Blog

by Rodríguez-Muñoz M, Cortés-Montero E, Onetti Y, Sánchez-Blázquez P, Garzón-Niño J.

Excerpt from the article published in Biomolecules. 2021 Nov 12;11(11):1681. doi: 10.3390/biom11111681. PMID: 34827679; PMCID: PMC8615847.

Editor’s Highlights

• Sigma-1 Receptors couple alpha2delta-1 proteins to NR1 subunits, enhancing NMDAR activity.
• Alpha2Delta-1, sigma-1 and NR1 trimers remain stable even in the presence of low levels of calcium, and they are much less sensitive to the effect of σ1R ligands.

Abstract

Nerve injury produces neuropathic pain through the binding of α2δ1 proteins to glutamate N-methyl-D-aspartate receptors (NMDARs). Notably, mice with a targeted deletion of the sigma 1 receptor (σ1R) gene do not develop neuropathy, whereas mice lacking the histidine triad nucleotide-binding protein 1 (Hint1) gene exhibit exacerbated allodynia. σ1R antagonists more effectively diminish neuropathic pain of spinal origin when administered by intracerebroventricular injection than systemically. Thus, in mice subjected to unilateral sciatic nerve chronic constriction injury (CCI), we studied the participation of σ1Rs and HINT1 proteins in the formation of α2δ1-NMDAR complexes within the supraspinal periaqueductal gray (PAG). We found that δ1 peptides required σ1Rs in order to interact with the NMDAR NR1 variant that contains the cytosolic C1 segment. σ1R antagonists or low calcium levels provoke the dissociation of σ1R-NR1 C1 dimers, while they barely affect the integrity of δ1-σ1R-NR1 C1 trimers. However, HINT1 does remove δ1 peptides from the trimer, thereby facilitating the subsequent dissociation of σ1Rs from NMDARs. In σ1R^-/- mice, CCI does not promote the formation of NMDAR-α2δ1 complexes and allodynia does not develop. The levels of α2δ1-σ1R-NMDAR complexes increase in HINT1^-/- mice and after inducing CCI, degradation of α2δ1 proteins is observed. Notably, σ1R antagonists but not gabapentinoids alleviate neuropathic pain in these mice. During severe neuropathy, the metabolism of α2δ1 proteins may account for the failure of many patients to respond to gabapentinoids. Therefore, σ1Rs promote and HINT1 proteins hinder the formation α2δ1-NMDAR complexes in the PAG, and hence, the appearance of mechanical allodynia depends on the interplay between these proteins.

Introduction

Persistent anomalous activation of glutamate N-methyl-D-aspartate receptors (NMDARs) typically accompanies different types of neuropathic pain, as characterized by tactile allodynia and hyperalgesia [1]. NMDARs are under allosteric regulation of different endogenous and exogenous molecules, of which the alpha2delta1 (α2δ1) protein plays a decisive role in the neuropathy promoted by the over activation of NMDARs [2]. Proteins of the α2δ family, α2δ1, α2δ2, α2δ3, and α2δ4 are derived from distinct genes (Cacna2d1-4) with different sequences. These α2δ genes encode a single precursor protein, which is post-translationally processed into two proteins in the endoplasmic reticulum (ER): a larger N terminal α2 protein that is essentially extracellular; and the smaller δ peptide in the carboxyl region that contains a transmembrane domain and an intracellular region [3,4]. Finally, the α2 and δ proteins are re-assembled via disulfide bonds [5]. The α2 protein is heavily glycosylated [3,6], and enzymatic deglycosylation of the reduced α2δ complex produces peptides of 100 and 17 kDa.

Murine α2δ1 is a 1067 residue protein of about 115 kDa found at high levels in the anterior cingulate cortex, amygdala, and periaqueductal gray (PAG), and at lower levels in the spinal cord (SC) [7]. As a consequence of nerve damage, α2δ1 proteins and α2δ1-NMDAR complexes augment considerably in the dorsal root ganglia (DRG) and SC, giving rise to neuropathic pain [8]. Gabapentinoids, such as gabapentin and pregabalin, are widely used to alleviate the symptoms of neuropathic pain and epilepsy [9,10,11]. These drugs bind to the α2 region of the α2δ1 and α2δ2 variants, but not to α2δ3 [12]. While, experimental α2δ1 overexpression potentiates NMDAR activity in spinal dorsal horn neurons, provoking pain hypersensitivity, disruption of the α2δ1 gene prevents nerve injury from enhancing NMDAR activity, suggesting that when coupled to NMDARs these α2δ1 proteins are the therapeutic target of gabapentinoids [8].

Notably, α2δ1 proteins and the type 1 sigma receptors (σ1Rs) physically interact with NMDARs to promote calcium permeation and ultimately, neuropathic pain [8,13]. Accordingly, σ1R antagonists alleviate neuropathic allodynia and inflammatory hyperalgesia in animal models of pain that involve NMDAR activation [14,15,16]. Similarly, σ1R^−/− mice do not develop allodynia in different paradigms of neuropathic pain, such as sciatic nerve chronic constriction injury (CCI) [17], paclitaxel induced pain [18], SC contusion injury [19], or spare nerve injury [20]. Mice lacking the histidine triad nucleotide-binding protein 1 (Hint1) gene display altered NMDAR activity [21] and they are more susceptible to CCI-induced mechanical hypersensitivity than their wild-type (WT) littermates. Moreover, HINT1 regulators can alleviate CCI-induced mechanical allodynia for several days in WT mice [22]. Hence, σ1Rs appear to promote and HINT1 proteins dampen NMDAR-mediated neuropathic pain. Both the σ1R and HINT1 protein are widely expressed in nervous tissue, detected at high levels in areas that are associated with pain control [23]. Furthermore, both these regulatory proteins bind to the NR1 subunit of the NMDAR that carries the C1 domain within the cytosolic C0-C1-C2(2′) tail [24]. This domain coordinates the activity of NMDARs with that of G-protein coupled receptors (GPCRs), such as the mu-opioid receptor (MOR) or cannabinoid type 1 receptor (CB1R) [25].

The relationship between α2δ1 proteins and NMDARs in nerve injury has been characterized extensively in the DRG and SC [8]. Nevertheless, drugs regulating σ1R or HINT1 activity efficiently alleviate neuropathic pain when administered by the intracerebroventricular (icv) route [22,25]. In fact, neuropathic pain persists even after spinal ascending nociceptive signals remit, suggesting a role for supraspinal neural structures in this syndrome. The periaqueductal gray (PAG) matter is a midbrain structure strongly implicated in the nociceptive and emotional aspects of pain processing. Specifically, the ventrolateral PAG controls upstream spinal nociceptive signals, regulating their strength in the midbrain and the dorsal SC (substantia gelatinosa) through inhibitory descending pathways [26]. This control may be impaired by spinal nerve injury, which causes upstream changes in PAG glutamatergic neurotransmission, with an upregulation of NMDARs and hypofunction of α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic receptors (AMPARs). These alterations reduce PAG descending pain inhibition and consequently, they prolong the duration of neuropathic pain [27]. Thus, we have studied the role of σ1R and HINT1 proteins in the formation of α2δ1-NMDAR complexes in this brain structure promoted by nerve injury.

Discussion

In response to nerve injury, the association of α2δ1 proteins with glutamate NMDARs increases, bringing about the onset and maintenance of neuropathic pain. These associations were previously evident at the spinal and DRG level [8]; however, pharmacological interventions at the supraspinal level efficaciously alleviated CCI neuropathy of spinal origin. Thus, our study reports the presence of such neuropathy-related α2δ1-NMDAR associations at the supraspinal PAG level that depend on the interplay between σ1Rs and HINT1 proteins.

In addition to the α2δ1 protein, the α2δ2 variant also binds to NMDARs; however, nerve damage barely induces the appearance of α2δ2-NMDAR complexes at a spinal level, and thus, they may have limited relevance in neuropathy [8,38]. Nevertheless, following spinal CCI there was an increase in α2δ1-NMDAR and α2δ2-NMDAR complexes in the PAG, although the highly selective σ1R antagonist S1RA disrupted both these associations. The α2δ1 and α2δ2 proteins both bind gabapentinoids [12,39], and they form complexes with NR1 C1 subunits. However, while σ1R and CaM bind to the (α2)δ1 peptide in a calcium-dependent fashion, their interaction with the (α2)δ2 peptide is less sensitive to calcium. Moreover, the (α2)δ1 but not the (α2)δ2 peptide binds to the HINT1 protein. It is possible that α2δ2 proteins play a different role in the onset and maintenance of neuropathy, although their regulation in the context of NMDARs may be similar to that of α2δ1 proteins.

The α2δ2 protein is much more strongly expressed in the PAG than in the SC, which might account for the presence of α2δ2-NMDAR complexes in this supraspinal structure. The antibody used in previous studies is directed against the α2(δ2) aa 850–865 sequence, an internal sequence that probably associates with sugars at N864 and that in turn may limit or even abrogate antibody binding. In our study, the polyclonal antiserum directed against a different and longer sequence of α2(δ2) (aa 550–800) revealed this α2δ2-NMDAR interaction. Because α2δ proteins are heavily glycosylated [3,6], the presence of sugars associated with their peptide sequence or even the variable sugar decoration at the same sequence, makes immunodetection complicated. In fact, the antibody directed against the internal sequence aa 527–576 of the α2(δ1) protein only detected the targeted protein when bound to NR2A, but not to NR2B subunits. Thus, the N terminal 1–15 α2(δ1) antibody labelled α2δ1 proteins associated with NR1, NR2A, and NR2B subunits in the absence of nerve injury, whereas the 527–576 α2(δ1) antibody mainly detected the association of α2(δ1) with NR2A but not with the NR1 or NR2B subunits. The CCI procedure enhanced the α2δ1-NMDAR associations, but again, the 527–576 α2(δ1) antibody indicated that the α2δ1 proteins bound to NR1 subunits are essentially contributed by NR2B subunits. This observation is consistent with previous reports suggesting that NMDARs containing NR2B subunits are those involved in neuropathic pain [40,41].

Glycosidase enzymes diminished the apparent size of the α2δ proteins [3,6] and accordingly, we found that PNGase F reduced the size of the NR2A/B-associated α2(δ1) proteins from 150 to 100 kDa. In the ER, glycosylation introduces a signal for protein membrane localization or exocytosis and in the particular case of α2δ, this modification is required for the functional membrane expression of calcium channels. Indeed, deglycosylation and glycosylation site-directed mutagenesis strongly reduces current densities without affecting the kinetic properties of such channels [6,42]. Therefore, the disparate glycosylation of α2δ1 proteins associated with NR2A or NR2B subunits may also modify their influence on NMDAR activity.

In cell expression systems, α2δ1 proteins bind to heterodimers of NR1 with NR2 subunits, but not to NR1, NR2A, or NR2B when expressed alone [8]. Thus, the NR1-NR2 dimer offers a binding surface to the extracellular α2(δ1) protein and to the transmembrane/cytosolic C terminal (α2)δ1 peptide, the latter being critical to stabilize the interaction. Remarkably, (α2)δ1 binds to the NR1 variant, which contains the cytosolic C1 segment intercalated between the C0 and C2(2′) regions, displaying no affinity towards the NR1 C0-C2(2′) variant. This preference is evident with GPCRs, which interact through their cytosolic C-terminus with NMDAR NR1 C1 subunits [25,43,44], and also in the tandem σ1R-HINT1 proteins [24,45], which connect GPCRs like the MOR and CB1R to NMDARs [25]. Thus, an external surface in the NMDAR provided by the NR1 C1 subunit when coupled to NR2A or NR2B subunits physically interacts with α2δ1 proteins. Collectively, these observations suggest an important role for NMDARs that contain NR1 C1 subunits in the impact of signals originated at GPCRs. In fact, NR1 C1 subunits are enhanced in depressive patients and they diminish in those affected by schizophrenia [46], augmenting five-fold in σ1R^−/− mice and about two-fold in HINT1^−/− mice. These changes do not affect the total NR1 levels, but they are compensated by fluctuations in the content of the C0-C2(2′) variant [21,25].

In the absence of nerve injury, the association of α2δ1 proteins with NMDARs does not promote noticeable neuropathic pain. Thus, HINT1 or (α2)δ1 at NR1 C1 subunits would barely alter the activity of NMDARs when triggered by regulators, such as glutamate, glycine, or D-serine. Enhanced GPCR signaling as a consequence of nerve lesion recruits PLCβ to activate PKCγ. This kinase acts on the NR1 C1 segment [22] to exchange HINT1 binding with that of σ1Rs, which now facilitates (α2)δ1 access to NR1 C1 subunits and stabilizes the α2δ1-σ1R-NMDAR interaction, augmenting calcium permeation [8,24]. HINT1 proteins and σ1Rs compete for binding to NR1 C1 subunits in a calcium-dependent manner. Thus, in the absence of nerve injury, the interplay between HINT1 proteins and σ1Rs determines the extent of NMDAR activation. A single NMDAR contains two NR1 subunits, which may be different variants: C0-C1-C2(2′) or C0-C2(2′). In this case, NMDARs exhibit intermediate deactivation kinetics and pharmacological properties compared to the respective NR1-NR2A/B or NR1 C1-NR2A/B receptors [47]. Thus, activation promoted by σ1Rs at C0-C1-C2 may be counterbalanced by inhibitory Ca²⁺-CaM at C0-C1 [36]. Regulation of NMDAR activity is also achieved by endogenous ligands of the σ1Rs. Agonists promote and antagonists dampen σ1R-NR1 C1 interactions that regulate the access of HINT1 proteins and of Ca²⁺-CaM to NR1 C1 subunits, thereby influencing the open probability of the NMDAR pore [24].

The (α2)δ1 peptide binds to the HINT1 protein, CaM, σ1R, and the NR1 C1 subunit. HINT1 and CaM at least partially, share their binding site on the (α2)δ1 peptide, and the σ1R and NR1 C1 bind to the last 30 aa of the (α2)δ1 C terminus. Thus, the (α2)δ1 peptide may also associate with a number of regulatory proteins so that they are immediately available when needed in the NMDAR compartment. Nerve damage augments the signaling activity of certain GPCRs [48], providing σ1Rs to bind to NR1 C1 subunits [24], sustaining calcium permeation and thus, the calcium available at the cytosolic side of the NMDAR pore. This mechanism promotes two opposing signaling pathways, the increase in Ca²⁺-CaM drives the release of HINT1 proteins from (α2)δ1 peptides to diminish the access of σ1Rs to NR1 C1 subunits and thus, NMDAR activity. However, σ1Rs can also bind to the (α2)δ1 C terminal region forming (α2)δ1-σ1R-NR1 C1 trimeric complexes that protect NMDAR activity. The number of these trimers would increase as the activity of PKCγ releases more of the HINT1 bound to NR1 C1 subunits [22,24], making these NMDAR subunits available to interact with (α2)δ1-σ1R.

In vitro, the trimer is more stable than the dimer when calcium levels decrease or when compared to the dissociative effect of σ1R antagonists. In this scenario, the trimer promotes NMDAR over activation and provokes the ensuing mechanical allodynia. This pro-nociceptive situation can be alleviated by HINT1 proteins removing (α2)δ1 peptides from the σ1R-NR1 C1 dimer, thereby increasing the dissociation of σ1R-NR1 C1 complexes as calcium diminishes or in the presence of σ1R antagonists. The (α2)δ1-NR1 C1 association observed in the absence of nerve damage/neuropathy may be mediated by CaM maintaining the NMDAR inhibitory (α2)δ1-CaM-NR1 C1 trimer even at low calcium levels. In this situation, the mobilization of HINT1 proteins would remove (α2)δ1 peptides following CaM separation from NR1 C1 at low calcium.

While CCI did not alter the NR1 C1 variant content in CD1 WT and 129 WT mice, NR2B subunit expression augmented and thus, there was an increase in the α2δ1 protein binding to NMDARs through NR2B subunits. The expression of total NR1 is similar in CD1 WT and CD1 σ1R^−/− mice, but in the latter, the NR1 C1 variant increases about five-fold [25]. Nevertheless, this increase does not facilitate access of α2δ1 proteins to NMDARs and thus, HINT1 binding to NR1 C1 subunits augments [24,25]. By increasing NR1 C1 subunit expression, CD1 σ1R^−/− mice may at least partially restore the interaction between GPCRs and NMDARs. These associations are facilitated by σ1Rs and they are further reduced by HINT1 transfer from GPCRs towards NMDARs in CD1 σ1R^−/− mice [25], favoring the formation of HINT1-NR1 C1 dimers. As mentioned, nerve injury does not promote α2δ1-NMDAR associations or cause mechanical allodynia in CD1 σ1R^−/− mice. In these mutant mice, (α2)δ1 peptides may be switched with HINT1 proteins at NR1 C1 subunits, yet such (α2)δ1-NR1 C1 complexes apparently exert no significant effect on NMDAR activity. Thus, σ1Rs certainly appear to be decisive to form neuropathy-related (α2)δ1-σ1R-NR1 C1 complexes.

HINT1 proteins couple weakly active NMDARs to certain GPCRs, such as MORs. In this context, the function of the GPCR activates the coupled NMDAR, which now separates to negatively regulate the signaling of the GPCR. In 129 mice with a targeted deletion of the HINT1 gene, GPCRs lack this negative feedback and thus, NR1 C1 levels may increase to restore this function. Thus, NR1 C1 and the neuropathy-related NR2B subunit increase two-fold in 129 HINT1^−/− mice, and α2δ1 proteins increase their association with these subunits, which influences NMDAR activity [21]. In this mutant mouse, CCI further increases the availability of the NR1 C1 variant and the formation of (α2)δ1-σ1R complexes, although the severe neuropathic syndrome exhibited by these mice [22] was accompanied by a drastic reduction in α2δ1-NMDAR complexes. Because, smaller fragments of α2(δ1) appeared in the PAG of CCI HINT1^−/− mice, proteolytic degradation of α2δ1 proteins may account for this reduction.

The binding of HINT1 to NR1 C1 subunits is not very dependent on calcium and σR1s hardly remove HINT1 proteins from NR1 C1 subunits. As mentioned above, this is facilitated by the PKCγ-mediated phosphorylation of the C1 region of NR1 subunits, which reduces the affinity of HINT1 binding to this cytosolic region and increases that of σ1Rs [24]. HINT1 binds in a zinc-dependent manner to cysteine-rich domains in the regulatory region of PKCγ and prevents its kinase activity [49]. Thus, PKC activity is enhanced in the absence of HINT1 [21], facilitating σ1R binding to NMDARs containing NR1 C1 subunits. This mechanism may account for the enhanced mechanical allodynia observed in HINT1^−/− mice after CCI surgery.

We have learned how alterations of proteins such as HINT1, σ1R, and NR1 C1 subunit may affect adaptive responses of NMDARs. Indeed, a series of human HINT1 mutants cause autosomal recessive axonal neuropathy with neuromyotonia (ARAN-NM) [50]. In most HINT1 mutants, interactions with a series of signaling proteins are impaired, NR1 C1 and σ1Rs included [51]. Motor neurons are enriched in σ1Rs [52] and autosomal recessive loss-of-function mutations in σ1Rs are primarily associated with distal hereditary motor neuropathy and amyotrophic lateral sclerosis/frontotemporal dementia [53,54]. Thus, HINT1 mutants may promote α2δ1 and σ1R mediated activation of NMDARs, and accordingly, amyotrophic lateral sclerosis could be treated with drugs reducing NMDAR activity [55]. Similarly, there are fewer NR1 C1 subunits in the prefrontal cortex of schizophrenic patients, while they increase in depressive individuals. These changes may alter the cross-talk between GPCRs and NMDARs, and also the capacity of α2δ1 proteins to activate this glutamate receptor [46]. Our present study reveals that in the PAG of CD1 σ1R^−/− mice, CCI recruits HINT1 proteins to reduce NMDAR activity, thereby enhancing descending pain control and abolishing the supraspinal perception of neuropathic pain. Alternatively, molecular and electrophysiological studies indicate that 129 HINT1^−/− mice exhibit higher NMDAR/AMPAR and NR2B/NR2A subunit ratios [21], and thus, CCI may promote severe σ1R-mediated hypofunction of PAG glutamate activity, which compromises descending pain control and enhances the supraspinal impact of allodynia [27].

The HINT1 protein reduces the formation of pro-allodynic (α2)δ1-σ1R-NMDAR complexes and thus, neuropathy is enhanced in the absence of HINT1. In this scenario, α2δ1 proteins undergo proteolysis, probably in an attempt to reduce the impact of pain mediated by NMDAR overactivity. Unfortunately, proteolysis of α2δ1 proteins may remove the gabapentinoid binding site from α2(δ1), and in fact gabapentinoids do not alleviate allodynia in 129 HINT1^−/− mice. This phenomenon may account for the large number of patients suffering neuropathy who are refractory to the beneficial effects of α2(δ1)-binding gabapentinoids, almost 50% [56]. Thus, selective σ1R antagonists may be the agents of choice to treat gabapentinoid-resistant neuropathy. The efficacy of systemic S1RA increases considerably when combined with morphine; however, this potentiation is not observed when both compounds are administered via the icv route. Hence, spinal MORs would appear to be more relevant than brain MORs in reducing CCI-induced neuropathy. Thus, at the supraspinal level, S1RA may collaborate with activated spinal MORs to alleviate neuropathic pain of spinal origin. Notably, anti-allodynia evoked by systemic administration of the σ1R antagonist S1RA is enhanced and it persists for longer when combined with low doses of memantine, a low affinity antagonist of NMDARs. Because NMDARs containing NR2B subunits are critical to regulate peripheral persistent inflammatory pain [57], NR2B specific antagonists may also alleviate mechanical allodynia.

In summary, our study suggests that the α2δ1-NMDAR association, and hence allodynia, depends on the interplay between σ1Rs and HINT1 proteins. Interestingly, recent reports suggest a potential therapeutic role for exogenous regulators of σ1R and HINT1 in the clinical management of neuropathic pain [17,22]. The possible use of such pharmacological interventions to alleviate the progression of this pain syndrome merits consideration.Go to: