Epilepsy is a neurological disorder characterized by recurrent, unprovoked seizures resulting from an imbalance between excitatory and inhibitory neural systems that afflicts ~1% of the general population (1, 2). Temporal lobe epilepsy (TLE) is characterized by an epileptogenic focus within the temporal lobe of the brain, most commonly within the mesial temporal structures, and more frequently so within the hippocampus (3). Seizures originating from the mesial temporal structures in mesial TLE (mTLE) present a high rate of intractability, for which suitable patients can be offered surgical treatment (4, 5).

Metabotropic glutamate receptor type 5 (mGluR5) is a G-protein coupled receptor expressed as a dimer primarily in the periphery of the post-synaptic terminal (6, 7). Each protomer includes a large extracellular Venus flytrap domain that contains the orthosteric binding site for glutamate. When glutamate binds to the receptor, a conformational change is induced that activates the G protein, initiating multiple signaling cascades, including several involved in gene regulation, enzymatic function, and release of intracellular Ca²⁺ stores (6, 7). mGluR5 has been shown to mediate neuronal excitability and promote synaptic plasticity under normal physiological conditions (8). In particular, the receptor is involved in the establishment of long-term potentiation and depression (9, 10).

Various lines of evidence have pointed to mGluR5 as a protein of interest in the study of epileptogenesis. The receptor has been implicated in the generation of ictal events and persistent neuronal hyperexcitability (11, 12). Furthermore, abnormal expression of mGluR5 has been identified in rat models of the disorder and in mTLE patients, through in vivo and ex vivo studies (13–19). Immunohistochemistry (IHC) analyses in mTLE surgical specimens have demonstrated mGluR5 upregulation in the hippocampus, independent of a neuropathological diagnosis of hippocampal sclerosis (13–15). In the pilocarpine rat model of mTLE, IHC and immunofluorescence data revealed decreased expression of hippocampal mGluR5 following status epilepticus, which was associated with a loss of mGluR5-dependent long-term depression (16, 17). In an amygdala-kindling rat model of TLE, reduced mGluR5 mRNA expression was detected in the hippocampus 24 h after the last kindled seizure (20).

The development of the positron emission tomography (PET) tracer [¹¹C]ABP688 expanded the investigation of the role of this neuroreceptor in epileptogenesis by providing a non-invasive means to evaluate whole brain mGluR5 expression in vivo. [¹¹C]ABP688 is a highly-selective and high-affinity non-competitive antagonist of mGluR5 that binds to its transmembrane allosteric site, comprising TMIII and TMVII (21–25). [¹¹C]ABP688 PET studies have demonstrated reduced [¹¹C]ABP688 binding potential (BP) within the epileptogenic hippocampi of mTLE patients, and within the hippocampus and amygdala of a pilocarpine-induced mTLE rat model (18, 19). Similarly, reduced [¹¹C]ABP688 BP was found within the borders of focal cortical dysplasia, another intrinsically epileptogenic lesion frequently diagnosed in epilepsy patients (26).

Reconciling hippocampal mGluR5 in vivo [¹¹C]ABP688 PET and ex vivo IHC data to date remains challenging, as these techniques significantly differ in their methods of receptor quantification, each providing a snapshot with very different spatial resolution. Whereas reduced in vivo [¹¹C]ABP688 BP might at first seem in disagreement with ex vivo reports of increased immunoreactivity within the same brain regions, the dynamic conformational changes taking place in vivo must be recalled. It has been speculated that in vivo decreases in [¹¹C]ABP688 BP in mTLE hippocampi may represent downregulation or internalization of mGluR5, unavailability of the [¹¹C]ABP688 allosteric binding site due to conformational changes induced by persistent tissue hyperexcitability, or excess glutamate-induced changes to its binding affinity that could take place despite an upregulation and overexpression of the receptor (18).

To further shed light on the results of [¹¹C]ABP688 PET studies, as well as to further support a role of mGluR5 in epileptogenesis, we conducted saturation autoradiography with [³H]ABP688 in surgically resected hippocampi of drug-resistant mTLE patients. We compared mTLE hippocampi with non-epileptic post-mortem hippocampal specimens from necropsy, to evaluate changes in mGluR5 density and [³H]ABP688-mGluR5 binding affinity ex vivo.

This study was approved by the institutional REB.

One control specimen (male, age 61yo) was excluded from all analyses due to abnormally high [³H]ABP688 binding in the 0.5 nM condition (approximately equal to the 2nM condition), constituting an outlier. There were no significant differences in sex distribution [χ(1)2 = 2.264, p = 0.132] and age [t_(6.132) = 0.416, p = 0.692] between control and mTLE groups. Both Age [r₍₄₎ = −0.5061, p = 0.3056] and PMD [r₍₄₎ = −0.1533, p = 0.7718] were uncorrelated with B_max in the control group, supporting their use as a control for the mTLE surgical samples.

In this study we deployed gold standard techniques to quantify mGluR5 density in hippocampal tissue using quantitative autoradiography. We found reduced density in mTLE hippocampi as compared to controls. mTLE did not impose changes on [³H]ABP688 affinity for mGLUR5. [³H]ABP688 density observed in mTLE hippocampi in this ex vivo study corroborates previous observations describing reduced [¹¹C]ABP688 BP in vivo in pilocarpine-treated rats and in the epileptogenic hippocampi of mTLE patients using PET (18, 19). The present results corroborate in vivo [¹¹C]ABP688 PET findings of reduced availability of mGluR5 in the epileptogenic hippocampus in mTLE. In the context of previous observations from independent studies using mGluR5 IHC, membrane-expressed receptor pools with and without conformational changes, as well as internalized receptors, may contribute to the observation of increased immunoreactivity in the tissue.

The magnitude of the B_max reduction, of over 40% as compared to control specimens, is well aligned with in vivo observations in mTLE patients using [¹¹C]ABP688 PET (18). In addition, the lack of a significant difference in K_D values suggests the binding affinity of [³H]ABP688 for mGluR5 is similar between mTLE and control specimens. Lam and colleagues suggested that excess glutamate binding to mGluR5 may reduce [¹¹C]ABP688's affinity for its allosteric binding site, contributing to reduced BP in vivo (18). Although our [³H]ABP688 autoradiography protocol includes a pre-incubation step that removes endogenous ligands from the tissue, brain cells have been flash-frozen immediately after surgery, and mGluR5 conformation therefore remains static as a snapshot at time of resection. However, the preservation of K_D indicates that, at the time of resection, the pool of receptors to which [³H]ABP688 binds in the tissue had not undergone any glutamate-induced conformational changes affecting [³H]ABP688' affinity for mGluR5's allosteric site.

Although reduced [³H]ABP688 binding could be at first interpreted as reduced mGluR5 expression in mTLE hippocampi, previous mGluR5 IHC studies have shown increased immunoreactivity which indicates increased mGluR5 expression instead (13, 14). Even though neither necropsy controls nor patients' specimens had mGluR5 IHC to support that, this apparent discrepancy can be easily explained by the techniques applied. First, the fixation and permeabilization steps of IHC studies permit detection of internalized as well as membrane-bound receptors (18, 28). The antibodies used in the studies by Kandratavicius, Notenboom and their respective colleagues that demonstrated mGluR5 upregulation targeted the cytoplasmic domain of the receptor, whereas ABP688 targets the transmembrane domain (13, 14, 23). If a conformational change affecting the availability of the transmembrane allosteric binding site, but not the C-terminal antibody binding site, occurred, then it would be possible to observe decreased [¹¹C] and [³H]ABP688 binding with PET and autoradiography despite upregulation of mGluR5. As such, [³H]ABP688 binding could be interpreted as a marker of a specific pool of mGLUR5 receptors.

Notably, intracellular mGluR5 has recently emerged as an important modulator of synaptic plasticity (29–31). For example, local dendritic calcium increases and protein-synthesis-dependent LTD, but not LTP, were shown to be mediated by intracellular mGluR5 (32). However, further research is needed to discern whether intracellular mGluR5 expression is altered in mTLE.

As previously discussed, [³H]ABP688 requires the allosteric site in the TM domain to be accessible for binding, and membrane expressed mGluR5 might undergo conformational changes that renders the binding site unavailable for the ligand. In addition, due to chemical and physical differences between the cytoplasmic and membrane milieux, it is unknown whether [³H]ABP688 can detect mGluR5 that is dissociated from the cell membrane, thus remaining unclear if it can bind to internalized mGluR5 (33).

The decreases in [¹¹C] and [³H]ABP688 binding observed in vivo and ex vivo seem instead to reflect a scenario in which mGluR5 allosteric binding sites are not available for ligand binding. Access into the binding pocket within the allosteric site for mGluR5 negative allosteric modulators has been shown to be narrow and restricted (34). Additionally, an inverse relationship between extracellular glutamate levels and [¹¹C]ABP688 binding has been observed (35). Subsequently, glutamate-induced conformational changes that lead to occlusion of this binding site may explain reduced availability in a brain region that is under recurrent hyperexcitability, such as the epileptic focus.

Furthermore, mGluR5 is expressed as a constitutive dimer, and the dimer structure of mGluRs has been shown to influence the activity of allosteric modulators (36). Indeed, agonist binding has been shown to lead to a rearrangement at the 7TM dimer interface of mGluRs (37). Therefore, it is conceivable that an excess glutamate-induced change to mGluR5 dimerization state could elicit changes in the availability of the allosteric binding site. Furthermore, the monomer-dimer ratio of mGluRs has been shown to be affected in other neurological disorders and could well occur in epilepsy as well (38, 39). If the dimerization of mGluR5 were essential for ABP688 binding, then the observed decreases could also reflect a shift in the ratio of monomeric and dimeric mGluR5 in mTLE.

The interpretation of these results should take into considerations a few methodological limitations. The present study did not evaluate region-specific differences in mGluR5 density and binding affinity due to an incomplete view of hippocampal subregions in mTLE specimens. Future studies could investigate regional differences to identify possible hippocampal subfield-specific mechanisms underlying mTLE. Additionally, emerging evidence that mGluR5 has a variety of functions within the cell also requires further research into how intracellular pathways may be involved in epileptogenesis or neuroprotection. Most of the current literature on mGluR5 has focused on the function of the receptor when it is associated with the cell membrane, so exploring how this protein behaves when located intracellularly could offer new insight into this topic.

This study has demonstrated that [³H]ABP688 binding, but not binding affinity, is altered in the hippocampus of mTLE patients. We propose that these reductions signify the internalization of mGluR5 or a loss of allosteric binding site availability related to conformational alterations of mGluR5.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by MNI/McGill MUHC REB. The patients/participants provided their written informed consent to participate in this study.

Study design and data analysis: MZ, LM, PR-N, and EK. Data acquisition: MZ and AA. Manuscript preparation: MZ and EK. Manuscript revision and approval: All authors. All authors contributed to the article and approved the submitted version.

This study was supported by operating funds from the Savoy Foundation (www.savoy-foundation.ca; Pilot project grant to EK, PR-N, and MSc studentship to MZ).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.