H ip p o c a m p aPl la s tic ityln v o lv e sE x te n s...

lournal of Molecular NeurosuenceCopyright O 1998 Humana Press Inc.All rights of any nature whatsoever reservedISSN0895-8696 / 98 / t0:75-98 / $L4.00

Hippocampal Plasticity lnvolves Extensive Gene Inductionand Multiple Cellular Mechanisms**

Dana Hevronirl Amir Rattnerrt Marsha Bundmanrl Doron LederfeinrlAwni Gabarahrt Miriam Mangelus,l Michael A. Silvermanrt Hilla Kedar/

Cathy Naorrt Masayo Kornucrs Tamar Hanochr2 Rony Segerr2Lars E. Theill,s EIly Nedivirt Gal Richter-Levin,a and Yoav Citril

l Departments of Hormone Research and 2Membrane Research and Biophysics, The Weizmann lnsti-tute of Science, Rehovot 76100, lsrael;3Molecular Biology Department, Amgen Center, Thousand

Oaks, CA 91 320-1 789; and aDepartment of Psychology, University of Haifa, Haifa 31905, Israel**x

Received January 1 5, 1997; Revised January 7, 1998; Accepted January 10, 1998

AbstractLong-term plasticity of the central nervous system (CNS) involves induction of a set of genes

whose identity is incompletely characterized. To identify candidate plasticity-related genes(CPGs), we conducted an exhaustive screen for genes that undergo induction or downregula-tion in the hippocampus dentate gyrus (DG) following animal treatment with the potentglutamate analog, kainate. The screen yielded 362 upregulated CPGs and 41 downregulatedtranscripts (dCPGs). Of these, 66 CPGs and 5 dCPGs are known genes that encode for a varietyof signal transduction proteins, transcription factors, and structural proteins. Seven novel CPGspredict the following putative functions: cpg2-a dystrophin-like cytoskeletal protein; cpg4 -aheat-shock protein: cpg16-a protein kinase; cpg2}-a transcription factor; cpg21-a dual-speci-ficity MAP-kinase phosphatase; and cpg30 and cpg38-two new seven-transmembrane domainreceptors. Experiments performed in vitro and with cultured hippocampal cells confirmed theability of the cpg-2l product to inactivate the MAP-kinase. To test relevance to neural plasticity,66 CPCs were tested for induction by stimuli producing long-term potentiation (LTP). Approxi-

*Address to which all correspondence and reprint requests should be sent Department of Biological Regulation, The WeizmannInstitute of Science, Rehovot 76100,Israel

**This article is dedicated to the memory of our mentor, fuiend, and colleague, Dr. Yoav Citri" who died tragically in a roadaccident on December 13.1995.

***Present addresses: Dr. Hevroni-QBl Enterprises, P.O. Box 74l,Nes-Ziona74106, Israel. Dr. Rattner-Department ofMolecnlar Biology and Medicine, Johns Hopkins University School of Medicine, Baltimore,MD 21205. Dr. Bundman-{enter forMolecular Neurobiology, University of Hamburg, D-20246 Hamburg, Gerrnany. Dr. Lederfein-{enter for Neurobiology andBehavior, College of Physicians and Surgeons of Columbia University, New York, NY 10032. Dr. Silverman-Department ofNeuroscience, University of Virginia School of Medicine, Charlottesvill e ,Y A 22908 . Dr. Nedivi-{old Spring Harbor Laboratory,Cold Spring Harbor,NY 11724.

-':{-

Journal of Molecular Neuroscience 75 Volume 10. 1998

76 Hevroni et al.

mately one-fourth of the genes examined were upregulated by LTP. These results indicate thatan extensive genetic response is induced in mammalian brain after glutamate receptor activa-tion, and imply that a significant proportion of this activity is coinduced by LTP. Based on theidentified CPGs, it is conceivable that multiple cellular mechanisms underlie long-term plastic-ity of the nervous system.

Index Entries: Dentate gyrus; gene expression, glutamate; hippocampus; neural plasticity;slmapsei signal transduction.

Introduction

Plasticity is a property of the brain that under-lies learning and memory processes, and is alsoinvolved in the refinement of neuronal connec-tions during nervous system development (Shatz,1990; Kandel and O'DeIL,1992; Bliss and Collinridge,1.993). Plasticity changes occur at the synapse,where they are expressed as activity-dependentchanges in the efficacy of synaptic transmission.Many plasticity-related changes in the mamma-lian central nervous system (CNS) are induced byactivation of excitatory glutamate receptors. Denoao activation of gene transcription participatesin long-term plasticity, and glutamate is likely tocause induction of the relevant genes (Goelet et al.,1986; Sheng and Creenberg, 1990; Huang et al.,1994; Nguyen et al., 1994). The cyclic-AMP-(cAMP) activated transcription factor, CREB, hasbeen implicated in memory-related plasticitychanges (Burtchuladze et aI.,1994). CREB and theproducts of many immediate-early genes (e.g.,c-fos and zif269) induce expression of downstreamgenes whose products are likely to produce syn-aptic changes (Frank and Greenbery, 1994). Thenature of these downstream candidate plasticity-related genes (CPCs) may illuminate the types ofprocesses responsible for long-term plasticitychanges.

The dentate gyrus (DG) of the hippocampus isan appropriate source for cloning plasticity-relatedgenes, because it exhibits marked plasticity fol-lowing electrophysiological or pharmacologicalstimulation (Ben-Ari and Represa,1990; Bliss andCollinridge, 1993). Activation of DG neurons bythe potent glutamate analog, kainate, results inprominent, plasticity-related, physiological andmorphological changes (Tauck and Nadler, 1985;Ben-Ari and Repres a, 1990 ; Sloviter, 1992). Impor-

lournal of Molecular Neuroscience

tantly, the granular neurons of the DG are not sus-ceptible to excitotoxic damage induced by kainateactivation (Ben-Ari, 1985; Smeyene et al., 7993).Many studies have examined the involvement ofspecific genes, primarily immediate-early genes(IEGs) (Cole et al., 1989; Sheng and Greenberg,L99};Morgan and Curran, 1991), and genes encod-ing neurotrophic factors (Gall, 1988) in hippocam-pal plasticity. Other studies used differentialcDNA cloning to isolate genes involved in plastic-ity (Qian et al., 1993; Yamagata et al., L994a,b;Lyford eta1.,1995).

To obtain an unbiased and comprehensive viewof the population of activity-modulated genes, weemployed a highly sensitive differential cDNAcloning protocol. Initially, 1000 cDNA clones of akainate-activated, subtracted DG cDNA librarywere screened, and 52 kainate- induced CPGcDNAs cloned (Nedivi et al., 1993). Here wepresent a 10-fold larger screen.Individual screen-ing of approx 10,000 subtracted cDNA clones, byusing a "clone-by-clone" sftategy, resulted in theidentification of 362 different CPGs. By contrast,41 transcripts whose transcription is apparentlydownregulated by glutamate were identified,implying that extensive gene induction, ratherthan repression, plays a dominant role in affectingneural plasticity. The identity of the cloned CPGsand dCPGs appears to reflect involvement of mul-tiple biochemical mechanisms in the response toglutamate.

Experimental Procedures

The "Clone-by-Clone" Screening StrategyThe "clone-by-clone" screen for CPGs com-

prised the following steps:

Volume 10. 1998

Cand i date P I asti c ity-Rel ated Cenes

1,. Constructing a subtractedlibraryfrompoly(A)+RNA of kainate-activated DG as we previ-ously described (Nedivi et al., 1993).BriefIy,ubiquitous activated DG cDNA sequenceswere subtracted by hybridization to excesspoly(A)* RNA derived from whole rat brain.

2. Picking phages (in }"-ZAP) from the unamplifiedsubtracted library and converting them indi-vidually into plasmids (pBluescript). PlasmidDNA was digested to release the cDNA insert,separated on duplicate agarose gels, andtransferred into nitrocellulose filters (see Table1; 10,015 library clones were thus processed).

3. Primary screen: hybridizing the duplicateblots to high concentrations of cDNA probes.One probewas prepared from poly(A)+ RNAof kainate-activated DG, and the second fromnonactivated DG.

4. Identifying clones that exhibited more pro-nounced reaction with the kainate-activatedprobe, as compared to the control, nonacti-vated, cDNA probe (Table 1: positives on 1oscreen). In a similar fashion, clones exhibit-ing a reduced hybridization signal with thekainate-activated probe (downregulatedCPGs-dCPGs) were also selected.

6 .

Secondary screen: repeating the same typeof Southern screen (steps 2-4) Ior all posi-tive clones identified in the primary screen(Table 1: positives on 2o screen). The probesof the secondary screen were synthesizedindependently from different RNA prepara-tions than those used in the primary screen.Clones positive in the primary and second-ary screens were subjected to partial nucle-otide sequence analysis. Sequence comparisonsto available DNA databases, including ourdatabase of previously cloned CPGs (Nediviet a1.,1993), were used to determine whethera differentially expressed cDNA was new oran already known sequence.

77

and samples were mixed vigorously at 37"C for 3h. One and a half milliliters were heat-inactivated(20 min at70"C) and centrifuged (8000 rpm in anEppendorf microfuge). Supernatants were trans-ferred into fresh tubes (stored at 4'C) and used (10pL) to infect E. coli SOLR (Stratagene) bacterialcells (100 pL) for 15 min at 37oC. Bacteria (20-30pL) were plated on LB agar plates containingampicillin and incubated overnight at 37"C. Singlecolonies were then incubated for 7 h at 37"C. Analiquot (3 mL) of each culture was pelleted (1 minat 90009), and DNA prepared according to thealkaline lysis method (Sambrook et a1., 1989).

cDNA SynthesisPoly(A)+ RNA from DG of kainate-activated or

from control rats was treated with RQ-1 DNase I(Promega, Madison, WI), followed by proteinaseK (Sigma, St. Louis, MO). The RNA quality wastested by visualization on agarose gel, determina-tion of absorbance (260 /280 nm), and by Northernblot analyses (probed for induction of proen-kephalin and cpgl and normalized to glyceralde-hyde-3-phosphate-dehydrogenase RNA levels[Nedivi et al., 1993]). For cDNA synthesis, 5 pgpoly(A)* RNA were incubated in 19 pL water with2.5 FLgoligo-dT (15-mer) primer (Promega) at7}oCfor 10 min, followed by cooling on ice. Reversetranscriptase (Superscriptll, 200 U ; BRL, Bethesda,MD) was added for 30 min at 45"C, in a final reac-tion volume of 50 pL containing H-RT buffer(BRL),30 U RNasin (Promega), the four deoxy-nucleotides (0.5 mM each), 10 mM DTT, and 30pCi [cr-32P]dCTP (Amersham, Arlington Heights,IL). A second aliquot of Superscriptll (200 U) wasadded and the incubation continued for 30 addi-tional minutes. The reaction was stopped on icewith 10 pL NaOH (1 N) and 2 pL EDTA (0.5 M).The volume was then adjusted to 100 pL, and themixture incubated for 20 min at 68'C for RNAhvdrolvsis. Ten microliters of HCI 1 N and 2 Uyeast IRNA (Sigma) were added. A sample (1 pL)was used for radioactive counting, and the rest ofthe cDNA was purified on a Sephadex C-50 spincolumn. The eluate was sampled (1 pL for radio-active counting, and the amount of cDNA synthe-sized calculated. For probe labeling, 75 ng cDNAwere incubated in 15 pL water with randomhexamer primers (0.5 U, Boehringer; 5 min at 68"C)

5 .

Plasmid Rescue from Individual Library PhagesIndividual plaques were picked into L mL of

1x).dil buffer (10 mM Tris-HCl, pH 7.5,1.0 m1//,MgSOa) containing 50 pL chloroform and stored at4'C. Two hundred microliters of phage supernatantwas used to infect 200 mL of overnight culture ofEscherichia coli XLl,-blue cells, in the presence of 1pL ExAssist helper phage (Stratagene, La Jolla, CA).After 15 min at 37oC,0.3 mL LB solution was added

Iournal of Molecular Neuroscience Volume 10, 1998

7B

Table 1Summary of the "C1one-by-C1one" Screen of the Subtracted Kainate-Activated DG cDNA

Hevroni et al.

Library for CPGs% of total clones screened 7o of recombinants

Total clones screenedClones with cDNA insert (recombinants)Recombinants producing a signalPositive clones on 1o screenPositive clones on 2" screenIndependent CPGs

Known genesNew genes

Independent dCPGsKnown genesNew genes

10,015750059381243655JOZ

662924I536

10074.959.372.46.53.60.72.90.40.050.4

10079.21.6.68.74.80.93.90.50.060.5

and put on ice. The cDNA was then added to areaction mix (300 pL, final volume) containing 10mM DTT, 90 mM HEPES buffer, pH 6.6,10 mMMgClr, deoxynucleotides (dATP, dGTP, and dTTP)at 0.5 mM each, 1 pCi [o-32p]dctP (Amersham,3000 Cilmmol), and 100 U Klenow DNA poly-merase (Amersham). The reaction was incubatedovernight at room temperature, 200 pg herringsperm carrier DNA were then added, and the mix-ture spun through a Sephadex C-50 column.

DNA HybridizationThe Southemblot filters were prehybridized over-

night at 42"C in 50% formamide, 5X SSCPE (0.6 MNaCl,75mM Na-citrate, 65mM KH2Pq, and 5 mMEDTA, pH7.2),0.5% SDS,5X Denhardt's solution(Sambrook et a1., 1989), and 0.5 mg/mL herringsperm DNA. The prehybridization solution was thenreplaced by a fresh solution containing labeledcDNA probes at 50-100 x 106 cpm/ml. The radioac-tivity of kainate-activated DG cDNA probe used forhybridization was equal to that of the control probe.Hybidizatronwas for 48hat42oC in a shakingwaterbath. Filters were washed three times (15 min each) atroom temperature and then three times (1.5 h each) at68'C, in 0.1X SSC, 0.1% SDS. Filters were blotted andexposed to X-ray films at-70'C, with an intensifyingscreen, for 16 h and then for 7 d.

Nucleotide SequencingThe autoradiograms of duplicate blots were

compared. Clones that appeared to be kainate-

Journal of Molecular Neuroscience

act ivated in two independent screens weresequenced (ABI automatic DNA sequencer) usingT3 and T7 primers for pBluescript. The sequencesobtained were compared to all databases and toour own database of CPG sequences.

In Vivo LTP

cDNA clones were randomly selected for LTPtests once their full-length nucleotide sequencinghad been completed. LTP was induced in the hip-pocampal DG granule cells via high-frequencystimulation (HFS) of the perforant path. MaleWistar rats (250-300 g) were anesthetized withUrethane (1..5 g/kg, ip), and surgery and electro-physiological recordings performed in a Kopf ste-reotaxic apparatus with the plane of the skulloriented horizontally. The stimulation intensity oftest pulses (100 ps, 0'.066H2) was adjusted to yielda population spike amplitude of about 1.0 mV andbaseline responses collected for 15 min. Both the ini-tial slope of the positive-going excitatory postsynap-tic potentiation (EPSP) and the amplitude of thegranule cell population spike were measured. HFS,at an intensity sufficient to elicit the maximumgranule cell population spike, consisted of 10 rep-etitions (1/min) of 50 pulses delivered in 5 bursts,each of 25-ms duration and 400 Hz; bursts wereseparated by 1 s (Jeffery et a1.,1990). DG EEG wasmonitored before, during, and after each HFSburst. No afterdischarges accompanied the induc- ition of long-term potentiation (LTP) in any of therats used in this study. Fol lowing HFS, theresponses to test pulses were monitored for :*-

Volume 10, 1998

Cand i d ate P I asti c ity- Re I ated G enes

60 min and the rats sacrificed 30 min, t h,3 h, or6 h later (two rats at each time-point). Only ani-mals that showed at least a 20% increase in theEPSP slope following HFS were analyzedby insitu hybridization. The means of increases inEPSP slope and in population spike amplitudewere 35.8 !15.6% and 143.8 + 26.3"/", respectively(n = 8). The brains were removed and kept frozenat-70"C until sectioning for in situ hybridizationanalysis. Blocking of the NMDA receptor wasperformed by injection of (RS)-2-carboxy-pip er azin-4-yl)-propyl-1 -phosphonic acid (CPP)(10 mglkg) 20 minprior to LTP stimulation.

ln Situ Hybridization AnalysisAnalyses were conducted essentially as described

(Hogan et a1., 1986). RNA probes were preparedin both sense and antisense orientations usinglinearized pBluescript SK templates of the CPGcDNAs, and T3 or T7 RNA polymerase. Onlyclones that showed increased hybridization on theLTP-stimulated side in two rats and at the sametime-point were considered LTP-induced'

Construction of a CPG2| Catalvtic SiteMutantMutant CPG21 (mCPG21) was generated by in

vitro mutagenesis, using the oligonucleotide: 5'-GTCCTGGTTCACCTCGAGGCCGGGGTCTCC-3'. For bacterial expression, the full-length openreading frame (1'L52 bp) was amplified by PCR,using the oligonucleotides: S'-CCCGAATTCATATGAAGGTCACGTCGCTC-3' and S'-CCCAACCTTGCAGGATGTGGCTGTGGC-3'. The ampli-fied DNA fragment was subcloned into pGEX2Tto be expressed and purified as glutathione S-transferase (GST) fusion protein (Guan and Dixon,1991). For expression in neurons, full-length wild-type and mutant cpg2l were subcloned through anXbaI cloning sites into a mammalian expressionvector downstream of an LTR promoter.

Bacterial Expression and In Vitro AssaysOf CPG2|

' Bacterially expressed ERK-2 was purified, acti-vated by MEK, in the presence of [y-32P]AtP and

\ purified on DE-52 minicolumns essentially as


79

described (Seger et aL,7994). The specific activityof ERK-2 was 300 nmol/min/mg. For dephospho-rylation reactions, the 32P-labeled ERK-2 (1 pg, 100pL) was incubated with 100 pL (4 pg) of eitherCPG21 or its mutant, and dephosphorylationreaction terminated by boiling in gel samplebuffer. Proteins were separated on sodiumdodecyl sulfate-polyacrylamide gel electrophore-sis (SDS-PAGE) gels that were dried and exposedto an X-ray film. For enzymatic inactivation reac-tions, a preparation of active ERK-2 (20 n9,10 pL)was incubated with 10 pL (10 ng) of GST-CPG21,or its mutant form. An aliquot (12.5 pL) wasremoved after various time intervals and sub-jected to phosphorylation reaction with myelinbasic protein, as described (Seger eI aL,1994).

Primary H i ppocampal Cultu resand TransfectionsPreparation of primary cultures of hippocampal

neurons (E 19), transfections, and chloramphenicolacetyltransferase (CAT) assays were performed aswe previously described (Rattner et al., 1993)'Quantitation of CAT activity was done using aphosphorimager and expressed in histogramsshowing percent conversion of laC-labeled chloram-phenicol to its acetylated products. Transfectionswere conducted L d after neurons were culfured invitro. Glutamate (1 mM) or MK-801 (2 St"M) treat-ments were for 12 h, beginning immediately aftertransfection. The c-fos promoter plasmids usedfor transfections were previously described (Cra-ham and Gilman, 1.991,).

Results

Strategy for Genetic ScreenCPGs were operationally defined as DC tran-

scripts that undergo induction 6 h after ip injec-tion of the potent glutamate analog, kainate(Nedivi et al., 1993). To clone kainate-inducedtranscripts, we extracted RNA from the DC oftreated rats, and then eliminated, through DNAhybridization, transcripts that were not affectedby the drug. This enabled construction of a sub-tracted cDNA librarv, whose clones were indi-

Volume 10, 1998

BO

\ oner-rtmfr cpgsonQro*IDrenascinI

f f i I II synaecann 5 I IGnspza-z

I

N I f f iH. cpg2I E E

Hevroni et al.

Ea-actinin

Journal of Molecular Neuroscience Volume 10. 1998

Cand i d ate Pl asti c ity-Rel ated Ce nes

vidually screened by differential hybridization tocDNA pools prepared from treated and untreatedanimali (see Experimental Procedures). Over 104clones of the subtracted kainate-activated cDNAllbrary were screened, and the results are summa-rized in Table 1. Approximately 4.8"/, of the tran-scripts that were examined conformed to thecriteria of CPGs. Partial nucleotide sequencingindicated that the majority of the isolated CPCsrepresented unknown genes. Full-length nucle-otide sequencing of seven of the novel transcriptsprovided clues to their function, because theybelonged to known protein families or contain rec-ognizable protein motifs (Table 1). In addition,41transcr ipts ( termed dCPGs) that underwentdownregulation on kainate treatment were iso-lated. To test directly the functional basis for iden-tification of the isolated CPGs, we comparedby insituhybridization the expression patterns of 79 ofthe transcripts in brains of kainate-treated anduntreated animals. Of the 79 clones tested, 68showed detectable transcripts by in situ hybrid-ization. We attribute the lack of detectable signalwith the 11 remaining clones to the low abundanceof some transcripts cloned by the extremely sensi-tive "clone-by-clone" method. Examination of 38clones for transcriptional induction by kainic acidtreatment of two animals revealed that mostclones were induced. Therefore, the other 30clones were tested only in one animal. Overall,gene induction was observed with 65 of the 68detectable transcripts (95%), demonstrating anunusually low percentage of false positives in thedifferential screen. Figure 1 depicts examples of lnsitu hybridization analyses performed with somerepresentative CPGs. Although each CPC dis-played a unique pattern of spatial expression, inall cases, induction in the DG was prominent.High-power magnification localized the inducedtranscripts in the DG to the granular layer (Fig. 1I-

B 1

N). Owing to the large size of the CPG group, onlyten of the 4L dCPGs isolated in the primary screenwere analyzedin one animal by in situhybridiza-tion (data not shown). This analysis, which wasperformed 6 h after kainate treatment, confirmedkainate-induced downregulation of 5 dCPGs atthe selected time interval, suggesting that the realnumber of dCPGs is lower than 41. However, eventhe maximal number of 41 is considerably lessthan that of the upregulated CPGs, implying thatgene induction, rather than gene repression,mediates neural plasticity.

Classification of the Known CPGsThe known CPGs may be tentatively classified,

on the basis of their deduced protein structures,into several main categories (Table 2). The majorgroup is comprised of proteins that are involvedin transduct ion of extracel lular s ignals, andincludes neurotrophic factors, membrane receP-tors, protein kinases and phosphatases, heat-shockproteins, and transcription factors. Another majorgroup encodes structural proteins, that presum-ably play a role in synapse generation and mainte-nance. The observed broad pattern of gene activitymay represent a pleiotropic cellular response thatreflects one or several signaling cascades. Some ofthe CPGs tested for their kinetics of inductionshowed a time-course of induction consistent witha sequential genetic response (Fig. 2): The IEGzif26B, encoding a transcr ipt ion factor/ wasinduced first, and levels of expression peaked atthe earliest time-point tested. cpgL, encoding a pre-dicted trophic factor,was induced somewhat laterand rapidly returned to basal levels. The proteaseinhibitor TIMP (16C8) was induced early, but itslevel of induction was maintained persistently forat least 48 h, consistent with a possible role in syn-aptic elaboration processes (Monard, 1988).

Fig. 1 . (previous) /n situ hybridization analyses of selected CPGs. (Top) The indicated CPC cDNAs were used tosynthLsize probes for in situ hybridization analysis of control (left) and kainate-activated (right) coronal rat brainsections. The identities of the selected CPGs (A-H) are as follows: ARPP-21 , a cAMP-regulated brain phosphopro-tein; cpg30, a seven-transmembrane span protein; PAt-2, an inhibitor of tissue plasminogen activator-2; Tenascin,an extracellular matrix protein; cr-actinin, a cytoskeletal protein; Syndecan, a heparin-sulfate proteoglycan coreprotein; hspTO-2, a heat-shock protein; and cpg20, a putative transcription factor related to Zif6B. (Bottom) Bright-iield rni.tographs of the DC granular layer of sections prepared from kainate-treated rats showing variabilities in thedegree of induction. Silver grains represenl in situ hybridization patterns specific tor Tenascin (1,l\, PAI-2 (K, L), andcpg20 (M, N) photographed at magnifications of .l0X (1, K, and M)or 20X (J, L, and N).

Journal of Molecular Neuroscience Volume 10. 1998

B2 Hevroni et al.

Table 2Known CPGs and Their Possible Relationship to Plasticity'

CPG Reference CPG ReferenceIEGs

Transcription factors*c-junjunB"zif268"c-fos"fosB*CREM*BTE*Nur77Nor-2*cpg20, *E4BP4*EGR3Siqh*NGFI-CLRF.1RNR-1*TIS1.7IKB

Heat-shock proteins"cpg4, "hsp27"hsp70 and*hsp70-2

Trophic factors*rpg1*MARC*MCP-1*BDNFActivin*PC3PEDF

Signal transduction*MyD718*ME491*ARPP-2I.Metallothionein*rheb*STAT3*IGF-BP3*RC3

Kitabayashi et al., 1990Kawakami et a1.,1,992

Sheng and Greenberg, 1990Curran et a1.,7987

Zerial et a1.,7989Foulkes eta1.,7997Imataka et a1.,1992Ohkura et a1.,1994

Petropoulos et al., 1995Cowell et aL.,7992

Patwardhan et a1.,7997Della et a1.,7993

Crosby et a1.,7997Hsu et a1.,1991,

Scearce et a1.,1993Kaneda eta1.,7992Tewari et a1.,7992

Caestel et a1.,1993Powell and Watts. 1990

Retrograde messengers*Cox2 cyclooxygenase O'Banion et al.,'1,992*Heme oxygenase Muller et a1.,1987*Arginino succinate lyase Amaya et al., 1988

Cenes possibly involvedin synaptic remodeling

Synaptic transmission-related

Protein kinases*cpg76, *FLT*PKR

*dy,PE

Finnerty et aL.,1,993Mellor et al.,7994

Civelli et al., 1985Noda et al.,1982

Criado et al., 1988Pragnell eta1.,7991

Kojima eta1.,1992

Kojima et a1.,1992Youssoufian et al., 1990

Saga et al.,\997Bhatet a1.,1991,

Turley eta1.,7993Julien et al., 1985

Link et a1.,1,995Lyford et a1.,1995

]ohnston et a1., L992Davis et al.,1993

Edwards eta1.,1986Antalis et al., 1988

Zmudzka et aI.,7986Yamaguchi et a7.,7987Majumdar etaL,7997

*Clathrin Kirchhausen et a1.,1987hsc70 Chappell et a1.,7986*Secretogranin Fiicher etal.,I990COMT Salminen et al.. 1990cpg22, aPDZ domain Brakeman et a1.,7997

protein relatedto Homer

*cpg30, a new 7-TM receptor, Salton,1991*cpg38, a new 7-TM receptor,

Trophic factors and signaltransduction proteins"VGF

*nAchR-e*B-DHPR

Structural proteins"cpg2, a putative new

cytoskeletal protein,*Syndecan

Ryudocan"a-actinin*Tensscina-B-crystallin*Fibrilarin*NF-L*Arc/Arg3.1

"SC2*ABGP

Protease inhibitors*16C8*PAl,2

DNA replication*DNApol.B

HMG2

" S - ad eno sy lmethio ninedecarboxylase

* Cn I b i n d i n /Irl e u r o n a t i nProtein phosphatases

*PTP1.B*cpg27 , BSM-1

Nedivi eta1.,7993Kulmburg et a1.,1992

Yoshimura et a1.,1997Maisonpierr e et al., 1990

Fanget a1.,1996Bradbury eta1.,1991,

Steele et aL..1993

Abdollahi et a1.,1991Rapp et a1.,7990

Ehrlich and Greengard, 7991.Andersen et aI.,7986

Yamagata et a1.,7994bRipperger et a1.,1995Shimasaki eta1.,L989

Watson et a1.,7990Pajunen et al., 1988

Lomri et a1.,1989

Guan et a1.,1990accession number D45412

oClones analyzed for induction by LTP are indicated by asterisks and those that scored positive appear in bold.

Journal of Molecular Neuroscience Volume 10, 1998

Cand idate Pl asticitv-Related Cenes B3

Kainate(hrs): 3 6 12 24 48 0

<- cpgl<- zif268<- G3PD+ 1 6 C 8

Fig.2. Kinet ics of kainate induct ion of three CPCs. DG poly(A)+ RNA (0.5 pg)from control ( lane 5) or kainate-treatJd rats, that were sacrified at the indicated time after treatment, was used for Northern blot analysis' The filterwas first hybridized to a control probe, glyceraldehyde-3-phosphate-dehydrogenase (G3PD), and then to a mixtureof probes specific to the following three genes (seeTable 2): cpgl, zif269, and 16C8. Similar analyses were per-formed with additional CPCs and induction was found to be either transient (e.g., c-jun, c-fos, MyDl1B, hsp27, andpTPl B), intermediate (e.g., cpg4, CREM, and Clathrin), or persistent (e.g., Secretogranin, cpg2, and DYn).

65432

Importantly, kainate activation of the DC notonly induced, but also repressed gene expression.Out of the 4L downregulated genes (dCPGs) thatwere cloned and sequenced, only five are knowngenes. Notably, in situ hybridization confirmedthat one of the isolated dCPGs encodes a glutamatereceptor subunit (kainate-KA1). This finding prob-ably reflects a mechanism of protection fromoverstimulation by kainate. Interestingly, twoCPGs encode postsynaptic excitatory proteins(the nicotinic acetylcholine receptor, nAChR-e,

lournal of Molecular Neuroscience

and the B-subunit of the brain dihydropyridine-sensitive calcium channel, B-DHPR; Table 2).

CPG Induction by an LTP-Producing StimulusLTP is an important model of memory-related

synaptic plasticity (Bliss and Collinridge, 1993). T oaddress the relevance of CPGs to LTP, 58 of the74CPGs listed in Table 2 (marked by asterisks), aswell as B additional novel CPGs, were tested forinduction by an LTP-producing stimulus in vivo(Fig. 3, and data not shown). Of this grottp,IT

Volume 10, 1998

B4

5 1 0 t 5 2 0 2 5 3 0 3 5 4 0 4 5 5 0 5 5 6 0Time (min)

CPGs (15 marked by bold letters in Table 2, and2novel CPGs) exhibited greater induction in the


Hevroni et al.

r@nrffiirMllltffiHflilt5 1 0 1 5 2 0 2 5 3 0 3 5 4 0 4 5 5 0 5 5 6 0

Time (min)

stimulated (left side of the hippocampus; Fig. 3),than in the contralateral hippocampus. All of these

Eooosanat)U

oI

oI

Edo

F ig. 3. CPG induct ion by LTP-producing st imul i . (Top) Potent ia t ion of f ie ld potent ia l responses is shown asaveraged data (n = 6) recorded in the granular layer of the dentate gyrus. Populat ion spike (PS) ampl i tude( t r iangles) and in i t ia l EPSP s lope (c losed c i rc les) are expressed as a percentage of the mean basel ine responsesbefore appl icat ion of h igh- f requency st imulat ion (HFS) to the per forant path (marked by arrows at t ime zero) .(Bot tom) ln s i tu hybr id izat ion analyses of representat ive CPCs. LTP was conducted on the le f t hemisphere,whereas the r ight hemisphere served as a contro l . Data presented are f rom rats sacr i f ied t h fo l lowing HFS. Brains l ices were hybr id ized wi th r iboprobes speci f ic for the indicated CPGs (see Table 2) .

Volume 10, 1998

C a n d i d ate P I asti c ity- Re I ated G e n es

LTP-induced transcripts displayed elevated levelst h after HFS, and some maintained high levels foras long as 6 h after stimulation (data not shown)' Inorder to confirm that the observed induction byLTP occurs through activation of the N-methyl-o-aspartate (NMDA) receptor, control rats weretreated with the NMDA receptor blocker CPPprior to LTP stimulation. Seven CPGs (MyD11B'ipg20, BDNF , EGR, VGF , ARPP2L , and c-iun) weteteited by in situhybridization, and none of themexhibited a significant induction after CPP injec-tion (data not shown), in accordance with the pre-diction that CPG inductionby LTP depends on anactive NMDA receptor.

Novel CPGs: A Putative TranscriptionFactor and a Protein KinaseComplete nucleotide sequences were deter-

mined for 15 of the cloned novel CPGs. Thededuced protein sequences of 7 of the L5 CPGscontained sufficient homologies to known pro-teins, thereby enabling prediction of their cellularfunctions (Table 2 and Fig. a). These are cpg2, adystrophin-like cytoskeletal protein, cpg4, I Pte-sumed heat-shock protein, cpg20, a potential tran-scription factor, cpg16, a putative protein kinasedistlntly related to calcium and calmodulin-(CaM)-dependent kinases, cpg21, a predicted proteinphosphatase related to mitogen-activated proteinkinase (MAPK) phosphatases, and two putativereceptors, each containing seven-transmembranedomiins, that are encoded by cpg30 and cpg3B. Ahumanhomolog of CPG30 hasbeen recently iden-tified as a pituitary receptor involved in growthhormone release (Howard et aI.,1996).

To exemplify their relationships to known pro-tein families, the primary structures of three novelCPGs are compared with their closest relatives(Fig. a). The full-length amino acid sequence ofCPC20 predicts a new protein with three zinc fin-gers homologous to the transcription factors Sp-1indZlf268/Egr-1 (boxed in Fig.4A). The latter isprobably the most thoroughly studied gene thatiesponds to LTP-producing stimuli (Cole et al',1989; Silva and Giese, 1994).Similarly, we observedthat cpg20 undergoes upregulation by an LTP-pro-ducing stimulus (Fig.3B). However, the zinc fin-gers of CPG20 are more related to those 9! Sp-flhan to the fingers of Zif26B/Egt-1', and all resi-

Jaurnal of Molecular Neuroscience

B5

dues determining DNA sequence specificity areshared, suggesting that CPG20 and Sp-1,, but notZtf268/Egr-1, bind to.similar or identical sequenceelements of genomic DNA. Recently, two possiblehuman orthologs of cpg2.0 were isolated as growthresponse transcripts (Blok et al., 1995; Subramaniamet a1.,1995).

In situhybridization analyses, which are not pre-sented, indicated that cpgL6 is transcriptionallyactive in several brain regions, including the cortexand the hippocampus. The primary structure ofCPG16 predicts a protein kinase function, becauseall of the 11 subdomains shared by the proteinkinase superfamily (Hanks, 1991) ate conserved inCPG16 (Fig. aB). Specifically, CPG16 is more closelyrelated to ser/thr-specific kinases than to tyrosine-specific protein kinases. Although, the closest rela-tives of CPG16 are all members of the calcium/calmodulin-dependent protein kinases (CaM kin-ases, including CaM kinases I and II, and myosinlight-chain kinase), the C-terminallyJocated cal-modulin binding region of established CaM kinases(Maedor er a1.,1993) differs from the correspondingregion of CPG16 (Fig. B). Indeed, experiments per-foimed with a bacterially expressed CPG16 fusionprotein indicated that CPG16 may not function asan ordinary CaM kinase (M.A.S. and R. S., manu-script in preparation): First, the protein did not bindto an agarose-immobilized calmodulin. Second,although CPG16 undergoes autophosphorylationin vitro, as expected for many ser/thr-specifickinases, its activity was not affected by two CaMkinase inhibitors, calmidazolium and W-7. Finally,in vitro assays have shown that the catalytic activityof CPG16 is independent of calcium and calmoduliruand when expressed in COS-7 cells, a calcium iono-phore exerted no effect on the enzymatic activity ofCpCfO. In conclusion, despite structural related-ness, CPG16 is apparently regulated by a mecha-nism distinct from that controlling the enzymaticfunction of CaM kinases.

CPG21, a MAP-Kinase Phosphatase Capableof T ranscri ption RegulationTranscripts of cpg21 are undetectable in rat

brain, except for the CA2 region of the hippocam-pus. Treatment with kainaic acid elevated tran-scription of cpg2l in the DG,CAZ, and CA1 regions@a1a not shown). The deduced amino acid

Volume | 0, 1998

B6 Hevroni et al.

A CPG2 O MLNFGASLQQASEGKMELT SEKSKEGAHPWDKAEQsDFEAVEALMsMscDWKsHFKKYLE 6 O

CPG2 O NRPVTPVSDTSEEDSLLPGTPDLQTVPAFCLTPPYSPSDFEPSQGSNLTAPAPPTGHFRS 12 O

CPG2O LSDAAKPPSIAPFKEEEKSPLAAPPLPKAQATSVIRHTADAQLCNHQSCPVKAASILNYQ 180

CPG2O DNSFRRRTHINVEATRKNIPCAAVSPNRPKPEPSTAANGAEKAGTAPYDFAVPSSETVIC 240

C PG 2 U RS SQPAPTS PVQKSWMSS PTVSTGGVPPLPVICQMVPLPANNSLWTWPSS PPSQPPA 3 O O

CPG2 O VCSPVLFMGTQVPKGTVMFWPQPWQSPKPPWSPNGTRLSPIAPAPGFSPSAARWPQ 3 6 O

CPG2OsP1

z I F 2 5 8

CPG2 OsPt_

L r E Z O 6

o 6 u3 6 0

KLNDIALPPATASAQ 480

EI CPg16 MLELIEVNGTPGSQIJSTPRSGKSPSPSPTSPGSI'RKQRISOHGGSSTSLSI camdk-GR MLKVTVPSC

#I

CPg15 STKVCSSMDENDGPGEEESDEGFQIPATITERYKVGRTIGDGNFAWKECCamdK-GR PSSP- - -\,rTSSTENLVPDYWIDGSKRDPLSDFFE- ESEL-R-ATSI -YR -

mlcK-dicdi MTEVEKI-EFKEEL-R-A.SI -YLG

i l ' i l * t lIERSTAREYAIKIIKKSK CRGKEHMIQNEVSILRRVKHPNIWLIEEMD 149KQKG-QKP-- - -vL- -T VDK KIVRT- IGV. I , -LS- - - - IK-K- IFE 105ENKQ-KQR- - I -V-N-SELGKDY-KNt KM- -D- -KK-N- - - - rA-K-LF- ' l 5

V VIa-l-t-1-VPTELYLVMELVKGGDLFDATTSTSKYTERDASGMLYNLASAIKYLHSLN 1 9 9T- - - rs- -L- - -T- -E- - -R-VEKGY-S - - - -ADAVKQTLE-VA- - -ENG 155T-EK-y- - - - - -T - -E- - -K-vEKcs-s -A- -ANLVKKTVS -vc- - -cL- 12 5

VIb V I I V I I I

IVHRDIKPENLLVYEHQDGSKSLKLGDFGLATIVDGPLY TVCGTPTW 2 47- - -- -L- - - - - - -ATPAPDA- - - rA- - - -SK- -EHQWMK- - - - - -c-C 204- - -- -L- - - - - - ITKSKENHLEVATA- - - -SK-rGOT-VMO-A- - - -S - - 174

" i l t lAPEIIAETGYGLKVDIWAAGVITYILLCGFPPFRGSGDDQEVLFDOILMG 2 97- - - - -RGCA- -pE- -M-SV- r- - - - - - - - - E- -yD ERGDQFM-RR- -NC 2 53- - -VLNA-G-DKE--M-Sr - - - - - - - - - - - - - -y - DTVp- r -EQ-MEA 222

50

100) t2 5

cpg16Camdk-GR

mlck-d icd i

cpg16Camdk-cR

ml-ck-d icd i

cpg15Camdk-cR

mlck-d icd i

cpgL6Camdk-cR

mlck-d icd i

X I

"on* DDGLpE 347CamdK-GR EYY- IS-W_-E- -LN- -DLVKKLIWDPKK-LTTF-A-Q-_- -TGKA A 301

mlcK- -d icd i NYE--EE--GGI-KE--DF-GKL-v_D-sK-LN-TNA_N-- - IJKSNN s 270

Cpg16 NEHQLSVAGKIKKHFN?GPKPSSTAAGVSVIATTALDKERQVFRRRRNQD 3 97CamdK-GR _ FVHMDTAQ_KLQE- -ARR-LKAAVKA.vAssRLGsAsSsHTNIoEs _ KA 3 5 1

mlck-dicdi -r! ' t Ixl|I ' -M-EYIVERR-N-ENWLTKRTFQ 3Ol.

Cpg16 VRGRYKAQPAPPELNSESEDYSPSSSETVRSPNSPF 43

Fig. 4. Predicted amino acid sequences of selected CPCs and comparisons with their closest homologs. Dashesindicate identical residues; gaps were introduced for maximal alignment. (A) CPC20, a zinc-finger protein of the Sp1/Egr family. The full- length nucleotide sequence of the cpg20 cDNA was determined, and the predicted amino acidsequence is shown. The boxed sequences indicate the three zinc fingers of CPG20, and their sequences are (continued)

FSCSWKGCERRFARSDELSRHR-M-N-SY-cK- -T- - - - -Q- -K

- G ] 1 6


Cand i d ate P I asti c itv- Re I ated C en es

fr cPG21v h\'rH3

MKP_1PAC-l-

MKVTSLDGRRLRKMLRKEAEARCWLDCRPYLAFAASSVRGSLM/NLNSWL52M - - - - - - - - - - - - - - - - - - A - - - - - - N - - - - - 5 2

MWEVGI _ -AGG- -AL -_ER _AQ_LL - -__SFF__N-GHTV-_V '_RFST I_R 53MPIAMGLETACE-ECAA_GAL__EPR__E-TLL_ - - ' _F__ -cRsH-RAARP_PW_ALLR50

B7

Fig. 4. (cont inued) compared to the corresponding zinc f ingers of the proteins Sp1 and Zif268 (also cal led Egr-1).Asterisks denote the conserved cysteine and histidine residues of the zinc fingers. (B) CPGl 6, a novel kinase relatedto ser ine- and threonine-specif ic protein kinases. The deduced amino acid sequence of CPG16 is compared to i tsclosest k inases: the calcium/calmodul in-dependent protein kinase Gr (Ohmstede et al . , 1991), and the Dictyostel iumdiscodeum myosin l ight-chain kinase (mlck-dicdi) . Subdomains of CPG15 shared with other kinases (numbered l-Xl) (Hanks, i 991) arelndicated. Aster isks mark amino acids conserved through the whole family of k inases. Residuenumbers are indicated in the r ight column. (C) CPC21, a protein phosphatase. The deduced amino acid sequenceof CpC2l is al igned with the fol lowing PTPases: hVH3 (human) (Kwak and Dixon, 1995), MKP-1 (Keyse andEmsi le, 1992; Sin et al . , 1 993), and pnL-t (Ward et al . , 1gg4). Residue nu mbers are shown in the r ight colu m n.The pTpase signature sequence is indicated (st ippled box)and the Cys (bold)to Leu act ive si te mutant of CPG2T(nCPG21) is shown.

CPG2 1 RRARGG AVSARYVLADEAARARLLQEGGGGVAAVWLDQGSRHWOKLREESAARWLTSh w r 3 - - - - - - - - P - -M K P - 1 _ - - K - - M G L E H I V P N T E L _ G . - _ A - A Y T P _ - L _ _ E R - A A L D G R K R D G T L A L A A G AP A c - ] - _ _ - P _ T P - A A L A C L - P _ R - L _ _ - _ G R _ E L - R A _ _ _ - E S _ A S V T E _ P P D G P A H L L - A A

CPG2 1 GAYCTFPTSVLAPVPTHATVAELHRSPVATATSCh v H 3 - - - - - - - A - - - - - - - - - s - - s - - s - - - - - - - - - -MKP_ 1 TV.NFP-S] _ _HPTNSALNYLQ_ _ IT_SP_ _

CPG21. LLACLSAGP RVYFLKGGYETFYSQYPECCVDAKPISQEKLEGERGLLSQCGKPILSVAY 170h ! T l 3 - - - - - P - - - - - - E - - - - - - - V - - - - - - - I - S - - A - I - - - - - - \ ' \ ' t ' l - S - L 7 0MKp-1 -CREARS TQ-F--Q----A-SASC--L-SKQSTPMGLS- PLSTSVPDSAESGC-SCS 165

MM ̂ - -R__FKS_OTYC_DL_SE-PAQ A - P PAGAENSN_DPR 168H A L - r - \ 2 n U Y K U - - r r - L

CPG2 1 RPAYDQGGPVEILPFLYLGSAYHASKCEFLANLHITALLNVSRRTSEACTTHLHYKWIPV 234h v H 3 - - - - - M - - - - - - - - - - 2 3 0M K p - 1 T - r , - - - - - - - - - - s - - - - - - - - - - - R K D M - D A - G - - - - r - - - A N C P N H F E G - Y Q - - S - - - 2 2 5p A c - 1 v - r - - - - - - - - - - - Y - - - - - C N - S - D L Q G - Q A C G - - - V - - - - A S C P N H F E G L F - - - S - - - 2 2 8

mCPG21 r: 'cpG2 1 EDSHTADISSHFOEAIDFIDCVREEGGIffLy'HgEACVqRSPTICMAYLMKTKQFRLKEAF 2 9 0h v H 3 - - K - - - : : : : q - . - . - I : . : - - - - 2 9 0MKP- 1 - -N-K- - - - -W-N- - - - - - -STKDA- -R-.q- :cQ::1- . - . -A- - -L- - - -R-NR\,X-D- - - 285pAC- 1 - -NQMVE--AW- - - - -S- - - S-KNS- -R.- : - : - :QQ-.- I : . . :A- - -L- - - TQSHRV- -D- - - 288

CPG21 EYIKQRRSWSPNFGFMGQLLQYESEILPSTPTPQPPSCQGEAASSTFIGHLQTLSPDMQ 350h v H 3 D - - - - - - - M - - - - - - - - - - - - N G - S L - - - - - - - - - - - - 3 5 0MKP-1 -FV - - - - - r r - - - - s - - - - - - -F - -QV- A -H -SA- -G-PAMAVLDRGT-TT 336p A i - - 1 n F \ / - - - - c - T - - - - S - - - - - . - L - T O V - C H 3 1 8

l l ll l 1r 0 8L L 7

J d q

3 8 4361

blocks of conserved sequences, are common to otherprotein tyrosine phosphatases (PTPases), includingtwo related dual-specificity phosphatases thatdephosphorylate and inactivate MAPK, namelyMKP-1 (Keyse and Emsile, 1992;3net a1.,1993) andPAC-1 (Ward et al., 199 4) (approx 45% identity).

In order to examine the predicted phosphataseactivity of CPG21, and its expected specificity toMAPK, we expressed a full-length cDNA oI cpg21

sequence of CPG21 predicts a 384-residue-longprotein with a high degree of homology to two hu-man dual-specificity protein phospatases, bVH3(94% identlty) and 823 (84% identity) (Kwak andDixon, 1995). The two human proteins differ intheir C-termini, but they share a cysteine-contain-ing canonical phosphatase motif that slightly dif-fers from that of CPG21 (Fig. 4C, and data notshown). This signature motif, as well as additional


BB Hevroni et al.

ATime cPG21(min) 15 30 60 0

BM60

MW

87

48

33

: =aO F( E <y Eo - E< c L

csXSRE

r J

10 20Time (min)

+ m -wt

cpg2l:G lu :

+++ + +

n[:IhI

T4 0 +".[J

+ - +

1 2 3 4Fig. 5. Funct ional analyses of CPC21 in vi t ro and in l iv ing cel ls. (A) Dephosphorylat ion of MAPK. Bacter ial ly

expressed MAPK was phosphorylated in vitro by using [y-32p]ntp and MAPK-kinase (MEK). The radioactively labeledMAPK was incubated with a bacterially expressed wild-type CPC21, or with an active site mutant (m), for theindicated time intervals, and the reaction mixtures resolved by gel electrophoresis (mol-wt markers are indicated inkilodaltons), followed by autoradiography. (B) Inactivation of MAPK. MAPK (ERK-2) was phosphorylated by MEK asin A, except that nonradioactive ATP was used. The active enzyme was purified over an ion-exchange minicolumnand incubated at 37"C with either a GST-CPC21 fusion protein (closed circles), or with an active site mutant (squares)for the indicated time intervals. For control, the wild-type fusion protein was incubated with 5 mM vanadate (opencircles). Kinase assays were performed at the end of incubation by using myelin basic protein as a substrate. (C)Inac t iva t ion o f SRE-media ted t ranscr ip t ion in h ippocampal ce l l s . Pr imary cu l tu res o f ra t h ippocampal ce l l s(embryonic d 19) were transfected with a c-fos promoter construct (1 pg,/dish), ei ther a plasmid containing a 356-bp-long 5'-sequence of c-fos (marked wt), or a plasmid containing 5 copies of SRE (sxsR} fused to a (coi t inued)

1 11 0

; * t t l * * { p

5 6 7 8


C a n d i d ate P I asti c i ty- Re I ated G en es

in bacteria and tagged it with a GST moiety. Ascontrol, we attempted to block the enzymaticactivity by introducing a cysteine to leucinereplacement in the catalytic core of CPG21 (Fig.4C). When incubated with a phosphorylated MAPK(ERK-2), purified preparations of CPG21', but not itsmutated version (denoted mCPG21), rapidly dephos-phorylated the kinase (Fig. 5A). Because MAPK phos-phorylation was performed by using MEK an MAPKkinase that activates MAPK dephosphorylation byCPG21.is expected to inactivate MAPK. lrdeed, a bac-terially expressed GST-CPG21 fusion protein inhib-ited the ability of MAPK to phosphorylate myelinbasic protein in vitro (Fig. 5B). As expected, the cata-lytically impaired mutant of CPG21 exerted no effecton MAPK activity (Fig. 5B). Vanadate, a potent tyro-sine PTPase inhibitor, partial$ abolished the effect ofCPG21, consistent with the possibility that CPG21functions as a dual-specificity PTPase.

Interestingly, similar in vitro assays that tested theability of CPG21 to inactivate the MAPK-related Jun-N-terminal kinase (JNK or SAPK) detected no effecton the ability of purified JNK1 to phosphorylate arecombinant Jun protein in vitro (data not shown).Taken together, the assays performed with CPG21indicated that this protein can act in vitro as aMAPK-specific phosphatase, whose catalytic siteshares similarity with the related MKP-1 and PAC-1PTPases (Sun et al., 1993; Ward et aI., 1994).

We next attempted to examine the effect of CPG21,on MAPK in living cells. As an experimental system,we selected primary cultures of hippocampal neu-rons, because glutamate activation of the NMDAreceptor in these cells was shown to cause phos-phorylation of MAPK (Bading and Greenberg,1991) and induction of c-fos expression (Bading etal., 1.993), through the serum response element(SRE) of its promoter. The adequacy of the hippoc-ampal culture system was confirmed in initial stud-ies, whose results are not presented, demonstratingCPG1L expression only on treatment with glutamate.Irr addition, we found that cpg2l induction was pre-

B9

ceded by upregulation of c-fos, and it could beblocked by the NMDA receptorblocker MK-801.

Two previously described CAT reporter con-structs (Graham and Gilman, 1991,) were employedto follow SRE-mediated control of cios expressionby CPG2L: a 356-bp-long upstream region of c-foscontaining a single SRE (denoted wild-type, wt), andanother construct carrying five copies of SRE in frontof a 56Sp-long minimal promoter of c-fos (SXSRE).Both plasmids, when transfected into hippocampalneurons, induced basal expression of the reporterCAT enzyme. Cell treatment with glutamate signifi-cantly enhanced CAT expression (Fig. 5C), confirm-ing coupling of NMDA receptors to SRE (Bading etaI., 7993). However, cotransfection of hippocam-pal cells with an expression vector drivtng cpg21reduced basal, as well as glutamate-induced expres-sionof CATfromeither aWT c-fos promoterorthe5XSRE construct (Fig. 5C). An expression vectorencoding a catalytically inactive CPG21 exertedonly a small inhibitory effect on glutamate-in-duced expression from the 5XSRE construct (Fig'5C), in line with a model attributing transcriptioncontrol to inactivation of MAPK-mediated regula-tion of SRE.

It is relevant to note that the inhibitory activity ofCPG21, was more prominent when SXSRE, ratherthan a full-length promoter, was used as a reporter(Fig. 5C). In addition, inhibition by CPC21 wasobserved even when the NMDA receptor pathwaywas blocked by MK-80L. Together with the incom-plete shutdown of transcription by CPG2L, theseobservations may reflect an ability of CPGZ1' to blockalso non-NMDA receptor-coupled pathways. Inadditioru these results suggest involvement of ele-ments other than SRE in c-fos induction by glutamate .Nevertheless, the relatively late induction of cpg2L(induction is observed after c-fos expression is reducedto basal level, data not shown) and its biochemicalfunctions, implicate this novel PTPase in a mecha-nism that downregulates neural resPonses to exces-sive glutamate-induced gene expression.

56-bp-long 5, sequence of c-fos. In both plasmids, a CAT coding sequence was fused to the promoter region.Where indicated, cul tures were cotransfected with ei ther an expression vector encoding CPC2! (5 pg/dish, lanes1,2,5, and 6, or 9 pg in lane 9), or an act ive si te mutant (marked m, 9 [rg), and cel ls were treated with ei therglutamate (1 mM,lanes marked +), orthe NMDA receptor blocker MK-801 (5lLM,lanes marked -) , immediatelyiol lowing plasmid transfect ion. CAT act iv i ty was determined 12 h later by the abi l i ty of the corresponding cel lextracts to acetylate labeled chloramphenicol . Results are expressed in arbi trary units ( lower panels). The experi-ment was repeated three l imes.


90

Discussion

Conventional differential cDNA cloning may allowidentification of candidate genes involved in produc-ing phenotypic changes. The major problem of thisapproach is sensitivity; genes that are not highlyexpressed may evade detection. By contrast, the"clone-by-clone" strategy yielded a large numberof differentially expressed genes. Two features ofthis strategy are prominent (Table 1): First, high sen-sitivity of detection; almost B0% of the library'srecombinants produced a signal in the differentialscreen. Second, the strategy enables high yield of dif-ferentially expressed independent cDNA clones. Thislarge number is probably owing to the high sensitiv-ity of the screery and to faithful representation of theDG RNA within the subtracted library.

Regardless of its possible technological importance,the present study may be significant to the molecularmechanisms that underlie neural plasticity. In thiscontext it is relevant to ask whether all the identifiedCPGs are involved in plasticity, or if they mediatedisparate processes that merely reflect glutamateexcitotoxicity (Ben-Ari, 1985). For example, plastic-ity and excitotoxicity were shown to activate thesame set of genes (e.g., c-fos, c-jun, zif269) (Smeyeneet a1.,1993; Dragunow et aI.,1994).In addition, plas-ticity and cell death are two overlapping processesthat take place during CNS development (Shatz,1990;Raff et a1., 1993). The following lines of evidence sup-port involvement of a significant fraction of the iso-lated CPGs in plasticity, rather than in glutamateexcitotoxicity:

1. Several CPGs identified bv the present screenwere previously associaiea wittr plasticity.These include IEGs, genes encoding neuro-trophins, neuropeptides, signal transduc-tion proteins, and dendritic components(Yamagata et al., 1993, 1994a,b; Silva andGiese, 1994;Lyford et a1.,1995).

2. A subset of six CPGs are specifically inducedin visual cortex in response to light, and arealso naturally induced during postnatal corti-cal development (Nedivi et aI.,1996). One ofthese, CPG15/neuritin, encodes a glycosyl-phosphatidylinositol-anchored protein that pro-motes neurite outgrowth in primary embryonichippocampal cultures (Naeve et a1.,1997), andenhances dendritic arbor growth of tectal cells


Hevroni et al.

in vivo (Nedivi et aI., 1997). Another, CPG22, isa PDZ protein related to Homer, a dendriticprotein that selectively binds metabotropicglutamate receptors (Brakeman et a1.,1997).

3. CPG induction appears to be localized prima-rily to the grantilar cells of the DG, rather thanto basket or glial cells (note that in Fig. 1,tenascin is the exception, being one of the fewCPGs induced in glia). Since granular neuronsare resistant to kainate toxicity (Nadler, 1978;Ben-Ari, 1985), it is likely that CPG inductionin these neurons is not related to excitotoxicifibut rather to the propensity of granular cells toexhibit massive mossy fiber sprouting, as wellas physiological changes, on kainate activation(Tauck and Nadler,1985; Ben-Ari and Represa,1990; Sloviter,1992).

4. Seventeen out of 66 CPGs that were testedundergo upregulation in response to an LTP-producing stimulus (Fig. 3).

We find it especially important that a significantproportion of the identified CPGs undergo inductionduring LTP. This fraction is likely to be a minimalestimate because LTP was induced under anesthesiathat partially blocks glutamate receptor activation(Demmer et aI.,1993; Frank and Greenberg,7994).Moreover, the following six CPGs: BTE, cpg20,W D LL B, ARP P -21, V GF, and cr-actinin are describedhere for the first time as genes coinduced with LTPin vivo. Notably, not only regulatory genes, such aszif69 or cpg20, but also genes encoding structuralproteins, such as cr-actinin, the vesicle protein VGF(Possenti et al., 1989), and the dendritic proteinArc/ Arg3.1 (Link et al., 1995; Lyford et aL, 1995),are induced by LTP-producing stimuli. This sup-ports the notion that morphological alterations playa role in longerJasting LTP (Bailey and Kandel,1993;Bliss and Collinridge, 1993; Huang et al., 1994).

Owing to its unbiased nafure, our screening strat-egy may identify molecular mechanisms whose par-ticipation in neural plasticity is currently not fullyappreciated. Examples include the putative dual-specificity phosphatase encodedby cpg21 and a newlydescribed protein kinase, CPG16, whose closest rela-tives are CaM kinases and a myosin light-chainkinase. Both CaM kinase tr and protein phosphataseshave been implicated in LTP and in long-term depres-sion (LTD) (Silva et a1.,L992;Mulkey et al., 1994;Tho-mas et aI., 1994; Mayford et al., 1995). However,ftrnctional analyses of CPG21, imply that this kinase is

Volume 10. 1998

Cand i date P I asti c itv-Rel ated Genes

regulated by cAMP levels, rather than by the calcium/calmodulin axis (M. A. S. and R. S., manuscript inpreparation). Nevertheless, the nature of the majorityof CPGs appears to be consistent with existing mod-els of neural plasticity (Goelet et aI.,1986; Turley etal., 1993; Thomas et aI., 199 4). \Mhat follows is a briefdescription of these models and their possible rela-tionships to CPGs (Fig. 6). Not included are signaltransduction molecules, such as transcription factorsand cytoplasmic signaling molecules, since thesemay contribute to all models.

1. Synaptic elaboration-Pre- or postsynapticelaboration is expected to change pattems andstrengths of synaptic connections (Bailey andKandel, 1993; Bliss and Collinridge, 1993;Huanget aI., 1994). Accordingly, most of the isolatedCPGs fall into this category, including severaltrophic factors (e.g., BDNF and activin), cyto-skel-etal and other skuctural proteins (e.g., qmdecan,Arc/ Argp.L, ot-actinin, and fibrillarin), and twoprotease inhibitors (TIMP and PAI-2).

2. Presyraptic release---changes in the amount ofreleased neurotransmitter would affect sy:rapticefficacy (Bekkers and Stevens, 7990; Bliss andCollinridge, 1993; Arancio et al., 1995). The twonew seven-transmembrane domain receptors(CPG30 and CPG38) could be involved in modu-lation of the amount of transmitter released in amanner similar to that of other modulatoryreceptors (Aramori and Nakanishi, \992). Up-regulation of the vesicle proteins clathrin,HSC70, secretogranin, and VGF, which wereidentified as CPGs, may lead to an increasednumber of secretory vesicles and thereby elevatequantal release.

3. Postsyraptic receptors--<hanges in the numberor properties of post-synaptic receptors andchannels can affect rymaptic plasticity (Davies eta7, 1989 ;Manabe et aJ., 1992;Bliss and Collinridge,1993). The e-subunit of the nicotinic acetylcholinereceptor and the i--type Ca2* channel subunit,BDHP& which we found to be upregulated byneural activity, maybe involved in modulationof s1'naptic action.

4. Retrograde messengers-evidence fot "tetro-grade" communication between post- andpresynaptic neurons during synaptic pro-cesses has been reported (Bliss and Collinridge,1993; Malenka,1994). Three major candidateswere proposed as retrograde messengers (Will-iams et a1.,1989;Stevens and Wang, 1993;Zhuo


et a1.,7993): arachidonic acid (AA) and the gasesNO and CO. Lrterestingly, three of the identifiedCPGs encode enzymes involved in AAmetabolism, CO qmthesis, and arginine (a pre-cursor of NO) metabolism.

In view of the unexpectedly large number of tran-scripts that were identified as CPGs, it is relevant toask how genes are involved in the response tokainate. Lr our previous study (Nedivi et al., 1993),this was estimated to be in the range of 500-L000. Thepresent study verifies this predictiory because aftericreening 10,000 library clones, 362 CPGs, and41 dCPGs, a significant fraction (-30-50%) of the pre-dicted sum have already been cloned. Similarly, ourLTP tests predict that the number of CPGs involvedin LTP is in the range of 100-200 genes/ implyingextensive gene induction. By contrast, only 5-10% ofthe genes that were modulated by glutamate recep-tor ictivation underwent downregulation (termeddCPGs). Of these, five are knovrn Senes/ including asubunit of a kainate receptor (Ga11, 19BB), and two areCa2+ binding proteins, 51008 and calmodulin. Down-regulation of these genes couid reflect a protective re-sponse of the granular neurons to strong kainateactivation. Interestingly, the dCPG group includesonly two transcription factors, one of which is the p50subunit of NF-rB (Liou and Baltimore, 1993). More-over, we idenffied the gene encoding the inhibitorysubunit (kB) of NF-rB as a CPG' It is worth notingthat NF-rcB is a potent activator of gene expression inprimary cultures of hippocampal neurons (Rathrer etal, 1993). Possibly, this dual regulation may ensurekainate-induced reduction in NF-rB activity. It willbe interesting to see whether the dCPGs are genes thatrequire NF-kB activity for expression. Regardlessof the identities of dCPGs, one obvious conclu-sion is that gene activation, rather than repres-sion, is the more prevalent resPonse to the strongactivation of the glutamate receptor, which weemployed in this study.

Obviously, an unequivocal assignment of anygiven CPG to neuronal plasticity will require fur-ther experiments, involving different and com-plementing approaches. Howevet, expression inthe exci-totoxicity-resistant granular neurons/induction of a substantial portion of the identi-fied CPGs by both LTP-producing stimuli (Fig.3), and under physiological conditions such asvisual excitation (Nedivi et al., L996), togetherwith the nature of the known CPG products, are

9 1

Volume 10, 1998

- > . ' -

o . 9 =P P ;

" n 9 X "= (v :s

€ u 3

H E o-- >-poH o i

o . - =

. , C J JY - o *. ! = Fq ' i

\-/ it o

Y 9 +" : i \ J

( d : f- o EP 6 .

v ' =

. _ : : c

o . Q o

* 9 O

> . to b 6L . : = o

: a v ' =

F O b D V; 1 , , f , v

t s ! : ^. r g - s> ) l * No lT c.r> o . Y =c = = :' - u : L :\ w : v . _' r = u - o :y = e . :A - V L

o t s 6 . :E d b o ;L - > Jcg .)v > qJ- i E c j o

E e z 6U L. - 1 - V

3 € 9 , ;^ 9 - -

r r A n U. F C P Yo z . i 3 o o- A a ^ L

> . u = -o u - c H

. - w Pv o ' = . 0o C * U- o . 9 9 EA v - . -

> . = o I@ ( U Y i :

. € L t to o c _ vo * 9 u

o- ci -y

$u$giHg**- fl*** $*srr

tl$ f;|,

E$eeflgEaEEflg

o o <z o r {

olEIrllPIslol.clolol>l

rrlolEIEI€l!t^;lg

FI

€eo o;tpe go oE Eo=

j* {+.FF o- o* sirEE -m

E€o 9<selgfiEIqoli6^ g o . FHE3

ef rflgg€EEE-sdsEc$u**sfrnsEn

92

Cand i date P I asti c ity- Rel ated G enes

all consistent with involvement of a significant frac-tion of the entire CPG populationinneural plasticity.

Acknowledgments

We thank Debora Steiner, Yonathan Levav, andAnat Hanuno for help with the gene scteen, SaraEimerl, Rainer Kramer, and Ann Jensen for con-tributing to this study, Menahem Segal for helpwith LTP experiments, Amgen's EST Program forDNA sequence analysis, Mike Gilman for plas-mids and advise, and Eli Keshet and Ahuva Itinfor help wlth in situ hybidization. We also thankRafael Malach, Michael Walker, Yinon Ben-Neriah,Yechezkel Ben-Ari, Michael Rosbash, and MichaelSchramm for comments on the manuscript. Thisstudy was supported by grants from Amgen Inc., TheIsrael Academy of Sciences and Humanities, and TheIsrael Ministry of Science and the Arts. The laboratoryof R. S. is supported by the P. Godfrey Foundation inchildren's diseases and a Samuel and Isabelle Fried-man Career Development Chair. E. N. was supportedby an Eshkol fellowship and by the Israel CancerResearch Fund (ICRF).Y. C.was supported by anICRF Research Career Development Award.

GenBank Accession Numbers

The accession numbers described in this articleare U78857 (cp 91 5) and U7 887 5 (cp g20).

References

Abdollahi A., Lord K. A., Hoffman L. B., and LiebennannD. A. (1991) Sequence and expression of a cDNAencoding MyD118: a novel myeloid differentia-tion primary response gene induced by multiplecytokines. Oncogene 6, 1,65-167.

Amaya Y., Matsubasa T., Takiguchi M., KobayashiK., Saheki T., Kawamoto S., and Mori M. (1988)Amino acid sequence of rat argininosuccinatelyase deduced from cDNA. l. Biochem. (Tokyo) 103,177-1.81..

Andersen R. D., Birren B. W., Taplitz S. ]. andHerschman H. R. (1986) Rat metallothionein-Lstructural gene and three pseudogenes, one ofwhich contains 5'-regulatory sequences. Mol.CeIl. Biol. 6,302-3L4.


93

Antalis T. M., Clark M. A., Barnes T., Lehrbach P. R.,Devine P. L., Schevzov G., et al. (1988) Cloningand expression of a cDNA coding for a humanmonocyte-derived plasminogen activator inhibi-Lor. Proc. Natl. Acad. Sci. USA 85, 985-989.

Aramori L and Nakanishi S. (1992) Signal transduc-tion and pharmacological characteristics of ametabotropic glutamate receptor, mGluRl, intransfected CHO cells. Neuron 8,757-765.

Arancio O., Kandel E. R., and Hawkins R. D. (1995)Activity-dependent long-term enhancement oftransmitter release by presynaptic 3',5'-cyclicGMP in cultured hippocampal neurons. Nature376,74-80.

Bading H., Ginty D.D., and Creenberg M. E. (1993)Regulation of gene expression in hippocampalneurons by distinct calcium signaling pathways.Science 260,181-1,86.

Bading H. and Greenberg M. E. (1991) Stimulation ofprotein tyrosine phosphorylation by NMDAreceptor activation. Science 253, 912-91'4,

Bailey C. H. and Kandel E. R. (1993) Structuralchanges accompanying memory storage. Annu.Reo. Phy siol. 55, 397 426.

Bekkers ]. M. and Stevens C. F. (1990) Presynapticmechanism for long-term potentiation in the hip-pocampus. N ature 346, 724-729.

Ben-Ari Y. (1985) Limbic seizure and brain damageproduced by kainic acid: Mechanisms and rel-evance to human temporal lobe epilepsy. Neuro-science 14,375403.

Ben-Ari Y. and Represa A. (1990) Brief seizure epi-sodes induce long-term potentiation and mossyf ibre sprout ing in the hippocampus. TrendsNeurosci. 13,312-3L8.

Bhat S. P., Horwitz J., Srinivasan A. and Ding L.(1991) Alpha B-crystallin exists as an independentprotein in the heart and in the lens. Eur.I. Biochem.202,775-787.

Bliss T. V. P. and Collinridge G. L. (1993) A synapticmodel of memory: long-term potentiation in thehippocamp us. N atur e 361, 3L-39.

Blok L. J., Grossmann M. E., Perry ]. E., and TindalllD. J. (1995) Characterization of an early growthresponse gene, which encodes a zinc fingertranscription factor, potentially involved in cellcycle re gula tion. Mol. Endo cr inol. 9, L610-1,620.

Bradbury A., Possenti R., Shooter E. M., and TironeF. (1991) Molecular cloning of PC3, a putativelysecreted protein whose mRNA is induced bynerve growth factor and depolarization. Proc.NatI. Acad. Sci. USA 88, 3353-3357.

Volume 10, 1998

94

Brakeman P. R., Lanahan A. A., O'Brien R., Roche K.,Barnes C. A., Huganir R. L., et al. (1997) Flomer: aprotein that select ively binds metabotropicglutamate receptors. N sture 386, 284-288.

Burtchuladze R., Frenguelli 8., Blendy J., Cioffi D.,Schutz D., and Silva A. I. (994) Deficient long-term memory in mice with targeted mutation ofthe cAMP-responsive element-binding protein.CeI l79,59-68.

Chappell T. G., Welch W. J., Schlossman D. M., Pal-ter K. B., Schlesinger M. ]., and Rothman J. E.(1986) Uncoating ATPase is a member of the 70kilodalton family of stress proteins. CeII 45,3-73.

Civelli O., Douglass |., Goldstein A., and Herbert E.(1985) Sequence and expression of the rat prody-norphin gene. Proc. Natl. Acad. Sci. USA 82,4291-4295.

Cole A. j., Saffen D. W., Baraban J. M., and Worley P.F. (1989) Rapid increase of an immediate earlygene messenger RNA in hippocampal neurons bysynaptic NMDA receptor activation. Nature 340,474476.

Cowell I. G., Skinner A., and Hurst H.C. (7992)Tran-scriptional repression by a novel member of thebZIP family of transcription factors. Mol. Cell. Biol.12,3070-3077.

Criado M., Witzemann V., Koenen M., and SakmannB. (1988) Nucleotide sequence of the rat muscleacetylcholine receptor epsilon-sub unit. NucleicAcids Res.16,10,920.

Crosby S. D., Puetz J. J., Simburger K. S., Fahrner T.1., and Milbrandt I. $991) The early response geneNGFI-C encodes a zinc finger transcriptional acti-vator and is a member of the GCGGGGGCG (GSG)element-binding protein family. Mol. Cell Biol.1l,3835-3841.

Curran T., Gordon M. 8., Rubino K. L., and SambucettiL. C.0987\ Isolation and characterization of the c-fos(rat) cDNA and analysis of post-translationalmodification in vitro. Oicogene 2, Zg-8+.

Davies S. N., Lester R. A., Reymann K. G., andCollingridge G. L. (1989) Temporally distinct pre-and post-synaptic mechanisms maintain long-term potentiation. Nature 338, 500-503.

Davis j. Q., Mclaughlin T., and Bennett V. (1993)Ankyrin-binding proteins related to nervous sys-tem cell adhesion molecules: candidates to pro-vide transmembrane and intercellular connectionsin adult brain. I. Cell. Biol.121,121.-1.33.

Della N. G., Senior P. V., and Bowtell D. D. (1993) Iso-lation and characterisation of murine homologuesof the Drosophila seven in absentia gene (sina).D ea elopment 117, 7333-7343.


Hevroni et al.

DemmerJ., Dragunow M., Lawlor P. A., MasonS. E.,Leah J. D., Abraham W. C., et al. (1993) Differentialexpression of immediate early genes after hippoc-ampal long-term potentiation in awake rats. BrainRes. Mol. Brain Res.17,279-286.

Dragunow M., Beilharz E., Sirimanne E., Lawlor P.,Williams C., Bravo R., et al. (1994) Immediate-earlygene protein expression in neurons undergoingdelayed deattu but not necrosis, following hypoxic-ischaemic injury to the young rat brain. Brain Res.Mol. Brain Res. 25, 1913.

Edwards D. R., Waterhouse P., Holman M. L., andDenhardt D. T. (1986) A growth-responsive gene(16C8) in normal mouse fibroblasts homologousto a human collagenase inhibitor with erythroid-potentiating activity: evidence for inducible andconstitutive transcripts. Nucleic Acids Res. 14,8863-8878.

Ehrlich M. E. and Greengard P. (7997) Characteiza-tion of rat ARPP-21 mRNA: sequence analysis, tis-sue d.istribution, and regulation. l. Neuroihem. 57,1985-1997.

Fang I., Yin W., Smiley E., Wang S. Q., and Bonadio J.(1996) Molecular cloning of the mouse activin bEsubunit gene. Biochem. Biophys. Res. Commun.228,669-674.

Finnerty H., Kelleher K., Morris G. E. Bean, K., MerbergD. M., Kriz R., et al. (1993) Molecular cloning ofmurine FLT and FLT4. Oncogene 8,2293-2298.

Fischer C. R., GutierrezJ., Hsu C. M., Iacangelo A.,and Eiden L. E. (1990) Sequence analysis, tissuedistribution and regulation by cell depolarization,and second messengers of bovine secretogranin II(chromogranin C) mRNA. l. Biol. Chem. 265,9208-9213.

Foulkes N. S., Borrelli E., and Sassone-Corsi P. (1991)CREM gene: Use of alternative DNA-bindingdomains generates multiple antagonists of cAMP-induced transcription. Cell 64, 739-749.

Frank D. A. and Gleenberg M. E. (1994) CREB: Amediator of long-term memory from molusks tomammals. Cell 79,5-8.

Gaestel M., Gotthardt R., and Muller T. (7993) Struc-ture and organisation of a murine gene encodingsmall heat-shock protein Hsp25. Gene 128,279-283.

Gall C. (1988) Seizures induce dramatic and distinctlydifferent changes in enkephalin, dynorphin, andCCK immunoreactivities in mouse hippocampalmossy fibers. l. Neurosci. 8,7852-7862.

Goelet P., Castellucci V. F., Shacher S., and Kandel E.R. (1986) The long and the short of long-termmemory-a molecular framework. Nature 322,419422.

Volume 10. 1998

Cand i d ate P I asti citv-Re I ated G enes

Graham R. and Gilman M. (1991) Distinct protein tar-gets for signals acting attlrre c-fos serurn resPonse ele-ment. S cience 25T' 189-792.

Guan K., Haun R. S., Watson S. J., Geahlen R. L., andDixon J. E. (1990) Cloning and expression of a pro-tein-tyrosine-phosphatase. Proc. Nstl. Acad' Sci. USA8Z 1501-1505.

Guan K. L. and Dixon J. E. (1991) Eukaryotic proteinsexpressed in Escehrichia coli: an improved thrombincleavage and purification procedure of fusion pro-teins with glutathione S-transferase. Anrtl- Biochem.192,262-267.

Hanks S. K. (1991) Eukaryotic protein kinases. Cur. Op.Struct. Biol. 1., 369-383.

Hogan B., Constantini F., and Lacy E. (1986) Manipulat-ing the Mouse Embryo: A Laboratory Manual. ColdSpring Harbor Laboratory, Cold Spring Harbor, NY.

Howard A., Feighner S., Cully D., Arena )., Liberator P.,Rosenblum C., et al. (1996) A receptor in pituitaryand hypothalamus that functions in growth hor-mone releas e. Science 273, 97 +-977 .

Hsu ]. C., LazT., Mohn K. L', and Taub R. (1991) Identi-' fication of LRF-1, a leucine-zipper protein that is rap-

idly and highly induced in regenerattngliver. Proc.NatL Acnd. Sci. USA 88,3511-3515.

Huang Y. Y., Li X. C., and Kandel E. R. (1994)cAMP contributes to mossy fiber LTP by initiat-ing both a covalently mediated phase and mac-romolecular synthesis-dependent late phase.CeII79,69-79.

Imataka H., Sogawa K., Yasumoto K., Kikuchi Y., SasanoK., Kobayashi A., et al. (1992) Two regulatory pro-teins that bind to the basic transcription element(BTE), a GC box sequence in the promoter region ofthe rat P-4501A1 gene. EMBO 1.11,36631671'.

Jeffery K. J., Abraham W. C., Dragunow M., and MasonS. E. (1990) Induction of fos-like immunoreactivityand the maintenance of long-term potentiation in thedentate gyrus of unanesthestetued rcts. MoI. BrainRes. & 267-274.

Johnston L G., Rush S. J., Gurd ]. W., and Broen I. R.(1992) Molecular cloning of novel mRNA using anantibody directed against syraptic glycoproteins. /.N eurosci. Res. 32, 159-1'66.

Julien J. P., Ramachandran K., and Grosveld F. (1985)Cloning of a cDNA encoding the smallest neuro-filament protein from the rct. Biochem. Biophys. Acta82s,398404.

Kandel E. R. and O'Dell T.I. $992) Are adult learningmechanisms also used for development? Science 258,24T245.


95

Kaneda N., Oshimura M., Chung S. Y., and Guroff G.(1992) Sequence of rat TIS11 cDNA, an immediateearly gene induced by growth factors and phorbolesters. Gene l1.8, 289-297.

Kawakami Z.,Kitabayashi I., Matsuoka T., GachelinG., and Yokoyama K. (1992) Conserved structuralmotifs among mammalian junB genes. NucleicAcids Res. 20,91'4.

Keyse S. M. and Emsile E. A. (1992) Oidative stress andheat shock induce a human gene encoding a proteintyrosine phosphatase . Nnture 359' 64+447 .

Kirchhausen T., Harrison S. C., Chow E. P., MattalianoR. ]., Ramachandran K. L., Smart J., et al' (1987)Clathrin healy chain: molecular cloning and com-plete primary shucture. Proc.Nntl. Acad. Sci. USA84'8805-8809.

Kitabayashi I., Saka F., Gachelin G., and Yokoyama K.(1990) Nucleotide sequence of rat c-jun proto-oncogene. Nucleic Acids Res. L$ 3400.

Kojima T., Shworak N. W., and Rosenberg R. D. (1992)Molecular cloning and expression of two distinctcDNAs encoding heparan sulfate proteoglycan coreproteins from a rat endothelial cell ltne.l. BioI. Chem.267,48704877.

KuJrnburg P. A., Huber N. E., Scheer B. j., Wrann M., andBaumruker T. (1992) Immr.rnoglobulin E plus antigenchallenge induces a novel intercrine,/chemokine inmouse mast cells. /. Exp. Med.176,177T1778.

KwakS. P. and DixonJ. E. (1995) Multiple dual specific-ity protein tyrosine phosphatases are expressed andregulated differential$ in liver cell line s. | . Biol. Chem.270,115G11.60.

Link W., Konietzko U., Kauselmann G., Krug M',Schwanke B., Frey Y., and Kuhl D. (1995) Soma-todendritic expression of an immediate early gene isregulated by synaptic activity. Proc' Natl. Acad' Sci.usA92,573+5738.

Liou H.-C. and Baltimore D. (1993) Regulation of NF-kB/rel transcription factor and IkB inhibitor system.Curr. Op. Cell 8io1. 5, 477487.

Lomri N., Perret C., Gouhier N., and Thomasset M.(1989) Cloning and analysis of calbindin-D28K cDNAand its expression in the centralnervoussystem.Garc80,87-98.

Lyford G. L., Yamagata K., Kaufmann W. E., Bames C.,Sanders L. K., Copeland N. G., et al. (1995) Atc, agrowth factor- and activity-regulated gene, encodes anovel cytoskeleton-associated protein that is enrichedin neuronal dendrites. Neuron L4' 433-445.

Maedor W. E., Means A. R., and Quiocho F. A. (1993)Modulation of calmodulin plasticity in molecular

Volume 10, 1998

96

recognition on the basis of x-ray structurcs. Science262,777U1721.

Maisonpierre P. C., Belluscio L., Squinto S., Ip N. Y.,Furth M. E., Lindsay R. M., and Yancopoulos G. D.(1990) Neurotrophin-3: a neurotrophic factor rela-ted to NGF and BDNF. Science 247,1446-145I.

Majumdar A., Brown D., Kerby S., Rudzinski I., PolteT., Randhawa 2., and Seidman M. M. (1991)Sequence of human HMG2 cDNA. Nucleic AcidsRes,19,6643.

Malenka R. C. (1994) Synaptic plasticity in the hippoc-ampus: LTP and LTD. Cell78, 535-538.

Manabe T., Renner P., and Nicoll R. A. (1992) Posfsynaptic contribution to long-term potentiationrevealed by the analysis of miniature synaptic cur-rents. Nature 355, 50-55.

Mayford M., Wang J., Kandel E. R., and O'Dell T. J.(1995) CaMKII regulates the frequency-responsefunction of hippocamppal synapses for the pro-duction of both LTD and LTP. CeII gt,897-904.

Mellor H., Flowers K.M., Kimball S. R., and JeffersonL. S. (1994) Cloning and characterization of acDNA encoding rat PKR, the double-strandedRNA-dependent eukaryotic initiation factor-2kinase. Biochem. Biophys. Acta 1219,699-696.

Monard D. (1988) Cel l -der ived proteases andptotease inhibitors as regulatorJof neurite out-growth. T r ends N eur o s ci.'1,1, 54I-5 44.

Morgan J. I. and Curran T. (1991) Stimulus-transcrip-tion coupling in the nervous system: involvementof the inducible proto-oncogenes/os and jun. Ann.Rea. N eurosci. 14, 421457.

Mulkey R.M., Endo S., Shenolikar S., and Malenka R.C. (1994) Involvement of a calcineurin/inhibitor-1phosphatase cascade in hippocampal long-termdepression. N ature 369, 486488.

Muller R.M., Taguchi H., and Shibahara S. (1957)Nucleotide sequence and organization of the ratheme oxygenase gene. /. Biol. Chem.262,6Z95-6802.

Nadler I.V. (1978) Kainic acid as a tool for the studyof temporal lobe epileps y . Ltfe S ci. 29, 202I-20 42.

Naeve G. S., Ramakrishnan M., Kramer R., HevroniD., Citri Y., and Theill L. E. (1997) Neuritin: A geneinduced by neural activity and neurotrophins thatpromotes neuritogenesis. Proc. NatI. Acad. Sci. USA94,2648-2653.

Nedivi E., Fieldust S., Theill L. E., and Hevroni D. (1996)A set of genes expressed in response to light in theadult cerebral cortex and regulated during develop-ment. Proc. Natl. Acad. Sci. U9A99,2048-2053.

Nedivi E., Hevroni D., Naot D., Israeli D., and CitriY. (1993) Numerous candidate plasticity-related


Hevroni et al.

genes revealed by differential cDNA cloning.Nature 363,778-722.

Nedivi E., Wu G.-Y., and Cline H. T. (7997) cpgli, acandidate plasticity-related gene involved in den-dritic arbor growth. Soc. Neurosci. Abstract 29,607.

Nguyen P. V., Abel T., and Kandel E. R. (1994)Requirement of a critical period of transcriptionfor induction of a late phase of LTP. Science 265,7104-1107.

Noda M., Teranishi Y., Takahashi H., Toyosato M.,Notake M., Nakanishi S., and Numa S. (1982) Iso-lation and structural organization of the humanpreproenkephalin gene. Nature 297, 437434.

O'Banion M. K., Winn V. D., and Young D. A. (1992)cDNA cloning and functional activity of a gluco-corticoid-regulated inflammatory cyclooxygenase.Proc. NatI. Acad. Sci. USA 89,4888-4892.

Ohkura N., Hijikuro M., Yamamoto A., and Miki K.(1994) Molecular cloning of a novel thyroid/ste-roid receptor superfamily gene from cultured ratneuronal cells. Biochem. Biophys, Res. Commun.20S,1959-1965.

Ohmstede C. A., Bland M. M., Merrill B. M., andSahyoun N. (1991) Relationship of genes encod-ing Ca+2 / calmodulin-dependent protein kinaseGr and calsperin. Proc. NatL Acad. Sci. llSA 88,5784-5788.

Pajunen A., Crozat A., Janne O. A., Ihalainen R.,Laitinen P. H., Stanley 8., et al. (1988) Structureand regulation of mammalian S-adenosylmethionine decarboxylase. l. Biol. Chem. 263,17,04U17,049.

Patwardhan S., Gashler A., Siegel M. G., Chang L.C., ]oseph L. J., Shows T. B., et al. (7991) EGR3, anovel member of the Egr family of genes encod-ing immediate-early transcription factors. Oncogene6,917-929.

Petropoulos D., Part A., Ochoa M., Zakin M., andLamas E. (1995) NOR-2 (neuron-derived orphanreceptor), a brain zinc finger protein, is highlyinduced during liver regeneration. FEBS Lett.372,273-278.

Possenti R., Eldridge J. D., Paterson B. M., Grasso A.,and Levi A. (1989) A protein induced by NGF inPC12 cells is stored in secretory vesicles andreleased through the regulated pathway. EMBO I.8,2217-2223.

Powell M. J. and Watts F. Z. (1990) Isolation of a geneencoding a mitochondrial HSP70 protein fromSchizosaccharomyces pombe. Gene 95, 105-110.

Pragnell M., SakamotoJ.,lay S. D., and Campbell K.P. (1991) Cloning and tissue-specific expression of

Volume 10, l99B

Can d i d ate P I asti c ity-Rel ated C e nes

the brain calcium channel beta-subunit. FEBS Lett.291,253-258.

Qian 2., Gilbert M. E., Colicos M. A., Kandel E. R.,and Kuhl D. (1993) Tissue-plasminogen activatoris induced as an immediate early gene during sei-zure, kindling and long-term potentiation. Nnture36't,453457.

Raff M. C., Barres B. A., Burne ]. F., Coles H. S.,Ishizaki Y., and Jacobson M. D. (1993) Programmedcell death and the control of cell survival: lessonsfrom the nervous system. Science 262' 695-700.

Rapp G., Freudenstein J., Klaudiny I1, Mucha ].,W"mp" F., Zimmer M., and Scheit K. H. (1990)Charicterization of three abundant mRNAs fromhuman ovarian granulosa cells. DNA CeIl Biol.9'479-485.

Rattner A., Korner M., Walker M. D., and Citri Y.(1993) NF-kB activates the HIV promoter in neu-rons. EMBO I. 12, 42674267 .

Ripperger J. A.,Fritz S., Richter K., Hocke G. M.,Lottspeich F., and Fey G. H. (1995) Transcriptionfactois Stat3 and Stat5b are present in rat livernuclei late in an acute phase response and bindinterleukin-6 response elements. l. Biol. Chem.270,29,998-30,006.

Saga Y., Tsukamoto T., Jing N', Kusakabe M., andSakakura T. (1991) Murine tenascin: cDNA cloning,structure and temporal expression of isoforms'Gene 104,177-185.

Salminen M., Lundstorm K., Tilgmann C., SavolainenR., Kalkkinen N., and Ulmanen I. (1990) Molecularcloning and characterization of rat liver catechol-O-methyltransferase. Gene 93, 241'-247.

Salton S. R. (1991) Nucleotide sequence and regula-tory studies of VGF, a nervous system-specificrn(Ne that is rapidly and relatively selectivelyinduced by nerve growth factor. l. Neurochem. 57,991-996.

Sambrook j., Fritsch E. F., and Maniatis T. (1989)Molecular Cloning: A Lnboratory Manual. ColdSpring Harbor Laboratory, Cold Spring Harbor,NY.

Scearce L. M., LazT. M., Hazel T. G., Lau L. F., andTaub R. (1993) RNR-1, a nuclear receptor in theNGFI-B/Nur77 family that is rapidly induced inregenerating liver. l. Biol. Chem. 268, 8855-8861.

Segei R., Seger D., Reszka A. A., Munar E. S.,Eldar-Finkelman H., Dobrowolska G., Jensen A.M., et al. (1'994) Overexpression of mitogen-activated protein kinase kinase (MAPKK) andits mutants in NIH3T3 cells. /. Biol. Chem.269,25,699-25,709.


97

Shatz C. J. (1990) Impulse actir-ifi- and the patterningof connections during CNS der-elopment. Neuron5,745-756.

Sheng M. and Greenberg M. E. (1990) The regulationand function of c-fos and other immediate earlygenes in the nervous system' Neurort 4,477185.

Shimasaki S., Koba A., Mercado M., Shimonaka M',and Ling N. (1989) Complementary DNA struc-ture of the high molecular weight rat insulin-likegrowth factor binding protein (IGF-BP3) and tis-Jue distribution of its mRNA. Biochem. Biophys.Res. Commun. L65, 907 -972.

Silva A. |. and Giese K. P. (1994) Plastic genes are in!Curr. Opin. Neurobiol. 4,473420.

Silva A. ]., Stevens C. F., Tonegawa S., and Wang Y'(7992) Deficient hippocampal long-term potentia-tion in a-Calcium-Calmodulin Kinase II mutantmice. Science 257 ' 20L-206.

Sloviter R. S. (1992) Possible functional consequencesof synaptic reorganization in the dentate gyrus ofkainate-treated rats. N eurosci. Lett. 137, 91'-96.

Smeyene R. J., Vendrell M., Hayward M., Baker S. J',Miao G. G., Shilling K., et al. (1993) Continuous c-/os expression precedes Programmed cell death'Nature 363, L66-169.

Steele F. R., Chadet G.1.,]ohnson L. V., and TombranT. J. (1993) Pigment epithelium-derived factor:neurotrophic activity and identification as a mem-ber of the serine protease inhibitor gene family.Proc. NatL Acad. Sci. U9A90,7526-1530-

Stevens C. F. and Wang Y. (1993) Reversal of long-term potentiation by inhibitors of haem oxyge-nase. N nture 364, 147 -149.

Subramaniam M. , Har r is S . A . , Ours le r M. J ' ,Rasmussen K., Riggs B. L., and Spelsberg T. C'(7995) Identification of a novel TGF-beta-regu-lated gene encoding a putative zinc finger pro-tein in human osteoblasts. Nucleic Acids Res. 23'4907-4972.

Sun H., Charles C. H., Lau L. F., and Tonks N. K'(1993) MKP-1 (3CH134), an immediate early geneproduct, is a dual specificity phospahatase thatdephosphorylates MAP kinase in vivo. Cell 75,487-493.

Tauck D. L. and Nadler J. V. (1985) Evidence of func-tional mossy fiber sprouting in hippocampal for-mation of kainic acid-treated rats. Neuroscience 5,701,6-1022.

Tewari M., Mohn K. L., Yue F. E., and Taub R. (7992)Sequence of rat RL/IF-1 encoding an lkappaB,and comparison with related proteins containingnotch-like repeats. Nucleic Acids Res. 20,607.

Volume 10, 1998

9B

Thomas K. L., Laroche S., Errington M. L., Bliss T. V.,and Hunt S. P. $99$ Spatial and temporal changes insignal transduction pathways during LTP . Neuron 13,737-745.

Turley S. J., Tan E. M., and Pollard K. M. (1993) Molecu-lar cloning and sequence analysis of U3 snoRNA-associated mouse fibrillarin. Biochem. Biophas. Acta1216,119-122.

Ward Y., Gupta S., Jensen P., Waterman M., and Kelly K.(1994) Control of MAP kinase activation by the mito-gen-induced threoine/tyrosine phosphatase PAC-1.Nature 367,651-654.

WatsonJ. B., Battenberg E.F., Wong K. K., Bloom F. E.,and Sutcliffe J. G. (1990) Subtractive cDNA cloning ofRC3, a rodent cortex-enriched mRNA encoding anovel 78 residue protein. /. Neurosci. Res.26,g9Z408.

Williams j. H., Errington M. L., Lyrrch M. A., and Bliss T.V. (1989) Arachidonic acid induces a long-term activ-ity-dependent enhancement of synaptic transmissionin the hippocampus. Nature 34'1",739-742.

YamagataK., AndreassonK. I., KaufmannW. E., BamesC. A., and Worley P. F. (1993) Expression of amitogen-inducible cyclooxygenase in brain neu-rons: regulation by synaptic.activity and gluco-corticoids. N eur on 11., 371186.

Yamagata K., Kaufmann W. E., Lanahan A. A., PapavlouM., Bames C. A., Andreasson K. L, and Worley P. F.(1994a) Egr3/Pilot, a zinc-finger transcription factor, israpidly regulated by activity in brain neurons andcolocalizes with Egrl/2ifl.68. Learning E Memory l,140-152.

Yamagata K., Sanders L. K., Kaufman W. E., Yee W.,Bames C. A., Nathans D., and Worley P. F. (1994b)

Hevroni et al.

rheb, a growth factor- and syraptic activity-regu-lated gene, encodes a novel Ras-related protein. /.Biol. Chem. 269, 16,333-16,339.

Yamaguchi M., Flirose F., Hayashi Y., NishimotoY., and Matsukage A. (1987) Murine DNA poly-merase beta gene: mapping of transcription ini-tiation sites and the nucleotide sequence of theputat ive promoter region. MoI. CII I . Biol . 7,2012-201,8.

Yoshimura T., Takeya M., and Takahashi K. (1991)Molecular cloning of rat monocyte chemoattractantprotein-1 (MCP-1) and its expression in rat spleencells and tumor cell lines. Biochem. Biophys. Res.Commun.174,504-509.

Youssoufian H., McAfee M., and Kwiatkowski D.J.(1990) Cloning and chromosomal localization ofthe human cytoskeletal alpha-actinin gene revealslinkage to the beta-spectrin gene. Am.l. Hum. Genet.47,62-71.

Zerial M., Toschi L., Ryseck R. P., Schuermann M.,Muller R., and Bravo R. (1989) The product of a novelgrowth factor activated gene, fos B, interacts withJUN proteins enhancing their DNA binding activ-Ity. EMBO I.8,805-813.

ZhuoM., Small S. A., Kandel E. R., and Hawkins R. D.(1993) Nitric oxide and carbon monoxide produceactivity-dependent long-term symaptic enhancementin hippocampus. Science 260, 1946-1950.

ZmudzkaB. Z.,SenGupta D., Matsukage A., CobianchiF., Kumar P., and Wilson S. H. (1986) Structure of ratDNA polymerase beta revealed by partial aminoacid sequencing and cDNA cloning.Proc. Nqtl. Acad.Sci. USA 83,5106-5110.


H ip p o c a m p aPl la s tic ityln v o lv e sE x te n s...

Documents

Transcript of H ip p o c a m p aPl la s tic ityln v o lv e sE x te n s...