Research Articles, Development/Plasticity/Repair

Concentration-Dependent Bidirectional Modification of Evoked Synaptic Transmission by Gadolinium and Adverse Effects of Gadolinium-Based Contrast Agent

Odgerel Zorigt, Hiroki Yasuda, Takahito Nakajima and Yoshito Tsushima
Journal of Neuroscience 23 April 2025, 45 (17) e1622242025; https://doi.org/10.1523/JNEUROSCI.1622-24.2025

Abstract

Gadolinium-based contrast agents (GBCAs) for magnetic resonance imaging (MRI) are gadolinium chelates and can leave gadolinium in brain regions after administration, causing damage to brain tissues. However, the exact effects of gadolinium on synaptic function and the underlying mechanisms have not yet been elucidated. Here, we report that gadolinium differentially modulates evoked and spontaneous synaptic transmission and induces bidirectional changes in the efficacy of evoked synaptic transmission in the mouse hippocampus in a concentration-dependent manner. Low-concentration gadolinium (100 μM) modestly potentiated evoked field excitatory postsynaptic potentials (fEPSPs), while high-concentration gadolinium induced group 1 metabotropic glutamate receptor (mGluR)-, endocannabinoid (eCB)-, and purinergic P2Y1 receptor (P2Y1R)-dependent, presynaptically expressed long-term depression (LTD). Higher concentration of gadolinium (1,000 μM) also induced NMDAR- and mGluR-independent, partially P2Y13R-dependent, postsynaptically expressed LTD. Low-concentration gadolinium greatly increased miniature excitatory postsynaptic current (mEPSC) frequency, while high-concentration gadolinium much more robustly increased its frequency and amplitude. Finally, we found that evoked EPSCs were not affected by a macrocyclic GBCA, gadoterate meglumine (Gd-GOTA, Magnescope). However, evoked EPSCs were enhanced by a linear GBCA, gadopentetate dimeglumine (Gd-DTPA, Magnevist), at 100 μM, a clinically relevant concentration in the human brain after repeated clinical GBCA administration and in the cerebrospinal fluid in the rodent brain during experimental GBCA administration. Thus, evoked and spontaneous synaptic transmissions are independently modulated by gadolinium. Furthermore, Gd-GOTA effectively chelated gadolinium; however, Gd-DTPA had side effects on the evoked synaptic transmission, presumably because it did not completely chelate gadolinium.

Significance Statement

Gadolinium is used in gadolinium-based contrast agents (GBCAs), gadolinium chelates, for magnetic resonance imaging examination. Herein, we report influences of gadolinium and GBCAs on synaptic transmission. High-concentration gadolinium (500–1,000 μM) induces metabotropic glutamate receptor-, endocannabinoid-, and purinergic receptor-dependent long-term depression and simultaneously enhances spontaneous glutamate release. In contrast, gadolinium enhances evoked synaptic transmission at 100 μM, which is the concentration observed in the human patient brain after repeated GBCA administration. Gadoterate meglumine (Magnescope, 100 μM), a macrocyclic GBCA, did not affect synaptic transmission. However, gadopentetate dimeglumine (Magnevist, 100 μM), a liner GBCA, enhanced synaptic transmission, suggesting that gadopentetate dimeglumine does not fully chelate gadolinium, which can have a negative effect on brain function.

Introduction

Gadolinium-based contrast agents (GBCAs) are gadolinium chelates that are widely used in magnetic resonance imaging (MRI) to detect tumors or inflammatory lesions. GBCAs are hydrophilic and soluble in blood, and the blood–brain barrier prevents them from entering the central nervous system. However, recent reports suggest that frequent GBCA administration could cause gadolinium deposition within specific brain areas, such as the basal ganglia, dentate nucleus in the cerebellum of human patients (Kanda et al., 2014), and the lateral cerebellar nucleus and hippocampus of rodents (Kartamihardja et al., 2016; Habermeyer et al., 2020). Moreover, continuous MRI image acquisition revealed that GBCAs moved into the cerebrospinal fluid (CSF) in the perivascular spaces 3–5 h after intravenous administration (Jost et al., 2017). The glymphatic system may also be involved in gadolinium deposition (Taoka and Naganawa, 2018). Based on these reports, the exposure of neurons to GBCAs appears to be more frequent than previously thought; therefore, the effects of GBCAs on brain tissue should be examined.

To date, only a few studies have described the effects of free gadolinium ions on synaptic transmission. Gadolinium increases the frequency of miniature endplate potentials (mEPPs) in frog neuromuscular junctions (Molgo et al., 1991) and miniature excitatory postsynaptic currents (mEPSCs) in a calcium-independent manner in cultured hippocampal neurons (Capogna et al., 1996; Lei and MacDonald, 2001). Gadolinium can induce the exocytosis of synaptic vesicles without calcium at presynaptic terminals (Molgo et al., 1991; Waseem et al., 2008). In contrast, gadolinium depresses evoked EPPs by inhibiting presynaptic voltage-dependent calcium channels (Molgo et al., 1991). In addition, a high concentration of gadolinium activates group 1 metabotropic glutamate receptors (mGluR) types 1 and 5 expressed in oocytes (Kubo et al., 1998). Group 1 mGluRs are located at the postsynaptic sites and mediate postsynaptically and presynaptically expressed long-term depression (LTD; Huang et al., 2008; Yasuda et al., 2008, 2020; Luscher and Huber, 2010; Castillo et al., 2012; Sanderson et al., 2022). However, it has not yet been determined whether gadolinium activates postsynaptic group 1 mGluRs to induce LTD. Therefore, we examined whether gadolinium induces mGluR-dependent and mGluR-independent LTD in the hippocampus and found concentration-dependent bidirectional effects of gadolinium on evoked synaptic transmission and independent modification of spontaneous synaptic transmission. Group 1 mGluRs induce endocannabinoid (eCB)-dependent LTD (Huang et al., 2008; Yasuda et al., 2008; Castillo et al., 2012), and we also report that cannabinoid receptor 1 (CB1R) and 2-arachidonylglycerol (2-AG), a major eCB in the hippocampus, is responsible for higher-concentration gadolinium-induced, mGluR-dependent, presynaptically expressed LTD.

CB1R on presynaptic terminals is associated with Gi/o-protein and reduces transmitter release by inhibiting cAMP signaling (Castillo et al., 2012). In contrast, CB1Rs on astrocytes are associated with the Gq-protein, elevate intracellular calcium levels through IP3 receptor activation, and release gliotransmitters, including ATP (Eraso-Pichot et al., 2023). Purinergic P2Y receptors (P2YRs) are also located on astrocytes and are associated with the Gq-protein; additionally, ATP and eCBs can induce production and release of 2-AG from astrocytes (Eraso-Pichot et al., 2023). Therefore, the release of 2-AG from postsynaptic sites may trigger the robust amplification of their release from astrocytes through CB1R and P2YRs. We found that the involvement of P2Y1R in high-concentration gadolinium-induced LTD was associated with an increase in the paired-pulse ratio (PPR), which implies a decrease in glutamate release. We also found that NMDAR- and mGluR-independent P2Y13R were partially involved in gadolinium-induced LTD at high concentrations, without changes in the PPR. P2Y13R is expressed in microglia (Stefani et al., 2018), and we discuss the role of microglia in gadolinium-induced postsynaptically expressed LTD.

Finally, we investigated whether synaptic transmission is affected by gadopentetate dimeglumine (Gd-DTPA; Magnevist), a linear GBCA, and gadoterate meglumine (Gd-DOTA; Magnescope), a macrocyclic GBCA, at clinically relevant concentrations, i.e., concentrations found in the human brain after repeated GBCA administration (Bower et al., 2019) and the CSF during experimental administration of GBCAs in mice (Jost et al., 2017). Here, we report a significant influence of Gd-DTPA on evoked synaptic transmission.

Materials and Methods

Animal use and all experimental procedures were approved by the Ethical Committee for Animal Experiments of Gunma University (#14-030) and the Animal Care and Use Committee of Saga University (#30-048-0). All experiments were performed in accordance with committee guidelines.

Hippocampal slice preparation

Extracellular field excitatory postsynaptic potentials (fEPSPs) were recorded using hippocampal slices from 12- to 18-week-old male ddY mice (Japan SLC), as described previously (Yasuda and Mukai, 2015; Yasuda et al., 2020). The mice were deeply anesthetized with isoflurane in a bell jar and decapitated, and their brains were quickly removed from the skull. The hippocampi were dissected and cut into 400-μm-thick slices using a vibrating microtome (VT1200S, Leica) in ice-cold carbogenated (95% O2/5% CO2) normal artificial cerebrospinal fluid (ACSF) containing the following (in mM): 119 NaCl, 2.5 KCl, 26.2 NaHCO3, 1 NaH2PO4, 4 CaCl2, 4 MgSO4, and 11 glucose, pH 7.4. The slices were incubated for at least 2 h in a submersion-type incubation chamber containing the same carbogenated normal ACSF.

For whole-cell recordings, we prepared hippocampal slices from mice of the same age range using sucrose-based ACSF or N-methyl-d-glucamine (NMDG)-based ACSF for the protective recovery method (Tomioka et al., 2014; Ting et al., 2018). Hippocampal slices were cut in ice-cold carbogenated sucrose-based ACSF containing the following (in mM): 215 sucrose, 2.5 KCl, 26.2 NaHCO3, 1.6 NaH2PO4, 1 CaCl2, 4 MgSO4, 4 MgCl2, and 20 glucose, pH 7.4. The slices were incubated at room temperature in 50% sucrose-based ACSF and 50% normal ACSF for 30 min and then transferred to an incubation chamber containing normal carbogenated ACSF.

For the protective recovery method, the mice were initially anesthetized with isoflurane, followed by an intraperitoneal injection of ketamine (100 mg/kg) and xylazine (10 mg/kg). Then, pure oxygen was administered through a face mask. The mice had their thorax opened and transcardially perfused with ice-cold carbogenated protective NMDG-HEPES ACSF containing the following (in mM): 92 NMDG, 2.5 KCl, 1.2 NaH2PO4, 30 NaHCO3, 20 HEPES, 25 glucose, 5 ascorbate, 3 sodium pyruvate, 12 N-acetyl-ʟ-cysteine, 0.5 CaCl2, and 10 MgSO4. The mice were decapitated, and their brains were quickly removed from the skull, the hippocampi were dissected, and 400-μm-thick slices were cut using a vibrating microtome in ice-cold carbogenated NMDG-HEPES ACSF. Slices were transferred into a holding chamber containing 60 ml carbogenated NMDG-HEPES ACSF at 34°C, and a gradual Na+ spike-in procedure was carried out by adding 2 M NaCl dissolved in NMDG-HEPES ACSF as previously described (Ting et al., 2018). The slices were then incubated in carbogenated ACSF containing the following (in mM): 83 NaCl, 2.25 KCl, 1.13 NaH2PO4, 27 NaHCO3, 18 HEPES, 22.5 glucose, 4.5 ascorbate, 2.7 sodium pyruvate, 10.8 N-acetyl-ʟ-cysteine, 3.6 CaCl2, and 3.6 MgSO4 at room temperature.

Slice electrophysiology

Even a low concentration of gadolinium (100 μM) induced precipitation in normal ACSF (Fig. 1A), we used HEPES-buffered ACSF containing the following (in mM): 140 NaCl, 2.5 KCI, 1.3 MgSO4·7H2O, 10 HEPES, 11 glucose, 2.5 CaCl2, and 0.05 picrotoxin (pH 7.4 with NaOH, osmolality adjusted to 280–290 mOsm) to record synaptic transmission. Apparent gadolinium precipitation was not observed in HEPES-buffered ACSF with 100, 500, and 1,000 μM gadolinium (Fig. 1A). The gadolinium concentration in ACSF was quantified through inductively coupled plasma-optical emission spectrometry (ICP-OES) using Optima 5300 DVZ (PerkinElmer Japan G.K.). Gadolinium was detected at 342 nm and a linear calibration curve was made using diluted gadolinium standard solutions (Gd 1000, Fujifilm Wako Pure Chemical Corporation) to quantify gadolinium concentration in ACSF samples.

Figure 1.
Figure 1.

Low concentration of gadolinium (100 μM) presynaptically potentiates synaptic transmission. A, Appearance of normal ACSF with 100 μM gadolinium (normal) and HEPES-buffered ACSF with 100, 500, and 1,000 μM gadolinium (HEPES; right) and gadolinium concentrations in these HEPES-buffered ACSF quantified by ICP-OES (n = 3; left). Gadolinium precipitation occurred in normal ACSF; however, apparent precipitation did not occur with HEPES-buffered ACSF with 100, 500, and 1,000 μM gadolinium. B, C, Example (B) and summary (C) of the effects of 100 μM gadolinium on fEPSPs recorded in HEPES-buffered ACSF (n = 18 from 8 mice). Relationship between ratios of PPRs 40–50 min after gadolinium application to baseline PPR and the amplitude of synaptic potentiation by 100 μM gadolinium is shown in inset (C). These were significantly correlated (r = −0.778; p = 0.00014). D, Average time course of fEPSPs in HEPES-buffered ACSF without gadolinium application (n = 15 from 11 mice). fEPSP amplitude was stable during recordings for >60 min and PPRs were not changed between 0 and 60 min after start of recordings (inset). E, Summary of the effects of 10 μM MTEP and 5 μM YM 298198 on potentiation of synaptic transmission by 100 μM gadolinium (n = 9 from 6 mice). All the hippocampal slices used in this figure were cut in normal ACSF.

Hippocampal slices were transferred to a recording chamber and perfused with oxygenated (100% O2) HEPES-buffered recording ACSF at 30°C. Buffers except HEPES were omitted to prevent gadolinium-induced precipitation. fEPSPs were evoked with a stimulating glass electrode containing the same ACSF placed in the stratum radiatum and recorded in the CA1 region using a MultiClamp 700B amplifier (Molecular Devices). Data acquisition and analyses were performed using Igor Pro (WaveMetrics). Basal synaptic transmission was obtained at 0.05 Hz.

mEPSCs were recorded in voltage-clamp mode at −70 mV using a MultiClamp 700B in the presence of 1 μM tetrodotoxin (TTX) through a glass recording pipette with an internal solution containing the following (in mM):135 CsMeSO3, 10 HEPES, 0.2 EGTA, 8 NaCl, 4 MgATP, and 0.3 Na3GTP (pH 7.2 with CsOH, osmolarity adjusted to 275–285 mOsm). Spontaneous EPSCs (sEPSCs) were recorded simultaneously with evoked EPSCs in HEPES-buffered recording ACSF without TTX. Traces were digitized at 2 kHz and mEPSC and sEPSCs were analyzed using Mini Analysis software (Synaptosoft). The threshold mEPSC amplitude was set at 4 pA, and each mEPSC was initially detected automatically and visually verified. Evoked EPSC amplitudes were measured using Igor Pro.

Drugs

GdCl3·6H2O was purchased from Sigma-Aldrich. MTEP, MPEP, YM 298198, AM251, Orlistat, and MRS 2211 were purchased from Tocris Bioscience. d-APV and PPADS were purchased from Cayman Chemical. MRS 2179 was from Cayman Chemical and Abcam. Gd-DTPA (Magnevist) was purchased from Bayer and Selleck Chemicals. Gd-DOTA (Magnescope) was purchased from Guerbet. All other chemicals were purchased from Fujifilm Wako Pure Chemical Corporation or Sigma-Aldrich.

Experimental design and statistical analyses

Hippocampal slices were prepared from one mouse in a day. For fEPSP recordings, two hippocampal slices were placed in the recording chamber on one microscope, and two same type recording setups were used; therefore, fEPSPs were recorded from four hippocampal slices at the same time. EPSCs were recorded from one cell from one slice at one time using a blind patch-clamp method, and 0–2 cells were retrieved from a mouse in a day. We did not statistically calculate sample sizes; however, the data obtained in the present study were similar to those in papers previously published in our field. We usually used at least five mice for one group in each experiment in figures.

One-way ANOVA with the Tukey–Kramer post hoc test was used for multiple statistical comparisons, and Pearson’s correlation coefficient was calculated using EZR (Easy R) software (Kanda, 2013). The t test, paired t test, and analysis of correlation coefficients were performed using Excel. Statistical significance was defined as p < 0.05, and numbers of data (slices, cells, and mice) and statistical values (p value, t value for t test, or F value for ANOVA) were described in the corresponding text. Numbers of data and p values were described also in the figure legends. Results are reported here as the mean ± SEM.

Results

Low gadolinium concentration induces presynaptically expressed potentiation

Initially, we attempted to investigate the effects of a low concentration of gadolinium (100 μM for 20 min) on synaptic transmission. Unfortunately, white precipitation occurred in normal ACSF with 100 μM gadolinium (Fig. 1A); thus HEPES-buffered ACSF was used. We did not find apparent precipitation in HEPES-buffered ACSF with 100, 500, and 1,000 μM gadolinium (Fig. 1A). We also quantified gadolinium concentrations using ICP-OES, and almost all gadolinium added was detected in HEPES-buffered ACSF (Fig. 1A; 100 μM gadolinium ACSF, no vehicle, 119.7 ± 0.1 μM, n = 3, vehicle, 106.7 ± 0.5 μM, n = 3; 500 μM gadolinium ACSF, no vehicle, 514.8 ± 0.3 μM, n = 3, vehicle, 503.7 ± 0.6 μM, n = 3; 1,000 μM gadolinium ACSF, no vehicle, 974.4 ± 6.5 μM, n = 3, vehicle, 963.2 ± 2.7 μM, n = 3). Gadolinium at 100 μM potentiated fEPSPs in hippocampal slices cut in normal ACSF (Fig. 1B,C; 40–50 min after applying gadolinium, 126.7 ± 4.2%, n = 18 from 8 mice). Also, the amplitude of gadolinium-induced potentiation was significantly correlated with a decrease in PPR (Fig. 1C; r = −0.78; p = 0.00014), suggesting that a low concentration of gadolinium enhances glutamate release in the hippocampus. The enhancement of synaptic transmission is not a nonspecific drift of fEPSP amplitude because we recorded stable fEPSPs for >60 min without gadolinium in HEPES-based ACSF (Fig. 1D; 100.2 ± 1.5% of baseline 60 min after start of recordings, n = 15 from 11 mice). Also, PPRs were not significantly changed between 0 and 60 min after the start of recordings (Fig. 1D; 0 min, 1.51 ± 0.04; 60 min, 1.48 ± 0.03). Gadolinium has been previously reported to activate mGluRs (Kubo et al., 1998). mGluR5 is highly expressed (Ferraguti and Shigemoto, 2006) and mGluR1 is functional at excitatory synapses in the CA1 region (Gil-Sanz et al., 2008; Gasselin et al., 2017; Yasuda et al., 2020; see below). Therefore, we examined whether a low concentration of gadolinium activates group 1 mGluRs. MTEP (10 μM) and YM 298198 (5 μM), mGluR5 and mGluR1 antagonists, respectively, did not affect the enhancement (Fig. 1E; 130.5 ± 4.7%, n = 9 from 6 mice). Thus, a low concentration of gadolinium did not activate group 1 mGluRs.

High gadolinium concentrations induce group 1 mGluR- and eCB-dependent LTD

Next, we examined the effects of high gadolinium concentrations on synaptic transmission. Gadolinium at a high concentration (500 μM) depressed fEPSPs, and this effect persisted even after the washout of gadolinium in hippocampal slices cut in normal ACSF (Fig. 2A,E,F; 78.3 ± 4.3%, n = 16 from 10 mice), suggesting that a high concentration of gadolinium induces LTD. To examine the contributions of group 1 mGluRs, we applied 500 μM gadolinium in the presence of 10 μM MTEP, because mGluR5 is important for inducing mGluR-dependent LTD in the CA1 region (Fitzjohn et al., 1999; Huang et al., 2004; Luscher and Huber, 2010; Yasuda et al., 2020). MTEP significantly inhibited gadolinium-induced LTD (Fig. 2B,E,F; 98.7 ± 4.3%, n = 15 from 10 mice; F(3,55) = 6.82; p = 0.042; ANOVA with Tukey–Kramer test). However, MTEP did not restore synaptic potentiation induced by a low-concentration gadolinium (Fig. 1). mGluR1 is also involved in mGluR-dependent LTD in the CA1 region (Volk et al., 2006; Kumar and Foster, 2007; Luscher and Huber, 2010; Yasuda et al., 2020). We found that 5 μM YM 298198 also inhibited gadolinium-induced LTD (Fig. 2C,E,F; 100.5 ± 4.4%, n = 14 from 8 mice; p = 0.023; ANOVA with Tukey–Kramer test). Furthermore, MTEP and YM 298198 partially restored gadolinium-induced potentiation (Fig. 2D,E,F; 111.7 ± 8.1%, n = 14 from 11 mice; p = 0.00030; ANOVA with Tukey–Kramer test). These results suggest that high gadolinium concentration induces mGluR1- and mGluR5-dependent LTD and that direct inhibition of presynaptic voltage-dependent calcium channels by gadolinium is not necessarily fully responsible for LTD. Among all the data, the changes in fEPSP slopes 50 min after the start of 500 μM gadolinium application were significantly negatively correlated with PPRs 50 min after the start of gadolinium application compared with those at baseline (Fig. 2G,H; r = −0.56, p = 0.0000039). PPRs were significantly reduced when 500 μM gadolinium was applied in the presence of MTEP and YM 298198 (Fig. 2H; F(3,55) = 5.29; p = 0.0012; ANOVA with the Tukey–Kramer test). These results suggest that group 1 mGluR-dependent LTD induced by a high concentration of gadolinium is associated with decreased presynaptic transmitter release. If 500 μM gadolinium had induced group 1 mGluR-dependent LTD only, PPRs should have been elevated by 500 μM gadolinium and should not have changed when 500 μM gadolinium was applied in the presence of MTEP and YM 298198. However, the average PPR in the absence of MTEP and YM 298198 was near 100% (97.1%), while that in the presence of MTEP and YM 298198 was 88.1% (Fig. 2H). Therefore, we speculated that, similar to low concentration of gadolinium, high concentration of gadolinium also activated molecular mechanisms that elevated glutamate release simultaneously with other mechanisms that are activated by group 1 mGluRs and decreased glutamate release probability and that it induces presynaptically expressed synaptic potentiation.

Figure 2.
Figure 2.

High concentration of gadolinium (500 μM) induces group 1 mGluR-dependent LTD. A–D, LTD induced by 500 μM gadolinium for 20 min (A) and the effects on high-concentration gadolinium-induced LTD of 10 μM MTEP (B), 5 μM YM 298198 (C), and 10 μM MTEP and 5 μM YM 298198 (D). E, Average time course of gadolinium-induced LTD in the absence and presence of group 1 mGluR antagonists (control, n = 16 from 10 mice; MTEP, n = 15 from 10 mice; YM, n = 14 from 8 mice; MTEP + YM, n = 14 from 11 mice). F, Summary of gadolinium-induced LTD and the effects of group 1 mGluR antagonists on LTD at 50 min after the start of 500 μM gadolinium application. G, H, Relationship between ratios of PPRs 50 min after 500 μM gadolinium application to baseline PPRs and the amplitude of changes in fEPSPs. Averages are shown with standard errors (H). PPR ratios and changes in fEPSP slopes were significantly and negatively correlated among all data (r = −0.56, p = 0.0000039). All the hippocampal slices used in this figure were cut in normal ACSF. *p < 0.05, **p < 0.01, ***p < 0.001, one-way ANOVA with Tukey–Kramer test.

Previously, we reported that group 1 mGluRs activate eCB signaling and induce heterosynaptic eCB-dependent LTD of excitatory synaptic transmission, which is associated with the elevation of PPRs in the developing hippocampus and cortex (Huang et al., 2008; Yasuda et al., 2008). Therefore, we examined the effects of AM251, a CB1R antagonist, on LTD induced by high concentration of gadolinium. High-concentration gadolinium (1,000 μM) induced LTD in the presence of a vehicle (0.1% DMSO; Fig. 3A,G,H; 67.6 ± 4.4% of baseline 80 min after the start of gadolinium application, n = 16 from 10 mice), and the LTD was significantly reduced when 5 μM AM251 was applied (Fig. 3B,G,H; 86.5 ± 3.1%, n = 18 from 7 mice; F(5,79) = 17.8; p = 0.022; ANOVA with Tukey–Kramer test). We also tested the involvement of 2-AG, a dominant eCB in the brain (Katona and Freund, 2012; Araque et al., 2017) by applying Orlistat (tetrahydrolipstatin, THL), an inhibitor of diacylglycerol lipase-a, which synthesizes 2-AG (Peterfi et al., 2012). Orlistat (10 μM) significantly reduced LTD (Fig. 3C,G,H; 87.7 ± 3.8%, n = 13 from 6 mice; p = 0.027; ANOVA with Tukey–Kramer test), indicating that 2-AG was released and activated CB1Rs, and LTD was induced when a high concentration of gadolinium was applied. We verified that LTD induced by 1,000 μM gadolinium was reduced by MPEP (another mGluR5 antagonist, 10 μM) and YM298198 to the same extent as AM251 and Orlistat did (Fig. 3D,G,H; 87.8 ± 3.6%, n = 10 from 5 mice; p = 0.049; ANOVA with Tukey–Kramer test). The changes in fEPSP slopes 80 min after the start of 1,000 μM gadolinium application was significantly negatively correlated with PPRs 80 min after gadolinium application compared with those of the baseline among data with the vehicle only, AM251, Orlistat, and MPEP + YM298198 (Fig. 3I,J; r = −0.443, p = 0.00056). PPRs were also significantly reduced when 1,000 μM gadolinium was applied with AM251, Orlistat, and MPEP + YM298198 compared with those when 1,000 μM gadolinium was applied with 0.1% DMSO (Fig. 3I,J; vehicle, 103.2 ± 1.9%; AM251, 88.4 ± 2.1%, F(5,79) = 6.66, p = 0.000012; Orlistat, 93.5 ± 1.9%, p = 0.022; MPEP + YM, 93.3 ± 2.7%, p = 0.036; ANOVA with Tukey–Kramer test). These results suggest that eCB signaling is activated by high concentrations of gadolinium, which reduces presynaptic transmitter release. LTD remained suppressed by inhibitors of mGluRs or eCB signaling; however, PPR ratios were ∼90% after gadolinium application in the presence of these inhibitors (Fig. 3J), indicating that high concentrations of gadolinium also activated molecular mechanisms that enhanced glutamate release simultaneously with mGluR- and eCB signaling-dependent, presynaptically expressed LTD. We also speculate that postsynaptically expressed LTD was induced by high concentrations of gadolinium (see below).

Figure 3.
Figure 3.

LTD induced by high-concentration gadolinium is mediated by eCB and ATP signaling. A, LTD induced by 1,000 μM gadolinium for 20 min in the presence of a vehicle (0.1% DMSO). B–F, Effects on high-concentration gadolinium-induced LTD of 5 μM AM251 (B), 10 μM Orlistat (C), 10 μM MPEP and 5 μM YM298198 (D), 5 μM AM251 and 50 μM PPADS (E), and 50 μM PPADS (F). G, Average time course of gadolinium-induced LTD in the absence and presence of inhibitors (vehicle, n = 16 from 10 mice; AM251, n = 18 from 7 mice; Orlistat, n = 13 from 6 mice; MPEP + YM298198, 10 from 5 mice; AM251 + PPADS, 13 from 5 mice; PPADS, 15 from 5 mice). Inhibitors were applied for >60–90 min before gadolinium application. H, Summary of gadolinium-induced LTD and the effects of inhibitors on LTD 80 min after the start of 1,000 μM gadolinium application. I, J, Relationship between ratios of PPR 80 min after 1,000 μM gadolinium application to baseline PPR and the amplitude of changes in fEPSPs. Their averages are also shown with their standard errors (J). A regression line between PPR ratio and changes in fEPSPs was calculated from data in all conditions except AM251 + PPADS and PPADS application (r = −0.443, p = 0.00056). Note that the plots of averaged data on AM251 + PPADS and PPADS were far from the regression line (J). All hippocampal slices used in this figure were cut in normal ACSF. *p < 0.05, **p < 0.01, ***p < 0.001, one-way ANOVA with Tukey–Kramer test.

High-concentration gadolinium-induced presynaptic LTD requires P2Y1R and postsynaptic LTD partially requires P2Y13R

To investigate which receptor mediated LTD after mGluR and eCB signaling was blocked by inhibitors, we applied Orlistat and an NMDA receptor (NMDAR) antagonist, d-APV. However, 10 μM Orlistat and 50 μM d-APV did not completely inhibit LTD induced by 1,000 μM gadolinium (data not shown; 76.7 ± 8.1%, n = 4 from 2 mice); therefore, the involvement of NMDAR is less likely. Next, we tested the effects of an antagonist of purinergic P2Rs, because both P2X receptors (P2XRs) and P2YRs are involved in LTD in the hippocampus (Chen et al., 2013; Pougnet et al., 2014). Application of 5 μM AM251 + 50 μM PPADS, a P2 receptor antagonist, inhibited LTD induced by 1,000 μM gadolinium and restored small potentiation which was similar to that observed when 100 μM gadolinium was applied (Fig. 3E,G,H; 117.9 ± 6.8%, n = 13 from 5 mice, p = 0.0000000; ANOVA with Tukey–Kramer test). PPADS did not change PPRs in the presence of AM251 (Fig. 3I,J; 90.3 ± 2.9%), suggesting involvement of postsynaptic expression in high-concentration gadolinium-induced LTD. We also applied PPADS only and surprisingly, it completely inhibited LTD and restored potentiation (Fig. 3F–H; 113.3 ± 4.8%, n = 15 from 5 mice, p = 0.0000000; ANOVA with Tukey–Kramer test). PPADS significantly reduced PPRs (Fig. 3I,J; PPADS, 91.7 ± 1.2%; p = 0.0020; ANOVA with Tukey–Kramer test), suggesting that PPADS inhibits both mGluR/eCB signaling-dependent, presynaptically expressed LTD and postsynaptically expressed LTD. PPADS is an inhibitor of P2XRs and P2YRs; however, involvement of P2XRs is less likely because ATP-induced currents through P2X2R and P2X4R, main P2XRs in the hippocampus (Rubio and Soto, 2001; Pougnet et al., 2014), are inhibited by 100 μM gadolinium (Lee et al., 2001; Kawate et al., 2009).

To ensure that mGluR-dependent presynaptically expressed LTD was CB1R- and P2R-dependent, we applied a group 1 mGluR agonist DHPG at 30 μM, which induces presynaptically expressed LTD at this concentration (Fitzjohn et al., 2001; Sanderson et al., 2022). Administration of 30 μM DHPG induced small LTD (Fig. 4A,D,E; 84.9 ± 2.5% of baseline 60 min after the start of DHPG administration, n = 13 from 6 mice) and elevated PPRs (Fig. 4F,G; 109.8 ± 0.8%), suggesting the involvement of decreased glutamate release. AM251 and PPADS inhibited LTD induced by 30 μM DHPG (Fig. 4B–E; AM251, 101.5 ± 2.1%, n = 15 from 7 mice, F(2,39) = 8.68; p = 0.0031; PPADS, 103.7 ± 4.7%, n = 14 from 4 mice, p = 0.0016; ANOVA with Tukey–Kramer test) and significantly reduced PPRs with 30 μM DHPG (Fig. 4F,G; AM251, 100.8 ± 1.0%, F(2,39) = 16.6; p = 0.000039; PPADS, p = 0.000028; ANOVA with Tukey–Kramer test). These results indicate that presynaptically expressed LTD induced by 30 μM DHPG requires CB1R and P2R activity. We also examined whether cannabinoid receptor (CBR) agonist-induced LTD is P2R dependent. WIN55212-2 (WIN; 2 μM), a CBR agonist, induced LTD, which was inhibited by PPADS (Fig. 5A–D; vehicle, 69.8 ± 2.5%, n = 14 from 6 mice; PPADS, 95.0 ± 4.3%, n = 14 from 6 mice, t(26) = −5.08, p = 0.000027; t test). WIN also elevated PPRs, which was significantly reduced by PPADS (Fig. 5E; vehicle, 116.4 ± 2.0%; PPADS, 103.5 ± 1.5%, t(26) = 5.21, p = 0.000020; t test). Therefore, P2R activity is required for the suppression of glutamate release by CBRs. In summary, high concentrations of gadolinium activate P2Rs that induce postsynaptically expressed LTD and simultaneously activate group 1 mGluRs, which leads to 2-AG release, CB1R activation, and inhibits presynaptic glutamate release together with P2Rs.

Figure 4.
Figure 4.

LTD induced by low-concentration group 1 mGluR agonist is mediated by CB1Rs and P2Rs. A–C, LTD induced by 30 μM DHPG for 20 min (A) and the effects of 5 μM AM251 (B) and 50 μM PPADS (C) on low-concentration DHPG-induced LTD. D, Average time course of low-concentration DHPG-induced LTD in the absence and presence of AM251 and PPADS (vehicle, n = 13 from 6 mice; AM251, n = 15 from 7 mice; PPADS, n = 14 from 4 mice). The inhibitors were administered for 60–90 min before gadolinium application. E, Summary of low-concentration DHPG-induced LTD and the effects of AM251 and PPADS on LTD 60 min after the start of 500 μM gadolinium application. F, G, Relationship between ratios of PPRs 60 min after 500 μM gadolinium application to baseline PPRs and the amplitude of changes in fEPSPs. Their averages are also shown with their standard errors (G). PPR ratios and changes in fEPSP slopes were significantly negatively correlated (r = −0.601, p = 0.0000253). All the hippocampal slices used in this figure were cut in normal ACSF. **p < 0.01, ***p < 0.001, one-way ANOVA with Tukey–Kramer test.

Figure 5.
Figure 5.

P2Rs are required for cannabinoid-dependent LTD. A, B, LTD induced by 2 μM WIN55212-2 (WIN) for 20 min (A) and the effect of 50 μM PPADS on WIN-induced LTD (B). C, Average time course of low-concentration WIN-induced LTD in the absence and presence of PPADS (vehicle, n = 14 from 6 mice; PPADS, n = 14 from 6 mice). PPADS was applied for >90 min before the application of gadolinium. D, Summary of WIN-induced LTD and effects of PPADS on LTD 60 min after WIN application. E, Relationship between ratios of PPRs 60 min after 500 μM gadolinium application to baseline PPRs and the amplitude of changes in fEPSPs. Their averages are also shown with their standard errors. PPR ratios and changes in fEPSP slopes were significantly and negatively correlated (r = −0.601, p = 0.0000253). All the hippocampal slices used in this figure were cut in normal ACSF. **p < 0.01, ***p < 0.001, one-way ANOVA with Tukey–Kramer test.

PPADS at 50 μM is effective against P2Y1R and P2Y13R (Ralevic and Burnstock, 1998; von Kugelgen and Hoffmann, 2016). Initially, we examined the effects of MRS 2179, a P2Y1R antagonist, on LTD. High gadolinium concentration (1,000 μM) induced robust LTD, which was blocked by 60 μM MRS 2179 (Fig. 6A,B,D,E; vehicle, 51.7 ± 3.4% of baseline 80 min after the start of gadolinium application, n = 21 from 11 mice, MRS 2179, 98.8 ± 5.9%, n = 14 from 5 mice; F(2,54) = 27.65; p = 0.0000000; ANOVA with Tukey–Kramer test). MRS 2179 also reduced PPRs (Fig. 6F,G; vehicle, 106.7 ± 1.3%, MRS 2179, 87.7 ± 2.3%; F(2,37) = 34.2; p = 0.0000000; ANOVA with Tukey–Kramer test). These results indicate that presynaptically expressed LTD induced by high gadolinium concentration is mediated by P2Y1R. Next, we examined the involvement of P2Y13Rs in LTD. MRS 2211, a P2Y13R inhibitor, slightly reduced LTD induced by high gadolinium concentration (Fig. 6C–E; 30 μM, 66.7 ± 4.8%, n = 12 from 5 mice, 60 μM, 67.1 ± 6.9%, n = 10 from 5 mice, all data, 66.9 ± 4.0%; p = 0.025; ANOVA with Tukey–Kramer test); however, the PPR remained unchanged (Fig. 6F,G; 107.1 ± 2.0%). Therefore, P2Y13R may be involved in postsynaptically expressed LTD induced by higher concentrations of gadolinium.

Figure 6.
Figure 6.

LTD induced by high concentration of gadolinium is inhibited by P2Y1R and P2Y13R antagonists. A–C, LTD induced by 1,000 μM gadolinium for 20 min (A) and the effects of 60 μM MRS 2179, a P2Y1R antagonist (B), and 60 μM MRS 2211, a P2Y13R antagonist (C), on gadolinium-induced LTD. D, Average time course of high-concentration gadolinium-induced LTD in the absence and presence of MRS 2179 and MRS 2211 (vehicle, n = 21 from 11 mice; MRS 2179, n = 14 from 5 mice; MRS 2211 30 μM, n = 12 from 5 mice, 60 μM, n = 10 from 5). The inhibitors were administered for >90 min before gadolinium application. E, Summary of high-concentration gadolinium-induced LTD and the effects of MRS 2179 and MRS 2211 on LTD 80 min after the start of 1,000 μM gadolinium application. F, G, Relationship between ratios of PPRs 80 min after gadolinium application to baseline PPRs and the amplitude of changes in fEPSPs. Their averages are also shown with their standard errors (G). The regression line was calculated for the vehicle and MRS 2179 data. All the hippocampal slices used in this figure were cut in normal ACSF. *p < 0.05, ***p < 0.001, one-way ANOVA with Tukey–Kramer test.

High concentration of gadolinium simultaneously induces LTD of evoked EPSCs and potentiation of spontaneous synaptic transmission

We found that low and high concentrations of gadolinium potentiate and reduce evoked fEPSPs, respectively, and we examined whether they have the same effects on spontaneous synaptic transmission. We recorded mEPSCs in the presence of 1 μM TTX, and 100 μM gadolinium dramatically increased mEPSC frequency without changing mEPSC amplitude 50 min after the start of gadolinium application (Fig. 7A–C; amplitude, baseline, 10.55 ± 0.99 pA, gadolinium, 10.86 ± 1.03 pA; frequency, baseline, 1.33 ± 0.31 Hz, gadolinium, 16.92 ± 6.15 Hz, n = 8 from 5 mice; t(7) = −2.65, p = 0.33; paired t test). Surprisingly, 500 μM gadolinium increased mEPSC amplitude (Fig. 7D,E; baseline, 8.47 ± 0.56 pA; gadolinium, 11.70 ± 0.99 pA; n = 8 from 6 mice; t(7) = −3.37, p = 0.012; paired t test). It more robustly increased mESPC frequency (Fig. 7D,F; baseline, 1.48 ± 0.22 Hz; gadolinium, 34.03 ± 3.64 Hz; t(7) = −9.29, p = 0.000035; paired t test). The increase in mEPSC amplitude between 100 and 500 μM gadolinium was not statistically significant (Fig. 7G; 100 μM, 107.6 ± 13.1%; 500 μM 139.9 ± 11.1%); however, the increase in mEPSC frequency induced by 500 μM gadolinium was more than that induced by 100 μM (Fig. 7H; 100 μM, 1,177.1 ± 313.8%; 500 μM, 2,598.8 ± 449.2%; t(14) = −2.59, p = 0.21, t test), suggesting that 500 μM gadolinium potentiates spontaneous miniature synaptic transmission >100 μM gadolinium.

Figure 7.
Figure 7.

Low and high concentrations of gadolinium increase mEPSC frequency in slices cut in sucrose-based ACSF. A, mEPSCs before and 50 min after 100 μM gadolinium application. B, C, Summary of the effects of 100 μM gadolinium on mEPSC amplitude (B) and frequency (C; n = 8 from 5 mice). Low concentration of gadolinium robustly increased mEPSC frequency (t(7) = −2.65, p = 0.033; paired t test); however, it did not significantly affect mEPSC amplitude. D, mEPSCs before and 50 min after 500 μM gadolinium application. E, F, Summary of the effects of 500 μM gadolinium on mEPSC amplitude (C) and frequency (D; n = 8 from 6 mice). High concentration of gadolinium robustly increased mEPSC frequency (t(7) = −9.29, p = 0.000035; paired t test). It also increased mEPSC amplitude (t(7) = −3.37, p = 0.012; paired t test). G, H, Ratios of mEPSC amplitude (G) and frequency (H) 50 min after gadolinium application to those at baseline. 500 μM gadolinium increased mEPSC frequency significantly >100 μM gadolinium, although 100 μM gadolinium potentiated fEPSPs (t(14) = −2.59, p = 0.21, t test) and 500 μM gadolinium depressed them. All hippocampal slices used in this figure were cut in sucrose-based ACSF. *p < 0.05, ***p < 0.001.

Here, we investigated whether a high concentration of gadolinium induces LTD of evoked EPSCs as it induced LTD of fEPSPs and recorded evoked EPSCs and sEPSCs simultaneously in the absence of TTX. Gadolinium (500 μM) significantly increased sEPSC amplitude (Fig. 8A,B; baseline, 11.12 ± 0.83 pA; gadolinium, 14.04 ± 0.85 pA; n = 10 from 8 mice; t(9) = −5.16, p = 0.00060; paired t test) and robustly increased sESPC frequency (Fig. 8A,B; baseline, 2.26 ± 0.34 Hz; gadolinium, 47.98 ± 3.98 Hz; t(9) = −11.87, p = 0.00000085; paired t test). On the other hand, simultaneously recorded evoked EPSC amplitude was depressed (Fig. 8A,C; 59.8 ± 7.7% of baseline 50 min after the start of gadolinium application). To determine whether LTD expression is presynaptic, we analyzed the coefficient of variation (CV) of the evoked EPSC amplitude before and 50 min after gadolinium application (Yasuda et al., 2008; Tomioka et al., 2014). Suppose that each synapse has a release probability of p with a quantal size of q and that the number of activated synapses is N; then, the amplitude of the evoked EPSCs is Npq, and CV−2 is equal to Np/(1 − p). If there is a decrease in p, the CV−2 plot must be below the diagonal line and not affected by postsynaptic changes in q. All points plotted were on or below the diagonal line (Fig. 8D), suggesting that there is a decrease in release probability at presynaptic sites with high-concentration gadolinium-induced LTD. Thus, a high concentration of gadolinium modulated the evoked EPSCs and sEPSCs differently. Evoked EPSCs are induced by presynaptic action potentials. However, sEPSCs and mEPSCs do not require presynaptic activity. Therefore, we tested the possibility that LTD was only induced in activated pathways and that LTD was absent in nonactivated pathways. We recorded fEPSPs in two pathways and stopped stimulation for 40 min from the start of 500 μM gadolinium application in one (inactive) pathway. High-concentration gadolinium (500 μM) induced LTD in both active and inactive pathways (Fig. 9; active pathway, 70.0 ± 4.3% of baseline 60 min after the start of gadolinium application; inactive pathway, 75.0 ± 4.7%, n = 11 from 4 mice), suggesting that a high concentration of gadolinium induces LTD of evoked EPSCs in an activity-independent manner. This implies that the possibility that sEPSCs and mEPSCs were not depressed because they did not need presynaptic action potentials was less likely. Thus, a high concentration of gadolinium differentially modulates evoked and spontaneous synaptic transmission.

Figure 8.
Figure 8.

High concentration of gadolinium induces LTD of evoked EPSCs and simultaneously potentiates spontaneous synaptic transmission. A, Example of evoked EPSCs and sEPSCs before (1) and 50 min after 500 μM gadolinium application (2). Arrows indicate the time of afferent stimulation. Time course of the amplitude of evoked EPSCs was shown below. B, Summary of the effects of 500 μM gadolinium on sEPSC amplitude and frequency (n = 10 from 8 mice). As with mEPSC, sEPSC amplitude was mildly and its frequency was robustly enhanced by 500 μM gadolinium (amplitude, t(9) = −5.16, p = 0.0006; frequency, t(9) = −11.87, p = 0.00000085; paired t test). C, Average time course of gadolinium-induced LTD of evoked EPSCs. D, Normalized plots of 1/CV2 versus normalized amplitude of EPSCs before and 50 min after the start of gadolinium application. Mean values ± SEM are also shown. All hippocampal slices used in this figure were cut in sucrose-based ACSF. ***p < 0.001.

Figure 9.
Figure 9.

High concentration of gadolinium induces LTD of fEPSPs both in active and inactive pathways. Two stimulating electrodes were placed in the hippocampal slices and fEPSPs were recorded from two pathways using a recording electrode. Stimulation was continued during 500 μM gadolinium application in one pathway (active pathway) and ceased 40 min after the start of gadolinium application in another pathway (inactive pathway). High concentration of gadolinium (500 μM) induced LTD in both active and inactive pathways (n = 11 from 4 mice). All the hippocampal slices used in this figure were cut in normal ACSF.

Evoked EPSCs is affected by Gd-DTPA, a linear GBCA, but not by Gd-GOTA, a macrocyclic GBCA, at clinically relevant concentration

Finally, we examined whether clinically used GBCAs have side effects on synaptic transmission. The distribution of GBCAs and the time course of their signal intensity in rat brains after intravenous administration has been previously evaluated using MRI (Jost et al., 2017). When gadopentetate dimeglumine (Gd-DTPA, Magnevist), a linear GBCA, and gadoterate meglumine (Gd-GOTA, Magnescope), a macrocyclic GBCA, were administered at a dose of 1.8 mmol gadolinium/kg, the signal intensities in the rat brain ventricles reach maximum 10–20 min after administration, last for ∼20 min, and largely decreased 240 min after administration (Jost et al., 2017, their Fig. 4a). The concentration of gadolinium in the CSF 270 min after GBCA administration was ∼25 μM (Jost et al., 2017, their Fig. 5a). The maximal signal intensity in the ventricles 10–20 min after administration was approximately three times (gadoterate = Gd-GOTA) and five times (gadopentetate = Gd-DTPA) higher than that at 240 min (Jost et al., 2017, their Fig. 4a). Suppose gadolinium concentration in the CSF at 240 min after GBCA administration is the same as that at 270 min, gadolinium concentration could be 75–125 μM at 10–20 min after GBCA administration. Furthermore, 100 μM is a “clinically relevant” concentration because the estimated gadolinium concentration in the human brain tissue of autopsy patients who repeatedly received GBCA administration is 127 μM (Bower et al., 2019). Therefore, we investigated the effects on evoked and spontaneous EPSCs of GBCAs at a concentration of 100 μM and applied them for 20 min.

Initially, we applied 100 μM gadolinium to hippocampal slices cut in sucrose-based ACSF. Gadolinium (100 μM) significantly increased sEPSC frequency without affecting sEPSC amplitude 60 min after the start of gadolinium application, similar to its effects on mESPCs (Fig. 10A,B; amplitude, baseline 10.41 ± 0.49 pA, gadolinium, 11.30 ± 0.81 pA; frequency, baseline, 2.11 ± 0.50 Hz, gadolinium, 22.68 ± 5.33 Hz; n = 9 from 7 mice; t(8) = −3.92, p = 0.0044; paired t test). However, although 100 μM gadolinium potentiated fEPSPs (Fig. 1), it depressed evoked EPSCs, (Fig. 10C,D; 62.7 ± 12.5% of baseline 60 min after the start of gadolinium application). We analyzed the CV of evoked EPSC amplitude before and 60 min after the start of 100 μM gadolinium and CV−2 plots of cells in which EPSCs were depressed were approximately on the diagonal line (Fig. 10E). These results suggest that depression of EPSCs induced by 100 μM gadolinium was associated with a decrease in the number of active synapses (=N). The possibility that depression of evoked EPSCs was caused by the depletion of presynaptic glutamate due to frequent spontaneous synaptic transmission is less likely because the CV analysis did not suggest a decrease in release probability (p).

Figure 10.
Figure 10.

Potentiation of evoked EPSCs by low concentration of gadolinium is not replicated in hippocampal slices cut in sucrose-based ACSF. A, Example of sEPSCs before (1) and 60 min after 100 μM gadolinium application (2). Arrows indicate the time of afferent stimulation. B, Summary of the effects of 100 μM gadolinium on sEPSC amplitude and frequency (n = 9 from 7 mice). A low concentration of gadolinium significantly increased sEPSC frequency (t(8) = −3.92, p = 0.0044; paired t test). C, D, Example (C) and average time course (D) of evoked EPSC amplitude before and after 100 μM gadolinium application in hippocampal slices cut in sucrose-based ACSF. E, Normalized plots of 1/CV2 versus normalized amplitude of EPSCs before and 60 min after the start of gadolinium application. Mean values of all data except a potentiated cell were calculated and plotted together with SEM. All hippocampal slices used in this figure were cut in sucrose-based ACSF. **p < 0.01.

Next, we attempted to replicate low-concentration gadolinium-induced potentiation of EPSCs by using hippocampal slices prepared using the NMDG protective recovery method (Ting et al., 2018). We have previously reported robust presynaptic long-term potentiation of EPSPs in oriens lacunosum-moleculare (OLM) cells in adult mouse hippocampal slices cut using this method (Tomioka et al., 2014). Administration of gadolinium at 100 μM modestly elevated sEPSC frequency without affecting sEPSC amplitude 60 min after the start of gadolinium application in the hippocampal slices cut in NMDG-HEPES ACSF (Fig. 11A,B; amplitude, baseline, 8.83 ± 0.35 pA, gadolinium, 9.83 ± 0.72; n = 9 from 9 mice; frequency, 1.60 ± 0.25 Hz, gadolinium, 3.21 ± 0.63 Hz; t(8) = −2.66, p = 0.029; paired t test). When we applied Gd-DTPA (Magnevist), a linear GBCA or Gd-GOTA (Magnescope), a macrocyclic GBCA, at the same concentration, either GBCA did not change spontaneous synaptic transmission (Fig. 11C–F; Gd-DTPA, amplitude, baseline, 10.49 ± 0.76 pA, Gd-GTPA, 9.23 ± 0.33 pA; frequency, baseline, 1.54 ± 0.19 Hz, Gd-GTPA, 1.44 ± 0.16 Hz; n = 11 from 10 mice; Gd-GOTA, amplitude, baseline, 8.95 ± 0.37 pA, Gd-GOTA, 8.25 ± 0.40 pA; frequency, baseline, 2.10 ± 0.33 Hz, Gd-GOTA, 1.69 ± 0.19 Hz; n = 11 from 11 mice).

Figure 11.
Figure 11.

sEPSC frequency is modestly elevated by low concentration of gadolinium but not by GBCAs in hippocampal slices cut using the NMDG protective recovery method. A, sEPSCs before and 60 min after 100 μM gadolinium application. B, Summary of the effects of 100 μM gadolinium on sEPSC amplitude and frequency (n = 9 from 9 mice). A low concentration of gadolinium modestly increased mEPSC frequency in hippocampal slices cut in NMDG-HEPES ACSF (t(8) = −2.66, p = 0.029; paired t test). However, the increase is smaller than that in hippocampal slices cut in sucrose-based ACSF (Fig. 7B). C, sEPSCs before and 60 min after application of 100 μM Gd-DTPA (Magnevist), a linear GBCA. D, Summary of the effects of 100 μM Gd-DTPA on sEPSC amplitude and frequency (n = 11 from 10 mice). E, Example of sEPSCs before and 60 min after application of 100 μM Gd-GOTA (Magnescope), a macrocyclic GBCA. F, Summary of the effects of 100 μM Gd-GOTA on sEPSC amplitude and frequency (n = 11 from 11 mice). All hippocampal slices used in this figure were cut using NMDG protective recovery method. *p < 0.05.

Potentiation of EPSCs by a low concentration of gadolinium was not replicated in hippocampal slices cut in sucrose-based ACSF (Fig. 10A,C,D); however, 100 μM gadolinium robustly potentiated EPSCs in hippocampal slices prepared using the NMDG protective recovery method (Fig. 12A–C; 289.8 ± 36.0% of baseline 60 min after the start of gadolinium application; n = 9 from 9 mice). The CV−2 plots were on or above the diagonal line (Fig. 12D), suggesting that an increase in release probability (p) was associated with potentiation induced by a low concentration of gadolinium. We examined the effects of low concentration of GBCAs on EPSCs. Gd-DTPA (100 μM) weakly potentiated EPSCs (Fig. 12A–C; 177.6 ± 19.5%; n = 11 from 10 mice; F(2,28) = 16.96, p = 0.0062; ANOVA with Tukey–Kramer test). Similar to the data for gadolinium, CV−2 plots of the data for Gd-DTPA were also on or above the diagonal line (Fig. 11D), suggesting that DTPA did not completely chelate gadolinium and that it possibly unchelated gadolinium-induced potentiation of EPSCs. On the other hand, 100 μM Gd-GOTA did not potentiate ESPCs (Fig. 12A–C; 95.5 ± 13.3%; n = 11 from 11 mice; p = 0.0000087; ANOVA with Tukey–Kramer test), suggesting that GOTA fully chelated gadolinium.

Figure 12.
Figure 12.

Low concentration of gadolinium robustly potentiates evoked EPSCs and potentiating effects of gadolinium are partially and fully inhibited in linear and macrocyclic GBCAs, respectively, in hippocampal slices cut using the NMDG protective recovery method. A, B, Example (A) and average time courses (B) of evoked EPSC amplitude before and after application of 100 μM gadolinium, Gd-DTPA, and Gd-GOTA. C, Summary of the effects of 100 μM gadolinium (n = 9 from 9 mice), Gd-DTPA (n = 11 from 10 mice), and Gd-GOTA (n = 11 from 11 mice) on evoked EPSC amplitude. Low concentration of gadolinium (100 μM) robustly potentiated evoked EPSCs in hippocampal slices cut in NMDG-HEPES ACSF. Potentiation remained partially uninhibited when Gd-GTPA was applied (F(2,28) = 16.96, p = 0.0062; ANOVA with Tukey–Kramer test); however, it was almost completely inhibited when Gd-GOTA was applied (p = 0.0000087; ANOVA with Tukey–Kramer test). D, Normalized plots of 1/CV2 versus normalized amplitude of EPSCs before and 60 min after the start of gadolinium or Gd-DTPA application. Mean values ± SEM are also shown. All hippocampal slices used in this figure were cut using NMDG protective recovery method. *p < 0.05, **p < 0.01, ***p < 0.001.

Discussion

Gadolinium at low concentrations (5–50 μM) increases frequency of mEPSCs in cultured hippocampal neurons (Capogna et al., 1996; Lei and MacDonald, 2001), and we verified the concentration-dependent increase in m/sEPSC frequency induced by gadolinium. In contrast, the evoked fEPSP amplitude was slightly enhanced by a low concentration of gadolinium (100 μM), although the same concentration of gadolinium robustly elevated mEPSC frequency. Furthermore, the present study is the first to show that higher concentrations of gadolinium (500–1,000 μM) induce group 1 mGluR/eCB/P2Y1R-dependent, presynaptically expressed, and partially P2Y13R-dependent, postsynaptically expressed LTD. Finally, we examined adverse effects of GBCAs on synaptic transmission in slices prepared using the NMDG protective recovery method, in which free gadolinium at a clinically relevant concentration (100 μM; Bower et al., 2019), induced robust potentiation of EPSCs. We found that 100 μM Gd-GOTA did not affects EPSCs, but 100 μM Gd-DTPA caused modest potentiation, suggesting that gadolinium may not be completely chelated in Gd-GPTA. The effects of free gadolinium on synaptic transmission are summarized in Table 1.

View this table:
Table 1.

Summary of the effects of low and high concentrations of gadolinium on evoked and spontaneous synaptic transmission

Differential modulation of evoked and spontaneous synaptic transmission by gadolinium

We found that the generation of 2-AG and activation of CB1R were responsible for approximately half of the higher-concentration gadolinium-induced LTD (Fig. 3). eCB is released by group 1 mGuR activation, which activates CB1R and depresses presynaptic glutamate release at excitatory synapses (Huang et al., 2008; Yasuda et al., 2008; Peterfi et al., 2012; Araque et al., 2017). CB1R-dependent LTD is associated with a decrease in the frequency of mEPSCs and sEPSCs without a decrease in their amplitude (Misner and Sullivan, 1999; Azad et al., 2003; Kellogg et al., 2009). However, 500 μM gadolinium elevated both the amplitude and frequency of mEPSCs (Fig. 7) and sEPSCs (Fig. 8), although it induced LTD of the evoked EPSCs (Fig. 8). These results suggest that gadolinium differently modulates evoked and miniature/spontaneous synaptic transmission. Presynaptic CB1Rs decrease transmitter release by inhibiting presynaptic voltage-dependent calcium channels, activating potassium channels, and inhibiting the cyclic AMP/protein kinase A signaling and release machinery (Azad et al., 2003; Yasuda et al., 2008; Castillo et al., 2012; Araque et al., 2017). In contrast, gadolinium induces calcium-independent exocytosis of synaptic vesicles, presumably by directly activating the release machinery (Molgo et al., 1991; Waseem et al., 2008), and elevates miniature/spontaneous EPSC frequency. Gadolinium and CB1Rs may act on different molecules at presynaptic sites, and the direction of changes in evoked and spontaneous EPSCs may differ. Spontaneous and evoked release occur in molecularly heterogeneous active zones with different vesicle pools within the same synapse, and individual active zones are independently regulated (Melom et al., 2013; Walter et al., 2014; Kavalali, 2015; Guzikowski and Kavalali, 2021). Additionally, some synapses preferentially induce evoked or spontaneous release (Peled et al., 2014; Walter et al., 2014; Kavalali, 2015). Therefore, gadolinium and eCB-CB1R may affect different active zones and synapses.

P2Rs are required for both high-concentration gadolinium-induced presynaptically and postsynaptically expressed LTD

Gadolinium (1,000 μM) induced mGluR/eCB/P2Y1R-dependent, presynaptically expressed, and partially P2Y13R-dependent, postsynaptically expressed LTD. MRS 2179, a P2Y1R inhibitor, robustly reduced the amplitude of higher concentrations of gadolinium-induced LTD by ∼50% of baseline and reduced the PPR, whereas MRS 2211, a P2Y13R inhibitor, only slightly reduced LTD and did not affect the PPR (Fig. 6D–G). In contrast, the amplitude of mGluR/eCB/P2YR-dependent, presynaptically expressed LTD induced by a higher concentration of gadolinium was ∼20–25% of the baseline, and the amplitude of P2Y-dependent, postsynaptically expressed LTD was ∼25% of the baseline (Fig. 3G,H). Therefore, MRS 2179 may also inhibit LTD which is not associated with changes in PPRs, and P2Y1R may also be required to induce postsynaptically expressed LTD.

Gadolinium directly activates mGluRs (Kubo et al., 1998); however, it is not clear whether gadolinium directly activates P2YRs, which are also G-protein-coupled, seven-transmembrane receptors similar to mGluRs, or induces ATP release from astrocytes. However, LTD induced by WIN55212-2 was inhibited by PPADS (Fig. 5). Therefore, CBRs could induce the release of ATP in the hippocampus. In contrast to CB1Rs on presynaptic terminals, which are coupled to Gi-proteins, CB1Rs on astrocytes are coupled to Gq-proteins, which activate astrocytic CB1Rs, mobilize intracellular calcium, and release ATP (Rasooli-Nejad et al., 2014; Baraibar et al., 2023; Eraso-Pichot et al., 2023). Because P2Y1R is robustly expressed in astrocytes and elevates the intracellular calcium concentration through IP3 receptors, P2Y1R release ATP and eCB from astrocytes (Eraso-Pichot et al., 2023) and induce presynaptically expressed, input-nonspecific LTD (Chen et al., 2013).

In contrast, P2Y13R is expressed in the microglia (Fields and Burnstock, 2006; Zarrinmayeh and Territo, 2020). Microglia are involved in some forms of LTD (Wu et al., 2015; Innes et al., 2019), and microglial complement receptor 3 (CR3) induces NMDAR- and mGluR-independent postsynaptically expressed LTD (Zhang et al., 2014). Higher concentrations of gadolinium also induced NMDAR- and mGluR-independent LTD without changes in PPRs, and this action is dependent on P2YRs, including P2Y13R (Figs. 3, 6); therefore, these could be the same type of LTD. LTD is an activity-dependent reduction in synaptic transmission and an essential step in synaptic pruning; microglia are involved in synaptic pruning as well as in some forms of LTD (Innes et al., 2019). However, the role of microglial P2YRs in LTD and synaptic pruning remains unclear. For example, microglial processes also involve P2Y12R (Fields and Burnstock, 2006; Zarrinmayeh and Territo, 2020), which could be important in these phenomena; however, further investigation is required.

Gd-GTPA at clinically relevant concentrations modestly potentiates synaptic transmission

Low concentrations of gadolinium potentiated evoked fEPSPs in hippocampal slices cut in normal ACSF (Fig. 1); however, it depressed evoked EPSCs in hippocampal slices cut in sucrose-based ACSF (Fig. 10). CV analysis suggested that 100 μM gadolinium silenced active synapses (Fig. 10E), although the underlying mechanism is unknown. Low gadolinium concentration also largely potentiated evoked EPSCs in the hippocampal slices cut using the NMDG protective recovery method (Fig. 12). Previously, we reported robust LTP of EPSPs associated with an increase in presynaptic release probability in inhibitory OLM neurons in hippocampal slices prepared using the NMDG protective recovery method (Tomioka et al., 2014). We recorded EPSPs in OLM neurons in the perforated patch configuration using gramicidin to avoid LTP washout, and we recorded EPSCs in the whole-cell configuration in the present study. However, we speculate that presynaptic potentiation may be well preserved in slices prepared using the NMDG protective recovery method for unknown reasons.

After intravenous administration of GBCAs, gadolinium is deposited in the brains of humans and rodents (Ray et al., 1996; McDonald et al., 2015), and the extent of gadolinium deposition depends on administration dose and time (Kanda et al., 2014; Robert et al., 2015). We attempted to estimate the side effects of GBCAs on synaptic transmission using hippocampal slice preparations. We chose 100 μM GBCA for 20 min application because the gadolinium concentration at the maximal plateau of gadolinium signal intensities in the ventricles of the rat brain is close to 100 μM and the plateau continues for ∼20 min after intravenous injection of GBCAs at 1.8 mmol Gd/kg (Jost et al., 2017, their Fig. 4a). Moreover, 100 μM can be considered a “clinically relevant” concentration, as this is very close to that in the brains of autopsy patients who repeatedly received GBCA administration (Bower et al., 2019). Gd-GOTA (100 μM) did not affect evoked EPSC amplitude; however, 100 μM Gd-GTPA partially induce potentiation of evoked EPSCs (Fig. 12). GOTA is a macrocyclic gadolinium chelator with a logarithm of the conditional thermodynamic stability constant (Kcond) at 19.3 and high (=slow) kinetic stability (Idee et al., 2009). Kcond describes the equilibrium between gadolinium and the chelator at physiological pH, and kinetic stability describes the speed at which gadolinium is released from GBCAs. On the other hand, GTPA is a linear gadolinium chelator with a log Kcond of 17.7 and lower (=faster) kinetic stability (Idee et al., 2009), indicating that Gd-GTPA has lower affinity for gadolinium and is less stable than Gd-GOTA. Free gadolinium ions were more abundant when the same concentration of Gd-GTPA was applied, which caused potentiation of the evoked EPSCs (Fig. 12). Linear GBCAs have lower kinetic stabilities and reduce the viability and mitochondrial function of human neuronal cell lines more severely than macrocyclic GBCAs, which have higher kinetic stabilities (Bower et al., 2019). Frequent administration of gadodiamide (Omniscan), a linear GBCA, causes binding of gadolinium with perineuronal net protein aggrecan in the deep cerebellar nucleus of mice, although gadobutrol (Gadovist), a macrocyclic GBCA, does not cause gadolinium binding (Habermeyer et al., 2020). Gadolinium is persistently retained in the cerebellum of a neuroinflammation mouse model after linear Gd-GTPA administration; however, gadolinium is cleared in 40 d after a macrocyclic GBCA administration (Anderhalten et al., 2022). Furthermore, the administration to perinatal mice of GBCAs, especially a linear GBCA, resulted in elevated anxiety, impaired motor skills, and disrupted memory in adults (Khairinisa et al., 2018). Therefore, we had better avoid clinical use of linear GBCAs.

Data Availability

The data generated in this study are available from the corresponding author upon request.

Footnotes

  • This work is supported by JSPS KAKENHI Grant Number JP26430003 to H.Y. and Shimadzu Grants (#459) from Shimadzu Science Foundation to T.N. We thank Erika Shinchi (Analytical Research Center for Experimental Sciences, Saga University) for technical support on inductively coupled plasma-optical emission spectroscopy, which is supported by the MEXT Project for promoting public utilization of advanced research infrastructure (JPMXS0422400025).

  • ↵*O.Z. and H.Y. are the co-first authors.

  • The authors declare no competing financial interests.

  • Correspondence should be addressed to Hiroki Yasuda at yasuda{at}cc.saga-u.ac.jp.

References

  1. ↵
    1. Anderhalten L, et al.
    (2022) Different impact of gadopentetate and gadobutrol on inflammation-promoted retention and toxicity of gadolinium within the mouse brain. Invest Radiol 57:677–688. https://doi.org/10.1097/RLI.0000000000000884 pmid:35467573
    OpenUrlCrossRefPubMed
  2. ↵
    1. Araque A,
    2. Castillo PE,
    3. Manzoni OJ,
    4. Tonini R
    (2017) Synaptic functions of endocannabinoid signaling in health and disease. Neuropharmacology 124:13–24. https://doi.org/10.1016/j.neuropharm.2017.06.017 pmid:28625718
    OpenUrlCrossRefPubMed
  3. ↵
    1. Azad SC,
    2. Eder M,
    3. Marsicano G,
    4. Lutz B,
    5. Zieglgansberger W,
    6. Rammes G
    (2003) Activation of the cannabinoid receptor type 1 decreases glutamatergic and GABAergic synaptic transmission in the lateral amygdala of the mouse. Learn Mem 10:116–128. https://doi.org/10.1101/lm.53303 pmid:12663750
    OpenUrlAbstract/FREE Full Text
  4. ↵
    1. Baraibar AM,
    2. Belisle L,
    3. Marsicano G,
    4. Matute C,
    5. Mato S,
    6. Araque A,
    7. Kofuji P
    (2023) Spatial organization of neuron-astrocyte interactions in the somatosensory cortex. Cereb Cortex 33:4498–4511. https://doi.org/10.1093/cercor/bhac357 pmid:36124663
    OpenUrlCrossRefPubMed
  5. ↵
    1. Bower DV,
    2. Richter JK,
    3. von Tengg-Kobligk H,
    4. Heverhagen JT,
    5. Runge VM
    (2019) Gadolinium-based MRI contrast agents induce mitochondrial toxicity and cell death in human neurons, and toxicity increases with reduced kinetic stability of the agent. Invest Radiol 54:453–463. https://doi.org/10.1097/RLI.0000000000000567
    OpenUrlCrossRefPubMed
  6. ↵
    1. Capogna M,
    2. Gahwiler BH,
    3. Thompson SM
    (1996) Presynaptic inhibition of calcium-dependent and -independent release elicited with ionomycin, gadolinium, and alpha-latrotoxin in the hippocampus. J Neurophysiol 75:2017–2028. https://doi.org/10.1152/jn.1996.75.5.2017
    OpenUrlCrossRefPubMed
  7. ↵
    1. Castillo PE,
    2. Younts TJ,
    3. Chavez AE,
    4. Hashimotodani Y
    (2012) Endocannabinoid signaling and synaptic function. Neuron 76:70–81. https://doi.org/10.1016/j.neuron.2012.09.020 pmid:23040807
    OpenUrlCrossRefPubMed
  8. ↵
    1. Chen J, et al.
    (2013) Heterosynaptic long-term depression mediated by ATP released from astrocytes. Glia 61:178–191. https://doi.org/10.1002/glia.22425
    OpenUrlCrossRefPubMed
  9. ↵
    1. Eraso-Pichot A,
    2. Pouvreau S,
    3. Olivera-Pinto A,
    4. Gomez-Sotres P,
    5. Skupio U,
    6. Marsicano G
    (2023) Endocannabinoid signaling in astrocytes. Glia 71:44–59. https://doi.org/10.1002/glia.24246 pmid:35822691
    OpenUrlCrossRefPubMed
  10. ↵
    1. Ferraguti F,
    2. Shigemoto R
    (2006) Metabotropic glutamate receptors. Cell Tissue Res 326:483–504. https://doi.org/10.1007/s00441-006-0266-5
    OpenUrlCrossRefPubMed
  11. ↵
    1. Fields RD,
    2. Burnstock G
    (2006) Purinergic signalling in neuron-glia interactions. Nat Rev Neurosci 7:423–436. https://doi.org/10.1038/nrn1928 pmid:16715052
    OpenUrlCrossRefPubMed
  12. ↵
    1. Fitzjohn SM,
    2. Kingston AE,
    3. Lodge D,
    4. Collingridge GL
    (1999) DHPG-induced LTD in area CA1 of juvenile rat hippocampus; characterisation and sensitivity to novel mGlu receptor antagonists. Neuropharmacology 38:1577–1583. https://doi.org/10.1016/S0028-3908(99)00123-9
    OpenUrlCrossRefPubMed
  13. ↵
    1. Fitzjohn SM,
    2. Palmer MJ,
    3. May JE,
    4. Neeson A,
    5. Morris SA,
    6. Collingridge GL
    (2001) A characterisation of long-term depression induced by metabotropic glutamate receptor activation in the rat hippocampus in vitro. J Physiol 537:421–430. https://doi.org/10.1111/j.1469-7793.2001.00421.x pmid:11731575
    OpenUrlCrossRefPubMed
  14. ↵
    1. Gasselin C,
    2. Inglebert Y,
    3. Ankri N,
    4. Debanne D
    (2017) Plasticity of intrinsic excitability during LTD is mediated by bidirectional changes in h-channel activity. Sci Rep 7:14418. https://doi.org/10.1038/s41598-017-14874-z pmid:29089586
    OpenUrlCrossRefPubMed
  15. ↵
    1. Gil-Sanz C,
    2. Delgado-Garcia JM,
    3. Fairen A,
    4. Gruart A
    (2008) Involvement of the mGluR1 receptor in hippocampal synaptic plasticity and associative learning in behaving mice. Cereb Cortex 18:1653–1663. https://doi.org/10.1093/cercor/bhm193
    OpenUrlCrossRefPubMed
  16. ↵
    1. Guzikowski NJ,
    2. Kavalali ET
    (2021) Nano-organization at the synapse: segregation of distinct forms of neurotransmission. Front Synaptic Neurosci 13:796498. https://doi.org/10.3389/fnsyn.2021.796498 pmid:35002671
    OpenUrlCrossRefPubMed
  17. ↵
    1. Habermeyer J, et al.
    (2020) Comprehensive phenotyping revealed transient startle response reduction and histopathological gadolinium localization to perineuronal nets after gadodiamide administration in rats. Sci Rep 10:22385. https://doi.org/10.1038/s41598-020-79374-z pmid:33372182
    OpenUrlCrossRefPubMed
  18. ↵
    1. Huang CC,
    2. You JL,
    3. Wu MY,
    4. Hsu KS
    (2004) Rap1-induced p38 mitogen-activated protein kinase activation facilitates AMPA receptor trafficking via the GDI.Rab5 complex. Potential role in (S)-3,5-dihydroxyphenylglycene-induced long term depression. J Biol Chem 279:12286–12292. https://doi.org/10.1074/jbc.M312868200
    OpenUrlAbstract/FREE Full Text
  19. ↵
    1. Huang Y,
    2. Yasuda H,
    3. Sarihi A,
    4. Tsumoto T
    (2008) Roles of endocannabinoids in heterosynaptic long-term depression of excitatory synaptic transmission in visual cortex of young mice. J Neurosci 28:7074–7083. https://doi.org/10.1523/JNEUROSCI.0899-08.2008 pmid:18614676
    OpenUrlAbstract/FREE Full Text
  20. ↵
    1. Idee JM,
    2. Port M,
    3. Robic C,
    4. Medina C,
    5. Sabatou M,
    6. Corot C
    (2009) Role of thermodynamic and kinetic parameters in gadolinium chelate stability. J Magn Reson Imaging 30:1249–1258. https://doi.org/10.1002/jmri.21967
    OpenUrlCrossRefPubMed
  21. ↵
    1. Innes S,
    2. Pariante CM,
    3. Borsini A
    (2019) Microglial-driven changes in synaptic plasticity: a possible role in major depressive disorder. Psychoneuroendocrinology 102:236–247. https://doi.org/10.1016/j.psyneuen.2018.12.233
    OpenUrlCrossRefPubMed
  22. ↵
    1. Jost G,
    2. Frenzel T,
    3. Lohrke J,
    4. Lenhard DC,
    5. Naganawa S,
    6. Pietsch H
    (2017) Penetration and distribution of gadolinium-based contrast agents into the cerebrospinal fluid in healthy rats: a potential pathway of entry into the brain tissue. Eur Radiol 27:2877–2885. https://doi.org/10.1007/s00330-016-4654-2 pmid:27832312
    OpenUrlCrossRefPubMed
  23. ↵
    1. Kanda Y
    (2013) Investigation of the freely available easy-to-use software ‘EZR’ for medical statistics. Bone Marrow Transplant 48:452–458. https://doi.org/10.1038/bmt.2012.244 pmid:23208313
    OpenUrlCrossRefPubMed
  24. ↵
    1. Kanda T,
    2. Ishii K,
    3. Kawaguchi H,
    4. Kitajima K,
    5. Takenaka D
    (2014) High signal intensity in the dentate nucleus and globus pallidus on unenhanced T1-weighted MR images: relationship with increasing cumulative dose of a gadolinium-based contrast material. Radiology 270:834–841. https://doi.org/10.1148/radiol.13131669
    OpenUrlCrossRefPubMed
  25. ↵
    1. Kartamihardja AA,
    2. Nakajima T,
    3. Kameo S,
    4. Koyama H,
    5. Tsushima Y
    (2016) Distribution and clearance of retained gadolinium in the brain: differences between linear and macrocyclic gadolinium based contrast agents in a mouse model. Br J Radiol 89:20160509. https://doi.org/10.1259/bjr.20160509 pmid:27459250
    OpenUrlCrossRefPubMed
  26. ↵
    1. Katona I,
    2. Freund TF
    (2012) Multiple functions of endocannabinoid signaling in the brain. Annu Rev Neurosci 35:529–558. https://doi.org/10.1146/annurev-neuro-062111-150420 pmid:22524785
    OpenUrlCrossRefPubMed
  27. ↵
    1. Kavalali ET
    (2015) The mechanisms and functions of spontaneous neurotransmitter release. Nat Rev Neurosci 16:5–16. https://doi.org/10.1038/nrn3875
    OpenUrlCrossRefPubMed
  28. ↵
    1. Kawate T,
    2. Michel JC,
    3. Birdsong WT,
    4. Gouaux E
    (2009) Crystal structure of the ATP-gated P2X(4) ion channel in the closed state. Nature 460:592–598. https://doi.org/10.1038/nature08198 pmid:19641588
    OpenUrlCrossRefPubMed
  29. ↵
    1. Kellogg R,
    2. Mackie K,
    3. Straiker A
    (2009) Cannabinoid CB1 receptor-dependent long-term depression in autaptic excitatory neurons. J Neurophysiol 102:1160–1171. https://doi.org/10.1152/jn.00266.2009 pmid:19494194
    OpenUrlCrossRefPubMed
  30. ↵
    1. Khairinisa MA,
    2. Takatsuru Y,
    3. Amano I,
    4. Erdene K,
    5. Nakajima T,
    6. Kameo S,
    7. Koyama H,
    8. Tsushima Y,
    9. Koibuchi N
    (2018) The effect of perinatal gadolinium-based contrast agents on adult mice behavior. Invest Radiol 53:110–118. https://doi.org/10.1097/RLI.0000000000000417
    OpenUrlCrossRefPubMed
  31. ↵
    1. Kubo Y,
    2. Miyashita T,
    3. Murata Y
    (1998) Structural basis for a Ca2+-sensing function of the metabotropic glutamate receptors. Science 279:1722–1725. https://doi.org/10.1126/science.279.5357.1722
    OpenUrlAbstract/FREE Full Text
  32. ↵
    1. Kumar A,
    2. Foster TC
    (2007) Shift in induction mechanisms underlies an age-dependent increase in DHPG-induced synaptic depression at CA3 CA1 synapses. J Neurophysiol 98:2729–2736. https://doi.org/10.1152/jn.00514.2007
    OpenUrlCrossRefPubMed
  33. ↵
    1. Lee JH,
    2. Chiba T,
    3. Marcus DC
    (2001) P2x2 receptor mediates stimulation of parasensory cation absorption by cochlear outer sulcus cells and vestibular transitional cells. J Neurosci 21:9168–9174. https://doi.org/10.1523/JNEUROSCI.21-23-09168.2001 pmid:11717350
    OpenUrlAbstract/FREE Full Text
  34. ↵
    1. Lei S,
    2. MacDonald JF
    (2001) Gadolinium reduces AMPA receptor desensitization and deactivation in hippocampal neurons. J Neurophysiol 86:173–182. https://doi.org/10.1152/jn.2001.86.1.173
    OpenUrlCrossRefPubMed
  35. ↵
    1. Luscher C,
    2. Huber KM
    (2010) Group 1 mGluR-dependent synaptic long-term depression: mechanisms and implications for circuitry and disease. Neuron 65:445–459. https://doi.org/10.1016/j.neuron.2010.01.016 pmid:20188650
    OpenUrlCrossRefPubMed
  36. ↵
    1. McDonald RJ,
    2. McDonald JS,
    3. Kallmes DF,
    4. Jentoft ME,
    5. Murray DL,
    6. Thielen KR,
    7. Williamson EE,
    8. Eckel LJ
    (2015) Intracranial gadolinium deposition after contrast-enhanced MR imaging. Radiology 275:772–782. https://doi.org/10.1148/radiol.15150025
    OpenUrlCrossRefPubMed
  37. ↵
    1. Melom JE,
    2. Akbergenova Y,
    3. Gavornik JP,
    4. Littleton JT
    (2013) Spontaneous and evoked release are independently regulated at individual active zones. J Neurosci 33:17253–17263. https://doi.org/10.1523/JNEUROSCI.3334-13.2013 pmid:24174659
    OpenUrlAbstract/FREE Full Text
  38. ↵
    1. Misner DL,
    2. Sullivan JM
    (1999) Mechanism of cannabinoid effects on long-term potentiation and depression in hippocampal CA1 neurons. J Neurosci 19:6795–6805. https://doi.org/10.1523/JNEUROSCI.19-16-06795.1999 pmid:10436037
    OpenUrlAbstract/FREE Full Text
  39. ↵
    1. Molgo J,
    2. del Pozo E,
    3. Banos JE,
    4. Angaut-Petit D
    (1991) Changes of quantal transmitter release caused by gadolinium ions at the frog neuromuscular junction. Br J Pharmacol 104:133–138. https://doi.org/10.1111/j.1476-5381.1991.tb12397.x pmid:1686201
    OpenUrlCrossRefPubMed
  40. ↵
    1. Peled ES,
    2. Newman ZL,
    3. Isacoff EY
    (2014) Evoked and spontaneous transmission favored by distinct sets of synapses. Curr Biol 24:484–493. https://doi.org/10.1016/j.cub.2014.01.022 pmid:24560571
    OpenUrlCrossRefPubMed
  41. ↵
    1. Peterfi Z, et al.
    (2012) Endocannabinoid-mediated long-term depression of afferent excitatory synapses in hippocampal pyramidal cells and GABAergic interneurons. J Neurosci 32:14448–14463. https://doi.org/10.1523/JNEUROSCI.1676-12.2012 pmid:23055515
    OpenUrlAbstract/FREE Full Text
  42. ↵
    1. Pougnet JT,
    2. Toulme E,
    3. Martinez A,
    4. Choquet D,
    5. Hosy E,
    6. Boue-Grabot E
    (2014) ATP P2X receptors downregulate AMPA receptor trafficking and postsynaptic efficacy in hippocampal neurons. Neuron 83:417–430. https://doi.org/10.1016/j.neuron.2014.06.005
    OpenUrlCrossRefPubMed
  43. ↵
    1. Ralevic V,
    2. Burnstock G
    (1998) Receptors for purines and pyrimidines. Pharmacol Rev 50:413–492. https://doi.org/10.1016/S0031-6997(24)01373-5
    OpenUrlAbstract/FREE Full Text
  44. ↵
    1. Rasooli-Nejad S,
    2. Palygin O,
    3. Lalo U,
    4. Pankratov Y
    (2014) Cannabinoid receptors contribute to astroglial Ca(2)(+)-signalling and control of synaptic plasticity in the neocortex. Philos Trans R Soc Lond B Biol Sci 369:20140077. https://doi.org/10.1098/rstb.2014.0077 pmid:25225106
    OpenUrlCrossRefPubMed
  45. ↵
    1. Ray DE,
    2. Cavanagh JB,
    3. Nolan CC,
    4. Williams SC
    (1996) Neurotoxic effects of gadopentetate dimeglumine: behavioral disturbance and morphology after intracerebroventricular injection in rats. Am J Neuroradiol 17:365–373.
    OpenUrlPubMed
  46. ↵
    1. Robert P,
    2. Lehericy S,
    3. Grand S,
    4. Violas X,
    5. Fretellier N,
    6. Idee JM,
    7. Ballet S,
    8. Corot C
    (2015) T1-weighted hypersignal in the deep cerebellar nuclei after repeated administrations of gadolinium-based contrast agents in healthy rats difference between linear and macrocyclic agents. Invest Radiol 50:473–480. https://doi.org/10.1097/RLI.0000000000000181 pmid:26107651
    OpenUrlCrossRefPubMed
  47. ↵
    1. Rubio ME,
    2. Soto F
    (2001) Distinct localization of P2X receptors at excitatory postsynaptic specializations. J Neurosci 21:641–653. https://doi.org/10.1523/JNEUROSCI.21-02-00641.2001 pmid:11160443
    OpenUrlAbstract/FREE Full Text
  48. ↵
    1. Sanderson TM,
    2. Ralph LT,
    3. Amici M,
    4. Ng AN,
    5. Kaang BK,
    6. Zhuo M,
    7. Kim SJ,
    8. Georgiou J,
    9. Collingridge GL
    (2022) Selective recruitment of presynaptic and postsynaptic forms of mGluR-LTD. Front Synaptic Neurosci 14:857675. https://doi.org/10.3389/fnsyn.2022.857675 pmid:35615440
    OpenUrlCrossRefPubMed
  49. ↵
    1. Stefani J,
    2. Tschesnokowa O,
    3. Parrilla M,
    4. Robaye B,
    5. Boeynaems JM,
    6. Acker-Palmer A,
    7. Zimmermann H,
    8. Gampe K
    (2018) Disruption of the microglial ADP receptor P2Y(13) enhances adult hippocampal neurogenesis. Front Cell Neurosci 12:134. https://doi.org/10.3389/fncel.2018.00134 pmid:29867367
    OpenUrlCrossRefPubMed
  50. ↵
    1. Taoka T,
    2. Naganawa S
    (2018) Gadolinium-based contrast media, cerebrospinal fluid and the glymphatic system: possible mechanisms for the deposition of gadolinium in the brain. Magn Reson Med Sci 17:111–119. https://doi.org/10.2463/mrms.rev.2017-0116 pmid:29367513
    OpenUrlCrossRefPubMed
  51. ↵
    1. Ting JT,
    2. Lee BR,
    3. Chong P,
    4. Soler-Llavina G,
    5. Cobbs C,
    6. Koch C,
    7. Zeng H,
    8. Lein E
    (2018) Preparation of acute brain slices using an optimized N-methyl-D-glucamine protective recovery method. J Vis Exp 132:53825. https://doi.org/10.3791/53825 pmid:29553547
    OpenUrlCrossRefPubMed
  52. ↵
    1. Tomioka NH, et al.
    (2014) Elfn1 recruits presynaptic mGluR7 in trans and its loss results in seizures. Nat Commun 5:4501. https://doi.org/10.1038/ncomms5501
    OpenUrlCrossRefPubMed
  53. ↵
    1. Volk LJ,
    2. Daly CA,
    3. Huber KM
    (2006) Differential roles for group 1 mGluR subtypes in induction and expression of chemically induced hippocampal long-term depression. J Neurophysiol 95:2427–2438. https://doi.org/10.1152/jn.00383.2005
    OpenUrlCrossRefPubMed
  54. ↵
    1. von Kugelgen I,
    2. Hoffmann K
    (2016) Pharmacology and structure of P2Y receptors. Neuropharmacology 104:50–61. https://doi.org/10.1016/j.neuropharm.2015.10.030
    OpenUrlCrossRefPubMed
  55. ↵
    1. Walter AM,
    2. Haucke V,
    3. Sigrist SJ
    (2014) Neurotransmission: spontaneous and evoked release filing for divorce. Curr Biol 24:R192–R194. https://doi.org/10.1016/j.cub.2014.01.037
    OpenUrlCrossRefPubMed
  56. ↵
    1. Waseem TV,
    2. Lapatsina LP,
    3. Fedorovich SV
    (2008) Influence of integrin-blocking peptide on gadolinium- and hypertonic shrinking-induced neurotransmitter release in rat brain synaptosomes. Neurochem Res 33:1316–1324. https://doi.org/10.1007/s11064-007-9585-5
    OpenUrlCrossRefPubMed
  57. ↵
    1. Wu Y,
    2. Dissing-Olesen L,
    3. MacVicar BA,
    4. Stevens B
    (2015) Microglia: dynamic mediators of synapse development and plasticity. Trends Immunol 36:605–613. https://doi.org/10.1016/j.it.2015.08.008 pmid:26431938
    OpenUrlCrossRefPubMed
  58. ↵
    1. Yasuda H,
    2. Mukai H
    (2015) Turning off of GluN2B subunits and turning on of CICR in hippocampal LTD induction after developmental GluN2 subunit switch. Hippocampus 25:1274–1284. https://doi.org/10.1002/hipo.22435
    OpenUrlCrossRefPubMed
  59. ↵
    1. Yasuda H,
    2. Huang Y,
    3. Tsumoto T
    (2008) Regulation of excitability and plasticity by endocannabinoids and PKA in developing hippocampus. Proc Natl Acad Sci U S A 105:3106–3111. https://doi.org/10.1073/pnas.0708349105 pmid:18287074
    OpenUrlAbstract/FREE Full Text
  60. ↵
    1. Yasuda H, et al.
    (2020) PKN1 promotes synapse maturation by inhibiting mGluR-dependent silencing through neuronal glutamate transporter activation. Commun Biol 3:710. https://doi.org/10.1038/s42003-020-01435-w pmid:33244074
    OpenUrlCrossRefPubMed
  61. ↵
    1. Zarrinmayeh H,
    2. Territo PR
    (2020) Purinergic receptors of the central nervous system: biology, PET ligands, and their applications. Mol Imaging 19:1536012120927609. https://doi.org/10.1177/1536012120927609 pmid:32539522
    OpenUrlCrossRefPubMed
  62. ↵
    1. Zhang J,
    2. Malik A,
    3. Choi HB,
    4. Ko RW,
    5. Dissing-Olesen L,
    6. MacVicar BA
    (2014) Microglial CR3 activation triggers long-term synaptic depression in the hippocampus via NADPH oxidase. Neuron 82:195–207. https://doi.org/10.1016/j.neuron.2014.01.043
    OpenUrlCrossRefPubMed
Back to top

In this issue

Email
Print
View Full Page PDF
Citation Tools
Respond to this article
Concentration-Dependent Bidirectional Modification of Evoked Synaptic Transmission by Gadolinium and Adverse Effects of Gadolinium-Based Contrast Agent
Odgerel Zorigt, Hiroki Yasuda, Takahito Nakajima, Yoshito Tsushima
Journal of Neuroscience 23 April 2025, 45 (17) e1622242025; DOI: 10.1523/JNEUROSCI.1622-24.2025
Twitter logo Facebook logo Mendeley logo

Jump to section

Keywords

Responses to this article

Respond to this article

Jump to comment:

No eLetters have been published for this article.