Brain Insulin Receptors and Spatial MemoryCORRELATED CHANGES IN GENE EXPRESSION, TYROSINE PHOSPHORYLATION, AND SIGNALINGMOLECULES IN THE HIPPOCAMPUS OF WATER MAZE TRAINED RATS*

(Received for publication, August 5, 1999, and in revised form, September 16, 1999)

Weiqin Zhao‡§, Hui Chen¶, Hui Xu‡, Elizabeth Moore‡, Noam Meirii, Michael J. Quon¶,and Daniel L. Alkon‡

From the ‡Laboratory of Adaptive Systems, NINDS, and the ¶Hypertension-Endocrine Branch, NHLBI,National Institutes of Health, Bethesda, Maryland 20892 and the iInstitute of Animal Science,The Volcani Center, P. O. Box 6 Bet Dagan 50250, Israel

Evidence accumulated from clinical and basic re-search has indirectly implicated the insulin receptor(IR) in brain cognitive functions, including learning andmemory (Wickelgren, I. (1998) Science 280, 517–519). Thepresent study investigates correlative changes in IR ex-pression, phosphorylation, and associated signalingmolecules in the rat hippocampus following water mazetraining. Although the distribution of IR proteinmatched that of IR mRNA in most forebrain regions, adissociation of the IR mRNA and protein expressionpatterns was found in the cerebellar cortex. After train-ing, IR mRNA in the CA1 and dentate gyrus of the hip-pocampus was up-regulated, and there was increasedaccumulation of IR protein in the hippocampal crudesynaptic membrane fraction. In the CA1 pyramidal neu-rons, changes in the distribution pattern of IR in partic-ular cellular compartments, such as the nucleus anddendritic regions, was observed only in trained animals.Although IR showed a low level of in vivo tyrosine phos-phorylation, an insulin-stimulated increase of in vitroTyr phosphorylation of IR was detected in trained ani-mals, suggesting that learning may induce IR functionalchanges, such as enhanced receptor sensitivity. Fur-thermore, a training-induced co-immunoprecipitationof IR with Shc-66 was detected, along with changes in invivo Tyr phosphorylation of Shc and mitogen-activatedprotein kinase, as well as accumulation of Shc-66, Shc-52, and Grb-2 in hippocampal synaptic membrane frac-tions following training. These findings suggest that IRmay participate in memory processing through activa-tion of its receptor Tyr kinase activity, and they suggestpossible engagement of Shc/Grb-2/Ras/mitogen-acti-vated protein kinase cascades.

Because insulin and insulin receptors (IRs)1 were found inthe central nervous system, their role(s) in brain function hasbeen a subject attracting many researchers’ interest. IR is

widely distributed in the brain (1), with particularly high con-centrations in neurons and much lower levels in glia (2, 3).Although IR mRNA is largely localized in neuronal somata,abundant IR protein is found in both cell bodies and synapses,including the dendritic field of the hippocampal CA1 pyramidalcells, the adrenergic terminals in hypothalamus, and mem-branes of dendrodendritic synaptosomes from the rat olfactorybulb (3, 4). Although the major molecular structure and most ofthe properties of IR in the central nervous system are indistin-guishable from those in peripheral tissues, some differencesexist between central nervous system IRs and peripheral IRs.Both the a and b subunits, for example, of IR in the centralnervous system are slightly lower in molecular weight thanthose in the periphery (5). Unlike peripheral IR, the brain IRdoes not undergo down-regulation after exposure to high con-centrations of insulin (6, 7). In addition, insulin appears toregulate glucose metabolism only in glia cells but not in neu-rons (3, 8, 9). These findings have led to speculation that theneuronal IRs mediate functions different from those regulatedby their peripheral and glial counterparts.

IR is a glycoprotein with an a2b2 tetrameric structure. Bind-ing of insulin to the extracellularly located a subunits results inautophosphorylation of the b subunits at tyrosine residues1158, 1162, and 1163, located in the cell interior (10). Auto-phosphorylation stimulates the intrinsic tyrosine kinase activ-ity of IR, which is believed to initiate the biological actions ofinsulin (10). Like other growth factor receptors, phosphoryl-ated IR triggers subsequent biological responses by activatingdifferent cellular signal transduction cascades. These includeinteraction of the receptor with IR substrate proteins and avariety of Src homology domain 2- and 3-containing proteins(11–13). One of such pathways involves Shc, an adapter protein(14, 15) that mediates association of the receptor with theGrb-2zSOS protein complex (16). This process is known to bringabout activation of Ras (16, 17) that in turn triggers the mito-gen-activated protein kinase kinase/mitogen-activated proteinkinase (MAPK) pathway leading to regulation of nuclear tran-scription (18). It is found that the phosphotyrosine bindingdomain of Shc is responsible for the IR-mediated mitogenicsignaling (19).

Evidence from functional studies demonstrates that insu-lin/IR may play a modulatory role in brain synaptic transmis-sion. Consistent with its enriched distribution in adrenergicterminals, insulin is found to promote central catecholaminer-gic activities by releasing both epinephrine and norepinephrine(20), inhibiting synaptic reuptake of norepinephrine (21), andaltering catecholomine kinetics (22). In the hippocampus, in-sulin is reported to enhance a1 adrenergic receptor activity,leading to stimulation of membrane phosphoinositol turnover

* The costs of publication of this article were defrayed in part by thepayment of page charges. This article must therefore be hereby marked“advertisement” in accordance with 18 U.S.C. Section 1734 solely toindicate this fact.

The nucleotide sequence(s) reported in this paper has been submittedto the GenBankTM/EBI Data Bank with accession number(s) M29014.

§ To whom correspondence should be addressed: Laboratory of Adapt-ive Systems, NINDS, National Institutes of Health, Bethesda, Mary-land 20892. Tel.: 301-402-0514; Fax: 301-402-2281; E-mail: wqzhao@helix.nih.gov.

1 The abbreviations used are: IR, insulin receptor; MAPK, mitogen-activated protein kinase; RT, reverse transcription; PCR, polymerasechain reaction; PBS, phosphate-buffered saline; IGF-1R, insulin-likegrowth factor-1 receptor; IGFR, insulin-like growth factor receptor.

THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 274, No. 49, Issue of December 3, pp. 34893–34902, 1999Printed in U.S.A.

This paper is available on line at http://www.jbc.org 34893

by guest on April 30, 2020

http://ww

w.jbc.org/

Dow

nloaded from

and diacylglycerol production (23), two important second mes-sengers involved in PKC activation.

At the behavioral level, insulin/IR has been implicated in atleast two major brain functions: feeding behavior (3, 24) andcognition, including learning and memory (25). A role for insu-lin/IR in learning and memory is supported by several findings.First, IR is highly concentrated in neurons of specific brainregions, such as hippocampus, amygdala, and some corticalareas (1, 2). Secondly, data from behavioral-pharmacologicalstudies show that injection of streptozotocin, a diabetes-induc-ing compound, into the brain induces significant memory im-pairment (26). More importantly, defects in insulin action inboth periphery and the brain have been found in Alzheimer’sdisease (27, 28), and insulin has been shown to reduce phos-phorylation of tau protein in human neuronal cultures by in-hibiting activity of glycogen synthase kinase-3, hence promot-ing association of tau to microtubules (29). Finally, impairmentof brain cognitive functions in diabetes mellitus, presumablydue to disruption of glucose metabolism, is well documented(30–34).

Despite the large body of evidence from clinical and animalstudies suggesting involvement of insulin/IR in learning andmemory, actual correlated changes of IR during memory proc-essing have never been reported. In this study, we investigatechanges in IR expression, Tyr phosphorylation, and certainIR-associated signaling molecules in the rat hippocampal neu-rons following a water maze training experience.

EXPERIMENTAL PROCEDURES

Water Maze Training—Male 60–90 day-old Wistar rats (200–250 g)were housed in standardized conditions as described elsewhere (35). Toadapt rats to the experimental environment and behavioral activity, allrats were subjected, in the first day of experiments, to 2 min of swim-ming in a 1.5 m (diameter) 3 0.6 m (depth) pool, with water tempera-ture set at 21 6 1 °C. On the following day, rats were trained in afour-trial water maze task, each trial lasting up to 2 min. Duringtraining, rats learned to escape from water by finding an unseen rigidplatform submerged about 1 cm below the water surface in a fixedlocation. The escape latency during each trial was measured as anindicator of learning. In order to assess short-term and long-term bio-chemical changes, rats were sacrificed at 1 and 24 h, respectively, aftertraining. For the controls, rats were given four swimming trials, butwithout the platform present in the pool. The length of each swimmingtrial for each control rat was yoked to that of each trained rat toequalize nonlearning components, such as locomotor activity andstress. These swimming controls were also sacrificed at 1 and 24 h,respectively, after the last swimming trial. All rats were sacrificed bydecapitation, and their hippocampi were rapidly removed, frozen on dryice, and stored at 280 °C before use. Hippocampi from a naı̈ve groupwere also collected and used for basal controls. For in situ hybridizationand immunohistochemistry studies, the whole brain was rapidly re-moved following decapitation and frozen on dry ice. In total, 15 ratswere used for each group.

The experiments were carried out under the guidelines of the Na-tional Institutes of Health regulations for the Care and Use of Animalsfor Scientific Purposes.

Preparation of the IR Riboprobe for in Situ Hybridization—A frag-ment of cDNA corresponding to bases 760–1140 of the rat IR mRNA(GenBankTM accession number M29014) was synthesized by reversetranscription (RT)-PCR from the rat hippocampus (see under “RT-PCR”below). This fragment shared no homology with any other proteins inthe rat brain and was subcloned into the SrfI cloning site of pPCR-Script Amp SK(1) cloning vector (Stratagene) to produce a template forriboprobe synthesis. The sense and antisense riboprobes were synthe-sized by in vitro transcription with T7 or T3 RNA polymerase (Ambion)respectively in the presence of 1 mg of IR DNA template, 0.5 mM NTPsand 2 mM [a-35S]UTP (.1000 Ci/mmol, NEN Life Science Products).The transcribed product was purified on a Sephadex G-25 spin column(5 Prime 3 3 Prime, Inc., Boulder, CO), and the final labeling to theprobe was assessed by scintillation counting.

Preparation of Brain Sections—Brains from naive, trained, andswimming control rats were sectioned at 12 mm in a cryostat at 220 °C.

Sections were collected on silanated glass slides (Digene) and dried atroom temperature before being returned to 280 °C for storage.

In Situ Hybridization Histochemistry—Brain slices were fixed for 5min in 4% formaldehyde, acetylated, and dehydrated in graded ethanol.The IR riboprobe (1 3 106 cpm in 50 ml) was applied to each slideholding three sections and hybridized at 55 °C for 30 h in a mixturecontaining 20 mM Tris-HCl, pH 7.4, 1 mM EDTA, 300 mM NaCl, 50%formamide, 10% dextran sulfate, 13 Denhardt’s solution, 4 mg/mlsalmon sperm DNA, 10 mg/ml yeast total RNA, 10 mg/ml yeast tRNA,100 mM dithiothreitol, 0.1% SDS, 0.1% NTS. Slices hybridized with thesense riboprobe or pretreated with RNase before hybridization with theantisense probe were used as negative controls. Following high strin-gency posthybridization washes and RNase treatment, brain sectionswere dehydrated in graded ethanol and then subjected to film andliquid emulsion autoradiography.

Immunohistochemistry—Frozen brain sections from naı̈ve, swim-ming control, and trained rats were fixed for 5 min in 4% formaldehyde.After washing with phosphate-buffered saline (PBS), and preincubationwith 0.5% bovine serum albumin in PBS for 1 h, the slices were incu-bated overnight at room temperature in the same medium containingthe primary antibody against rat insulin receptor-b (Santa Crutz Bio-technology) diluted to 1:400. In control slices, the primary antibody wasomitted. Slices were washed with PBS, followed by incubation withbiotinylated anti-rabbit IgG diluted to1:400 for 1 h at room tempera-ture. Following a wash process, signals were visualized with the avidin-biotin-peroxidase technique (Elite kit, Vector Laboratories), in which39,39-diaminobenzidine was used as chromogen.

RT-PCR—Adult rats were killed by decapitation, and their cerebralcortex, hippocampus, and cerebellar cortex were rapidly dissected andfrozen on dry ice. The total RNA from each above region was extractedusing RNA Isolator (Genosys). Following removal of genomic DNA withDNase treatment, the RNA samples were subjected to a reverse tran-scription reaction to synthesize single strand cDNAs using the firststrand cDNA synthesis mix (Novagen). Amplification of a 372-base pairIR cDNA fragment was performed on a DNA thermal cycler 480 (Per-kin-Elmer) through a 25-cycle PCR (94 °C 1 min, 50 °C 1 min, and 74 °C2 s) with primers 59-CTCTGTCCGCATCGAGAA-39 (forward) and 59-CCACACACTGACCATCCA-39 (reverse). To control for any variabilityin sample processing, primers (forward, 59-AGGTGCTCAACAACATG-GAG-39; reverse, 59-TACCAGAGGCCACAGTAGCT-39) synthesizing a183-base pair rat phosphoglycerate kinase 1 cDNA fragment were in-cluded in the PCR amplification. The final PCR products were sepa-rated on a 2% agarose gel and visualized with ethidium bromide.

Preparation of Homogenate, S3, and P2M Fractions—The generalcytosolic and crude synaptic membrane fractions were prepared bycentrifugation according to the procedure of Rostas et al. (36) withminor modifications. Briefly, the frozen tissues from different brainregions were homogenized respectively in precooled Buffer A (0.32 M

sucrose, 2 mM Hepes, pH 7.4, and 1 dose of protease inhibitor mixturepurchased from Roche Molecular Biochemicals) with a glass-Teflonhomogenizer. The homogenate was diluted to 10% (w/v) with Buffer A,and centrifuged at 1000 3 g for 5 min. The supernatant (S1) waspreserved, whereas the pellet (P1) was washed with Buffer A andcentrifuged at 1000 3 g again for 5 min. The supernatant was combinedwith S1 and centrifuged at 10,000 3 g for 20 min. The supernatant (S2)was centrifuged at 100,000 3 g for 1 h to produce a general cytosolicfraction (S3). The pellet (P2) was osmotically lysed (10 ml/g originaltissue) on ice for 30 min in Buffer B (2 mM Hepes, pH 7.4, 50 mm Ca21,and 1 dose of protease inhibitor mixture), and then centrifuged at20,000 3 g for 30 min. The pellet (P2M) from this step was collected asa crude synaptic membrane fraction, redissolved in a small volume of 2mM Hepes, pH 7.4, containing a dose of protease inhibitor mixture, andstored at 280 °C until use. To investigate training-induced changes,hippocampi from four rats were combined to generate three independ-ent pools from a total of 12 rats in each group.

Protein concentrations were measured using the Bio-Rad Proteinassay reagent (Bio-Rad).

Insulin Stimulation of the IR or Insulin-like Growth Factor-1 Recep-tor (IGF-1R) Overexpressed in NIH 3T3 Cells—The human IR orIGF-1R that was stably transfected into the NIH 3T3 cell (37, 38) wasused as positive control for the rat hippocampal IR. Tyrosine phospho-rylated IR or IGF-1R was generated by stimulating the transfected cellswith insulin (100 nM at 37 °C for 2 min). Cells were washed withice-cooled PBS and immediately frozen in liquid nitrogen. Cells werethen lysed on ice for 30 min in a lysis buffer (50 mM Tris, pH 7.5, 1%Triton X-100, 300 mM NaCl, and 1 dose of protease inhibitor mixture)and briefly centrifuged at 8000 rpm (4 °C). The supernatant was col-

Spatial Training-induced Changes in Insulin Receptor34894

by guest on April 30, 2020

http://ww

w.jbc.org/

Dow

nloaded from

lected as cell lysates. Protein concentrations were assessed using theBCA reagent (Pierce).

Immunoblotting and Immunoprecipitation Processes—Proteins fromeach subcellular fraction were separated by 4–20% gradient SDS-PAGE. The resolved proteins were transferred to a 0.45 mM nitrocellu-lose membrane. Blocked with 5% milk powder in 0.01 M PBS, pH 7.5,the membrane was incubated with a given primary antibody (such asantibody against IR, Sch, Grb-2, MAPK, or phosphotyrosine) at 4 °Covernight with constant shaking. On the following day, the membranewas washed and incubated with a secondary antibody conjugated withhorseradish peroxidase at room temperature for 1 h. The immunoreac-tive signal was then revealed with the enhanced chemiluminescenceprocess. Alternatively, a target protein was immunoprecipitated fromsubcellular fractions with its specific antibody and then detected withimmunoblotting procedures. If the protein of interest was expected tohave a molecular mass similar to that of the heavy chain of IgG (around55 kDa), the primary antibody was covalently cross-linked to agarosebeads in a cross-link reaction using the CarboLinkTM kit (Pierce). Thisprocess retained IgG molecules on agarose beads during sample dena-turing, thereby preventing the target protein signal(s) from beingmasked by IgG in immunoblotting.

In Vitro Tyrosine Phosphorylation of Hippocampal Proteins—In vitrophosphorylation of hippocampal S3 and P2M proteins was carried outin a total volume of 50 ml of reaction mixture containing 50 mM Tris-HCl, pH 7.5, 1 mM MgCl2, 2 mM sodium vanadate, 2 mM EGTA, 0.1 mM

phenylmethylsulfonyl fluoride, 10 mg/ml leupeptin, 6 1 mM CaCl2 6 0.1mM insulin, and 50 mg of S2 or P2M proteins. Although Mn21 has beenknown to be a potent effector for in vitro tyrosine phosphorylation, itwas eliminated from the present reactions, as it reacted with sodiumvanadate to form yellow precipitates. The phosphorylation reaction wasinitiated by addition of ATP to a final concentration of 100 mM. Afterincubating at 37 °C for 3 min, the reaction was terminated by adding 50ml of SDS reducing sample buffer. The mixture was boiled for 10 minfollowed by separation by a 4–20% gradient SDS-PAGE. The extent of

tyrosine phosphorylation was detected in Immunoblotting using amonoclonal anti-phosphotyrosine antibody (Py20, Upstate).

Data Analysis—All biochemical experiments were repeated at leastthree times for each pool of hippocampal tissues. In situ hybridizationhistochemistry images and signals from immunoblotting were analyzedby measuring the optic mean density using the NIH Image program.Values from the swimming controls and trained animals were normal-ized against values from naı̈ve animals. Data from three pooled sampleswere subjected to one- or two-way analysis of variance.

RESULTS

Distribution of IR mRNA and Protein in the Rat Brain—Consistent with previous findings, our in situ hybridizationhistochemistry results showed that IR mRNA is widely ex-pressed in the rat brain (Fig. 1A). The highest density of IRmRNA signals was revealed in the cerebellar cortex and cho-roid plexus of the third and lateral ventricles. Concentrated IRmRNA signals were also detected in forebrain areas, such asanterior olfactory nucleus, pyriform cortex, thalamic and hypo-thalamic nuclei, hippocampal formation, amygdaloid nucleus,and the cerebral cortex. IR mRNA levels in the hippocampus,cerebral, and cerebellar cortex were also measured using RT-PCR, which again revealed highest levels of IR PCR product inthe cerebellum (Fig. 2A), indicating a high concentration of IRmRNA in this area.

When IR protein levels in the rat brain were measured withimmunoblotting and immunohistochemistry, the polyclonal an-ti-IR-b antibody detected a 95-kDa protein band from subcel-lular preparations of the rat brain that migrated slightly fasteron SDS gel than the human IR transfected in NIH 3T3 cells

FIG. 1. Distribution of IR mRNA revealed by in situ hybridization. Adult rat brain sections were hybridized with an IR riboprobe labeledwith 35S as described under “Experimental Procedures.” Signals were visualized with film autoradiography. Brain areas were defined accordingto the Rat Brain Atlas (48). AO, anterior olfactory nucleus; FCtx, frontal cortex; PrL, prelimbic cortex; M1, primary motor cortex; M2, secondarymotor cortex; VO, ventral orbital cortex; DTr, dorsal transition zone; Cpu, caudate putamen (striatum); AI, agranular insular cortex; Acb,accumbens nucleus; Tu, olfactory tubercle; Pir, piriform cortex; LV, lateral ventricle; AM, anteromedial thalamic nucleus; VA, ventral anteriorthalamic nucleus; Rt, reticular thalamic nucleus; MPO, medial preoptic nucleus; VLH, ventrolateral hypothalamic nucleus; AVDM, anteroventralthalamic nucleus; VM, ventromedial thalamic nucleus; LH, lateral hypothalamic area; VMH, ventrolateral hypothalamic nucleus; CA1, CA1 areaof hippocampus; CA3, CA3 area of hippocampus; CA4, CA4 area of hippocampus; DG, dentate gyrus; Chp, choroid plexus; SNR, substantia nigra,reticular part; AmgN, amygdaloid nucleus; CCtx, cerebellar cortex; Bt, brain stem.

Spatial Training-induced Changes in Insulin Receptor 34895

by guest on April 30, 2020

http://ww

w.jbc.org/

Dow

nloaded from

(Fig. 2B). The ratio of IR protein to mRNA was much lower inthe cerebellar cortex than in the hippocampus and cerebralcortex, (Fig. 2C). In all three areas, IR protein was highlyconcentrated in the P2M fraction. The disassociation of mRNAand protein expression in distinct brain regions was confirmedby results from immunohistochemistry studies. Fig. 3A showsthe liquid emulsion results, which demonstrate cellular distri-bution of IR mRNA. The dark grains represent the IR mRNAsignals, a high concentration of which is present in both Pur-kinje (indicated by arrows) and granule cells of the cerebellarcortex (bottom panel), whereas the pyramidal cell layer of thehippocampal CA1 area had lower levels of IR mRNA signals(top panel). Fig. 3B shows that immunoreactivity (indicated bydarkness of the brown staining) of IR protein in the CA1 py-ramidal cells (top panel) was stronger than that in Purkinjeand granule cells of the cerebellar cortex (bottom panel). Inaddition, the cerebral cortical neurons and the choroidal epi-thelium also showed high IR immunoreactivity (data notshown).

Changes in Expression of IR mRNA and Protein followingWater Maze Training—We trained rats in a four-trial watermaze task for 1 day in order to distinguish cellular eventsassociated with memory formation with those trigged by otherprocesses such as recall, which usually results from multi-daytraining tasks. After training, rat maze learning (Fig. 4A) wasdemonstrated by significantly reduced escape latency (F3,59 540.36, p , 0.001). Training induced an up-regulation of IRmRNA in the CA1 and dentate gyrus regions of the hippocam-pus (Fig. 4B). Compared with swimming controls trained ani-mals showed 30% and 22% increases in IR mRNA in the CA1area at 1 h and 24 h after training. A two-way analysis of

variance showed significant training (F1,8 5 14.04, p 5 0.006),but not time effects. Similarly, 30–40% increases were found indentate gyrus after training (F1,8 5 14.51, p 5 0.005), whereasno significant changes were found in CA3 region.

Changes in IR protein levels after training were determinedwith immunoblotting (Fig. 4C). No significant changes in IRprotein were found in the total homogenate fraction (Fig. 4C,a). A decrease (19%) in IR immunoreactivity in the S3 fractionwas apparent 1 h after training (F1,8 5 18.55, p 5 0.0026, Fig.4C, b). On the other hand, a significant increase in IR proteinwas detected in the P2M fraction at 1 and 24 h after training(F1,8 511.36, p , 0.01, Fig. 4C, c). Similar results were alsoobtained following immunoprecipitation of IR from P2M frac-tions from each group of animals (data not shown).

Fig. 4D illustrates immunocytochemical changes in IR pro-tein in hippocampal CA1 pyramidal cells after training. Thebrown staining indicates immunoreactivity of IR in the CA1neurons. For swimming controls, the majority of cells (; 80%)in the CA1 area showed moderate IR immunostaining, with adistribution pattern similar to that of naı̈ve animals (see Fig.3). A small population of cells (,20%) showed weak staining.After training, however, an increase in IR immunostaining(dark brown) was observed in some of these cells. Table I showsa semiquantitative analysis of training-induced IR immunore-activity changes. Although the proportion of the cells withweak staining was not changed, 30–38% of the CA1 neuronsshowed strong immunostaining after training, compared with3–5% in the swimming controls. In addition, IR immuno-staining appeared to be clustered within specific intracellularcompartments after training, compared with an even distribu-tion in the cytoplasm of the cell in swimming controls. In-creased staining was also seen in dendritic areas.

Changes in Tyrosine Phosphorylation of Hippocampal Pro-teins—To investigate a possible involvement of the IR tyrosineprotein kinase activity and its downstream signal cascades inlearning and memory, we first measured changes in overalltyrosine phosphorylation of the hippocampal proteins. Proteinconcentrations of the tissue preparations across each groupwere equalized to 1 mg/ml, of which 15 ml was resolved onSDS-PAGE. Tyr phosphorylation was measured in immuno-blotting with a monoclonal antibody against phosphotyrosine(Py20). As shown in Fig. 5, more than 10 Tyr-phosphorylatedmajor protein bands were revealed in hippocampal homoge-nates. One of these bands (P60) showed a decrease in phospho-rylation at 1 and 24 h after training, whereas increasing inphosphorylation in the swimming controls. A two-way analysisof variance indicated significant training (F1,8 5 37.57, p ,0.001) effects. In the S3 fraction, dephosphorylation of P60 alsooccurred at 1 h after training (F1,8 5 34.5 p , 0.001). A signif-icant increase in Tyr phosphorylation was found in a P36 bandin trained animals at both 1 and 24 h after training (F1,8 528.5, p , 0.001). In the P2M fraction, the most obvious train-ing-specific changes occurred to the P66 and P52 bands, both ofwhich showed significant increases in tyr phosphorylation at 1and 24 h after training (P66: F1,8 5 10.55, p 5 0.01; P52: F1,8 516.6, p , 0.01). In addition, several other P2M proteins showedmarked increases in Tyr phosphorylation at 24 h after training.These included P120, P95, P75, P 32, P25, and P10. The higherintensity of those protein bands was not due to greateramounts of protein loaded on the gel, as protein staining afterelectrophoresis showed similar amount of proteins across eachlane (data not shown).

In Vitro Phosphorylation of IR and Its Changes after Train-ing—Because the above immunoblotting results indicated alow basal level of IR Tyr phosphorylation, we then examined invitro regulation of IR Tyr phosphorylation in hippocampal P2M

FIG. 2. Comparison of IR mRNA and protein in different brainregions. A shows the RT-PCR result indicating that the cerebellarcortex (CCtx) possesses the highest IR mRNA level among the threebrain regions examined. B shows immunoblotting results for IR proteinlevels in subcellular fractions from the three above brain regions. Theprotein concentration in the cerebellar cortex is lower than that in thecerebral cortex (CrCtx) and hippocampus (HP). The right lane is apositive control showing the human IR that is overexpressed in the NIH3T3 cells. Results from RT-PCR and immunoblotting of brain homoge-nates were subjected measured by densitometry. The densitometryratio of IR protein to IR mRNA for different regions is shown in C.

Spatial Training-induced Changes in Insulin Receptor34896

by guest on April 30, 2020

http://ww

w.jbc.org/

Dow

nloaded from

fractions in the absence and presence of Ca21. In the presenceof ATP, Mg21, and Na3VO4, a 95-kDa band was heavily phos-phorylated (Fig. 6a, 3). This band co-migrated on SDS-PAGEwith the human IR (Fig. 6a, lane 5) and IGF-1R (Fig. 6a, 1). Thepresence of Ca21 (1 mM) in the reaction markedly reduced itsTyr phosphorylation (Fig. 6a, lane 4). To distinguish IR andIGF-1R, both of which migrated to a similar position on theSDS gel, immunoprecipitation was carried out with anti-Tyr(P)antibody following in vitro phosphorylation reactions, and theprecipitates were detected by immunoblotting with anti-IR(Fig. 6b) and IGF-1R (Fig. 6c) antibodies. There was littlecross-reaction between IR antibody and IGF-1R (Fig. 6b, lane1), or between IGF-1R antibody and IR (Fig. 6c, lane 5), indi-cating that these antibodies were target-specific. A major IR-bprotein was detected from the immunoprecitates after phos-phorylation (Fig. 6b, lane 3). The amount of precipitated IR-bwas markedly reduced when Ca21 was added to the phospho-rylation reaction (Fig. 6b, lane 3). The phosphorylated 95-kDaband was found to also include IGF-1R (Fig. 6c, lane 3), thephosphorylation of which, however, was not affected by Ca21

(Fig. 6c, lane 4). Results from lane 2 in all three panels of Fig.6 indicated that IR and IGF-1R were predominantly in a de-phosphorylated state in untreated synaptic membranefractions.

We next measured whether in vitro phosphorylation of IRwas changed as a function of training, because it may reflectalterations of the properties and sensitivity of the receptor toits ligand during memory processing. A significant increase ofin vitro Tyr phosphorylation was found in P180 (F1,8 5 12.04,p , 0.01) but not P95 after training (Fig. 7a1). The presence ofinsulin in the reaction stimulated the overall phosphorylationof P180, with significantly stronger signals (F1,8 529.12, p ,

0.001) detected in the trained animals (Fig. 7a2). Insulin treat-ment did not change Tyr phosphorylation of P95 from naı̈veand swimming control rats, but it significantly enhanced thatfrom trained animals (F1,8 5 17.48, p , 0.001). In the presenceof Ca21, the overall phosphorylation of P180 was markedlyincreased (Fig. 7b1), but no significant differences were de-tected among the trained and control rats. The overall Tyrphosphorylation of P95, on the other hand, was markedly re-duced in the presence of Ca21. Samples from the swimmingcontrols showed particularly weaker phosphorylation signals(Fig. 7b1) compared with that from other groups (F1,8 5 5.3,p 5 0.05). When Ca21 and insulin were both present in thereaction (Fig. 7b2), increases in phosphorylation of P180 (F1,8 520.11 p 5 0.002) and P95 (F1,8 5 18.37, p 5 0.003) were shownonly in trained animals. In addition, phosphorylation of P66was also increased after training (Fig. 7, b1 and b2). Theseresults suggest that sensitivities of P180 and P95 to insulinwere increased after water maze training.

Changes in Shc Protein and Its Interaction with IR—Toidentify the 66- and 52-kDa proteins in P2M fractions thatshowed increased Tyr phosphorylation after training (Fig. 5,P2M). Freshly prepared S3 and P2M samples from trained andcontrol rats were separated on SDS gels and immunoblottedwith anti-Shc antibody. This antibody detected Shc-66 and -52,localized mainly in the cytosolic fractions of the hippocampus.Although no apparent change was seen in the S3 fraction,significant increases in amounts of Shc-66 (F1,8 5 31.8, p ,0.001) and Shc-52 (F1,8 5 14.42, p 5 0.005) were found in theP2M fractions from trained animals (Fig. 8A).

A immunoprecipitation experiment was performed to deter-mine possible in vivo interactions between IR and Shc. Equalamounts of P2M proteins from each group were subjected toimmunoprecipitation with a goat anti-IR-b antibody cross-linked to agarose beads. The immunoprecipitate was then blot-ted with anti-Shc antibody following SDS-PAGE and transferprocesses. Shc-66 was co-precipitated with IR by the IR-b an-tibody from trained animals (Fig. 8B). Similarly, P2M proteinswere immunoprecipitated with anti-Shc antibody followed byimmunoblotting with anti-IR-b antibody. A strong immunore-active band was only detected by IR-b antibody in samples fromtrained animals (Fig. 8B). The co-immunoprecipitation of Shcwith IR suggests that these two proteins may be associated invivo after training.

Changes in Grb-2 Protein after Training—Because Grb-2

FIG. 3. Levels of IR mRNA and protein in the hippocampal CA1cells and the cerebellar cortex: IR mRNA in situ hybridizationsignals in hippocampal CA1 pyramidal cells and cerebellar granule and Purkinje cells. IR mRNA in situ hybridization signals inhippocampal CA1 pyramidal cells and cerebellar granular and Purkinje cells were visualized with liquid emulsion autoradiography (A) as blackgrains. B shows IR protein immunohistochemical staining in these two areas. The darkness of the brown staining correlates with the concentrationof IR protein. PL, pyramidal cell layer; GL, granule cell layer; ML, molecular layer. Arrows indicate the cerebellar Purkinje cells.

TABLE IImmunoreactivity of IR in the hippocampal CA1 pyramidal cells from

swimming control and trained rats

ImmunoreactivitySwimming control Water maze-trained

1a 11 111 1 11 111

%

Cell Proportion1 h after training 18 6 5 80 6 11 3 6 1 15 6 3 55 6 8 30 6 924 h after training 17 6 3 78 6 12 5 6 1 14 6 4 48 6 9 38 6 8a 1, weak staining; 11, moderate staining; 111, strong staining.

Data are mean 6 S.E.; n 5 3.

Spatial Training-induced Changes in Insulin Receptor 34897

by guest on April 30, 2020

http://ww

w.jbc.org/

Dow

nloaded from

protein is known to be involved in IR-Shc signaling, its changein learning was assessed in the S3 and P2M fractions withimmunoblotting. Like the Shc proteins, Grb-2 is largely acytosolic localized protein (Fig. 8C). Although no changeswere seen in S3 fraction, a significant training-induced in-crease in Grb-2 immunoreactivity was shown in the P2M

fraction (Fig. 8C).Changes in MAPK after Training—Finally, we measured the

training-induced changes in MAPK, a protein known to be adownstream molecule to the IR/Shc/Grab-2/SOS cascade (16–19). No apparent changes in the total amount of MAPK weredetected in the S3 and P2M fractions by the Regular-MAPK

FIG. 4. Changes in IR mRNA and protein after training. A, escape latency of rats in water maze training (n 5 15; p , 0.001). B, rats weresacrificed at 1 and 24 h after training. IR mRNA signals in the hippocampus from trained and control rats were detected with in situ hybridizationhistochemistry, visualized with film autoradiography, and quantified with an image program. All data was normalized against that from naı̈veanimals and subjected to two-way analysis of variance. The bar graphs show average results (means 6 S.E.) in CA1, CA3, and dentate gyrus areas(n 5 3; ** p , 0.01). C, equalized amount of subcellular proteins (homogenate, S3, and P2M) from each group was resolved with SDS-PAGE. IRwas detected with a polyclonal anti-IRb antibody. The bar graphs summarize results from three independently pooled samples (means 6 S.E.; ** p ,0.01). D, immunohistochemical staining of IR protein in CA1 pyramidal cells after training. Left panels are representative samples from swimmingcontrols, whereas the right panels are from trained animals. The darkness of brown staining represents the level of IR protein in the cell. The arrowindicates dendritic area of a neuron. These results demonstrate that IR mRNA is up-regulated, and IR protein undergoes translocation after learning.

Spatial Training-induced Changes in Insulin Receptor34898

by guest on April 30, 2020

http://ww

w.jbc.org/

Dow

nloaded from

antibody (Fig. 9A). A significant increase, however, in the ac-tive form of MAPK (P-MAPK) was observed in the S3 fractionat 1 h after training (F1,8 5 15.32, p , 0.01) and in the P2M atboth 1 and 24 h after training (F1,8 5 42.25, p , 0.001),whereas the P2M from swimming controls showed a decreasedMAPK phosphorylation (Fig. 9B). When in vitro phosphoryla-tion was performed, the overall P-MAPK signal in the S3 waselevated (Fig. 9C, left panel) and further enhanced by additionof insulin to the reaction (Fig. 9D, left panel). In the P2M,insulin treatment during in vitro phosphorylation only in-creased the P-MAPK signal in the samples from trained ani-mals (Fig. 9D, right panel).

DISCUSSION

The present findings from in situ hybridization, immunohis-tochemistry, and immunoblotting experiments confirm that IRis abundantly distributed in the brain. Specific regional con-centrations of IR may reflect different IR functions associatedwith particular brain regions. An abundance of IR, for example,in areas such as olfactory bulb and thalamic nuclei is consistentwith its involvement in regulation of food intake. High concen-trations of IR in the hypothalamus and limbic system includingthe hippocampus, pyriform cortex and amygdala areas thatreciprocally connect and communicate among each other, sug-gests its role in emotion and higher cognitive functions, partic-ularly learning and memory. The very high density of IR in thechoroid plexus suggests that it may be required for transport ofglucose and peripheral insulin across the blood-brain barrier.

Although levels of IR protein match that of its mRNA in theforebrain areas, a striking disassociation between IR mRNA

and protein levels was revealed in the cerebellar cortex. Theseresults suggest differences in IR mRNA translational efficiencyand/or IR protein stability in different brain regions. For ex-ample, IR in the cerebral cortex and hippocampus may possessa high efficiency of translation, or enhanced stability, whereasin the cerebellar cortex very high levels of IR mRNA may berequired to supply a rapid turnover of the receptor. We believethat the distribution patterns of IR mRNA and protein reflect

FIG. 6. Effect of Ca21 on in vitro Tyr phosphorylation of hip-pocampal IR and IGFR. Hippocampal P2M proteins were phospho-rylated in the presence or absence of Ca21 (see under “ExperimentalProcedures”). Extent of Tyr phosphorylation was assessed with ananti-Tyr(P) antibody (a). Alternatively, the phosphorylated sampleswere immunoprecipitated by the anti-Tyr(P) antibody. The precipitateswere resolved with SDS-PAGE and blotted by either anti-IRb (b) oranti-IGF-1R antibody (c). It was shown that Ca21 inhibited Tyr phos-phorylation of IR but not IGF-1R. Lane 1, positive controls for P-IGF-1R; lane 2, nonphosphorylated P2M samples; lane 3, Tyr phosphoryla-tion in the absence of Ca21; lane 4, Tyr phosphorylation in the presenceof 1 mMCa21; lane 5, positive controls for P-IR.

FIG. 5. Tyr phosphorylation of hip-pocampal proteins after training.Equal amounts of total homogenate, S3,and P2M proteins from each group wereresolved on SDS-PAGE. Tyr phosphoryl-ation was detected with a monoclonal an-ti-Tyr(P) antibody. The results show thatwater maze training induced specific Tyrphosphorylation changes in both cytosolicand synaptic membrane proteins. N,naı̈ve animals; C1 and C24, swimmingcontrols, sacrificed at 1 and 24 h, respec-tively, after the last swimming trial; T1and T24, trained animals sacrificed at 1and 24 h, respectively, after training.3T3, NIH 3T3 cells overexpressed withIR.

Spatial Training-induced Changes in Insulin Receptor 34899

by guest on April 30, 2020

http://ww

w.jbc.org/

Dow

nloaded from

the significance of IR functions in different brain regions.The hippocampus, which contains high levels of IR, is criti-

cally involved in spatial memory processing. IR mRNA wasclearly up-regulated in the hippocampal CA1 and dentate gy-rus areas shortly following water maze training, suggestingthat synthesis of IR may be increased in these areas as a resultof learning. An increased IR in the P2M was accompanied by areduction of IR in the S3, suggesting a possible translocation ofIR to the synaptic membrane after training. Changes in thedistribution pattern of IR were also observed in the immuno-histochemical studies, in which IR became more concentratedin certain cellular compartments such as the nucleus and den-drites of the CA1 neurons after training. It is of interest thatthis change is only seen in a subpopulation of the pyramidalcells in the CA1 area, suggesting that only in specific neuronswas IR activated by training. This subpopulation specificity oflearning-induced changes of IR are consistent with numerouspreviously published biophysical and biochemical studies. Volt-age-dependent K1 currents, for example, were found in singleidentified neurons of the mollusk Hermissenda after Pavlovianconditioning (39). Related changes of K1 currents were foundin a subpopulation of CA1 pyramidal cells (40), and a subpopu-lation of H6 cerebellar cortical Purkinje neurons (41) afterPavlovian conditioning of the rabbit eyelid response. Translo-cation of PKC was demonstrated in a subpopulation of cellswith these same paradigms within these same regions, as wellas with rat spatial learning (42, 43). Finally, in a very recentstudy, a subset of CA1 pyramidal cells was found to havereduced inhibitory postsynaptic potentials only after spatialmaze learning (44). These findings, together with those of otherreports that changes in expression of neural cell adhesion mol-ecule (45) are found in a subset of hippocampal neurons after

learning, support the interpretation that learning of a partic-ular task involves only a subpopulation of neurons within arelevant region such as the hippocampus.

Given that IR is a receptor tyrosine kinase, autophosphoryl-ation of which is essential for its activation, changes in auto-phosphorylation of IR after learning would be expected if thereceptor is actively involved in memory formation. Althoughthere was a low level of in vivo phosphorylation of IR, aninsulin-stimulated in vitro phosphorylation of IR was detectedin the synaptic membrane fraction only from trained animals.Although these results did not necessarily reflect the in vivophosphorylation status of IR under trained or swimming con-trol conditions, they provide evidence that the molecular prop-erties of some component of the IR signaling pathway may bealtered by training. Interestingly, the in vitro phosphorylationof IR but not IGF-1R was markedly inhibited by presence of 1mM Ca21 in the reaction. It is unclear how changes in intracel-lular Ca21 due to increased synaptic activity during learningmight influence the IR phosphorylation. The complexity of boththe relevant pre- and postsynaptic biochemical cascades, how-ever, precludes a straightforward interpretation at this time,but it will certainly motivate intensive follow-up studies.

The activated IR signaling cascade after training appeared toinvolve Shc protein. Levels of shc-66 and -52 were increased,and both proteins were significantly phosphorylated in vivo inthe P2M only in trained animals. Co-immunoprecipitation ofShc with IR from trained animals suggests that activation ofShc during learning may be associated with IR PTK activity.Because Tyr phosphorylation of Shc leads to its specific asso-ciation with the Grb-2zSOS complex, our detection of a training-induced accumulation of Grb-2 in the P2M fraction suggeststhat such an event may occur in water maze training. Finally,

FIG. 7. Changes in in vitro Tyr phos-phorylation of P95 and P180 aftertraining. Equal amounts of P2M pro-teins were phosphorylated in the absenceor presence of Ca21 and insulin, followedby measurement of phosphorylation withan anti-Tyr(P) antibody in Western blots.a1, Tyr phosphorylation in the absence ofCa21 and insulin; a2, Tyr phosphoryla-tion in the presence of insulin but theabsence of Ca21; b1, Tyr phosphorylationin the presence of Ca21 but absence ofinsulin; b2, Tyr phosphorylation in thepresence of both Ca21 and insulin. N,naı̈ve animals; C1 and C24, swimmingcontrols sacrificed at 1 and 24 h, respec-tively, after swimming trials; T1 and T24,trained animals sacrificed at 1 and 24 h,respectively, after training (n 5 3; ** p ,0.01). The increases in in vitro phospho-rylation of IR suggest that training mayinduce changes in molecular properties ofIR.

Spatial Training-induced Changes in Insulin Receptor34900

by guest on April 30, 2020

http://ww

w.jbc.org/

Dow

nloaded from

a training-specific activation of MAPK was also detected. Acti-vation of MAPK was also previously reported following in-creases in intracellular Ca21 (46) and retrieval of spatial mem-ory (47). Although no evidence, either from previous work orthe present study, directly identifies the upstream events as-sociated with activation of MAPK, our results showed that thetraining-induced MAPK phosphorylation was further en-hanced by in vitro insulin treatment of the P2M fraction, sug-gesting an increased sensitivity of this signaling pathway toinsulin. Although it is tempting to speculate that during learn-ing IR may have triggered a Ras/MAPK cascade mediated byShc and Grb-2, further studies are needed to identify the pre-cise links among changes in IR, Shc/Grb-2, and MAPK duringspatial memory formation.

Apart from Shc, a 180-kDa P2M protein, which co-migratedwith the insulin receptor substrate-1 (data not shown), alsoshowed an increased insulin-sensitive in vitro Tyr phosphoryl-ation after training. More detailed studies of this protein fol-lowing training are under investigation.

Taken together, our results reveal, for the first time, that

spatial training induces a series of changes in IR of the hip-pocampus, including IR gene expression, protein translocation,and Tyr phosphorylation. Because most of these changes weredetected in synaptic membrane fractions, IR may play a role inregulation of synaptic activities (such as neurotransmissionand/or synaptic plasticity) during memory formation.

REFERENCES

1. Werther, G. A., Hogg, A., Oldfield, B. J., Mckinley, M. J., Figdor, R., Allen,A. M., and Mendelsohn, F. A. O. (1987) Endocrinology 121, 1562–1570

2. Marks, J. L., King, M. G., and Baskin, D. G. (1991) in Molecular Biology andPhysiology of Insulin and Insulin-like Growth Factors, pp. 459–470,Plenum Press, New York

3. Schwartz, M. W., Figlewicz, D. P., Baskin, D. G., Woods, S. C., and Porte, D.,Jr. (1992) Endocr. Rev. 13, 387–414

4. Marks, J. L., Maddison, J., and Eastman, C. J. (1998) J. Neurochem. 50,774–781

5. Heidenreich, K. A., Zahniser, N. R., Berhanu, P., Brandenburg, D., and Olef-sky, J. M. (1983) J. Biol. Chem. 258, 8527–8530

6. Zahniser, N. R., Goens, M. B., Hanaway, P. J., and Vinych, J. V. (1984)J. Neurochem. 42, 1354–1362

7. Boyd, F. T., Jr., and Raizada, M. K. (1983) Am. J. Physiol. 245, C283–C2878. Heidenreich, K. A., and Brandenburg, D. (1986) Endocrinology 118,

1835–18429. Clarke, D. W., Boyd, F. T., Kappy, M. S., and Raizada, M. K. (1984) J. Biol.

Chem. 259, 11672–1167510. Kahn, C. R., White, M. F., Shoelson, S. E., Backer, J. M., Araki, E., Cheatham,

B., Csermely, P., Folli, F., Goldstein, B. J., Huertas, P., Rothenberg, P. L.,Saad, M. J. A., Siddle, K., Sun, X. J., Wilden, P. A., Yamada, K., and Kahn,

FIG. 8. Changes in Shc and Grb-2 proteins after training. A,equal amounts of S3 and P2M proteins from each group resolved withSDS-PAGE were detected with anti-Shc antibody; B, P2M proteins fromeach group were immunoprecipitated (ip) with anti-IRb antibody fol-lowed by immunoblotting (ib) with anti-Shc antibody (left), or sampleswere immunoprecipitated with anti-Shc antibody and then immuno-blotted with anti-IRb antibody (right). These results indicate that Shcwas increased in the P2M, and it appeared to interact with IR duringlearning. C, equal amounts of S3 and P2M proteins resolved withSDS-PAGE were blotted with anti-Grb-2 antibody. An increased accu-mulation of Grb-2 was shown after training. N, naı̈ve animals; C1 andC24, swimming controls sacrificed at 1 and 24 h, respectively, afterswimming trials; T1 and T24, trained animals sacrificed at 1 and 24 h,respectively, after training; 3T3, positive controls for Shc proteins (n 53; ** p , 0.01).

FIG. 9. Activation of MAPK after training. Equal amounts of S3and P2M proteins resolved by SDS-PAGE were blotted with anti-regu-lar (Reg.) MAPK (A) and active (P) MAPK (B) antibodies. C, P-MAPKafter an in vitro phosphorylation reaction; D, P-MAPK after in vitrophosphorylation in the presence of insulin. These results demonstratethat MAPK was activated in the hippocampus after training (n 5 3; **p , 0.01).

Spatial Training-induced Changes in Insulin Receptor 34901

by guest on April 30, 2020

http://ww

w.jbc.org/

Dow

nloaded from

S. A. (1993) Recent Prog. Horm. Res. 48, 291–33911. Roth, R. A., Zhang, B., Chin, J. E., and Kavocina, K. (1992) J. Cell. Biochem.

48, 2–1812. Quon, M. J., Butte, A. J., and Taylor, S. I. (1994) Trends Endocrinol. Metab. 5,

369–37613. Cheatham, B., and Kahn, C. R. (1995) Endocr. Rev. 16, 117–14214. Pelicci, G., Lanfrancone, L., Grignani, F., McGlade, J., Cavallo, F., Forni, G.,

Nicoletti, I., Grignani, F., Pawson, T., and Pelicci, G. (1992) Cell 70, 93–10415. Bonfini, L., Migliaccio, E., Pelicci, G., Lanfrancone, L., and Pelicci, P. G. (1996)

Trends Biochem. Sci. 20, 257–26116. Rozakis-Adcock, M., McGlade, J., Mbamalu, G., Pelicci, G., Daly, R., Li, W.,

Ratzer, A., Thomas, S., Brugge, J., Pelicci, P. G., Schlessinger, J., andPawson, T. (1992) Nature 360, 689–692

17. Sasaoka, T., Ishiki, M., Sawa, T., Ishihara, H., Takata, Y., Imamura, T., Usui,I., Olefsky, J. M., and Kobayashi, M. (1996) Endocrinology 137, 4427–4434

18. Orban, P. C., Chapman, P. F., and Brambilla, R. (1999) Trends Neurosci. 22,38–44

19. Ricketts, W. A., Rose, D. W., Shoelson, S., and Olefsky, J. M. (1996) J. Biol.Chem. 271, 26165–26169

20. Sauter, A., Goldstein, M., Engel, J., and Ueta, K. (1983) Brain Res. 260,330–333

21. Raizada, M. K., Shemer, J., Judkins, J. H., Clarke, D. W., Masters, B. A., andLeRoith, D. (1988) Neurochem. Res. 13, 297–303

22. Masters, B. A., Shemer, J., Judkins, J. H., Clarke, D. W., LeRoith, D., andRaizada, M. K. (1987) Brain Res. 417, 247–256

23. Figlewicz, D. P., and Szot, P. (1991) Brain Res. 550, 101–10724. Woods, S. c., Chavez, M., Park, C. R., Riedy, C., Kaiyala, K., Richardson, R. D.,

Figlewicz, D. P., Schwartz, M. W., Porte, D., Jr., and Seeley, R. (1996)Neurosci. Biobehav. Rev. 20, 139–144

25. Wickelgren, I. (1998) Science 280, 517–51926. Lannert, H., and Hoyer, S. (1998) Behav. Neurosci. 112, 1199–20827. Craft, S., Peskind, E., Schwartz, M. W., Schellenberg, G. D., Raskind, M., and

Porte, D., Jr. (1998) Neurology 50, 164–16828. Craft, S., Neocomer, J., Kanne, S., Dagogo-Jack, S., Cryer, P., Sheline, Y.,

Luby, J., Dagogo-Jack, A., and Alderson, A. (1996) Neurobiol. Aging 17,123–130

29. Hong, M., and Lee, V. M. (1997) J. Biol. Chem. 272, 19547–1955330. Matthaei, S., Horuk, R., and Olefsky, J. M. (1986) Diabetes 35, 1181–118431. Stolk, R. p., Breteler, M. M. B., Ott, A. et al. (1997) Diabetes Care 20, 792–79532. Ott, A., Stolk, R. P., Hofman, A., Van Harskamp, F., Grobbee, D. E., and

Breteler, M. M. B. (1996) Diabetologia 39, 1392–139733. Leibson, C. L., Rocca, W. A., Hanson, V. A., Cha, R., Kokmen, E., O’Brien, P.

C., and Palumbo, P. J. (1997) Am. J. Epedemiol. 145, 301–30834. Biessels, G. J. (1999) J. Med. 54, 35–4535. Cavallaro, S., Meiri, N., Yi, C. L., Musco, S., Ma, W., Goldberg, J., and Alkon,

D. L. (1997) Proc. Natl. Acad. Sci. U. S. A. 94, 9669–967336. Rostas, J. A., Brent, V. A., Heath, J. W., Neame, R. L., Powis, D. A.,

Weinberger, R. P., Dunkley, P. R. (1986) Neurochem. Res. 11, 253–26837. Kadowaki, T., Bevins, C. L., Cama, A., Ojamaa, K., Marcus-Saumuels, B.,

Kadowaki, H., Beitz, L., Mckeon, C., and Taylor, S. I. Science 240, 787–79038. Kato, H., Faria, T. N., Stannard, B., Roberts, C. T., Jr., and LeRoith, D. (1993)

J. Biol. Chem. 268, 2655–266139. Alkon, D. L., Lederhendler, I., and Shoukinmas, J. J. (1982) Science 215,

693–69540. Coulter, D. A., Lo Turco, J. J., Kubota, M., Disterhoft, J. F., Moore, J. W., and

Alkon, D. L. (1989) J. Neurophysiol. 61, 971–98141. Schreurs, B. G., Gusev, P. A., Tomsic, D., Alkon, D. L., Shi, T. (1998)

J. Neurosci. 18, 5498–550742. Olds, J. L., Anderson, M. L., McPhie, D. L., Staten, L. D., and Alkon, D. L.

(1989) Science 245, 866–86943. Olds, J. L., Golski, S., McPhie, D. L., Olton, D., Mishkin, M., and Alkon, D.

(1990) J. Neurosci. 10, 3707–371344. Gusev, P. A., and Alkon, D. L. (1999) Abstr. Soc. Neurosci. 25, 162345. Fox, G. B., O’Connell, A. W., Murphy, K. J., and Regan, C. M. (1995)

J Neurochem. 65, 2796–279946. Rosen, L. B., Ginty, D. D., Weber, M. J., and Greenberg, M. E. (1994) Neuron

12, 1207–122147. Blum, S., Moors, A. N., Adams, F., and Dash, P. (1999) J. Neurosci. 19,

3535–354448. Paxinos and Watson, Rat Brain Atlas, (Paxinos, G., and Watson, C., eds)

Academic Press, San Diego, CA

Spatial Training-induced Changes in Insulin Receptor34902

by guest on April 30, 2020

http://ww

w.jbc.org/

Dow

nloaded from

Daniel L. AlkonWeiqin Zhao, Hui Chen, Hui Xu, Elizabeth Moore, Noam Meiri, Michael J. Quon andMOLECULES IN THE HIPPOCAMPUS OF WATER MAZE TRAINED RATSGENE EXPRESSION, TYROSINE PHOSPHORYLATION, AND SIGNALING Brain Insulin Receptors and Spatial Memory: CORRELATED CHANGES IN

doi: 10.1074/jbc.274.49.348931999, 274:34893-34902.J. Biol. Chem. 

  http://www.jbc.org/content/274/49/34893Access the most updated version of this article at

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/274/49/34893.full.html#ref-list-1

This article cites 46 references, 16 of which can be accessed free at

by guest on April 30, 2020

http://ww

w.jbc.org/

Dow

nloaded from