Research On Defect Of Synaptic Plasticity In PFC Of A Mouse Model Of Fragile X Syndrome

【Objectives】This research aims to find the mechanism on defect of synaptic plasticity inPFC of a mouse model of Fragile X syndrome and a possible way to rescuesynaptic plasticity in FMR1KO mice. Also, try to find a new way in medicaltreatment for FXS.【Methods】1Whole-cell recordingsCoronal prefrontal brain slices (300μm) from FMR1WT and KO mice,containing the anterior cingulate cortex (ACC), were prepared using standardmethods. Slices were transferred to submerged recovery chamber withoxygenated (95%O2and5%CO2) artificial cerebrospinal fluid (ACSF) atroom temperature for at least1h.Experiments were performed in a recordingchamber on the stage of an Axioskop2FS microscope with infrared DIC optics for the visualization of whole-cell patch clamp recording. In the present study,we recorded excitatory postsynaptic currents (EPSCs) from neurons in layer IIto III induced by the stimulation of layer V. AMPA receptor (AMPAR)-mediatedEPSCs were induced by repetitive stimulations at0.02Hz, and neurons werevoltage clamped at-70mV. After obtaining stable EPSCs for at least10min,LTP was induced by80pulses at2Hz, and then paired with postsynapticdepolarization at+30mV. Picrotoxin (100μM) was always present to blockGABAAreceptor-mediated inhibitory synaptic currents. To record the mEPSCs,0.5μM Tetrodotoxin (TTX) was added into the bath solution. The accessresistance was15to30M, which was monitored throughout the experiment.Data were discarded if access resistance changed by over15%during anexperiment.2Western blotPrimary cultures of prefrontal cortical neurons were prepared fromembryonic18days old (E18) mice as described previously. Cultures were usedfor experiments between10and14days in vitro. SKF81297(5μM), DL-AP3(10μM), and forskolin (5μM) were added in the culture medium for24h beforethe cells harvested. Surface protein samples were detected through abiotinylation assay, followed by Western blot analysis with the antibodies aspreviously described. Western blot analysis was performed as describedpreviously. Equal amounts of protein (50μg) from the cultures were separatedand electrotransferred onto NC membranes, which were probed with antibodiesfor mGluR5, GluR1, NR2A, NR2B, p-NR2B Tyr1472, and with β-actin orcadherin as loading control. The membranes were incubated with horseradishperoxidase conjugated secondary antibodies and bands were visualized using anECL system (Perkin Elmer).3Behavior testHorizontal locomotor activity was conducted with an open-field test system (Jiangliang, Shanghai, China). Mice were placed in the center of the box,allowed to freely explore for10min, and videotaped using a camera fixed abovethe floor and then analyzed with a video-tracking system.The elevated plus maze was tested after Horizontal locomotor activity. Micewere allowed to habituate to the testing room for2days before the test, andpretreated with gentle handling two times per day to eliminate their nervousness.For each test, individual animals were placed in the center square, facing anopen arm, and allowed to move freely for5min. Mice were videotaped using acamera fixed above the maze and analyzed with a video-tracking system. Thenumber of entries and time spent in each arm were recorded.In the Morris Water Maze test, mice were treated either with SKF81297alone or with SKF81297plus DL-AP3for5weeks. Mice were kept dry betweentrials in a plastic holding cage filled with paper towels. Starting locations werechanged every trial. The time required to escape onto the hidden platform wasrecorded as escape latency.Forty-five minutes after the subcutaneous injection of saline, DL-AP3,and/or SKF81297, each subject was placed in the center of the open field, andits activity was measured for30min.【Results】1Grp1mGluR antagonist rescues LTP facilitation by D1activation inFMR1KO miceThe PFC, including its cingulate region, plays an important role in learningand memory, drug addiction, and pain. First, we performed whole-cellpatch-clamp recordings in visually identified pyramidal neurons in layers II–IIIof cingulate region of PFC slices. LTP was induced by pairing presynapticstimulation with postsynaptic depolarization. The pairing training produced asignificant, long-lasting potentiation of synaptic responses in WT mice, but notin KO mice. Since Grp1mGluR-LTD is exaggerated in the FMR1KO mouse,we next examined the effects of mGluR1antagonist on LTP induction in the PFC. It has been reported that high dose of mGluR1antagonist,DL-2-amino-3-phosphonopropionic acid (DL-AP3) or(+)-alpha-methyl-4-carboxyphenylglycine (MCPG), reduced homosynaptic LTPin the hippocampus. In the present study, the slices were incubated withmGluR1antagonist DL-AP3at low dose of10μM at least for30min before theLTP induction was performed. At the concentration of10μM, DL-AP3did notalter the amplitude of LTP as compared to the pairing training only in the WTand KO mice. We next paired D1agonist application with an LTP inductionprotocol. Bath application of SKF81297for10min paired with the LTPinduction protocol significantly enhanced the amplitude of LTP, suggesting thatD1receptor activation can facilitate LTP induction in the WT mice. However, inFMR1KO mice SKF81297pairing training could not induce LTP,demonstrating that dopaminergic facilitation of LTP is impaired in FMR1KOmice.Next, D1agonist and mGluR1antagonist were applied simultaneously todetect their synergistic effects on the LTP induction. Bath application ofSKF81297(5μM) and DL-AP3for10min induced a significant LTP. Furthercomparison analysis found that the amplitude of the LTP was similar to the LTPinduced by SKF81297alone in the WT mice. However, simultaneousapplication of DL-AP3markedly rescued the LTP induction by SKF81297inthe FMR1KO mice. Additional D1receptor antagonist SCH23390for10minprior to, and during electrical stimulation totally blocked the LTP which wasrescued by synergistic application of SKF81297and DL-AP3in the KO mice.SKF81297or DL-AP330min had no effect on basal synaptic responses.Theses data indicate that lower dose of DL-AP3dose not alter the LTPinduction in the FMR1WT and KO mice; however, it rescues the LTPfacilitation of D1receptor activation in the FMR1KO mice. 2Adenylyl cyclase agonist rescues LTP facilitation by D1activation inFMR1KO miceIn our previous study, we found that D1receptor signaling is impaired, i.e.,the increase in cAMP caused by SKF81297is attenuated, accompanied by D1receptor hyperphosphorylation at serine sites in the PFC of FMR1KO mice. Todetect whether or not the increase of cAMP could mimic the function of D1receptor, the adenylyl cyclase activator, forskolin, was used. Many studies haveused forskolin to stimulate AMPAR trafficking and induce chemical LTP. In thepresent study, unlike in the hippocampus, LTP was not induced during theapplication of forskolin, which was not paired with the electrical stimulation inthe ACC, i.e, no chemical LTP was induced in the ACC. The amplitude of LTP issimilar to that of the control when the forskolin was paired with the electricalstimulation in the FMR1WT mice. Combining forskolin with DL-AP3did notresult in an increase in the amplitude of LTP compared with the case, in whichthe forskolin alone was used in the FMR1WT mice. Forskolin paired with theelectrical stimulation was unable to induce LTP in the FMR1WT mice.However, combining forskolin with DL-AP3partially rescued LTP in the FMR1KO mice, even if the amplitude was only118%±2.8%of the base line.To determine whether or not dopaminergic modulation of LTP requiredNMDAR activation, we applied a selective NMDAR antagonist, AP5, and foundthat LTP facilitation by SKF81297was completely blocked. Similarly, LTP wasabolished by BAPTA in the pipette solution, indicating that dopaminergicmodulation of LTP depended on the activation of NMDARs and elevatedpostsynaptic Ca2+concentrations.3Grp1mGluRs modulate plasticity by a postsynaptic mechanismIn the different regions of the hippocampus, both presynaptic andpostsynaptic mechanisms have been proposed to contribute to the expression ofLTP. To determine whether presynaptic and/or postsynaptic mechanisms areinvolved in Grp1mGluRs modulation of LTP, we measured paired-pulse facilitation (PPF) in the ACC. PPF is a phenomenon by which a second synapticstimulation of equal magnitude evokes a larger synaptic response than the first,and has been used as a tool to implicate presynaptic probability of transmitterrelease. PPF induced at five different intervals did not differ in WT and FMR1KO mice in the presence of DL-AP3or/and SKF81297in the ACC. The datasuggest that presynaptic mechanisms do not seem to be involved in the Grp1mGluRs modulation of LTP.4No changing of basal glutamatergic synaptic transmission by Grp1mGluRs inhibitionTo determine whether Grp1mGluRs inhibition affects basal glutamatergicsynaptic transmission in the ACC, we measured the miniature AMPAreceptor-mediated EPSCs (mEPSCs). As shown in Fig.4, no significantalteration was detected in the AMPA receptor-mediated mEPSCs frequency andamplitude among the control, DL-AP3or/and SKF81297treated slices fromFMR1WT and KO mice. Results indicate that Grp1mGluRs inhibition andactivation of D1receptor have no effect on the basal excitatory synaptictransmission in the ACC.5Expression of GluR1To determine the synergistic effect on the AMPA receptor expression andsurface insertion, we treated cultured PFC neurons with the D1receptor agonistsSKF81297and Grp1mGluRs antagonist DL-AP3. SKF81297alone orcombining with DL-AP3increased the total expression levels of GluR1in thecultures from FMR1WT mice. However, neither SKF81297nor DL-AP3affected the total expression levels of GluR1in cultures from FMR1KO mice.To further investigate the signaling involved in the synaptic plasticity, we testedthe effects of DL-AP3or/and SKF81297on AMPA receptor GluR1subunitssurface trafficking in cultured PFC neurons. Surface expression of GluR1wasincreased after treatment with SKF81297alone or combining with DL-AP3inthe cultures from FMR1WT mice. Contrast to the unchanged total levels of GluR1, surface expression of GluR1was increased after simultaneous treatmentof SKF81297and DL-AP3in the cultures from FMR1KO mice. This findingdemonstrates that inhibition of Grp1mGluRs can rescue the surface expressionof AMPA GluR1receptors by D1receptor in FMR1KO PFC neurons.Furthermore, the synergistic effects of DL-AP3and forskolin on AMPAreceptor GluR1subunits surface trafficking were detected in cultured PFCneurons. Forskolin alone or combined with DL-AP3increased the totalexpression levels of GluR1in the cultures from FMR1WT mice. However,neither forskolin nor DL-AP3affected the total expression levels of GluR1incultures from FMR1KO mice. The surface expressions of GluR1in the culturesfrom FMR1WT mice increase after treatment with forskolin alone andcombined with DL-AP3. In contrast with the unchanged total levels of GluR1,surface expressions of GluR1in the cultures from FMR1KO mice increasedafter simultaneous treatment of forskolin and DL-AP3. This findingdemonstrated the synergistic effects on the surface expressions of AMPA GluR1receptors in FMR1KO PFC neurons resulting from the inhibition of Grp1mGluRs and activation of adenylyl cyclase.6mGluR5, D1receptor, and NMDARs expression in the PFC culturesTo further explore the synaptic mechanisms behind the impairment ofdopaminergic modulation of LTP, we examined the expression of mGluR5andD1receptor in the ACC from WT and FMR1KO mice. As shown in Fig.6, therewas no difference in the levels of mGluR5, D1receptor, NMDA receptor NR2Asubunit, and NR2B subunit between the WT and FMR1KO mice. Thus, theimpairment of mGluR5and D1receptor modulation of synaptic potentiation isnot due to defect in levels of mGluR5, D1receptor, and NR2A-or NR2B-containing NMDA receptor.The function of NMDARs is regulated by its phosphorylation. Given thattyrosine phosphorylation of NMDARs contributes to LTP, we examined thetyrosine phosphorylation levels of NR2B subunits with anti-phosphotyrosine antibody to determine whether or not NMDARs are phosphorylated toSKF81297and DL-AP3. We found that the presence of SKF81297alone orSKF81297plus DL-AP3did not affect total NR2A or NR2B subunit expressionin the cultures from FMR1WT and KO mice. However, SKF81297alone orSKF81297plus DL-AP3significantly increased the level of phosphorylation ofthe NR2B subunit at Tyr-1472(p-NR2B-Tyr1472) in the cultures from FMR1WT mice. Meanwhile, SKF81297alone failed to increase p-NR2B-Tyr1472inthe cultures from FMR1KO mice. Synergistic addition of DL-AP3rescued thep-NR2B-Tyr1472by SKF81297in the cultures from FMR1KO mice. Thesedata implied that the increase of p-NR2B-Tyr1472contributed to the surfaceexpression of GluR1by SKF81297and DL-AP3in the KO cultures.7Synergistic effect of simultaneous treatments on the behavior of FMR1KO miceFragile X patients and animal models exhibit behavioral phenotypes that areconsistent with deficits in PFC function, such as hyperactivity, anxiety, andlearning abnormalities. We analyzed locomotor activity, anxiety, and learningabilities to characterize hyperactivity and learning deficit, the commonphenotypes in fragile X syndrome. We found increased locomotor activity inFMR1KO mice, which was consistent with our previous study. The locomotoractivity was not affected following SKF81297or/and DL-AP3subcutaneousinjection after forty-five minutes in WT mice. However, application of the D1receptor agonist SKF81297or simultaneous treatment of SKF81297andDL-AP3significantly reduced the distance traveled by FMR1KO mice. Thesedata indicate that FMR1KO mice exhibit hyperactivity and that simultaneoustreatment of D1agonist and Grp1mGluRs antagonist can reduce thehyperactivity phenotype.We then investigated anxiety-like behaviors with the elevated plus-maze test.We found no differences between FMR1KO and WT mice and no effect ofSKF81297alone or plus DL-AP3. These data indicate that anxiety-like behaviors are not detectable in this fragile X mouse model and suggest that thesebehaviors are probably not related to the D1receptor signaling pathway. Thesefindings are consistent with our previous study.Last, the learning ability was detected in the Morris Water Maze test. Micewere treated with SKF81297alone or plus DL-AP3for5weeks. Swim latencywas compared over the4days of the experiment, and there were significantdifferences in escape latency between genotype at day1. SKF81297or DL-AP3treatment alone did not change the swim latency in the FMR1WT and KO mice.Simultaneous treatment of SKF81297or DL-AP3also did not improve thelearning ability in the FMR1WT mice. However, Simultaneous treatment ofSKF81297and DL-AP3reduced the swim latency on the final day in the FMR1KO mice. Moreover, we compared the swimming speed during the water mazetest and found no differences among the groups. These findings clearly showthat simultaneous treatment of SKF81297and DL-AP3significantly improvesthe spatial leaning of FMR1KO mice.In summary, the present study, together with our previous study, reveals thatmGluR1inhibition is a useful strategy to recover D1receptor signaling in theFMR1KO mice, and that combination of Grp1mGluR antagonist and D1agonist is a potential drug therapy for the FXS.8Abnormal serotoninergic synaptic plastisty of FMR1KO mice ACCIn order to check if other neurotransmiting related system is normal, wechecked5-HT2A system, another important systern for synaptic plastisty. Wefound that bath application of R-96544or ketanserin significantly enhanced theamplitude of LTP. This suggests that inhibition of5-HT2A receptor canfacilitate LTP induction. However, this facilitation of LTP in WT mice byR-96544or ketanserin was lost in FMR1KO mice, demonstrating thatserotoninergic modulation of LTP is impaired in FMR1KO mice and thatFMRP is required for serotoninergic modulation of LTP in the ACC.We next applied a selective NMDAR antagonist, AP5(50μM), and found that LTP facilitation by R-96544(5μM) was completely blocked in the WTmice.Similarly, LTP was abolished by10mM BAPTA in the pipette solution,indicating that serotoninergic modulation of LTP is dependent on the activationof NMDARs and elevated postsynaptic Ca2+concentrations.To determine whether presynaptic and/or postsynaptic mechanisms areinvolved in serotoninergic modulation of LTP, we measured paired-pulsefacilitation (PPF) in the ACC. We found that PPF induced at five differentintervals did not differ in WT and FMR1KO mice in the presence of R-96544inthe ACC. The data suggest that presynaptic mechanisms do not seem to beinvolved in the serotoninergic modulation of LTP.To further explore the synaptic mechanisms behind the impairment ofserotoninergic modulation of LTP, we examined5-HT2A receptor andglutamate receptor subunits expression in the ACC from WT and FMR1KOmice.There were no differences in5-HT2A receptor expression between the WTand FMR1KO mice. Furthermore, the total GluR1, NMDA receptor NR2A, andNR2B subunits expression were no differences between the WT and FMR1KOmice under basal conditions. Thus, the impairment of serotoninergic modulationof synaptic potentiation is not due to a defect in5-HT2A receptor and glutamatereceptors expression. To investigate the role of5-HT2A receptor signalingpathway in the synaptic plasticity, we tested the effects of R-96544on AMPAreceptor GluR1subunits surface trafficking in cultured PFC neurons in thepresence of1μM serotonin. Surface expression of GluR1was increased aftertreatment with5-HT2AR antagonist R-96544. This finding demonstrates thatinhibition of5-HT2A receptor can facilitate the surface expression of AMPAGluR1receptors. However, the surface expression of GluR1triggered byR-96544was impaired in FMR1KO PFC neurons. These results provide direct evidence that FMRP is involved in serotonin receptor signaling and contributesto GluR1surface expression induced by5-HT2A receptor inactivation.