BDNF-induced LTP is associated with rapid Arc/Arg3.1-dependent enhancement in adult hippocampal neurogenesis

Adult neurogenesis in the hippocampus is a remarkable phenomenon involved in various aspects of learning and memory as well as disease pathophysiology. Brain-derived neurotrophic factor (BDNF) represents a major player in the regulation of this unique form of neuroplasticity, yet the mechanisms underlying its pro-neurogenic actions remain unclear. Here, we examined the effects associated with brief (25 min), unilateral infusion of BDNF in the rat dentate gyrus. Acute BDNF infusion induced long-term potentiation (LTP) of medial perforant path-evoked synaptic transmission and, concomitantly, enhanced hippocampal neurogenesis bilaterally, reflected by increased dentate gyrus BrdU + cell numbers. Importantly, inhibition of activity-regulated cytoskeleton-associated protein (Arc/Arg3.1) translation through local, unilateral infusion of anti-sense oligodeoxynucleotides (ArcAS) prior to BDNF infusion blocked both BDNF-LTP induction and the associated pro-neurogenic effects. Notably, basal rates of proliferation and newborn cell survival were unaltered in homozygous Arc/Arg3.1 knockout mice. Taken together these findings link the pro-neurogenic effects of acute BDNF infusion to induction of Arc/Arg3.1-dependent LTP in the adult rodent dentate gyrus.

Scientific RepoRts | 6:21222 | DOI: 10.1038/srep21222 is required for stable expansion of the actin cytoskeleton at perforant path synapses. Inhibition of Arc/Arg3.1 synthesis with anti-sense oligodeoxynucleotides therefore blocks development of stable LTP 17 .
Here, we asked whether BDNF-LTP is associated with enhanced hippocampal NPC proliferation. BDNF was locally and unilaterally infused in the DG of adult male rats (Fig. 1). Immediately prior to BDNF, Arc/Arg3.1 anti-sense oligodeoxynucleotides (ODNs) ("ArcAS") or scrambled Arc ODN sequence control ("ArcASScr") were also administered locally and unilaterally. We hypothesized that if BDNF-mediated pro-neurogenic actions were mediated by LTP induction, inhibition of Arc/Arg3.1 translation would block these effects. New evidence supports the involvement of Arc/Arg3.1 in the regulation of hippocampal neurogenesis 10,23 . While Arc/Arg3.1 induction in 'older' adult-born dentate granule cells with mature functional synapses is required for LTP consolidation, Arc/Arg3.1 expression in newborn cells (prior to synapse development) strongly correlates with long-term survival and differentiation towards a neuronal phenotype 10 .
We report that acute unilateral infusion of BDNF is associated with a bilateral increase in BrdU+ cell numbers in the DG. Moreover, unilateral infusion of ArcAS blocked bilateral changes in proliferation, indicating that the BDNF-induced proliferative response is a consequence of the unilateral induction of Arc/Arg3.1-dependent BDNF-LTP. We also compared basal levels of proliferation between Arc/Arg3.1 knockout and wild-type mice and found no significant differences. Together these data suggest that exogenous BDNF enhances neurogenesis indirectly by a mechanism involving LTP induction in the adult granule cell population. The data do not rule out a role for Arc/Arg3.1 synthesis in progenitor cells in neurogenesis.

Results
Previous studies have shown that brief infusion of BDNF into the dentate gyrus induces LTP. In the present study, BDNF was infused immediately above the DG, 300μ m dorsal to the synaptic zone of medial perforant path fibers. Previous immunohistochemical analysis of BDNF distribution showed rapid delivery of BDNF to the dorsal dentate gyrus (within < 15 min of infusion onset) and clearance from the dentate gyrus in less than 1 hour; BDNF also diffuses along the cannula tract in the CA1 region, with variable spread to CA3 24 . BDNF infusion promotes the induction of LTP which lasts for at least 15 hours in anesthetized rats, rapidly activates the extracellular signal regulated protein kinase (ERK) 25,26 and requires transcription of new Arc mRNA and new Arc protein synthesis 17,24 . Accordingly, in the present experiment, BDNF infusion led to a robust increase in the fEPSP slope and population spike amplitude, an effect which was inhibited by ArcAS infusion but not ArcASScr (Fig. 2). Statistical analysis (general linear model for repeated measures) indicated a significant main effect of time (baseline vs infusion vs recording: F = 425.801 (Wilks' Lambda), p < 0.0001) and a significant interaction between time and treatment (Cytochrome C (CytC) vs BDNF vs BDNF+ ArcAS vs BDNF+ ArcASScr: F = 31.391, p < 0.0001). Subsequent pair-wise comparisons (Sidak's post-hoc) indicated a significant difference between animals receiving BDNF and CytC (a protein similar to BDNF in molecular weight and charge that was used to control for possible nonspecific effects of protein infusion; p < 0.001). The effect of BDNF infusion on LTP was completely inhibited by preceding ArcAS administration (BDNF vs BDNF+ ArcAS, p < 0.001) but not ArcASScr (BDNF vs BDNF+ ArcASScr, p > 0.05; CytC vs BDNF+ ArcASSc, p < 0.001). No differences were detected between CytC and BDNF+ ArcAS-treated animals (Fig. 2). These results show that infusion of exogenous BDNF in the DG was associated with the induction of robust LTP which requires the synthesis of new Arc protein. Blockage of Arc translation, through the administration of Arc AS (but not Arc ASScrambled) inhibited LTP induction.

Total labeled (IdU+ and BrdU+) cells.
Iodo-deoxy-Uridine (IdU) and Bromo-deoxy-Uridine (BrdU) are two thymidine analogs that integrate into cells undergoing DNA synthesis. Their incorporation into the DNA can be visualized through immunohistochemistry to provide an in vivo marker to identify proliferating cells. IdU IdU (230 mg/kg; i.p.) and BrdU (200 mg/kg; i.p.) were administered 24 and 4 hours, respectively, prior to sacrifice to time-stamp hippocampal newborn cells that underwent division. After a 20 min baseline, BDNF (or Cytochrome C) was infused, 20 minutes following BrdU administration, for 25 minutes at a speed of 80 nl/min. Subsequent BDNF-induced LTP was monitored and recorded for 195 minutes, before animals were sacrificed and tissue was collected for histological analysis. Arc/Arg3.1 antisense oligodeoxynucleotides (Arc AS) or Arc/ Arg3.1 scrambled oligodeoxynucleotides (Arc ASScr) were administered 90 minutes prior to BDNF infusion for 12.5 minutes at a rate of 80 nl/min to inhibit Arc/Arg3.1 translation and block BDNF-induced LTP. was given to adult male rats 24 hours prior to sacrifice and 20 hours prior to BrdU injection to determine basal proliferative rates prior to BDNF administration. BrdU was administered 4 hours prior to sacrifice to explore BDNF-induced effects on hippocampal NPC proliferation (Fig. 1). Importantly, the anti-IdU antibody does not discriminate between IdU and BrdU, recognizing both analogs and labeling cells which have incorporated IdU and/or BrdU. In contrast, the anti-BrdU antibody selectively detects BrdU without cross-reacting with IdU. As a result of primary antibody specificity, total number of labeled cells thus represents a combination of three distinct cell populations: IdU(only)+ cells, BrdU(only)+ and double IdU+ /BrdU+ cells. Total labeled cells (IdU(only)+ and BrdU+ ) were quantified by light microscopy across the entire rostral-caudal extent of the hippocampus using serial sets of coronal sections (Fig. 3a,b). Two-way ANOVA was run to determine inter-hemispheric differences (BDNF infusion vs control side), treatment effects (CytC vs BDNF vs BDNF+ ArcAS) and their interactions (treatment × infusion side). Interestingly, although BDNF was infused in only one hemisphere, no differences in total labeled cell numbers were found between stimulated and non-stimulated sides in any hippocampal region examined (hilus, SGZ and GCL) (Fig. 3d,e). A significant main effect of treatment was found in the SGZ (F 2,44 = 8.454, p < 0.001), with BDNF infusion markedly increasing total labeled cell numbers (t = 3.651, p < 0.001) while BDNF+ ArcAS administration prevented this BDNF-mediated proneurogenic effect (t = 3.500, p = 0.0011) (Fig. 3c). No effects were observed in the hilus or GCL.
Total numbers of labeled cells in the different hippocampal regions were subsequently used in combination with confocal imaging (Fig. 4) to calculate basal proliferative rates (total IdU(only)+ , total IdU+ /BrdU− /Arc− and total IdU+ /BrdU− /Arc+ cells) as well as treatment-induced changes on hippocampal proliferation (total BrdU+ , total BrdU+ /Arc− and total BrdU+ /Arc+ cells) for each individual rat.

Total BrdU+ cells (IdU-/BrdU+ and IdU+/BrdU+ cells).
As BrdU was given 20 minutes prior to BDNF infusion, total numbers of BrdU+ cells provided a direct measure of BDNF-mediated effects on hippocampal NPC proliferation. Our data showed a significant main effect of treatment in the SGZ (F 2,44 = 16.391, p < 0.001) but not in the hilus or GCL (Fig. 5a). Interestingly, no main effects were found between the infusion and the contralateral side (Fig. 5b,c). Post-hoc analysis revealed a significant increase of total BrdU+ cells in the SGZ (BDNF vs CytC, t = 5.369, p < 0.001) while concurrent ArcAS administration prevented this effect (BDNF+ ArcAS vs BDNF, t = 4.467, p < 0.001) (Fig. 5a). These results suggest that the increased bilateral numbers of BrdU+ cells in response to unilateral BDNF infusion depended on BDNF-LTP induction. This was confirmed by the observation that unilateral infusion of ArcAS blocked BDNF-induced pro-neurogenic actions, bilaterally.
Total proliferating and surviving BrdU-labeled cells in Arc/Arg3.1-deficient mice. In view of the importance of Arc synthesis in BDNF-LTP, we investigated levels of basal proliferation and total neurogenesis in Arc/Arg3.1-deficient mice. BrdU was injected in Arc/Arg3.1-deficient mice and wild-type controls 24 hours or 28 days prior to sacrifice to investigate effects associated with reduced Arc/Arg3.1 expression on proliferating and differentiated newborn hippocampal cells (Fig. 8a-d). We hypothesized that if Arc transcription was critical in hippocampal neurogenesis, Arc/Arg3.1-deficient mice would display a significant reduction in the rate of proliferation and/or newborn cell survival compared to wild-type mice with normal Arc expression. If Arc involvement in hippocampal neurogenesis was limited to its role in the generation and maintenance of LTP, its The latter images demonstrate that although Arc seems ubiquitously expressed in the DG, some IdU+ and/ or BrdU+ cells were clearly Arc negative. Representative confocal photomicrographs depicting two different proliferating cells positive for BrdU and/or IdU (g-k): the first cell is clearly double positive for IdU and BrdU (j,j 1 ) while the second is only positive for IdU (k,k 1 ). effects on proliferation and/or survival would be limited as these mice were not stimulated (neither chemically with BDNF infusion nor behaviorally). Our data illustrate a significant reduction in the number of BrdU-labeled cells between proliferation and survival experiments in the hilus (F 1,20 = 17.96, p > 0.001), SGZ (F 1,20 = 140.06, p > 0.001) and GCL (F 1,20 = 139.70, p > 0.001), due to the well-documented loss of newborn cells during the first month after cell division (Fig. 8e). Interestingly, no main effects of strain (wild-type vs Arc/Arg3.1-deficient mice) or interactions were found in any hippocampal region examined, indicating that Arc-mediated influences on hippocampal neurogenesis were indirect and linked to its role in the generation and maintenance of LTP.

Discussion
The present results suggest LTP induction is a central mechanism through which acute BDNF infusion modulates hippocampal neurogenesis. Previous work has implicated enhanced BDNF signaling in the behavioral and cellular efficacy of antidepressant drugs 11-13,27-30 . Notably, a single bilateral infusion of BDNF in the dentate gyrus has been found to produce an antidepressant-like effect in behavioral models of depression 13 . Here, we show Graphs illustrating averages of total numbers of IdU(only)+ or IdU+ /BrdU-cells in the hippocampal hilus, subgranular zone (SGZ) and granule cell layer (GCL) per group (n = 8/9) ± SEM (d) as well as average numbers of IdU(only)+ cells in the stimulated (BDNF-infused, e) and contralateral hemisphere (f). The * and + symbols represent significant effects compared to CytC-or BDNF-treated animals, respectively. One, two or three symbols represent p < 0.05, p < 0.01, p < 0.001, respectively. that acute unilateral BDNF infusion induces stable LTP in the perforant pathway and enhanced hippocampal cell proliferation exhibited by increased BrdU+ cell numbers. Importantly, although only infused unilaterally, BDNF led to enhanced proliferation, bilaterally. Inhibition of Arc/Arg3.1 translation, through local and unilateral Arc/Arg3.1 anti-sense oligodeoxynucleotide (ArcAS) infusion, blocked both BDNF-LTP induction and bilateral enhancement of hippocampal cell proliferation. This data provides the first evidence that the pro-neurogenic actions of acute BDNF administration are coupled to induction of LTP, illustrating a possible mechanism through which the effects of BDNF are translated, at a systems level, into beneficial behavioral actions.
Here, unilateral BDNF-LTP induction was associated with a bilateral increase in DG BrdU+ cell numbers (Fig. 5a-c). Blocking of the bilateral cell proliferation by unilateral ArcAS infusion, in turn, coupled this effect to unilateral BDNF-LTP induction. In previous work we have shown that the infused BDNF is restricted to the ipsilateral hippocampus and does not spread to the contralateral side 24 . LTP of perforant path-evoked transmission is itself a unilateral response coupled to NMDA receptor activation, BDNF-TrkB signaling and gene expression in the ipsilateral DG. LTP induction can have secondary effects in the network however, including bilateral changes in specific genes' expression 33 . Inter-hemispheric communication in response to unilateral LTP has been demonstrated in fMRI studies in rats and appears to be mediated by activation of the mossy cell-commissural pathway to the contralateral DG 34,35 . Enhanced mossy cell-commissural transmission has in turn been suggested to function in the bilateral coordination of the DG networks involved in pattern separation and sequence learning 36 . While the current results link unilateral BDNF-LTP to bilateral proliferation, the implicated mechanisms are unknown and we do not exclude a potential role for direct effects of exogenous BDNF on NPCs. and contralateral hemisphere (c). Graphs illustrating average total numbers of IdU+ /BrdU-/Arc-cells in the hippocampal hilus, subgranular zone (SGZ) and granule cell layer (GCL) per group (n = 8/9) ± SEM (d) as well as average numbers of IdU+ /BrdU-/Arccells in the stimulated (BDNF-infused, e) and contralateral hemisphere (f). The * and + symbols represent significant effects compared to CytC-or BDNF-treated animals, respectively. One, two or three symbols represent p < 0.05, p < 0.01, p < 0.001, respectively. In addition to increased bilateral BrdU+ cell numbers in the SGZ (Fig. 5a), unilateral BDNF infusion was accompanied by a corresponding bilateral reduction of IdU+ /BrdU− cells in the hilus, SGZ and GCL (Fig. 5d). This suggests BDNF not only promotes the de novo proliferation of neuroprogenitor cells (IdU-/BrdU+ cells), but also stimulates the re-proliferation of NPCs which divided less than 20 hours prior to BDNF infusion (IdU+ / BrdU− cells). This new proliferative event reduced the IdU+ /BrdU− cell population since proliferating cells that incorporated IdU (administered approximately 20 hours prior to BrdU) divided again in response to BDNF and incorporated BrdU, thus becoming double positive for both thymidine analogues (IdU+ /BrdU+ cells) (Fig. 1). Due to the impossibility of distinguishing single-(IdU− /BrdU+ ; cells that divided once in response to BDNF) and double-labeled cells (IdU+ /BrdU+ ; cells that proliferated twice in the 24 hours prior to sacrifice), both cell populations were labeled as BrdU+ cells. This categorization thus explains why BrdU+ cells increased in response to BDNF (Fig. 5a) while IdU+ /BrdU− cells declined (Fig. 5d).
Previously we reported Arc/Arg3.1 expression in BrdU+ cells from a very early post-mitotic age -as early as 1 day after cell division 10 . In the present study, detailed confocal analysis revealed that all cells proliferating in response to BDNF infusion expressed Arc/Arg3.1 (BrdU+ /Arc+ cells) (Fig. 6a). No changes were found in the total number of BrdU+ /Arc− cells (Fig. 6d). Under basal conditions, only 1-5% of the total DG BrdU+ cell population is double-labelled with Arc/Arg3.1. Remarkably, the percentage of Arc/Arg3.1-expressing BrdU+ cells rose to more than 30% after BDNF infusion (Fig. 6a). It is important to emphasize that this bilateral increase in BrdU+ /Arc+ cells was blocked by unilateral ArcAS infusion. We thus consider it likely that the bilateral effects associated with BDNF infusion are a consequence of unilateral BDNF-LTP induction. Moreover, there appear to be two distinct populations of BrdU+ NPCs, one that proliferates and expresses Arc/Arg3.1 in response to

BDNF-LTP induction (BrdU+ /Arc+ cells) and one that is insensitive to LTP induction (BrdU+ /Arc− cells).
The mechanism underlying the bilateral increase in Arc/Arg3.1 expression in a subpopulation of young NPCs is presently unknown. Previously, we examined effects of unilateral high frequency stimulation-induced LTP (HFS-LTP) on Arc/Arg3.1 expression in newborn neurons 10 and were surprised to find Arc/Arg3.1 labelling in early post-mitotic cells. Moreover no differences were found between the HFS-treated and contralateral DG in percentage of BrdU+ /Arc+ cells. Although we, then suggested that the lack of inter-hemispheric difference in BrdU+ /Arc+ cell numbers in response to HFS-LTP was caused by the refractory nature of newborn cells to synaptically-evoked Arc/Arg3.1 expression, we now hypothesize that both HFS-LTP and BDNF-LTP promote bilateral pro-neurogenic effects and Arc/Arg3.1 expression in newborn cells through a similar mechanism. Since, it has been shown that the percentage of hippocampal BrdU+ cells expressing Arc/Arg3.1 increases with their age 10 , BrdU+ /Arc+ cells might be generated by the division of a subgroup of older NPCs which are quiescent under basal conditions but capable of undergoing rapid proliferation in response to LTP induction at perforant path-granule cell synapses. Furthermore, since Arc/Arg3.1 expression in newborn cells is associated with their long-term survival and neuronal differentiation 10 , this would provide a mechanism by which LTP induction directs neurogenesis towards the generation of new neurons (in contrast to glial cells).
The present findings suggest a role for Arc/Arg3.1 in the mechanism underlying BDNF-mediated pro-neurogenic actions, possibly within the intracellular machinery which controls induction and/or maintenance of LTP. To further examine Arc/Arg3.1's role in hippocampal neurogenesis, we investigated NPC proliferation and newborn cell survival in Arc/Arg3.1-knockout mice. While these transgenic mice were generated almost a decade ago 18 , no studies to date have explored the consequences of Arc/Arg3.1 deficiency on hippocampal neurogenesis. We hypothesized that if Arc/Arg3.1 regulates basal rates of NPC proliferation, a general reduction in BrdU+ cells would be clearly detectable in Arc/Arg3.1-deficient mice compared to wild-types. Basal proliferation rates may be unaffected however if Arc/Arg3.1 works selectively in activity-dependent regulation of neurogenesis, such as in response to LTP induction. Our results appear to support the latter as no changes were found in the total number of DG BrdU+ cells in ArcArg3.1-deficient mice sacrificed after 24 hours or 28 days compared to wild-type controls (Fig. 8e). Although our data suggest a secondary involvement of Arc/Arg3.1 in neurogenesis possibly related to its role in LTP generation and maintenance, we acknowledge these findings were obtained using homozygous Arc/Arg3.1-knockout mice whereas previous studies using other knockout lines have shown that functional loss of a gene due to a null mutation can often be compensated for, e.g. by their paralog(s). Granted we do not have the data to either confirm or reject this possibility in our Arc/Arg3.1-deficient mouse model, it will be important to explore the role of Arc gene/protein in the regulation of basal hippocampal neurogenesis using a conditional Arc/Arg3.1-knockout mouse model.
Our current findings point toward LTP induction as a primary mechanism underlying BDNF-mediated pro-neurogenic actions. Moreover, they illustrate an important yet complex role for the immediate-early gene, Arc/Arg3.1 and its protein in the BDNF-LTP-neurogenesis interplay. We propose the following scenario: 1) BDNF-induced expression of Arc/Arg3.1 in mature DG cells is required for BDNF-LTP 17 ; 2) BDNF-LTP activates mechanisms/signaling which lead to rapid bilateral proliferation of NPCs and 3) Arc/Arg3.1 expression characterizes the progeny of a subpopulation of quiescent NPCs which proliferate in response to BDNF-LTP. Taken together, these findings could hold relevance for learning and memory processes given the central role of BDNF in their regulation 4,[37][38][39] . In the DG, acute BDNF signaling is implicated in pattern separation, a process by which similar experiences or events are transformed into discrete, non-overlapping memory representations, particularly during the consolidation of pattern-separated memories 40 . Acute local BDNF infusion facilitates this consolidation, a process requiring adult-born neurons 41 , although it is not yet known whether BDNF acts on mature or immature neurons. Notably, impaired pattern separation has been linked to increased anxiety. Failed pattern separation, perhaps consequent to impaired BDNF response/signaling and/or dysfunctional hippocampal neurogenesis, may result in the generalization of previously encountered aversive events to new "innocuous" experiences as seen in individuals with panic and post-traumatic stress disorder 42 . Remarkably, BDNF administration, both intra-hippocampal and peripheral, has been associated with antidepressant effects in behavioral models of anxiety and depression 12,13,43 , supporting the hypothesis that altered BDNF signaling and aberrant regulation of neurogenesis are central in the pathophysiology of these disorders. Our results add a new piece to the complex puzzle of BDNF signaling and point towards Arc/Arg3.1 as a possible intracellular mediator underlying BDNF-mediated stimulation of NPCs/immature DG neurons. Although this protein is commonly known as a direct modulator of glutamatergic synapses, recent work has shown that Arc/Arg3.1 also enters the nucleus and regulates gene expression 44 . It is tempting to speculate that Arc/Arg3.1 function in newborn immature cells, prior to synapse formation, involves such a nuclear mechanism. Further studies are needed however to define the relevance and specificity of Arc/Arg3.1 in the transduction of BDNF-mediated signals with particular emphasis on its involvement in the regulation of NPC proliferation and integration of newborn cells in the adult hippocampal circuitry.

Materials and Methods
Animals. Four-month old Male Sprague Dawley rats (Møllegårds Avlslaboratorium, Denmark; n = 33, 250-350 g), four-month old male transgenic Arc/Arg3.1-deficient mice (n = 12, 20 gr) and wild-type controls (n = 12, 20 gr) were used. Animals were individually housed with ad libitum access to food and water under climate-controlled conditions (22º ± 1 °C). They were maintained on a 12-hour light/dark cycle (light on at 07:00) and received at least 7 days to acclimatize to their new environmental conditions prior to the experiments. This investigation was designed to minimize the number of animals and suffering, and carried out in accordance with the Norwegian and German Regulation on Animal Experimentation, and the European Convention for the Protection of Vertebrate Animals used for scientific purposes. All experimental procedures were reviewed and approved by the local animal welfare body and performed according to University of Bergen Guidelines for the Care and Use of Laboratory Animals (project id:4307).
Homozygous Arc/Arg3.1-deficient mice were provided by Prof. D. Kuhl and generated as previously described 18 . Briefly, genomic fragments of Arc/Arg3.1 were isolated from a λ phage genomic library (AB-1) Scientific RepoRts | 6:21222 | DOI: 10.1038/srep21222 prepared from 129/Sv(ev) embryonic stem (ES) cells. A 4 kb fragment encompassing the promoter and 5′ UTR was subcloned into pBLUESCRIPT (Stratagene), and a 3.7 kb fragment covering the whole open reading frame (ORF) and 3′ UTR was subcloned into pZErO-1. These plasmids were used for generation of a targeting construct. For additional information on targeting, genotyping, and anatomical analysis readers are referred to Plath et al. 18 .
Experimental design. In this study, the effects of intra-dentate unilateral BDNF infusion on hippocampal NPC proliferation were examined in male rats. The role of Arc/Arg3.1 protein on BDNF-mediated proneurogenic actions was also investigated by locally infusing ArcAS prior to BDNF administration (Fig. 1). Finally, the effects associated with reduced Arc/Arg3.1 expression on adult hippocampal neurogenesis were explored using adult Arc/Arg3.1-deficient mice.
Intra-dentate BDNF infusion and electrophysiology. BDNF was locally and unilaterally infused in the DG as previously described 17,24,26 . Briefly, rats were anesthetized with urethane (1.4-1.8 g/kg i.p.) and BDNF was infused immediately above the dorsal DG, in deep stratum-lacunosum-moleculare of field CA1, approximately 300 μ m above the nearest medial perforant path-granule synapses and 700 μ m above the hilar recording site. Infusion solutions of BDNF (2 μ g/2 μ l; Alomone Labs., Jerusalem) and cytochrome C obtained from yeast (Sigma, St Louis, MO, USA) were dissolved in sodium phosphate buffer, pH 7.0 and delivered by infusion pump at a rate of 80 nl/min over 25 minutes. With a molecular weight and charge similar to BDNF, cytochrome C was infused as a protein control as it has no effect on basal synaptic transmission or several signal transduction pathways that have been studied 25,26,45,46 . This administration protocol resulted in an immediate and potent LTP induction (Fig. 2). Signals from the dentate hilus were amplified, filtered (1 Hz to 10 kHz) and digitized (25 kHz). Acquisition and analysis of field potentials was accomplished using DataWave Technologies WorkBench software (Longmont, CO, USA). Responses were normalized to baseline and statistics were based on average values obtained during the baseline (from minute − 20 to 0), infusion (from 0 to minute 25) and final 20 min period of recording (from minute 170 to 195) (Figs 1 and 2).

Oligodeoxynucleotides (ODNs): Arc anti-sense and Arc scrambled ODNs. Chimeric ODNs con-
taining phosphorothioate linkages between the three bases on the 5′ and 3′ ends and phosphodiester internal linkages were synthesized, HPLC purified, ultrafiltrated, and sterilized (BIOGNOSTIK ® GmbH, Göttingen, Germany). The main Arc/Arg3.1 antisense ODN (ArcAS) used was directed against a 20-mer sequence (bases 209-228) covering the Arc start site. The ArcAS ODN sequence used in our experiment was 5′ GTC CAG CTC CAT CTG CTC GC 3′ while the scrambled Arc ODN sequence was 5′ CCT GCT GAC CTC CGT ATG CC 3′ . Scrambled Arc ODN sequence (ArcASScr), containing the same base composition of ArcAS sequence in a randomized order, served as control. ODNs did not contain motifs such as G-quartets, kinase domains or zinc fingers, and search of the European Molecular Biology Laboratory databases revealed no potential off-target genes (with significant homology and open secondary structure). Arc antisense and scrambled ODNs were infused at a rate of 80 nl/min over 12.5 min (Fig. 1). Although in this study the effects of anti-sense ODNs on Arc translation was not directly demonstrated, these effects have been extensively investigated in three previous studies. Initially, we showed that BDNF-LTP was accompanied by the selective upregulation of Arc mRNA and protein in the dentate gyrus 26 . In a following study, the effect of actinomycin D on Arc upregulation was assessed 24 . This transcription inhibitor blocked the upregulation of Arc protein in the dentate gyrus and induction of BDNF-LTP, strengthening the link between Arc and BDNF-LTP. Arc protein levels were measured by Western blot analysis of homogenates obtained from micro-dissected dentate gyrus, CA1 and CA3 regions. In the final study, the effects of Arc anti-sense ODN infusion on Arc protein expression (by quantitative immunoblot analysis) and high-frequency stimulation-induced LTP were investigated 17 . Our data revealed that Arc protein expression was reduced to 55 ± 10% in response to Arc AS administration compared to ArcAS scrambled-treated controls 17 confirming the efficacy and specificity of ArcAS ODNs in blocking Arc translation and suggesting a critical role for Arc protein and its synthesis for LTP consolidation.
Tissue collection. Rats were sacrificed after 195 minutes of recording (Fig. 1) and transcardially perfused with 4% paraformaldehyde in 0.1 M sodium phosphate buffer (pH 7.4). Mice were perfused with the same fixative 24 hours or 28 days after receiving BrdU. Brains were removed and post-fixed in the same solution for 24 hours at 4 °C, before being transferred to sodium phosphate buffer (NaPB 0.02 M, pH 7.4) and stored at 4 °C. Following cryoprotection of the brains by overnight immersion in a 30% sucrose solution, 35 μ m coronal serial sections were prepared on a cryostat microtome. Sections were collected in NaPB with sodium azide and stored at 4 °C. Immunohistochemical stainings were performed using subsets of coronal sections.
Immunohistochemistry. Every sixth section throughout the rostral/caudal extent of the hippocampus (Bregma − 2 to − 6 in rats and − 1.2 to − 2.6 in mice) was collected and coded before immunohistochemical processing and analysis to ensure objectivity. All stainings were performed on free-floating sections under continuous agitation.

Image Analysis and Quantification. Quantification of IdU+ cells in rats and BrdU+ cells in mice was
conducted using a modified unbiased stereology protocol 10,48 . All labeled cells in the granule cell layer (GCL), subgranular zone (SGZ) and hilus were counted (at 200×) regardless of size or shape using a Nikon Eclipse 80i microscope (Tokyo, Japan) coupled to a Nikon DS-5M camera. Cells were considered as being in the subgranular zone if they were located in or adjacent to the SGZ. Cells located more than two cell widths away from the SGZ were considered hilar. To facilitate counting, cell clusters were examined at 400×. Quantification was conducted bilaterally and the total number of labeled cells was estimated by multiplying the number of cells counted in every sixth section by 6 and reported as the total numbers of labeled cells (mean and SE). As rats received both IdU and BrdU (Fig. 1), it should be noted that the IdU antibody from BD Biosciences recognizes both analogs, and so will label cells which have incorporated IdU and/or BrdU. This means that the total number of labeled cells in rats thus represents a combination of three distinct cell populations: IdU(only)+ cells, BrdU(only)+ and double IdU+ /BrdU+ cells.
For multi-labeling, percentages of IdU+ and/or BrdU+ cells were determined by analyzing, in each animal, 50-100 randomly selected labeled cells throughout the GCL and SGZ using a Leica TCS SP2 AOBS confocal microscope (Leica Microsystems, Heidelberg GmbH). Care was taken to verify double-labeling and to control for false positives by examining double-positive nuclei in their z-axis and rotating them in orthogonal x-y planes using a 40× objective (1 μ m steps). To exclude potential cross-bleeding between detection channels, double-labeling was imaged in sequential scanning mode.
Given the distinct specificity profiles of the two antibodies (IdU antibody's dual specificity for IdU/BrdU and the BrdU antibody's sole specificity for BrdU), we applied a triple-labeling protocol using species-specific secondary antibodies conjugated to different fluorophores to identify IdU-and/or BrdU-labeled cells. Following this protocol, BrdU+ cells were labeled by both green (488-conjugated) and red (647-conjugated) fluorescent secondary antibodies, whereas IdU(only)+ cells which did not proliferate a second time in response to BDNF infusion were labeled by the red fluorescent antibody only (Fig. 4). If an IdU(only)+ cell divided again in response to BDNF, it would also incorporate BrdU thus becoming double labeled (IdU+ /BrdU+ ). An important limitation of this protocol lies in the inability to discriminate between IdU+ cells which re-divided in response to BDNF and incorporated BrdU (double IdU+ /BrdU+ cells) and IdU-negative neuroprogenitor cells which proliferated only once in response to BDNF (BrdU(only)+ ). Since the BD IdU antibody cannot distinguish between IdU and BrdU, all BrdU-labeled cells (BrdU(only)+ and double IdU+ /BrdU+ ) will be both green and red. Finally, depending upon Arc/Arg3.1 expression (yellow; 555-conjugated antibody), cells which divided within the timeframe of our experiment (Fig. 1) could appear either double (IdU(only)+ /Arc+ ) or triple-positive (IdU+ /BrdU+ /Arc+ ) as illustrated in Fig. 4.