BDNF val66met Polymorphism Impairs Hippocampal Long-Term Depression by Down-Regulation of 5-HT3 Receptors

Brain-derived neurotrophic factor (BDNF) is a key regulator of neuronal plasticity and cognitive functions. BDNF val66met polymorphism, a human single-nucleotide polymorphism (SNP) in the pro-domain of BDNF gene, is associated with deficits in activity-dependent BDNF secretion and hippocampus-dependent memory. However, the underlying mechanism remains unclear. Here we show that in the BDNFMet/Met mouse line mimicking the human SNP, BDNF expression in the hippocampus was decreased. There was a reduction in the total number of cells in hippocampal CA1 region, while hippocampal expression of mRNAs for NR2a, 2b, GluR1, 2 and GABAARβ3 subunits were up-regulated. Although basal glutamatergic neurotransmission was unaltered, hippocampal long-term depression (LTD) induced by low-frequency stimulation was impaired, which was partially rescued by exogenous application of BDNF. Interestingly, 5-HT3a receptors were down-regulated in the hippocampus of BDNFMet/Met mice, whereas 5-HT2c receptors were up-regulated. Moreover, impaired LTD in BDNFMet/Met mice was reversed by 5-HT3aR agonist. Thus, these observations indicate that BDNF val66met polymorphism changes hippocampal synaptic plasticity via down-regulation of 5-HT3a receptors, which may underlie cognition dysfunction of Met allele carriers.


INTRODUCTION
Brain-derived neurotrophic factor (BDNF), a member of nerve growth factor family, plays an important role in the development and function of the central nervous system (CNS; Huang and Reichardt, 2001). In addition, it has been widely accepted that the main function of BDNF in the brain is to regulate synaptic plasticity. Exogenous application of BDNF facilitates theta burst stimulation (TBS) induced LTP in developing hippocampal slices (Figurov et al., 1996). Moreover, hippocampal LTP is impaired in heterozygous or homozygous BDNF-KO mice, and this impairment could be rescued by incubation with recombinant BDNF (Patterson et al., 1996). In the mammalian brain, BDNF is synthesized as the precursor of BDNF (proBDNF), which is then cleaved into mature BDNF by the tPA/plasmin system (Pang et al., 2004). A common single nucleotide polymorphism (SNP) rs6265 at nucleotide 196 (G/A) at codon 66 in the pro-domain of human BDNF gene converts the amino acid valine (Val) to methionine (Met). This amino acid substitution affects dendritic trafficking of pro-BDNF and alters the regulated secretion of BDNF (Egan et al., 2003;Chen et al., 2006). The allele frequency varies from 0.55% to 43.6% in Africa, Europe and Asia (Petryshen et al., 2010). Human subjects with this SNP exhibit deficits in short-term episodic memory, suggesting the importance of activity-dependent BDNF secretion. Growing evidence has implicated that Met allele carriers might have a higher susceptibility to Alzheimer's disease and multiple psychiatric disorders (Bekinschtein et al., 2014;Lim et al., 2014;Notaras et al., 2015). In animal studies, it has been shown that the val66Met polymorphism decreases neurotransmission and impairs NMDAR-dependent synaptic plasticity in the hippocampus (Ninan et al., 2010). However, in dorsal striatum, the polymorphism enhances glutamatergic transmission but impairs synaptic plasticity (Jing et al., 2017). Also it has been revealed that in the IL-mPFC pyramidal neurons of BDNF Met/Met mice, spiketiming-dependent plasticity (STDP) is absent, and NMDA receptor-mediated synaptic transmission is impaired (Pattwell et al., 2012).
Accumulating evidence suggests that serotonin  and its receptors are involved in many physiological or pathophysiological processes of cognitive function (Strüder and Weicker, 2001), as well as regulating synaptic plasticity. It has been revealed that 5-HT6 receptors are down-regulated in the cortical areas of patients with Alzheimer's disease, correlating with the progression of disease and memory impairment (Garcia-Alloza et al., 2004). Iontophoresis of 5-HT2a receptor antagonists attenuates the memory fields of putative pyramidal cells in monkeys, whereas activation of 5-HT2a receptors accentuates the spatial tuning of these neurons (Williams et al., 2002). Administration of 5-HT3 receptors agonist m-CPBG at low and high doses impairs short-term or long-term memory (Meneses, 2007). Moreover, application of 5-HT inhibits hippocampal long-term depression (LTD) and the effect could be blocked by 5-HT1a receptor antagonist (Normann and Clark, 2005). In vivo study, it has been found that 5-HT4 receptor agonist blocks learning-induced depotentiation of LTP; whereas 5-HT4 antagonist enhances the intermediate phase of LTD and converts short-term depression into persistent LTD (Kemp and Manahan-Vaughan, 2005). It is notable that high level of BDNF and 5-HT receptors are both expressed in the hippocampus (Wetmore et al., 1994;Barnes and Sharp, 1999). Growing evidence suggests a complex interaction between BDNF and 5-HT system. For example, expression of 5-HT receptors is altered in BDNF +/− mice, correlating with the enhanced aggressiveness and hyperphagia (Lyons et al., 1999). Application of exogenous BDNF results in down-regulation of 5-HT2a receptors (Trajkovska et al., 2009), while BDNF expression is increased by a deficiency of 5-HT2c receptors, (Hill et al., 2011). There is good evidence that antidepressants including selective serotonin reuptake inhibitors (SSRIs) enhance the expression of BDNF mRNA (Castrén, 2004). However, the precise impact of val66met polymorphism on hippocampal synaptic plasticity and its underlying mechanism about the cross-talk with 5-HT system have not been well understood. In this study, using BDNF Met knock-in mice (BDNF Met/Met ), we examined the effect of BDNF val66Met mutation on synaptic plasticity in CA1 region in the hippocampus and its mechanisms. We found that BDNF val66met polymorphism attenuates hippocampal LTD via the down-regulation of 5-HT3a receptors.

Animals
BDNF val66met polymorphism mouse strain was used in this study. All animals were on C57BL/6 background and male BDNF Met/Met and BDNF Val/Val mice derived from heterozygous BDNF Val/Met parents were used. The details of the targeting construct and homologous recombination were described previously (Chen et al., 2006). All animal experiments were performed with protocols approved by the Animal Ethics Committee of Tongji University. The genotypes of animals were analyzed by preparing tail DNAs and PCR assay.
All electrophysiological recordings were performed with an Axopatch-200B amplifier (Molecular Devices, Sunnyvale, CA, USA) at the sampling rate of 10 kHz and filtered at 5 kHz. Field excitatory postsynaptic potentials (EPSPs) were recorded using 1.5-3.5 MΩ glass pipettes filled with ACSF and placed in the stratum radiatum of the CA1 region. Data were acquired and subsequently analyzed using ClampFit 9.0 software (Molecular Devices, Sunnyvale, CA, USA). A bipolar stimulating electrode was positioned at the terminals of the Schaffer collaterals (SC). Input/output relation of EPSP was assessed by electrical stimulation with intensities ranging from 20 pA to 140 pA. Stimulus were delivered every 30 s and repeated for five times at each intensity. Paired-pulse responses were evoked at inter-stimulus intervals of 60, 80, 100, 200, 300, 400 and 500 ms using a stimulation intensity of 0.5 mA. The paired-pulse ratio is defined as the ratio of second population spike amplitude to the first population spike amplitude. For LTP and LTD experiments, electrical stimulation was delivered every 30 s in recordings, consisting of lowintensity, square-wave pulses (0.1 ms). Theta-burst LTP was induced by three bursts (each burst having 4 pulses at 100 Hz) delivered at 5 Hz TBS. NMDAR-dependent LTD was induced by 900 pulses at 1-Hz stimulation (low frequency stimulation, LFS). Baseline responses were recorded for 15 min prior to stimulation. Responses were subsequently recorded for an additional 60 min to monitor changes in synaptic transmission. The synaptic strength was determined by the slope from 10% to 90% of the rising phase of the field EPSP (fEPSP). The magnitude of LTP or LTD was quantified as the normalized average slope of the fEPSP taken from the last 15 min of recording.

Quantigene Assay
Hippocampus of 3 or 6-week-old mice were mechanically homogenized with the homogenization buffer of Quantigene 2.0 assay kit (Affymetrix, Santa Clara, CA, USA), which contains 1% protease Kinase K, followed by incubation at 65 • C for 3 h with 800 rpm shaking in the Eppendorf incubator. Supernatants were collected after centrifugation at 21,200 g for 10 min. Quantigene probe sets for BDNF and GAPDH were customized from Panomics/Affymetrix.
The detection was performed following manufacture's instruction of kit. Generally, after recovering the plates and reagents at RT, 60 µl of working probe sets were added to the plate and 40 µl of prepared tissue lysates were dispended into each well. Following incubation at 55 • C overnight and a three-times wash step, the plates were added with 100 µl pre-amplifier working regents and incubated for 1 h at 55 • C. Hybridized the 2.0 amplifier by adding 100 µl amplifier working reagent into each well after the wash step. Sealed the plates and incubated at 55 • C for 1 h. After wash, the plates were further incubated with 100 µl 2.0 probe working reagent at 50 • C for 1 h followed by complete wash steps. Signals were detected by Luminometer within 5-10 min after adding of 100 µl substrate into the wells. Signal of BDNF was normalized to signal of GAPDH in each sample tested.
ELISA Hippocampal tissues of 3 or 6-week-old mice were homogenized mechanically with lysis buffer at the ratio of 1:10 (w/v), followed by sonication for 10 s. Then placed on ice for 30 min. Supernatants were collected after centrifugation at 21,200 g for 15 min. Five microliter of supernatants from each sample were added for BDNF Elisa with another 1 µl used for protein concentration measurement (BCA assay).
BDNF ELISA was performed following manufacture's instruction provided in kits (Human BDNF Quantikine SixPak, R&D, Minneapolis, MN, USA). Generally, after recovering the plates and reagents at RT, 100 µl standard assay diluents were added to each well. Fifty microliter BDNF standards prepared in RD5K or 5 µl sample plus 45 µl RD5K was added. After 2 h incubation at 600 rpm shaking, washed the plate for three times. One-hundred microliter BDNF conjugates were added to each well followed by 1 h incubation at 600 rpm shaking. Plate was washed for five times. Colors were developed by adding 200 µl mixture of reagent A&B and stopped by 50 µl stop solution. Mixed to make a homogenous yellow color. Absorbance was detected by Biotek plate reader at 450 nM with background set at 540 nM. The BDNF level in each sample were calculated against the standard curve and normalized to respective total protein concentration.

Realtime-PCR
Hippocampal tissues were collected from 6-week-old BDNF Met/Met and BDNF Val/Val mice. mRNA was extracted using Trizol Reagent (Invitrogen, Carlsbad, CA, USA) according to manufacturer's instruction. cDNA was synthesis with ReverTra Ace qPCR RT Kit (Toyobo, Japan). The expression of target genes were further analyzed with SYBR Green Realtime PCR Master Mix (Toyobo, Japan) using StepOnePlus TM Real-Time PCR System (Applied Biosystems, Foster City, CA, USA). The following primers were used in qRT-PCR:

Down Regulation of BDNF and Decrease in Total Cell Number in BDNF Met/Met Mice
In the CNS including hippocampus, BDNF is secreted both constitutively and in a regulated manner in response to neuronal activity (Farhadi et al., 2000). It has been previously reported that val66met polymorphism selectively attenuated activity-dependent BDNF secretion, suggesting a possible underlying mechanism for impaired hippocampal function (Egan et al., 2003). To investigate the impact of BDNF val66Met polymorphism on the expression of BDNF in the hippocampus, we compared the levels of BDNF mRNA and protein between the WT and mutant mice. Using Quantigene assay, we found that expression of BDNF mRNA was down-regulated in the hippocampus of BDNF Met/Met mice (3 or 6-week-old, Figure 1A; P < 0.0001 for Quantigene assay; unpaired t-test). Moreover, ELISA analysis showed that BDNF protein was decreased in the hippocampus of BDNF Met/Met mice as well ( Figure 1B, P < 0.0001, unpaired t-test).
As BDNF plays a pivotal role in the development and neurogenesis of nervous system, we then examined whether decreased level of BDNF alters the number of neurons during development. In the CA1 region of hippocampus, the total cell count decreased by 17.58% in BDNF Met/Met mice, suggesting

Change of Glutamatergic and GABAergic Receptor Subunits in BDNF Met/Met Mice
BDNF has been implicated in regulating expression of glutamatergic and GABAergic receptors. Treatment of hippocampal neurons with BDNF regulates NR1, NR2a and NR2b NMDA receptor subunits (Caldeira et al., 2007) as well as GABA A receptors (Brünig et al., 2001). To further elucidate the effect of BDNF val66Met polymorphism on NMDA, AMPA and GABA receptor subunits, we examined mRNA expression of NMDA receptor subunits (NR1, NR2a, NR2b), AMPA receptor subunits (GluR1, GluR2), GABA A receptor subunits (β2 and β3), glutamic acid decarboxylases (GAD65 and GAD67) and GFAP (as a marker of glia cells) in hippocampus. As revealed by qPCR, we found that NR2a and NR2b subunits for NMDA receptors, GluR1 and GluR2 for AMPA receptors, as well as GABA A receptor β3 subunits were up-regulated in BDNF Met/Met mice (Figure 2), suggesting structure changes of these receptors, which may underlie synaptic transmission and plasticity remodeling.

LFS Induced LTD Was Impaired in BDNF Met/Met Mice
To investigate the impact of BDNF val66met polymorphism on synaptic transmission, we examined the basal synaptic function of WT and BDNF Met/Met mice. The input/output relation of fEPSP slopes and paired-pulsed ratio (for the intervals of 60, 80, 100, 200, 300, 400, 500 ms) in WT and BDNF Met/Met mice were not different (Figures 3A,B), suggesting the basal transmission is not altered in mice with BDNF val66met polymorphism.
We also compared the stimulation induced LTP and LTD in slices from WT and BDNF Met/Met mice. CA3-CA1 LTP was induced by SC TBS (TBS, three bursts, each having four pulses at 100 Hz, delivered at 5 Hz), whereas LTD was induced by low-frequency stimulation LFS (1 Hz, 900 pulses). We found that LTD was impaired in slices from BDNF Met/Met mice (WT: 0.63 ± 0.053 of baseline, N = 7 from four mice; BDNF Met/Met : 0.96 ± 0.031 of baseline, N = 6 from four mice; P < 0.001, unpaired t-test; Figures 3E,F). However, the change of early-phase TBS-LTP was not detected in slices from BDNF Met/Met mice during 1 h of recording (WT: 1.38 ± 0.066 of baseline, N = 7 from five mice; BDNF Met/Met : 1.39 ± 0.104 of baseline N = 5 from four mice; P > 0.05, unpaired t test; Figures 3C,D). These results suggest that the BDNF val66Met polymorphism attenuates LTD, but not early-phase LTP at the hippocampal CA3-CA1 synapses.

Exogenous BDNF Partially Rescued LTD in BDNF Met/Met Mice
Brain slices were then incubated with exogenous BDNF. BDNF (40 ng/ml, 30 min; 100 ng/ml, 60 min) was applied before LTD induction. Interestingly, the LTD deficit in slices of BDNF Met/Met mice was partially rescued. The magnitude of LTD in BDNF Met/Met slices was significantly increased after BDNF perfusion (BDNF 40 ng/ml, 30 min, WT: 0.63 ± 0.053 of baseline, N = 7 from four mice; BDNF Met/Met : 0.96 ± 0.031 of baseline, N = 6 from four mice; BDNF Met/Met +BDNF: 0.77 ± 0.028 of baseline, N = 7 from four mice, P < 0.05, one-way ANOVA; Figure 4), suggesting that impaired LTD may be partially due to BDNF deficiency.

Change of 5-HT Receptors Expression in BDNF Met/Met Mice
The fact that BDNF could only partially rescue the LTP impairment indicates that secondary mechanisms other than BDNF deficiency maybe involved. Given the complex interaction between BDNF and 5-HT system (Lyons et al., 1999;Deltheil et al., 2008;Trajkovska et al., 2009), to investigate the potential impact of BDNF val66met polymorphism on the serotonin system, the levels of serotonin receptor mRNA in the hippocampus were analyzed by qPCR. We examined the expression of 5-HT1aR, 5-HT1bR, 5-HT2cR, 5-HT3aR, 5-HT4R, 5-HT6R and 5-HT7R, which have close relevance with memory and cognition (Seyedabadi et al., 2014). The mRNA expression of 5-HT2cR was up-regulated, while that of 5-HT3aR and 5-HT4R was down-regulated in the hippocampus of BDNF Met/Met mice. Furthermore, the changes of 5-HT2cR and 5-HT3aR proteins were confirmed by western blot analysis. However, we failed to detect any diminishment in 5-HT4R protein ( Figure 5). Thus, we further focused on the role of 5-HT2cR and 5-HT3aR in LTD impairment in BDNF Met/Met mice.

DISCUSSION
In the present study, using BDNF Met/Met mice, we investigated the effect of BDNF val66met polymorphism on synaptic plasticity and the underlying mechanism. We demonstrated that BDNF expression and total cell number were down-regulated in the hippocampus in BDNF Met/Met mice. Expression of mRNAs for NR2a, 2b, GluR1, 2 and GABA A Rβ3 subunits were up-regulated by BDNF val66met polymorphism. Moreover, hippocampal LTD was impaired, which was partially rescued by exogenous BDNF and totally reversed by 5-HT3aR agonist.
BDNF val66met polymorphism is a common SNP in the BDNF gene in human. Met-BDNF pro-peptide exhibits an altered trafficking and secretion of BDNF , leading to a decrease in activity-dependent BDNF release (Egan et al., 2003). A previous study showed that in the whole brain lysates of 21-day-old mice, the level of BDNF was similar in WT and BDNF Met/Met mice; yet the number of BDNF-containing varicosities in BDNF-expressing cells was reduced in the hippocampus of BDNF Met/Met mice (Chen et al., 2006). In this study, using multiple assays, we found that the expression of BDNF mRNA and protein was attenuated specifically in the hippocampus of 6-7 week-old BDNF Met/Met mice. Thus, BDNF val66met polymorphism may selectively reduce BDNF levels in certain brain regions such as hippocampus. Given the crucial role of BDNF in neuronal survival during development, it is not surprising that the cell number in the hippocampus was attenuated as the consequence of BDNF deficiency.
It has also been reported that both early-phase and late-phase hippocampal LTP are impaired in 1-month-old BDNF Met/Met mice, whereas in 4-month old mice, only late-phase LTP is reduced (Ninan et al., 2010), which indicates that early-phase LTP in adult animal is less sensitive to BDNF deficiency due to val66met polymorphism. It may be the possible reason for the unaltered early-phase CA3-CA1 LTP we found in BDNF Met/Met mice (6-8 week-old) in this study. Moreover, earlier study revealed that LFS induced NMDAR-LTD, but not DHPG induced NMDAR-independent LTD was impaired in 1-month-old BDNF Met/Met mice (Ninan et al., 2010). Here we found that LFS induced LTD was attenuated by BDNF val66met polymorphism consistently. Furthermore, we found that acute treatment with exogenous BDNF rescued the LTD deficit but incompletely, implying the possibility of additional mechanisms involved. BDNF is thought to have cooperative effect with 5-HT system. A deficiency of BDNF in BDNF +/− mice results in the alteration of 5-HT receptor expression in lateral frontal cortex, hippocampus and hypothalamus (Lyons et al., 1999). Acute administration of BDNF in the hippocampus potentiates the uptake of 5-HT by activating TrkB (Deltheil et al., 2008). In this study, we found that hippocampal expression of 5-HT2cR was up-regulated, while 5-HT3aR was down-regulated in BDNF Met/Met mice. Although our result of 5-HT2cR overexpression is different from the previous report that BDNF deficiency down-regulated hippocampal 5-HT2cR in 6-9 month-old mice (Lyons et al., 1999), which may be due to the different age of animals used in experiments, these results further confirmed the cross-talk between BDNF and 5-HT receptors.
In serotonin receptor family, 5-HT3 receptors (5-HT3Rs) are the only ionotropic receptor (Chameau and van Hooft, 2006;Faerber et al., 2007), which are specifically expressed in cholecystokinin (CCK)-containing inhibitory interneurons in the hippocampus and mediate fast synaptic transmission in the raphe-hippocampal interneuron synapses. The activity of 5-HT3Rs located on postsynaptic somatodendritic regions or presynaptic terminals modulates firing or neurotransmitter release of interneurons, which consequently regulates pyramidal synaptic plasticity via GABAergic projection from CCK interneurons and thus modulates memory (Rondé and Nichols, 1998;Sudweeks et al., 2002). Injection of the 5-HT3R antagonist granisetron into hippocampus impairs spatial learning (Naghdi and Harooni, 2005). Granisetron could also reduce new learning associated with extinction of fearful memories (Park and Williams, 2012). Our previous study found that 5-HT3aR disruption inhibited AMPA receptors internalization and impaired LTD (Yu et al., 2014). However, in contrast to 5-HT3Rs, 5-HT2 receptors are G-protein coupled receptors, which are also highly expressed in the hippocampus (Hoffman and Mezey, 1989) and modulate cognitive function. It has been revealed that 5-HTr2c antagonist facilitates long term memory although does not affect short term memory (Meneses, 2007). Moreover, spatial learning but not hippocampal LTP was impaired in 5-HTr2c mutant mice (Tecott et al., 1998). In this study, we found that acute application of exogenous 5-HT3aR agonist completely rescued the impaired LTD in BDNF Met/Met mice, but 5-HT2cR antagonist did not reverse deficient LTD. These results suggest a role of the 5-HT3aRdependent signaling in the LTD deficit in BDNF val66met polymorphism, whereas 5-HT2cR may not involve in this modulation. As ligand-gated ion channels, 5-HT3aRs have more rapid effect compared with other G-protein coupled receptors in serotonin receptor family, which may be more conducive to mediate or regulate fast synaptic transmission and plasticity. The down-regulation of 5-HT3aR in CCK interneurons may change GABA release at inhibitory synapses from these interneurons to pyramidal cells, altering their membrane depolarization. The disruption of depolarization-induced calcium influx and Ca 2+ -dependent phosphatase may contribute to impairments in AMPAR internalization and LTD. However, besides 5-HT3aRs, the impact of pro-BDNF deficiency on LTD could not be ruled out. In conclusion, this study revealed that BDNF val66met polymorphism impaired hippocampal LTD by reducing BDNF expression and down-regulation of 5-HT3aRs, and provided new insights into the modification of 5-HT receptors by BDNF val66met polymorphism and its contribution to the cognitive dysfunction of Met allele carriers.

AUTHOR CONTRIBUTIONS
YH and RH conceived and designed the study; RH, YQ, D-NH, C-YL, Y-YJ and C-YZ performed experiments; RH, YQ and C-YZ conducted data analysis; W-HY and BL assisted in experiments, data analysis; RH, YQ and YH wrote the manuscript; BL and W-HY contributed to the revisions.