Dopaminergic DA1 signaling couples growth-associated protein-43 and long-term potentiation in guinea pig hippocampus

Brain Research Bulletin 64 (2005) 433–440

Dopaminergic DA1 signaling couples growth-associated protein-43 and long-term potentiation in guinea pig hippocampus Sanika Chirwaa,e,∗ , Adwoa Aduonuma , Jose Pizarrob , Jonathan Reasora , Yumiko Kawaia , Marissa Gonzalezb , Brenda S. McAdoryc , Emmanuel Onaivid , Edwin J. Barea-Rodriguezb a

Department of Physiology, Meharry Medical College, 1005 D.B. Todd Blvd, Nashville, TN 37208, USA Department of Biology, University of Texas at San Antonio, 6900 North Loop 1604 West, San Antonio, TX 78249, USA Department of Biological Sciences, Tennessee State University, 3500 John A. Merritt Boulevard, Nashville, TN 37209, USA d Department of Biology, William Patterson University, Wayne, NJ 07470, USA e Department of Pharmacology, Vanderbilt University, 23rd Ave South at Pierce, Nashville, TN 37232, USA b

c

Received 21 June 2004; received in revised form 2 September 2004; accepted 27 September 2004 Available online 10 November 2004

Abstract The basic goal of the project was to determine whether dopaminergic DA1 receptor (DA1 R) signaling couples growth-associated protein 43 (GAP-43; a putative “plasticity” protein) and long-term potentiation (LTP; an enduring form of synaptic plasticity). Thus, guinea pigs were prepped to stimulate the CA3 and evoke population spikes in the CA1 neurons in the hippocampus in vivo. Animals were injected with either saline or SCH23390 (a selective DA1 R antagonist), 1–2 h prior to recordings. It was found that tetanic stimulation (100 Hz, 1 s, three trains at 15 s intervals) readily produced early-LTP and late-LTP in the saline group. In contrast, none of the guinea pigs pre-treated with SCH23390 developed late-LTP, though early-LTP had been present. Furthermore, both GAP-43 mRNA and protein were up-regulated after LTP induction in the saline group. However, GAP-43 protein up-regulation was blocked in animals treated with SCH23390. Anti-GAP-43 immunoreactivity was intense in CA3/CA1 synaptic regions, whereas GAP-43 mRNA hybridization was localized to somatic layers in the hippocampus. Altogether, our results suggest that dopaminergic DA1 signaling partly couples GAP-43 and LTP. © 2004 Elsevier Inc. All rights reserved. Keywords: CA1; CA3; DA1 receptors; GAP-43; Hippocampus; Long-term potentiation

1. Introduction Activity-dependent increases in synaptic efficacy such as long-term potentiation (LTP) are studied extensively in view of their putative role in learning and memory [5]. Briefly, specific patterns of synaptic activation give rise to a relatively “short-lasting” form of potentiation (1–2 h; early LTP) or to a longer-lasting potentiation (beyond 2–3 h; late-LTP) depending on neural activation history [8,11,27,30]. There is a general consensus that LTP maintenance (i.e., late-LTP) is stabilized by new synaptic growth orchestrated by gene expression. Consequently, certain molecular products have ∗

Corresponding author. Tel.: +1 615 327 6934; fax: +1 615 327 5789. E-mail address: [email protected] (S. Chirwa).

0361-9230/$ – see front matter © 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.brainresbull.2004.09.012

been linked to LTP [28] and prominent among them is growthassociated protein-43 (GAP-43; also termed neuromodulin, B-50, F-1, or pp46: [4,18,22]). However, the signaling pathway coupling neural activity and GAP-43 expression and how this, in turn, is linked to LTP is unclear. With this in mind, we set out to determine if dopamine DA1 receptor (DA1 R) pathways couple GAP-43 expression and LTP maintenance. We focused on the dopaminergic transductory pathway since published data already implicate the involvement of DA1 R activation for LTP maintenance [10,12,23,30]. For example, selective DA1 R agonists (e.g., SKF38393) induce a slowly developing potentiation across CA3/CA1 synapses lasting beyond 6 h [12]. In contrast, selective DA1 R antagonists (e.g., SCH23390) block LTP maintenance [10,12,23,30]. Furthermore, mice lacking DA1 Rs develop early-LTP but not

434

S. Chirwa et al. / Brain Research Bulletin 64 (2005) 433–440

late-LTP [16]. Thus, it is reasonable to speculate that the effects of DA1 Rs activation may involve modulation of gene expression underlying LTP maintenance. We present in this report data that link DA1 R activation to both up-regulation of GAP-43 and the occurrence of late-LTP in the guinea pig hippocampus in vivo.

2. Methods Animals for this project, 26 male guinea pigs (3–4 weeks old; weight, 150–200 g), were obtained from Harlan Sprague–Dawley; this number of animals was required to attain statistical significance for most of the variables that we examined. The research animals were housed in groups of one to five per cage, within the animal care facility (ACF) at Meharry Medical College. The ACF maintains its animal rooms at 65–70 ◦ F and 40–70% relative humidity. Animals were kept on a 12 h light/12 h dark cycle. Food and water were supplied ad libitum. Animals were kept for at least 1–2 days in the ACF prior to use. Subsequently, guinea pigs were assigned to one of three basic groups, namely: (a) unstimulated saline controls, i.e., “UNS” group, n = 6, (b) stimulated saline controls, i.e., “SAL” group, n = 14, and (c) treated with DA1 R antagonist SCH23390, i.e., “SCH” group, n = 6. Animals in both the UNS and SAL groups were injected with 1 mL phosphatebuffered saline (PBS) as a single bolus i.p. In contrast, animals in the SCH group were injected with 1 mL each of SCH23390 (0.25 mg/kg dose i.p.). The SCH23390 dose was based on information from the literature [29,30]. Drug solutions were freshly prepared whereby SCH23390 (Tocris Neurochemicals) was first dissolved in dimethyl sulfoxide and then made to final volume with PBS. In each case, an interval of at least 1 h elapsed from each intraperitoneal injection, respectively, before initiating recordings as described below. 2.1. Electrophysiology The six guinea pigs in the UNS group were set aside and not processed for recordings. Rather, these animals served as unstimulated controls for the molecular studies described later. All other guinea pigs (i.e., SAL=14 and SCH=6) were processed for recordings as follows. One hour after the intraperitoneal injections, guinea pigs were anesthetized using urethane (1500 mg/kg body weight i.p.) and then secured in the stereotaxic apparatus. The top of the head was shaved and disinfected with alcohol wipes, after which, a subcutaneous bolus of 2% lidocaine local anesthetic (0.2–0.3 mL) was injected in the head region. Subsequently, an incision was made along the midline using a scalpel; then a circular skin patch (2–3 cm diameter) was excised to expose the skull. Hydrogen peroxide (30%) was swabbed across the exposed skull. The bregma was identified and this was used as the zero reference point for determining brain coordinates for electrode

placement. Small access holes (∼2 mm diameter) were made in the skull to allow lowering of bipolar tungsten electrodes (SNEX-15, Rhodes) to the left CA3 region (5.2 mm posterior, 5.0 mm lateral and −5.0 mm vertical relative to bregma) and the left CA1 region (4.0 mm posterior, 3.6 mm lateral and −4 mm vertical rise relative to bregma). To begin an experiment, stimulation was applied to activate the CA3 region and elicit population spikes in the CA1 area. Each electrode tip was lowered to its optimal location where maximal and stable population spikes were recorded. Evoked responses were monitored to confirm stability, a prerequisite for continuation of recordings. The stimulus parameters utilized were those that evoked 60–80% of ‘plateau’ population spikes in each experiment. After at least 30 min of stable baseline recording, animals from each group were tetanized (100 Hz, 1 s duration, three trains at 15 s intervals) to induce LTP across CA3/CA1 synapses in vivo. The brief trains were followed by resumption of low frequency (0.06 Hz) test stimulation that was continued intermittently (at ∼10 min intervals) for 180 min [7], at which time electrodes were withdrawn and guinea pigs euthanized to harvest either the whole brain or just hippocampus. Recordings were amplified via a two-channel AC differential pre-amplifier (band pass 0.1–10 kHz), whereas rectangular pulses were delivered through a photoelectric current isolation unit regulated by a stimulator (Grass Instruments). Evoked field potentials were viewed on a digital multi-channel oscilloscope (Hitachi), coupled with computer-assisted data acquisition and analysis (pCLAMP 8 software, Axon Instruments). The majority of the recordings were also taped on videocassette (multi-channel PCMVCR system, A.R. Vetter Company Inc.) for off-line analysis. Aseptic techniques were utilized in all surgical and recording procedures. In addition, items touching brain tissue (e.g., electrodes, spatulas, micropipettes) were pre-sprayed with ElectroZapTM , (Ambion) to de-activate RNAse and DNAse contaminants, subsequent to rinsing thoroughly with RNAsefree water. To maximize animal comfort, guinea pigs were blanketed with warming pads in order to maintain regular body temperature. Animal experiments were performed with strict adherence to protocols approved by the Institutional Animal Care and Use Committee at Meharry. 2.2. GAP-43 mRNA hybridization in situ After recordings, brains were harvested from eight of 14 of guinea pigs in the SAL group; these brains were processed to localize and quantify GAP-43 mRNA hybridization in situ. In addition brains from three of the six guinea pigs in the UNS group (i.e., not subjected to recordings) were also assessed for GAP-43 mRNA in situ and these served as nonstimulated controls. To begin analysis, 25 ␮m thick coronal sections were cut using a cryostat, placed onto polylysine precoated slides (Fisher) and kept frozen until further processing. The slides were allowed to warm up to room temperature and

S. Chirwa et al. / Brain Research Bulletin 64 (2005) 433–440

fixed for 20 min in 4% paraformaldehyde-EM grade (Electron Microscopy Sciences) in PBS. Sections were quickly washed with RNAse-free water. To diminish probe binding to protein moieties, and thereby reduce background hybridization, tissue sections were acetylated in 250 mL of 0.1 M triethanolamine (pH 8.0) containing 0.65 mL acetic anhydride and incubated for 10 min. The slides were then washed in PBS and placed in 0.85% saline for 5 min each. This was followed with dehydration by washing in an ascending ethanol (EtOH) series as follows: 60% EtOH (1 min), 80% EtOH (1 min), 95% EtOH (2 min), 99% EtOH (1 min), chloroform (5 min), 99% EtOH (1 min), 95% EtOH (1 min). Sections were air-dried and then used immediately for hybridization in situ. Hybridization solution (containing 1 × 105 to 3 × 105 cpm/␮L of either sense or antisense oligonucleotide probes labeled on their 3 end with terminal transferase and ␣33 PdATP by standard methods) was heated to 80 ◦ C, placed on ice for 1 min, and pipetted onto the sections, which were then cover-slipped. Slides were placed horizontally in a slide box containing a tissue soaked with PBS and then incubated in a hybridization oven at 42 ◦ C for 16–24 h. Next, coverslips were removed from brain sections by washing slides vertically twice in 5× SSC (SSC = 3.0 M NaCl and 0.3 M Sodium Citrate) at 50 ◦ C for 30 min. This was followed by a 30-min wash with slow agitation in 1× SSC at room temperature, two 30-min washes in 1× SSC at 53 ◦ C, and 1 h wash in 1× SSC at room temperature. Finally, slides were washed in deionized water for 5 min and dehydrated. Sections were air-dried and autoradiographed on ␤-hypermax film. Developed films were scanned and analyzed. 2.3. Oligonucleotide probes The oligonucleotides (synthesized by Operon Technologies) were based on reported DNA sequence databases as follows: sense strand; 5 -AGA AGG CAG GGG AAG ATA CCA CCA TGC TGT GCT GTA TGA GA-3 ; antisense strand; 5 -TCT CAT ACA GCA CAG CAT ATC TTC CCC TGC CTT CT-3 . A search in the National Center for Biotechnology Information data bank showed that these oligonucleotide sequences were specific for GAP-43 mRNA. The probes (100 ng) were labeled on their 3 end using terminal transferase (20 U/␮L) and ␣33 P -dATP (specific activity: >2.6 × 106 cpm/␮g) using standard protocols [31]. 2.4. GAP-43 Western blots For quantitative GAP-43 analysis, left and right hippocampi lobes were harvested after recordings in the remaining six of 14 guinea pigs in the SAL group, and all of the guinea pigs in the SCH group (n = 6). In addition, hippocampi lobes were obtained from the remaining three of six guinea pigs in the UNS group that had not undergone recordings. The harvested hippocampi lobes were transferred separately into RNAse-free small tubes filled with TRI reagent (Sigma Chemicals) and homogenized, and total protein was isolated

435

according to the instructions of the manufacturer (Sigma Chemicals). After homogenization the tissue suspensions in TRI reagent were centrifuged at 12,000 × g for 10 min at 4 ◦ C. Chloroform was added, and after vigorous agitation, the mixtures were incubated for 15 min. Next, the mixtures were centrifuged (12,000 × g for 15 min at 4 ◦ C) to produce three distinct phases. The aqueous upper phase, containing RNA, was recovered by isopropanol precipitation. The middle and lower phases, containing proteins, were isolated by alcohol precipitation. Protein content was determined by the Pierce Bicichoninic Acid (BCA) protein assay, using bovine serum albumin as a standard, and following protocols of the manufacturer (Pierce Biotechnology). Subsequently, hippocampus protein samples were solubilized in buffer containing 1% sodium dodecyl sulfate (SDS) and separated on 10% Tris–glycine Novex gels (Invitrogen). To begin with, a linear range for assessing GAP-43 content in hippocampus protein samples was established. Thus concentrations of 20, 10, 5, and 3 ␮g total protein from a non-stimulated guinea pig in the SAL group was run on gel (electrophoretic settings: cycle 1; 35 min, 190 V, 170 mA and cycle 2; 15 min, 190 V, 70 mA). The running buffer was composed of 50 mM 3-(N-morpholino)propane sulfonic acid, 50 mM Tris–base, 3.5 mM SDS, and 1 mM ethylenediaminetetraacetic acid (EDTA). Based on the linear portion of the plots that were determined, an appropriate hippocampal protein concentration was selected for use in subsequent quantitative assays of GAP-43. Thus, representative protein samples from each animal group were run in parallel along with a rat brain lysate control. In all cases, protein bands on the gels were transferred onto polyvinylidene fluoride (PVDF) membranes using the Novex XCell II blotting apparatus according to the protocols of the manufacturer (Novex). The transfer buffer was composed of 25 mM bicine, 25 mM bis–Tris, 1 mM EDTA, and 0.05 mM chlorobutanol (transfer settings: 4 h, 35 V, 185 mA). Two PVDF membranes were used per gel. The first membrane in contact with the gel was probed for GAP-43. The second membrane was processed with colloidal gold total protein stain (Bio-Rad). Gels were placed in coomassie blue overnight with continuous agitation on a shaker at room temperature. Subsequently, each gel was analyzed to check for residual protein bands resulting from incomplete gel-to-primary membrane transfer. If substantial protein residuals were detected on a gel, then its associated primary PVDF membrane was excluded from Western blot data analysis. To quantify GAP-43, each primary PVDF membrane was incubated in 0.1× casein solution (Vector Laboratorys) while gently agitating at room temperature for 10 min. The concentration of casein solution was increased to 1× (1%, v/v), and the membrane was incubated and agitated as above for 10 min. This was followed by a 90-min incubation of the blot in a 1:5000 dilution of anti-GAP-43 primary antibody (AB5220; Chemicon International). Subsequently, the membrane were washed for 5 min, four times, at room temperature in PBS and Tween-20 (PBST). Then the blot was incubated

436

S. Chirwa et al. / Brain Research Bulletin 64 (2005) 433–440

with goat anti-mouse IgG, peroxidase conjugated secondary antibody (1:100,000 dilution in 1× Casein) for 1 h at room temperature. This was followed by four 5-min washes at room temperature in PBST. Thereon, the remainder of this reaction was performed under subdued lighting. Each blot was immersed for 5 min in SuperSignal west femto substrate working solution, which contained equal volumes of SuperSignal west femto luminol/enhancer solution and SuperSignal west femto stable peroxidase solution (Pierce Biotechnology). The blots were removed from the solution and excess solution removed by holding the edge of the blot with forceps and allowing the fluid to drip. Then, the blot was placed in a black container and allowed to shake for at least 5 min before it was scanned, using a UV-detector system (Ultra-Violet Products Inc.). After the blot was placed in the detection chamber, the focus of the camera was adjusted before imaging. Thereafter, the blot was scanned for at least 60 s for visualization of the GAP-43 protein bands. 2.5. GAP-43 immunochemistry This assay was conducted to qualitatively discern where GAP-43 protein was localized in the hippocampus in situ. Thus immunocytochemistry was performed in a few 25 ␮m thick coronal sections remaining over from those utilized in the hybridization assay described earlier. Briefly, a few brain sections were treated for 30 min with 3% hydrogen peroxide in methanol to block endogenous peroxidases. The sections were washed in PBS with 1% Triton X-100 (pH 7.4; 30 min) followed by 30-min incubation in 10% normal goat serum (Chemicon) to block nonspecific binding. Sections were then incubated in a polyclonal anti-GAP-43 primary antibody (1:100; Chemicon) overnight at 4 ◦ C. The next day, sections were incubated for 60 min in a biotinylated secondary antibody and then in an avidin-conjugated tertiary antibody (Vector Laboratories). Following each incubation, sections were washed in PBS for 10 min. They were subsequently treated for 10–15 min with Vector VIP peroxidase substrate for color development, dehydrated through xylenes, mounted in permount, and cover-slipped. Control sections were treated identically to the experimental sections except that primary antibody was absent during the overnight incubation period. Tissue sections were analyzed using a light microscope.

3. Results For data analysis, the criterion for animal inclusion was that recordings exhibited stable evoked population spikes monitored for at least 30 min. Stimulation intensities greater than 1000 ␮A produced plateau responses whereas stimuli of 600–800 ␮A elicited 60–80% of plateau responses; these latter values were used in the experiments. The evoked field potentials typically comprised of an initial positive-going wave (PW) that was interrupted by a negative-going population spike (PS) followed by a second PW (Fig. 1). Briefly, the first PW primarily reflects synaptic depolarization whereas the second PW corresponds to inhibitory influences occurring in CA1 cells [1,2]. The population spike is due to the algebraic summation of synchronously discharging CA1 neurons, and we quantified it as ‘the vertical rise from peak negativity to the intersection point with a line drawn at a tangent between the two PWs’. The population spike amplitudes were similar among the different animal groups and exhibited amplitudes ranging between 9 and 11 mV. The evoked population spikes had typical ‘stimulus artifact to peak of population spike negativity’ onset latencies ranging between 10 and 12 m. The high frequency tetanic stimulation (100 Hz, 1 s, three trains at 15 s intervals) produced both early-LTP and late-LTP in the SAL group (population spike amplitude: pre-tetanus, 9.86 ± 0.39 mV; 60 min post-tetanus, 17.55 ± 0.84 mV; 180 min post-tetanus, 16.76 ± 0.79 mV; n = 14; values are means ± S.E.M. in this and subsequent entries), whereas guinea pigs in the SCH group (n = 6) exhibited early-LTP but not late-LTP (population spike amplitude: pre-tetanus, 9.93 ± 0.99 mV; 60 min post-tetanus, 15.78 ± 1.26 mV; 180 min post-tetanus, 11.22 ± 0.79 mV; F(4, 25) = 3.752, p < 0.05, ANOVA with Tukey–Kramer test; Fig. 1). Though not significantly dif-

2.6. Measurements and statistics Blot and autoradiographic data were analyzed using the Micro-Computer Imaging Device (Imaging Research Inc.) or the UV-Detection System (Ultra Violet Products Inc.). Both systems have softwares that facilitate semi-quantitative densitometric measurements from scanned gels, blots and/or autoradiographic films. In terms of statistics, one-way ANOVA was used for comparisons among samples; Tukey–Kramer test was the post hoc used to detect statistical differences (α = 0.05).

Fig. 1. Evoked population spikes showing LTP in guinea pig hippocampus in vivo. The basic recording comprised of an initial positive-going wave (PW) that was interrupted by a negative-going population spike (PS) followed by a second PW. Population spike amplitude was measured as the vertical rise from peak negativity to the intersection point with a line drawn at a tangent between the two PWs. It can be seen that early-LTP occurred in all treatment groups. However, late-LTP only occurred in the SAL group but not in guinea pigs in SCH group. This is evident when pre-tetanus population spikes are compared to post-tetanus population spikes at the 180 min.

S. Chirwa et al. / Brain Research Bulletin 64 (2005) 433–440

Fig. 2. Time plots of LTP across CA3/CA1 synapses in guinea pig hippocampus in vivo. Briefly, LTP was observed as significant population spike enhancements of at least 50% above pre-tetanus baseline responses. Both early-LTP followed by late-LTP was present in animals in the SAL group. In contrast, only early-LTP developed in guinea pigs in the SCH group. Though statistically insignificant, the “maximal” magnitude of early-LTP (astericks) observed in animals in the SAL group tended to be higher relative to that recorded in SCH group. Each point is a mean ± S.E.M.; number of animals: SAL = 14 and SCH = 6. Arrow indicates time for tetanic stimulation, 100 Hz, 1 s, repeated three trains at 15 s intervals.

ferent, the “maximal” magnitude of early-LTP observed in animals in the SAL group tended to be higher relative to that recorded in the SCH group (Fig. 2). After recordings, brains harvested from the first batch of guinea pigs were probed for GAP-43 mRNA. We found that GAP-43 mRNA hybridization in situ was largely localized to cellular layers in the hippocampus (Fig. 3A). In fact, matched transverse sections obtained from the same brain but probed with sense oligonucleotides had greatly diminished GAP-43 mRNA hybridization in situ. Furthermore, when some of the transverse sections were pre-treated with RNAse to digest RNA before being probed with antisense oligonucleotides, there was an attenuation in detected GAP43 mRNA hybridization in situ. These results indicated that the oligonucleotide probes utilized in our study were probably detecting GAP-43 mRNA in guinea pig hippocampus. Within this context, we noted that densitometric quantifications of autoradiographic films consistently showed a significant increase in GAP-43 mRNA hybridization in brains with LTP relative to unstimulated controls (F(21, 242) = 3.502, p < 0.05, ANOVA with Tukey-Kramer test; Fig. 3B). All the key regions (i.e., CA1, CA3, and dentate gyrus) showed upregulated GAP-43 mRNA hybridization. In all cases, the increase in GAP-43 mRNA hybridization occurred in both left and right hippocampus but overall, we did not detect significant differences in GAP-43 mRNA hybridization between the two sides. This is not entirely surprising since ipsilateral CA3 stimulation does simultaneously evoke population spikes in both left and right hippocampus [7,9], and tetanic stimulation produces bilateral LTP in the two sides [7]. The finding that GAP-43 mRNA was up-regulated, made us wonder if this was associated with an increase in newly synthesized protein. This appeared to be the case in view of

437

Fig. 3. (A) A transverse section probed with antisense oligonucleotides showed GAP-43 mRNA hybridization was localized to somatic layers in hippocampus. An adjacent section from the same brain that was probed with sense oligonucleotides had diminished GAP-43 mRNA hybridization. Similarly, GAP-43 mRNA hybridization was not detected in sections pre-treated with RNAse prior to being probed with antisense oligonucleotides. (B) GAP43 mRNA hybridization in situ. For analyses, three to six randomly selected transverve brain sections encompassing much of the dorsal hippocampus were processed, scanned and used to determine average densitometric values for each hybridization protocol per animal. Subsequently, data from the best section from a guinea pig brain were pooled with similar sections from other animals in the same group. The unstimulated brains comprised the reference group (n = 3). All other brains were tetanized and developed LTP. Overall densitometric scores were significantly different for both SAL group relative to the unstimulated group (see text).

the following findings. Hippocampal lobes harvested from the six guinea pigs in the SAL group (see METHODS) were examined for total GAP-43 after recordings. We found that hippocampal protein samples probed with rat polyclonal antibody to GAP-43 (Chemicon) yielded specific bands that matched GAP-43 bands from a rat brain lysate. In addition, total protein samples between 5 and 15 ␮g per lane yielded GAP-43 bands that fell within the linear concentration range (Fig. 4A). Based on this result, a 5 ␮g total protein concentration was selected and used in quantitative assays of GAP-43. We found that densitometric measurements from Western blots revealed significant increases in GAP-43 levels in brains with LTP relative to three unstimulated controls (F(2, 15) = 4.02, p < 0.05, ANOVA with Tukey–Kramer test; Fig. 4B). In contrast, animals pre-treated with SCH23390 exhibited unchanged GAP-43 levels (n = 6). Taken together, our data raised the prospect that DA1 R activation could be the putative signaling pathway coupling GAP-43 and LTP. The success we had quantifying GAP-43 expression in guinea pig hippocampus spurred us to qualitatively assess layers in the hippocampus staining positive to antiGAP-43 binding in situ. GAP-43 is largely localized to

438

S. Chirwa et al. / Brain Research Bulletin 64 (2005) 433–440

the same brains but not probed with anti-GAP-43 lacked immunoreactivity in situ thereby suggesting some level of specificity with our techniques.

4. Discussion

Fig. 4. Densitometric measurements from GAP-43 Western blots. Western blotting utilized 5 ␮g total protein each and was probed with rat polyclonal anti-GAP-43. The unstimulated animals comprised the reference group (UNS; n = 3), whereas guinea pigs in SAL group (n = 6 each) had been tetanized and developed LTP. Note the up-regulated levels of GAP-43 protein in brains with LTP, whereas levels of GAP-43 in the SCH23390-treated group (n = 6) remained similar to the unstimulated group. Asterick indicates significant differences relative to UNS group (F(2, 15) = 4.01, p < 0.05; ANOVA with Tukey–Kramer test).

presynaptic boutons [18]. So, we examined a few unused brain sections from the GAP-43 mRNA hybridization in situ study discussed above. We found GAP-43 immunoreactivity was intense in the afferented/dendritic regions in the hippocampus. In contrast, the cellular layers comprising of the cornu ammonis pyramidal cells or dentate gyrus granular cells [15] lacked major GAP-43 immunostaining (Fig. 5). Similarly, adjacent transverse sections obtained from

Fig. 5. (A) Immunostaining of GAP-43 was mainly found in the heavily afferented dendritic regions in the hippocampus. By contrast, the cellular region comprised of pyramidal cells (Pyr) or granule cells (Gra) lacked any major GAP-43 immunostaining. (B) GAP-43 immunoreactivity was absent in an adjacent transverse section from the same brain but not probed with primary antibody. The images were acquired at 6.3× magnification.

We have shown up-regulation of GAP-43 mRNA and protein in association with LTP in the guinea pig hippocampus. This is consistent with several reports in the literature that have linked these macromolecules to LTP in mice and rats [4,17]. In particular, we have presented new data indicating that the occurrence of late-LTP and up-regulation of GAP-43 protein in hippocampus is simultaneously blocked by SCH23390, a selective DA1 R antagonist [29]. There are two notable features that arise from our data as discussed below. First, our findings correspond well with what has already been described about DA1 R and LTP per se. For example, specific agonists of DA1 Rs induce a slowly developing enhancement of EPSPs in the CA1 region. This increase in synaptic efficacy starts 50 min after drug application and lasts beyond 6 h [12]. In contrast, it has been reported that although mice lacking the DA1 Rs develop early-LTP, they do not develop late-LTP [16]. Furthermore, investigators have found that the development of late-LTP is blocked by SCH23390 [12,23,30]. The literature also has information supporting the correlation between new GAP-43 mRNA expression and LTP maintenance. Thus, perforant path LTP in intact mouse hippocampal dentate gyrus, for example, is associated with increased GAP-43 mRNA expression in hilar cells; the increase is positively correlated with the level of enhancement [20]. Our data provide evidence that links both GAP-43 and LTP expression to DA1 R functions. Hence, we may reasonably infer that both GAP-43 expression and late-LTP development share a common modulatory pathway that partly serves to stabilize temporary synaptic changes and thereby extend LTP maintenance. Second, it is very unlikely that up-regulation of GAP-43 was due to translocation of protein from different regional or cellular domains because we quantified GAP-43 by assessing total protein isolated from the whole hippocampus. Rather, the observed increase in GAP-43 after induction of LTP reflected new synthesis, and this is supported by the concomitant up-regulation of GAP-43 mRNA documented in our study. However, we are unable to ascertain from our study if the increased GAP-43 mRNA expression reflected new gene translation per se in view of the following. The control of GAP-43 gene expression occurs through both transcriptional and post-transcriptional mechanisms, with the latter partly mediated by changes in the stability of GAP-43 mRNA [6,21]. For example, NGF treatment of PC12 cells increases the half-life of GAP-43 mRNA from 5–6 h to 30 h which allows the mRNA to accumulate and be translated into protein [24,25]. In fact cAMP-regulated phosphoprotein-19 (ARPP-19) binds to the region in the 3 end of GAP-43 mRNA

S. Chirwa et al. / Brain Research Bulletin 64 (2005) 433–440

implicated in regulating its half-life in an NGF-dependent manner [13]. Similarly the embryonic lethal abnormal vision (ELAV)-like RNA-binding protein, HuD, does increase (a) GAP-43 mRNA cytoplasmic stability, (b) GAP-43 gene expression, and (c) PKC-dependent neurite outgrowth in PC12 cells [19]. Interestingly neuronal ELAV-like genes undergo a sustained up-regulation in hippocampal pyramidal cells of only mice and rats that have learned a spatial discrimination paradigm [26]. This increase is also accompanied by enhanced expression of the GAP-43 gene. Taken together, our data suggest that DA1 R signaling partly couples GAP-43 and LTP. Clearly, more detailed analysis of intracellular signaling cascade(s) downstream to DA1 R activation is needed to verify this interaction. Studies are needed to further explore the specificity of DA1 R-mediated signaling pathways in coupling GAP-43 and LTP. Indeed we recorded the population spike but not the dendritic field EPSP in the present study. This was done because the population spike is a consequence of algebraically summed bioelectrical activities from both apical and basal dendrites. This was ideal since CA3 cells give rise to Schaffer collaterals that innervate mostly CA1 apical dendrites on the same side, but their commissural afferents form synapses with CA1 basal dendrites on the contralateral side [14]. However, two distinct forms of LTP have been described, namely, that which exhibits potentiation of the population spike only (i.e., E–S potentiation), and that which shows concurrent increases to both population spike and field EPSP (i.e., synaptic potentiation [3]). Future recordings should also examine dendritic field EPSPs to unequivocally confirm the role of DA1 R activation in coupling GAP-43 and synaptic LTP per se. Some of this work is in progress in the laboratory.

Acknowledgements This work was supported by the National Institutes of Health grants MH57067 to S. Chirwa, RR13646 to E. BareaRodriguez, graduate fellowship DA05846 to A. Aduonum, and program grant RR03032 to Meharry Medical College.

References [1] P. Andersen, J.C. Eccles, Y. Loyning, Pathways of post-synaptic inhibition in the hippocampus, J. Neurophysiol. 27 (1964) 608– 619. [2] P. Andersen, T.V.P. Bliss, K.K. Skrede, Unit analysis of hippocampal population spikes, Exp. Brain Res. 13 (1971) 208–221. [3] P. Andersen, S.H. Sundberg, O. Sveen, J.W. Swann, H. Wigstrom, Possible mechanisms for long-lasting potentiation of synaptic transmission in hippocampal slices from guinea pigs, J. Physiol. (Lond.) 302 (1980) 463–482. [4] L.I. Benowitz, A. Routtenberg, GAP-43: an intrinsic determinant of neuronal development and plasticity, TINS 20 (1997) 84–91. [5] T.V.P. Bliss, G. Collingridge, A synaptic model of memory and long-term potentiation in the hippocampus, Nature 361 (1993) 31– 39.

439

[6] I. Cantallops, A. Routtenberg, Activity-dependent regulation of axonal growth: post-transcriptional control of the GAP-43 gene by the NMDA receptor in developing hippocampus, J. Neurobiol. 41 (1999) 208–220. [7] S.S. Chirwa, J. Mack, H. Park, K. Dennis, A. Aduounum, An in vivo model for investigating bilateral synaptic plasticity across CA3/CA1 synapses in guinea pig dorsal hippocampus, J. Neurosci. Methods 110 (2001) 25–30. [8] S. Davis, T.V.P. Bliss, G. Dutrieux, S. Laroche, M.L. Errington, Induction and duration of long-term potentiation in the hippocampus of the freely moving mouse, J. Neurosci. Methods 75 (1997) 75– 80. [9] G.T. Finnerty, J.G.R. Jefferys, Functional connectivity from CA3 to the ipsilateral and contralateral CA1 in the rat dorsal hippocampus, Neuroscience 56 (1993) 101–108. [10] U. Frey, Y.-Y. Huang, E.R. Kandel, Effects of cAMP simulate a late stage of LTP in hippocampal CA1 neurons, Science 260 (1993) 1661–1664. [11] U. Frey, R.G. Morris, Synaptic tagging: implications for late maintenance of hippocampal long-term potentiation, Trends Neurosci. 21 (1998) 181–188. [12] Y.-Y. Huang, E.R. Kandel, D1/D5 receptor agonists induce a protein synthesis-dependent late potentiation in the CA1 region of the hippocampus, PNAS 92 (1995) 2446–2450. [13] N. Irwin, S. Chao, L. Goritchenko, A. Horiuchi, P. Greengard, A.C. Nairn, L.I. Benowitz, Nerve growth factor controls GAP-43 mRNA stability via the phosphoprotein ARPP-19, PNAS 99 (2002) 12427–12431. [14] D. Johnson, D.G. Amaral, Hippocampus, in: G.M. Shepherd (Ed.), The Synaptic Organization of the Brain, fourth ed., Oxford University Press, Oxford, 1998, pp. 417–458. [15] R. Lorente De No, Studies on the structure of the cerebral cortex. II. Continuation of the study of the ammonic system, J. Psychologie und Neurologie 46 (1934) 113–177. [16] H. Matthies, A. Becker, H. Schroeder, J. Kraus, V. Hollt, M. Krug, Dopamine D1-deficient mutant mice do not express the late phase of hippocampal long-term potentiation, Neuroreport 8 (1997) 3533–3535. [17] P.J. Meberg, E.G. Valcourt, A. Routtenberg, Protein F1/GAP-43 and PKC gene expression patterns in hippocampus are altered 1–2 h after LTP, Mol. Brain Res. 34 (1995) 343–346. [18] K. Meiri, H. Pfenninger, M. Willard, Growth-associated protein, GAP-43, a polypeptide that is induced when neurons extend axons, is a component of a subcellular fraction enriched in growth cones, PNAS 83 (1986) 567–570. [19] C.D. Mobarak, K.D. Anderson, M. Morin, A. Beckel-Mitchener, S.L. Rogers, H.H. Furneaux, P. King, N.I. Perrone-Bizzozero, The RNAbinding protein HuD is required for GAP-43 mRNA stability, GAP43 gene expression, and PKC-dependent neurite outgrowth in PC12 cells, Mol. Biol. Cell 11 (2000) 3191–3203. [20] U. Namgung, S. Matsuyama, A. Routtenberg, Long-term potentiation activates the GAP-43 promoter: selective participation of hippocampal mossy cells, PNAS 94 (1997) 11675–11680. [21] E. Nedivi, G.S. Basi, I.V. Akey, J.H.P. Skene, A neural-specific GAP43 core promoter located between unusual DNA elements that regulate its activity, J. Neurosci. 12 (2000) 691–704. [22] A.B. Oestreicher, P.N.E. De Graan, W.H. Gispen, J. Verhaagen, L.H. Scrama, B-50, the growth-associated protein-43: modulation of cell morphology and communication in the nervous system, Prog. Neurobiol. 53 (1997) 627–686. [23] N.A. Otmakhova, J.E. Lisman, D1/D5 dopamine receptor activation increases the magnitude of early long-term potentiation at CA1 hippocampal synapses, J. Neurosci. 16 (1996) 7478–7486. [24] N.I. Perrone-Bizzozero, V.V. Cansino, D.T. Kohn, Posttranscriptional regulation of GAP-43 gene expression in PC12 cells through PKC-dependent stabilization of the mRNA, J. Cell Biol. 120 (1993) 1263–1270.

440

S. Chirwa et al. / Brain Research Bulletin 64 (2005) 433–440

[25] N.I. Perrone-Bizzozero, R.L. Neve, N. Irwin, S. Lewis, L. Fischer, L.I. Benowitz, Post-transcriptional regulation of GAP-43 mRNA levels during neuronal differentiation and nerve regeneration, Mol. Cell Neurosci. 2 (1991) 402–409. [26] A. Quattrone, A. Pascale, X. Nogues, W. Zhao, W. Gusev, A. Pacini, D. Alkon, Post-transcriptional regulation of gene expression in learning by the neuronal ELAV-like mRNA-stabilizing proteins, PNAS 98 (2001) 11668–11673. [27] A. Routtenberg, Tagging the Hebb synapse, TINS 22 (1999) 255–256. [28] J.R. Sanes, J.W. Lichtman, Can molecules explain long-term potentiation? Nat. Neurosci. 2 (1999) 597–604.

[29] P. Seeman, H.H. Van Tol, Dopamine receptor pharmacology, Trends Pharmacol. Sci. 15 (1994) 264–270. [30] J.L. Swanson-Park, C.M. Coussens, S.E. Mason-Parker, C.R. Raymond, E.L. Hargreaves, M. Dragunow, A.S. Cohen, W.C. Abraham, A double dissociation within the hippocampus of dopamine D1/D2 receptor and b-adrenergic receptor contributions to the persistence of long-term potentiation, Neuroscience 92 (1999) 485– 497. [31] W. Wisden, B.J. Morris, S.P. Hunt, In situ hybridization with synthetic DNA probes, in: J. Chad, H. Wheal (Eds.), Molecular Neurobiology, Oxford University Press, 1991, pp. 205– 225.