A key role for TGF-β1 in hippocampal synaptic plasticity and memory

Transforming Growth Factor β1 (TGF-β1) is a well-known neuroprotective and neurotrophic factor demonstrated to play a role in synaptic transmission. However, its involvement in physiological mechanisms underlying synaptic plasticity and memory at hippocampal level has not been thoroughly investigated. Here, we examine the role of TGF-β1 in hippocampal long-term potentiation (LTP) and memory in adult wild type mice. Our data provide evidence that administration of exogenous TGF-β1 is able to convert early-phase-LTP into late-phase-LTP. Furthermore, we show that the block of the endogenous TGF-β1 signaling pathway by the specific TGF-β1 inhibitor SB431542, impairs LTP and object recognition memory. The latter impairment was rescued by administration of exogenous TGF-β1, suggesting that endogenously produced TGF-β1 plays a role in physiological mechanisms underlying LTP and memory. Finally, TGF-β1 functional effect correlates with an increased expression of the phosphorylated transcription factor cAMP-Responsive Element Binding protein.

Scientific RepoRts | 5:11252 | DOi: 10.1038/srep11252 hippocampal neurons with TGF-β 2, which was able to enhance postsynaptic currents and to increase the amplitude and frequency of miniature potentials 14 .
Based on these findings, here we investigate the role of TGF-β 1 in hippocampal synaptic plasticity and memory in wild type mice. For this purpose, we first examine the effect of exogenous TGF-β 1 on CA1 LTP and CREB phosphorylation. Then, we investigate the role of endogenous TGF-β 1 in hippocampal synaptic plasticity, recognition memory and p-CREB levels by blocking its signaling pathway with SB431542, a selective inhibitor of the activin-like kinase (ALK) TGF-β 1 receptors 15 .

TGF-β1 converts E-LTP in L-LTP in hippocampal CA1 area.
To evaluate whether TGF-β 1 might affect synaptic plasticity we performed electrophysiological experiments on hippocampal slices in vitro. We stimulated Schaffer collateral fibers by using a weak tetanus (1 tetanus consisting of a 10-burst train of 4 pulses at 100 Hz with the bursts repeated at 5 Hz) to produce E-LTP in CA1 stratum radiatum. Ten minutes perfusion of hippocampal slices with TGF-β 1 at a concentration of 10 ng/ml 7,16 before a weak tetanus produced a late-LTP (n = 7/7; 211.06 ± 15.92% vs. 125.08 ± 8.74% of baseline slope 120 min after tetanus in vehicle-treated slices; ANOVA for repeated measures of the post-tetanic time points: F (1,12) = 11.02, p = 0.006; Fig. 1a), whereas no effects were observed with lower concentrations of TGF-β 1 (n = 7 for each condition; 5 ng/ml: 146.15 ± 14.30% of baseline slope 120 min after tetanus; 1 ng/ml: 115.84 ± 9.19% of baseline slope 120 min after tetanus; ANOVA comparing the averages of the 120° min of LTP recording in slices treated with different doses of TGF-β 1 and vehicle: F (3,24) = 11.907, p < 0.0001; Bonferroni post hoc test: p < 0.0001 comparing TGF-β 1 10 ng/ml with vehicle or other concentrations used; Fig. 1b).
To investigate the molecular correlate of the TGF-β 1-induced increase in LTP, we turned to the transcription factor CREB, involved in various forms of synaptic plasticity and memory 4-6 . We used western blot to analyze phospho-CREB (p-CREB) levels on the same hippocampal slices used for electrophysiological experiments. In slices treated with TGF-β 1 for 10 min before a weak tetanus, a 56% increase in CREB phosphorylation was found (1.58 ± 0.07 vs. 1.01 ± 0.04; Bonferroni post hoc: p = 0.001; two-way ANOVA for the interaction between tetanization and TGF-β 1 treatment: F (2, 6) = 25.470, p = 0.001; Fig. 1e).
Selective inhibition of endogenous TGF-β1 signaling impairs object recognition memory. We then studied recognition memory, a task based on the natural tendency of rodents to explore unfamiliar objects which, as recently documented by an increasing number of studies, depends on hippocampus integrity 18,19 . For this purpose we performed Object Recognition Test (ORT) 20,21 . The first day (T1) mice (e) Western blot analyses shows that a treatment with TGF-β 1 for 10 min before a weak tetanus significantly increased CREB phosphorylation. Slices treated with TGF-β 1 alone show a slight increase of p-CREB, whereas no p-CREB changes are observed in vehicle-treated tetanized slices after a weak-tetanus or in slices treated with vehicle alone. β -tubulin expression is shown as an internal control. * significant difference (p < 0.05).

Discussion
TGF-β 1 is an anti-inflammatory cytokine, which has been demonstrated to have neurotrophic and neuroprotective properties 8, 17 . Here we provide novel evidence that TGF-β 1 has a role in physiological mechanisms underlying synaptic plasticity and memory in the hippocampus.
We first showed that treatment of hippocampal slices with TGF-β 1 converted E-LTP into L-LTP, consistent with previous reports indicating that TGF-β 1 induced long-term but not short-term facilitation in Aplysia ganglia 11 , where it also increased excitability and decreased the firing threshold of cultured sensory neurons 12 .
In our electrophysiological experiments, the effective dose of TGF-β 1 was 10 ng/ml, as in previous works indicating that 10 ng/ml of TGF-β 1 was able to significantly increase p-CREB in sensory neurons of Aplysia 7 , whereas we did not observe any effect with 1 ng/ml or 5 ng/ml. We also demonstrated that endogenous TGF-β 1 plays a role in LTP. Indeed, inhibition of TGF-β 1 signaling by SB431542, a selective inhibitor of ALK5 TGF-β type I receptor, impaired L-LTP induced by a strong tetanization, whereas it did not modify potentiation after a weak tetanus. Rescue experiments by the addition of TGF-β 1 concomitant to SB431542 restored LTP, confirming the role of endogenous TGF-β 1 signaling in LTP.
These electrophysiological findings prompted us to investigate the role of TGF-β 1 in long-term memory by ORT, a test relying on the function of the hippocampus 18,19 , the perirhinal and the medial temporal lobe cortices, widely used to evaluate the neural basis of memory 20,21 . As for LTP, the inhibition of TGF-β 1 signaling obtained after i.p. as well as intrahippocampal administration of SB431542 caused a severe impairment of recognition memory that was completely rescued by TGF-β 1, providing a novel role for TGF-β 1 in learning and memory. Few other studies have investigated the involvement of TGF-β 1 on hippocampal memory demonstrating that it was able to rescue cognitive impairment in pilocarpine-treated rats 22 and that TGF-β 1 hippocampal overexpression restored neurogenesis as well as recognition memory in models of LPS-induced inflammation 23 . Other members of the TGF-β superfamily have a similar role. Indeed, TGF-β 2 facilitated synaptic plasticity in rat hippocampal neurons 14 , while Activin has been shown to be required for hippocampal L-LTP, consolidation of long-term memory 24 , and induction of moderate LTP upon weak theta-burst stimulation by acting on NMDA receptors currents and spine density 25 .
Another relevant finding of this study is the modification of CREB phosphorylation, which has been shown to play a central role in LTP and memory [4][5][6] . Our in vitro and in vivo studies showed that the inhibition of TGF-β 1 signaling induced by SB431542 provoked a decrease of hippocampal p-CREB levels that was rescued by concomitant treatment with TGF-β 1. Moreover, TGF-β 1 treatment stimulated CREB phosphorylation in hippocampal slices after a weak tetanus. These findings correlated with the modifications of synaptic plasticity and memory and are consistent with previous works demonstrating that TGF-β 1 increased nuclear levels of p-CREB in Aplysia 7 , whereas TGF-β 2-induced long-term changes in excitatory and miniature post-synaptic currents in the hippocampus 14 .
In this work we also investigate the involvement of p-SMAD2, a TGF-β 1 downstream pathway effector. Indeed, members of the TGF superfamily act through a receptor complex constituted by the serine/ threonine receptors ALK/TGF-β type I receptor and TGF-β type II receptor (Tβ RII), strongly expressed in the CNS and in particular in the hippocampus 26 . TGF-β 1 binding to Tβ RII induces the assembly of type I and type II receptors into a complex, with the subsequent transphosphorylation of type I receptor by the type II receptor kinase. The consequent activation of type I receptor leads to phosphorylation of selected SMAD proteins that, in turn, translocate into the nucleus to regulate the expression of different target genes involved in cell proliferation and neuronal survival 6 . Here we confirmed that SB431542, a validated tool to evaluate cellular mechanisms mediated by endogenous TGF-β 1 15 , selectively induced an inhibition of p-SMAD2 signaling, which was rescued by TGF-β 1 treatment, suggesting a similar trend between p-SMAD2 and p-CREB levels. This connection between the two pathways could be relevant considering the role of TGF-β 1 in neurodegenerative disorders characterized by synaptic plasticity and memory loss, such as Alzheimer's Disease (AD). Indeed, an impairment of TGF-β 1 has been demonstrated in the AD brain and serum 17 and a nucleotide polymorphism of the TGF-β 1 gene has been associated with an increased conversion of mild cognitive impairment in AD and with an increased risk to develop Late-Onset AD 27 . Accordingly, TGF-β 1 signaling has been found to interact at different levels with Amyloid-beta peptide (Aβ ), a key factor in the pathogenesis of AD. TGF-β 1 protected against Aβ -induced neurodegeneration 16 and, on the other hand, a suppression of TGF-β 1 pathway amplified Aβ neurotoxicity in rat hippocampus 16,17 . Thus, the neuroprotective features of TGF-β 1 combined with its physiological activity on hippocampal synaptic plasticity and memory suggest that it might represent a new therapeutical strategy against neurodegenerative diseases characterized by an impairment of TGF-β signaling, such as AD.

Materials and Methods
Animals. All experiments have been performed on 5-months old C57BL/6 mice obtained from a breeding colony kept at the University of Catania. Mice were maintained with a controlled temperature (21 °C ± 1 °C) and humidity (50%) on a 12 h light/dark cycle, with ad libitum food and water. All animal experimentation was conducted in accordance with the guidelines laid down by the European Community Council (2010/63/EU). The experimental protocols have been approved by the University Institutional Animal Care and Use Committee (Project #204/2013, Palmeri/Puzzo).
Drugs. SB431542 13 (Tocris, Bristol, UK) was reconstituted in DMSO (20 mM), dissolved in PBS containing 2% DMSO and applied to hippocampal slices (20 μ M in ACSF) or administered via intraperitoneal injection at a concentration of 4.2 mg/kg according to previous studies 28 . Human TGF-β 1 (R&D system, Minneapolis, MN, USA) was reconstituted at 20 μ g/mL in sterile 4 mM HCl containing 1 mg/ mL bovine serum albumin, dissolved in ACSF for in vitro studies (1, 5 or 10 ng/ml) and in PBS for in vivo studies (1 μ g in 300 μ l PBS) 7,16 . We injected 38.4 ng of SB431542 and 3.3 ng of TGF-β 1 over 1 μ l for intrahippocampal treatment.
Electrophysiology. Electrophysiological recordings were performed on 400 μ m transverse slices, in CA1 hippocampal area, as previously described 29,30 . Slices were maintained in a recording chamber at 29 °C and perfused with artificial cerebrospinal fluid (ACSF composition in mM: 124.0 NaCl, 4.4 KCl, 1.0 Na2HPO4, 25.0 NaHCO3, 2.0 CaCl2, 2.0 MgCl2, 10.0 glucose) continuously bubbled with 95% O2 and 5% CO2. Field excitatory postsynaptic potentials (fEPSPs) were recorded by stimulating Schaffer collateral fibers with a bipolar tungsten electrode and recording in CA1 stratum radiatum with a glass electrode filled with ACSF. After evaluation of basal synaptic transmission, a 15 min baseline was recorded every minute at an intensity evoking approximately 35% of the maximum response. LTP was induced by 4 pulses at 100 Hz, with the bursts repeated at 5 Hz and 1 (weak tetanus) or 3 tetani (strong tetanus) of 10-burst trains. Responses were recorded for 2 hrs after tetanization and measured as fEPSP slope, expressed as percentage of baseline.
Infusion technique. Cannulas were implanted as previously described 31 . After anesthesia with Tiletamine + Zolazepam (60 mg/kg) and Medetomidine (40 μ g/kg), mice were implanted with a 26-gauge guide cannula into the dorsal part of the hippocampi (coordinates: posterior = 2.46 mm, lateral = 1.50 mm to a depth of 1.30 mm). The cannulas were fixed to the skull with acrylic dental cement (RelyX™ Unicem). After 6-8 days of recovery, drugs were injected bilaterally in a final volume of 1 μ l over 1 min through infusion cannulas that were connected to a microsyringe by a polyethylene tube. Mice were injected right before the training session (T1) of the ORT. During infusion, animals were handled gently to minimize stress. After infusion, the needle was left in place for another minute to allow diffusion. After behavioral testing, animals were killed and their hippocampi were removed, frozen, and then used for western blot analysis. In some animals a solution of 4% methylene blue was infused for histological localization of infusion cannulas.
Novel object recognition test. Novel object recognition test was performed and analyzed as previously described 21,30 . After three days of habituation, at day 4th mice underwent the training session (T1). After treatment with vehicle, TGF-β 1, SB431542 or TGF-β 1 + SB431542 (via i.p. thirty min before T1, or via intrahippocampal injections right before T1) mice were placed in the arena for 10 min and allowed to explore two identical objects. Twenty-four hours after T1 mice underwent the second trial (T2) to test memory retention. Mice were presented with two different objects, respectively a "familiar" (i.e. the one used for T1) and a "novel" object. Animal exploration -defined as the mouse pointing its nose toward the object from a distance not > 2 cm -was evaluated in T2 to analyze: (i) exploration of the novel and of the familiar object; (ii) discrimination (D) index calculated as "exploration of novel object minus exploration of familiar object/total exploration time"; (iii) latency to first approach to novel object; (iv) total exploration time.
Western blot. Western blot analyses was performed as previously described 30 on hippocampal slices previously treated as for electrophysiological experiment and stored in liquid nitrogen at 1 min after treatment or in hippocampi of animals previously treated with intrahippocampal injections. Tissues were then homogenized in RIPA buffer in the presence of a cocktail of protease inhibitors (Sigma P2714), serine/threonine phosphatase inhibitors (Sigma P0044) and tyrosine protein phosphatases inhibitors (Sigma P5726) and sonicated. Protein concentrations were determined by Bradford's method using bovine serum albumin as a standard. After blocking, membranes were incubated with primary antibodies overnight at 4 °C: rabbit anti-phospho-CREB (Ser133) (Millipore, Billerica, MA, USA; 1:1000), rabbit anti-phospho-SMAD2 (ser465/467) (Cell signaling Technology, Danvers, MA; 1:1000) and mouse anti-β -tubulin (Sigma-Aldrich; 1:500). Proteins were visualized using the enhancing chemiluminescence detection system SuperSignal (Pierce, USA). Chemiluminescence was detected by the Uvitec System (Cambridge, UK) and quantified by densitometric analysis in three different blots per experiment.

Statistical Analysis.
All experiments were blind with respect to treatment. Data were expressed as mean ± standard error mean (SEM). Statistical analyses was performed by using Systat software (Chicago, IL, USA). For LTP we used ANOVA with repeated measures comparing 120 minutes of recording after tetanus; in some experiments we used one-way ANOVA with Bonferroni post-hoc to compare the 120° min of recording among groups. For ORT we used one-way ANOVA with Bonferroni post-hoc for comparisons among groups, paired t-test to compare exploration of the novel vs. the familiar object in the same mouse, one sample t-test to compare D with zero. For western blot we used one-way ANOVA with Bonferroni post-hoc and two-way ANOVA for tetanus and treatment. The level of significance was set at p < 0.05.