 were considered positive, and were distinguished frombackground with statistical significance N99.9%. Differentialexpression was considered significant between the three experi-mental groups when the ratio of signals between the samesites on different membranes was more than ±1.8 (Genespringsoftware). A one-way ANOVA model and SignificanceAnalysis of Microarrays (SAM) (Tusher et al., 2001) wereused to compare the means of altered genes in the differentgroups also (every group taking three microarrays indepen-dently). Experimental analysis is based on guidelines ofMinimumInformation about a Microarray Experiment (MIAME) (Brazmaet al., 2001).

2.8. Semiquantitative RT-PCR analysis

Candidate genes from the array data analysis were selectedfor further investigation by RT-PCR based on the magnitude oftheir expression ratios or their functional similarity to highlyregulated genes. After extraction of total RNA, 5 μg RNA wasreverse transcribed for 1 h at 39 °C with 0.5 μg random primersand 200 units of MMLV reverse transcriptase (Promega,Madison, WI, USA), 25 units of RNasin ribonuclease inhibitor,and 500 μM dNTP (final concentration) in a 25 μL reaction.The resulting single-strand cDNAwas amplified by PCR usingspecifically designed primers for the selected genes. After aninitial 1 min denaturing cycle at 95 °C, an optimal number ofdenaturation, annealing, and polymerization cycles for thespecific gene were performed, followed by a final 10 minelongation step at 72 °C. To overcome possible variation that

Fig. 3. Contribution of reversed expression genes of herbal preparation-treated groupgroup, when upregulation (or downregulation) occurs the herbal treatment group. Threverse of animals treated with vehicle. The black block represents upregulated (49) areversed upregulated (15) and downregulated genes in herbal preparation treated maltered in preparation-treated animals into seven groups. The proportion of altered g

may arise from the PCR machine (Biometra, USA), reactionswere performed in duplicate or triplicate. To determine thenumber of cycles to use in the PCR reaction, a cycle numberwas chosen in the linear portion of the amplification process.In addition, parallel PCR reactions were performed with pairsof GAPDH-specific primers as internal controls for normal-izing variations in RNA aliquots and for gel loading. Thenumber of PCR cycles of GAPDH, ERCC3, CaMKIIβ andApoE genes was 25, 30, 35, and 30, respectively. Forward andreverse primers were: ERCC3, 5′-ACACACATTTATTGCCCA-CATGG-3′ and 5′-TTTTGCCCATGGCAGCTACAGCAGC-3′;CaMKII β, 5′-ATGGCCACCACGGTGACCTGCACCCG-3′and 5′-TGAAACCAGGCGCAGCTCTCACTGCAG-3′;ApoE, 5′-AGGATCTACGCAACCGACTC-3′ and 5′-GGCG-ATGCATGTCTTCCACTA-3′; GAPDH, 5′-TAAAGGG-CATCCTGAGCTACACT-3′ and 5′-TTACTCCTTCGAG-GCCATGTAGG-3′. PCR products were separated by gelelectrophoresis in 2.5%SeakemAgarose and 40mMTris–acetate,1 mM EDTA, pH 7.6. Molecular mass markers were included oneach gel and DNA was visualized by inclusion of 0.5 μg/mLethidium bromide and scanning with a Molecular DynamicsFluorimagerSI and analysis with ImageQuant v. 4.1 (AmershamBiosciences, USA). Individual RT reactions were excluded fromassessment if either were negative, or if both were weak.

3. Results

3.1. Infarct volumes and neurological scores

Infarct volumes of the different groups were used to evaluatethe effect of the herbal preparation on rehabilitation inpostischemia (Fig. 1A). The total infarct volumes in theherbal-treatment group were significantly smaller (P=0.028)than in the ischemic vehicle treatment group (F=16.23,Pb0.01, vs. vehicle; n=8). Similar results were achieved forneurological scores, with the herbal-treatment group exhibiting

. ‘Reverse’ refers to genes that are downregulated (or upregulated) in the vehiclee quantity of altered genes in animals treated with the herbal preparation is thend downregulated genes (51) in vehicle treated animals. The white block showsice compared with those of vehicle treated animals (A). Classification of genesenes in each group is shown (B).

Table 1Increases in hippocampus gene expression after herbal treatment of globalcerebral ischemia (results represent the mean fold increase over vehicle for threeseparate microarray hybridizations (Pb0.05, ANOVA)

Name Mean±SD

GenBank SwissProt

Signal transductionCa2+/calmodulin-dependent proteinkinase II beta subunit

2.24±0.39 X63615 P28652

14-3-3eta 2.77±0.22 U57311 P115767B2 neuroendocrine protein 3.32±1.26 X15830 P12961Cek 5 receptor protein tyrosine kinaseligand

2.21±0.45 U12983 P52795

LIM domain kinase 1 2.80±0.82 U15159 P53668Proenkephalin A precursor 2.53±1.02 M55181 P22005Frizzled homolog 7 2.34±0.63 U43320 Q61090

OncogenesThyroid hormone receptor alpha 1 3.80±1.13 X51983 P16416Transcription factor AP-1 1.87±0.37 J04115 P05627Tyro3 precursor 1.91±0.49 U18342 P55144c-Fms proto-oncogene 1.80±0.43 X68032 P09581Mothers against dpp protein 1 2.52±0.57 U58992 P70340H-ras proto-oncogene 3.92±0.93 Z50013 Q61411Nucleoside diphosphate kinase B 2.53±0.53 X68193 Q01768

Transcriptional regulation/DNA bindingAmino enhancer of split protein 1.87±0.51 L12140 Q06195Transcription factor NFAT 1isoform alpha

2.84±0.72 U02079 O88942

Sim transcription factor 3.26±0.83 U42554 Q61079Engrailed protein homolog 1.80±0.46 L12705 P09066Ets-related transcription factor 1.87±0.54 U58533 P70459

DNA synthesis, repair, and recombinationproteinsDNA excision repair protein ERCC3 1.84±0.49 S71186 P49135Xerodema pigmentosum group Acorrecting protein

2.81±0.74 X74351 Q64267

MmRad52 1.87±0.42 Z32767 P43352Translin 1.83±0.38 X81464 Q62348

Cytoskeleton and motility proteinsCathepsin D 2.61±0.35 X53337 P18242Nonmuscle cofilin (CFL1) 2.75±0.61 D00472 P15760Kinesin motor protein 1.81±0.41 U92949 O08613Neuronal kinesin heavy chain (NKHC) 2.80±0.63 X61435 P28738

Cell cycle regulationTob antiproliferative factor 1.82±0.32 D78382 Q61471

OthersErp72 endoplasmic reticulumstress protein

1.89±0.42 J05186 P08003

Integrin alpha-M 1.87±0.35 X07640 P05555Glutathione S-transferase Pi 1 2.13±0.28 D30687 P19157

Table 2Decreases in hippocampus gene expression after herbal treatment of globalcerebral ischemia (results represent the mean fold increase over vehicle for threeseparate microarray hybridizations (Pb0.05, ANOVA)

Name Mean±SD GenBank SwissProt

Signal transductionNociceptin precursor −2.19±0.73 D82866 Q64387Axonal membrane protein GAP-43 −2.24±0.39 J02809 P06837β-arrestin −1.86±0.36 M24086 P20443Protease-activated receptor 4 (GPCR) −1.88±0.42 AF080215 O88634Neurogenic locus notch homolog 1precursor

−2.27±0.38 Z11886 Q01705

Oncogenesc-Src proto-oncogene −1.83±0.46 M17031 P05480

Transcriptional regulation/DNA bindingDrosophila NK1 transcription factor-related locus 1

−1.94±0.48 X75385 P42580

Transcription factor SEF2 −2.77±0.65 D00926 P23881DNA synthesis, repair andrecombination proteinsMethyl-CpG-binding protein 2(MeCP2)

−2.46±0.32 AF072251 Q9Z2D6

Four and a half LIM domains 1 −1.83±0.34 U77039 P97447Cytoskeleton and motility proteinsKinesin-like protein KIF38 −1.88±0.27 U43317 Q61088

Cell cycle regulationCyclin C (G1-specific) −2.92±0.94 U62638 Q62447

OthersApolipoprotein E precursor −2.36±0.41 M12414 P08226Granzyme A −1.86±0.51 M13226 P110325-hydroxytryptamine receptor −2.85±0.35 S49542 P35363

323Z. Wang et al. / Vascular Pharmacology 47 (2007) 319–327

a significant improvement (Fig. 1B, F=12.34, Pb0.01, vs.vehicle; n=8).

3.2. Histological analysis

We assessed neuronal death histological 45 days after the lasttreatment. Ischemia induced extensive death of pyramidal cellsin CA1 and the few remaining pyramidal neurons were severelydamaged (Fig. 2C and D) resulting in cell loss that was differentfrom that in normal issue (Fig. 2A). Herbal treatment affordedprotection against ischemia as neuronal death or damage wasnot evident (Fig. 2E and F). These findings were validated byneuronal counts (Fig. 2G).

3.3. Gene expression changes on microarrays

Among groups of sham-operated, vehicle, herbal treatment,from 45.4% to 54.4% (534–640) transcripts out of 1263 (totalgenes) were present. The relative expression level of a givencDNA was assessed by ratio of signal obtained from at leastthree membranes (after normalizing to the global value of all thegenes provided on the membranes) to that from other threemembranes (N=3–5). The signal intensity of housekeepinggenes on the two membranes was similar. Upregulation ordownregulation of more than 1.8 fold occurred in 180,100 and46 genes in the sham-operated, vehicle and herbal-treatmentgroup, respectively. Of those genes in the vehicle-treatmentgroup that were differentially expressed relative to the sham-operated mice, 49 were upregulated and 51 genes weredownregulated. Interestingly, it was found that of thosesignificant 46 genes in the herbal-treatment group, 31 increasedexpression reversed the set of 51 downexpression genes in thevehicle treatment group and 15 decreased against 49 upexpres-sion genes in ischemic mice (Fig. 3A). Target genes in theischemic mouse hippocampus were both upregulated (60.78%)and downregulated (30.61%): the number of upregulated geneswas shown to be about double the number of downregulatedgenes.

3.4. Functional analysis

The 46 genes that showed differential expression in response totreatment with the herbal preparation can be classified into seven

Fig. 4. Proposed mechanism of herbal preparation. Schematic view of known regulated apoptosis pathways and involvement of genes altered in response to herbaltreatment. In our study, NFAT, AP-1, CaMKII β, 14-3-3eta, and Fz mRNAs were upregulated (upward unfilled arrows) and β-arrestin mRNA was downregulated(downward unfilled arrow), demonstrating the complex effects that the herbal preparation has on signaling networks.

324 Z. Wang et al. / Vascular Pharmacology 47 (2007) 319–327

groups (Fig. 3B). The majority of these (59%) were distributedbetween signal transduction (12 genes, 27%; this is consistentwith the data from Tanaka et al., 2002), oncogene (8 genes, 17%)and transcriptional regulation/DNA binding (7 genes, 15%).Differentially expressed genes in each of the seven classificationsare listed according to gene function and upregulation (Table 1) ordownregulation (Table 2) in mouse hippocampus treated withherbal preparation after global cerebral ischemia.

Based on this classification and analysis of the literature, weattempted to reconstruct the pathway that is modulated by ourherbal preparation. There is no exact match with known signal

Fig. 5. RT-PCR analysis of ERCC3, CaMKII β, ApoE and GAPDH in mousehippocampus from the three different treatment groups.

intracellular pathways; therefore, we propose a multiple herbal-activated brain plasticity regulatory pathway (Fig. 4).

3.5. RT-PCR

The altered expression of ERCC3, CaMKII β, and ApoE inresponse to the herbal preparation obtained in the cDNAmicroarray was verified by RT-PCR (Fig. 5A). The geneexpression profiles measured by these two independent methods(RT-PCR and microarray) were very similar, even though theabsolute value for eachmRNA level differed.Densitometry valueshave the means and standard deviations for ERCC3, Calcium/calmodulin dependent protein kinase IIβ (CaMKIIβ), ApoE fromsham, vehicle and herbal groups shown, respectively (Fig. 5B).

4. Discussion

The concept of rehabilitation pharmacology, which dates backmore than 20 years (Feeney and Sutton, 1987), proposes thatconventional physical, occupational, or speech and languagetherapy might be augmented if coupled with pharmacotherapy toenhance activity-dependent plasticity (Feeney, 1998). Trials ofother noradrenergic agonists (methylphenidate, IDOPS), levodo-pa, and fluoxetine have been conducted recently and providefurther proof of concept for the strategy of poststroke rehabili-tation pharmacotherapy (Nishino et al., 2001; Grade et al., 1998;Scheidtmann et al., 2001). We have used microarray analysis todetect transcripts that are altered in the global cerebral ischemicbrain in response to treatment with herbal preparation. Our resultsindicated the possibility that the changes in expression of thesegenes might be important for the neuroprotection properties ofthis herbal preparation for stroke rehabilitation.

4.1. Alteration of known genes potentially contributes to brainplasticity

Our data identified seven groups of genes whose expressionwasmodified significantly when the animals were treatedwith the

325Z. Wang et al. / Vascular Pharmacology 47 (2007) 319–327

herbal preparation. The observed changes in some genes are inagreement with published data (Jin et al., 2001; Keyvani et al.,2004). The mRNAs for transcription factor AP-1 and c-Fmsproto-oncogene were upregulated and nuclear transcription factorSEF2 and c-Src proto-oncogene was downregulated, as expectedfor focal brain ischemia. Increased expression of the DNAexcision repair protein ERCC3 suggests activation of anti-oxidantresponses, and downregulation of MeCP2 (Couvert et al., 2001;Shahbazian and Zoghbi, 2001); and apolipoprotein E precursormay have a neurological effect on mental rehabilitation. Theabsence of GAP-43 can protect neurons from death (Gagliardiniet al., 2000).Another neuronal gene showing downregulationwasthat for the 5-hydroxytryptamine receptor, which is an importantGPCR that mediates the effects of hallucinogens and is the targetof a number of commonly prescribed medications includingatypical antipsychotics, antidepressants, and anxiolytics (Xiaet al., 2003). Some neurotrophic factors, such asBDNF (Fernandoet al., 2002), were not found in this study; it is possible that thesefactors rescue neurons in the acute stage, but are not active 45 dayspostischemia. Nevertheless, BDNF induces activation of CaMKIIβ, a critical mediator of synaptic plasticity that was found to beupregulated in our study.

4.2. Multiple signaling pathways participate in brain plasticity

From our gene expression results, several signaling path-ways relevant to brain plasticity may be activated or inhibited.

4.3. CaMKII β

CaMKII β has received much attention because it is essentialfor NMDAR-dependent LTP (Lisman et al., 2002). WhetherCaMKII acts as a scaffold or as an enzyme, the direct involvementof NMDARs in the activation and binding of CaMKII can explainthe recruitment of activated CaMKII β to stimulated synapses(Sheng and Kim, 2002). The unique biochemical properties ofCaMKII β have made this enzyme one of the paradigmaticmodels of the much sought after ‘memory molecule’. Inparticular, the central participation of CaMKII β as a sensor ofCa2+ signals generated by activation of NMDA receptors afterthe induction of long-term plastic changes has encouraged the useof pharmacological, genetic, biochemical, and imaging tools tounveil the role of this kinase in the acquisition, consolidation, andexpression of different types ofmemory (Cammarota et al., 2002).Increased expression of CaMKII β in this study shows that thispathway actively contributes to the herbal preparation-inducedneuroplasticity of memory recovery.

4.4. GPCR (β-arrestins)

β-arrestins are versatile adapter proteins that form complexeswith most GPCRs following agonist binding and phosphory-lation of receptors by GRKs. β-arrestins not only function in themolecular switch required for GPCR desensitization, internal-ization, and desensitization, but also act as scaffolds totransducer and compartmentalize alternative signals (Stephen,2001; Arriza et al., 1992). β-arrestins interact with a wide

variety of endocytic and signaling proteins that now includeclathrin, c-Src (Luttrell et al., 1999), MAPKs (DeFea et al.,2000), and Raf. β-arrestins both facilitate GPCR-stimulatedMAP kinase activation and target active MAP kinases tospecific locations within the cell. Thus, their binding to GPCRsmight initiate a second wave of signaling and represent a novelmechanism of GPCR signal transduction. The mechanism bywhich the herbal preparation downregulates β-arrestin mRNAexpression is unclear. It may be a balance of the pharmacolog-ical effect from different pathways.

4.5. The immune system may potentially be involved in strokerehabilitation

Although the immune system is often regarded as autonomous,research in the last two to three decades has provided strongevidence that the CNS receives messages from the immunesystem, and those messages from the brain modulate immunefunctions. NFAT represents a group of transcription factors thatare implicated in the expression of Fas ligand and several cytokinegenes. Recent studies have confirmed that NFAT plays a criticalrole in the expression of IL-2 and IL-4 by T cells. NFATc4 is alsodiscovered in neurons, where it is believed to play an importantrole in long-term changes in neuronal function (Patrick et al.,2003). Interestingly, it is particularly sensitive to the secondmessenger systems activated by BDNF. In neurons and vascularsmooth muscle cells, NFAT is selectively activated by Ca2+influx through L-type Ca2+ channels. In cultured rat CA3-CA1hippocampus neurons, BDNF activated NFAT-dependent tran-scription via TrkB receptors. NFATc4 plays a crucial role inneurotrophin-mediated synaptic plasticity (Groth and Mermel-stein, 2003). In this study, NFAT 1 gene expression wasupregulated following the herbal treatment, suggesting it maybe a pharmacological target for treatment of ischemia (Fig. 5).

4.6. Only 46 genes (46%) appear to be prime targets fortherapeutic intervention

What percentage of the human genome is ‘druggable’? Only10%, 50% or far larger of those targets (Garber, 2003). Our dataprovide evidence that about half of the target genes (46 genes)are likely to have value in the treatment of ischemia.

Findings from animal neurophysiology, human behavioralneuroscience imaging studies, and technological advances offerthe prospect ofmore rational and effective rehabilitation strategiesfor victims of stroke (Bracewell, 2003): for example, sensory andmotor rehabilitation (Byl, 2003). Unfortunately, much of thisrehabilitation intervention lacks the solid foundation of empiri-cally derived evidence for its efficacy (Landers, 2004). Our dataemphasize that stroke rehabilitation involves a complex neuronalsignal network, and suggest that research in this area isparticularly challenging. An example of the multipathwayinvolvement is that NFAT activation can evoke two distinctbiological programs of gene expression, dependent or indepen-dent of NFAT-AP-1 cooperation (Macián et al., 2000), CaMKII(Yasutake et al., 2000), 14-3-3 protein (Andres et al., 2001), GAP-43 (Gagliardini et al., 2000), and NF-kB pathways (Stephenson

326 Z. Wang et al. / Vascular Pharmacology 47 (2007) 319–327

et al., 2000) were implicated in mechanisms associated withischemia. Data from our study, suggest that a number of othersignal transduction pathways, such as CaMKIIβ, NFAT, andGPCRs,may be responsible for the pharmacological effects of theherbal preparation. It may be possible to reconstruct networks ofgene regulation in multiple pathways (Zdraveski et al., 2000) toreveal complex responses to ischemia.

Brain plasticity originates from a balance of up and down-regulated genes (Yin and Yang). These advances significantlybroaden our framework for understanding the effects of herbalpreparations on pharmacological mechanisms and neuronalplasticity. Recent evidence from animal experiments indicatesthat these functional changes are accompanied by anatomicalchanges. Because plasticity of the brain plays a major role in therecovery of function after stroke, the knowledge of the principlesof plasticity may help to design strategies to enhance plasticitywhen it is beneficial, such as after brain infarction (Bütefisch,2004). Animal and human research over the past decades hasprovided increasing detail of the brain's capacity to reorganizeneural network architecture in order to adapt to environmentalneeds. Neuroplasticity has consequences for understanding bothpathological dynamics and therapeutic options (Elbert andRockstroh, 2004). Importantly, this study may partially explainthe mechanisms by which the herbal preparation interfered withthe ischemic process and will be essential to provide clues to thepathological mechanisms. The approach of chemical genetics toidentifymore genes involved inmultiple pathwayswill add to ourknowledge of ischemic treatment.

Acknowledgements

This work was supported by Hi-Tech Research andDevelopment Program of China (863), The National NaturalScience Foundation of China (90209015) and foundation of“Eleventh Five” National Key Technologies R&D Programme(2006BAI08B04-06).

References

Andres, C., Osef, T.K., Julia, M.U., 2001. Association of GABAB receptors andmembers of the 14-3-3 family of signaling proteins. Mol. Cell. Neurosci. 17,317–328.

Arriza, J.L., Dawson, T.M., Simerly, R.B., Martin, L.J., Caron, M.G., Snyder,S.H., Lefkowitz, R.J., 1992. The G-protein-coupled receptor kinases β-ARK1and β-ARK2 are widely distributed at synapses in rat brain. J. Neurosci. 12,4045–4055.

Ashburner, M., Ball, C.A., Blake, J.A., Botstein, D., Butler, H.J., Cherry, J.M.,Davis, A.P., Dolinski, K., Dwight, S.S., Eppig, J.T., Harris, M.A., Hill, D.P.,Issel-Tarver, L., Kasarskis, A., Lewis, S., Matese, J.C., Richardson, J.E.,Ringwald, M., Rubin, G.M., Sherlock, G., 2000. Gene ontology: tool for theunification of biology. The Gene Ontology Consortium. Nat. Genet. 25,25–29.

Bergado-Rosado, J.A., Almaguer-Melian, W., 2000. Cellular mechanisms ofneuroplasticity. Rev. Neurol. 31 (11), 1074–1095.

Bracewell, R.M., 2003. Stroke: neuroplasticity and recent approaches torehabilitation. J. Neurol. Neurosurg. Psychiatry 74, 1465.

Brazma, A., Hingamp, P., Quackenbush, J., Sherlock, G., Spellman, P.,Stoeckert, C., Aach, J., Ansorge, W., Ball, C.A., Causton, H.C., Gaasterland,T., Glenisson, P., Holstege, F.C.P., Kim, I.F., Markowitz, V., Matese, J.C.,Parkinson, H., Robinson, A., Sarkans, U., Schulze-Kremer, S., Stewart, J.,

Taylor, R., Vilo, J., Vingron, M., 2001. Minimum information about amicroarray experiment (MIAME)-toward standards for microarray data.Nat. Genet. 29, 365–371.

Bütefisch, C.M., 2004. Plasticity in the human cerebral cortex: lessons from thenormal brain and from stroke. Neuroscientist 10, 163–173.

Byl, N., 2003. Effectiveness of sensory and motor rehabilitation of the upperlimb following the principles of neuroplasticity: patients stable post stroke.Neurorehabil. Neural Repair 17, 176–191.

Cammarota, M., Bevilaqua, L.R., Viola, H., Kerr, D.S., Reichmann, B.,Teixeira, V., Bulla, M., Izquierdo, I., Medina, J.H., 2002. Participation ofCaMKII in neuronal plasticity and memory formation. Cell. Mol. Neurobiol.22 (3), 259–267.

Chen, J., Li, Y., Wang, L., Zhang, Z., Lu, D., Lu, M., Chopp, M., 2001.Therapeutic benefit of intravenous administration of bone marrow stromalcells after cerebral ischemia in rats. Stroke 32, 1005–1012.

Couvert, P., Bienvenu, T., Aquaviva, C., Poirier, K., Moraine, C., Gendrot, C.,Verloes, A., Andrès, C., Fevre, A.C.L., Souville, I., Steffann, J., Portes, V.,Ropers, H.H., Yntema, H.G., Fryns, J.P., Briault, S., Chelly, J., Cherif, B.,2001. MECP2 is highly mutated in X-linked mental retardation. Hum. Mol.Genet. 10, 941–946.

DeFea, K.A., Zalevsky, J., Thoma,M.S., Dery, O., Mullins, R.D., Bunnett, N.W.,2000. β-Arrestin-dependent endocytosis of proteinase-activated receptor 2 isrequired for intracellular β-targeting of activated ERK1/2. J. Cell Biol. 148,1267–1281.

Elbert, T., Rockstroh, B., 2004. Reorganization of human cerebral cortex:the range of changes following use and injury. Neuroscientist 10 (2),129–141.

Feeney, D.M., 1998. Rehabilitation pharmacology: noradrenergic enhancementof physical therapy. In: Ginsberg, M.D., Bogousslavsky, J. (Eds.),Cerebrovascular Disease: Pathophysiology, Diagnosis and Management.Blackwell Science, Oxford, UK, pp. 620–636.

Feeney, D.M., Sutton, R.L., 1987. Pharmacotherapy for recovery of functionafter brain injury. Crit. Rev. Neurobiol. 3, 135–197.

Fernando, G.P., Ying, Z., Roy, R.R., Molteni, R., Edgerton, V.R., 2002.Voluntary exercise induces a BDNF-mediated mechanism that promotesneuroplasticity. J. Neurophysiol. 88, 2187–2195.

Gagliardini, V., Dusart, I., Fankhauser, C., 2000. Absence of GAP-43 canprotect neurons from death. Mol. Cell. Neurosci. 16, 27–33.

Garber, K., 2003. Big plans for small molecules: NIH launches chemicalinitiative. J. Natl. Cancer Inst. 95 (23), 1740–1741.

Giaever, G., Flaherty, P., Kumm, J., Proctor, M., Nislow, C., Jaramillo, D.F.,Chu, A.M., Jordan, M.I., Arkin, A.P., Davis, R.W., 2004. Chemogenomicprofiling: identifying the functional interactions of small molecules in yeast.PNAS 101 (3), 793–798.

Goldstein, L.B., 1993. Basic and clinical studies on pharmacologic effects onrecovery from brain injury. J. Neural Transpl. Plast. 4, 174–192.

Goldstein, L.B., 1998. Potential effects of common drugs on stroke recovery.Arch. Neurol. 55, 454–456.

Grade, C., Redford, B., Chrostowski, J., Toussaint, L., Blackwell, B., 1998.Methylphenidate in early post stroke recovery: a double-blind, placebo-controlled study. Arch. Phys. Med. Rehabil. 79, 1047–1050.

Groth, R.D., Mermelstein, P.G., 2003. Brain-derived neurotrophic factoractivation of NFAT (Nuclear Factor of Activated T-Cells)-dependenttranscription: a role for the transcription factor NFATc4 in neurotrophin-mediated gene expression. J. Neurosci. 23, 8125–8134.

Jin, K.L., Mao, X.O., Eshoo, M.W., 2001. Microarray analysis of hippocampalgene expression in global cerebral ischemia. Ann. Neurol. 5, 93–103.

Kawase, M., Murakami, K., Fujimura, M., Morita-Fujimura, Y., Gasche, Y.,Kondo, T., Scott, R.W., Chan, P.H., 1999. Exacerbation of delayed cellinjury after transient global ischemia in mutant mice with CuZn superoxidedismutase deficiency. Stroke 30, 1962–1968.

Keyvani, K., Sachser, N., Witte, O.W., Paulus, W., 2004. Gene expressionprofiling in the intact and injured brain following environmental enrichment.J. Neuropathol. Exp. Neurol. 63 (6), 598–609.

Kumar, A., Soprano, D.R., Parekh, H.K., 2001. Cross-resistance to the syntheticretinoid CD437 in a paclitaxel-resistant human ovarian carcinoma cell line isindependent of the overexpression of retinoic acid receptor. Cancer Res. 61,7552–7555.

327Z. Wang et al. / Vascular Pharmacology 47 (2007) 319–327

Kyoungho, S., Heasuk, L., Sang, S.K., Gyeong, J.C., Wan, S.C., 2003.Flavonoid baicalein attenuates activation-induced cell death of brainmicroglia. J. Pharmacol. Exp. Ther. 305, 638–645.

Landers, M., 2004. Treatment-induced neuroplasticity following focal injury tothe motor cortex. Int. J. Rehabil. Res. 27 (1), 1–5.

Lee, C.K., Weindruch, R., Prolla, T.A., 2000. Gene-expression profile of theageing brain in mice. Nat. Genet. 25, 294–297.

Lisman, J., Schulman, H., Cline, H., 2002. The molecular basis of CAMKIIfunction in synaptic and behavioural memory. Nat. Rev., Neurosci. 3, 175.

Luttrell, L.M., Ferguson, S.S., Daaka, Y.,Miller,W.E.,Maudsley, S., Della, R.G.J.,Lin, F., Thierry, D., 1999. From global expression data to gene networks.BioEs-survivalsays 21, 895–899.

Macián, F., Carmen, G.R., Rao, A., 2000. Gene expression elicited by NFAT inthe presence or absence of cooperative recruitment of Fos and Jun. EMBO J.19, 4783–4795.

Michal, H., Eli-Berchoer, L., Wapinski, I., Friedman, N., Kodesh, E., 2004.Stress-related genomic responses during the course of heat acclimation andits association with ischemic-reperfusion cross-tolerance. J. Appl. Physiol.97, 1496–1507.

Mirnics, K., Middleton, F.A., Lewis, D.A., Levitt, P., 2001. Analysis of complexbrain disorders with gene expression microarrays: schizophrenia as a diseaseof the synapse. Trends Neurosci. 8, 479–486.

Nishino, K., Sasaki, T., Takahashi, K., Chiba, M., Ito, T., 2001. Thenorepinephrine precursor L-threo-3, 4-dihydroxyphenylserine facilitatesmotor recovery in chronic stroke patients. J. Clin. Neurosci. 8, 547–550.

Patrick, G.H., Chen, L., Julie, N., Rao, A., 2003. Transcriptional regulation bycalcium, calcineurin, and NFAT. Genes Dev. 17, 2205–2232.

Scheidtmann, K., Fries, W., Muller, F., Koenig, E., 2001. Effect of levodopa incombination with physiotherapy on functional motor recovery after stroke: aprospective, randomized, double-blind study. Lancet 358, 787–790.

Shahbazian,M.D., Zoghbi, H.Y., 2001.Molecular genetics of Rett syndrome andclinical spectrum of MECP2 mutations. Curr. Opin. Neurol. 14, 171–176.

Shaw, C.A., Lanius, R.A., van den Doel, K., 1994. The origin of synapticneuroplasticity: crucial molecules or a dynamic cascade? Brain Res. Rev. 19,241–263.

Sheng, M., Kim, M.J., 2002. Postsynaptic signaling and plasticity mechanisms.Science 298, 776–780.

Soriano, M.A., Tessier, M., Certa, U., Gill, R., 2000. Parallel gene expressionmonitoring using oligonucleotide probe arrays of multiple transcripts with

an animal model of focal ischemia. J. Cereb. Blood Flow Metab. 20,1045–1055.

Stephen, S.G.F., 2001. Evolving concepts in g protein-coupled receptorendocytosis: the role in receptor desensitization and signaling. Pharmacol.Rev. 53, 1–24.

Stephenson, D., Yin, T., Smalstig, E.B., 2000. Transcription factor nuclearfactor-Kappa B is activated in neurons after focal cerebral ischemia.J. Cereb. Blood Flow Metab. 3, 592–603.

Tanaka, H., Calderone, A., Jover, T., Grooms, S.Y., Yokota, H., Zukin, R.S.,Bennett, M.V.L., 2002. Ischemic preconditioning acts upstream of GluR2down-regulation to afford neuroprotection in the hippocampus CA1. PNAS99 (4), 2362–2367.

Trendelenburg, G., Prass, K., Priller, J., Kapinya, K., Polley, A., Muselmann, C.,Ruscher, K., Kannbley, U., Schmitt, A.O., Castell, S., Wiegand, F., Meisel,A., Rosenthal, A., Dirnagl, U., 2002. Serial analysis of gene expressionidentifies metallothionein-II as major neuroprotective gene in mouse focalcerebral ischemia. J. Neurosci. 22, 5878–5888.

Tusher, V.G., Tibshirani, R., Chu, G., 2001. Significance analysis of microarraysapplied to the ionizing radiation response. PNAS 98, 5116–5121.

Wang, Z., Zhou, J., Ju, Y., Zhang, H., Liu, M., Li, X., 2001. Effects of twosaponins extracted from the Polygonatum zanlanscianense pamp on thehuman leukemia (HL-60) cells. Biol. Pharm. Bull. 24, 159–162.

Wang, Z., Du, Q., Wang, F., Liu, Z., Li, B., Wang, A., Wang, Y., 2004.Microarray analysis of genes expression on herbal glycoside recipesimproving deficient ability of spatial learning memory in ischemic mice.J. Neurochem. 88, 1406–1415.

Whitney, L.W., Becker, K.G., Tresser, N.J., Caballero-Ramos, C.I., Munson, P.J.,Prabhu, V.V., Trent, J.M., McFarland, H.F., Biddison, W.E., 1999. Analysisof gene expression in multiple sclerosis lesions using cDNA microarrays.Ann. Neurol. 3, 425–428.

Xia, Z.Q., Gray, J.A., Compton-Toth, B.A., Roth, B.L., 2003. A directinteraction of PSD-95 with 5-HT2A serotonin receptors regulates receptortrafficking and signal transduction. J. Biol. Chem. 278, 21901–21908.

Yasutake, M., Kazunori, I., Taiichi, K., 2000. Two cis-acting elements in the3′untranslated region of α-CaMKII regulate its dendritic targeting. Nat.Neurosci. 11, 1079–1084.

Zdraveski, Z.Z., Mello, J.A., Marinus, M.G., Essigmann, J.M., 2000. Multiplepathways of recombination define cellular responses to cisplatin. Chem. Biol. 7(1), 39–50.