# Probing the ionotropic activity of glutamate GluD2 receptor in HEK cells with genetically-engineered photopharmacology

1. Neuroscience Paris Seine - Institut de Biologie Paris Seine (NPS - IBPS), CNRS, INSERM, Sorbonne Université, France
2. CNRS, Université de Paris, UPR 9080, Laboratoire de Biochimie Théorique, France
3. Institut de Biologie Physico-Chimique-Fondation Edmond de Rothschild, PSL Research University, France
Research Article

## Abstract

Glutamate delta (GluD) receptors belong to the ionotropic glutamate receptor family, yet they don’t bind glutamate and are considered orphan. Progress in defining the ion channel function of GluDs in neurons has been hindered by a lack of pharmacological tools. Here, we used a chemo-genetic approach to engineer specific and photo-reversible pharmacology in GluD2 receptor. We incorporated a cysteine mutation in the cavity located above the putative ion channel pore, for site-specific conjugation with a photoswitchable pore blocker. In the constitutively open GluD2 Lurcher mutant, current could be rapidly and reversibly decreased with light. We then transposed the cysteine mutation to the native receptor, to demonstrate with high pharmacological specificity that metabotropic glutamate receptor signaling triggers opening of GluD2. Our results assess the functional relevance of GluD2 ion channel and introduce an optogenetic tool that will provide a novel and powerful means for probing GluD2 ionotropic contribution to neuronal physiology.

## eLife digest

Neurotransmitters are chemicals released by the body that trigger activity in neurons. Receptors on the surface of neurons detect these neurotransmitters, providing a link between the inside and the outside of the cell. Glutamate is one of the major neurotransmitters and is involved in virtually all brain functions. Glutamate binds to two different types of receptors in neurons. Ionotropic receptors have pores known as ion channels, which open when glutamate binds. This is a fast-acting response that allows sodium ions to flow into the neuron, triggering an electrical signal. Metabotropic receptors, on the other hand, trigger a series of events inside the cell that lead to a response. Metabotropic receptors take more time than ionotropic receptors to elicit a response in the cell, but their effects last much longer.

One type of receptor, known as the GluD family, is very similar to ionotropic glutamate receptors but does not directly respond to glutamate. Instead, the ion channel of GluD receptors opens after being activated by glutamate metabotropic receptors. GluD receptors are produced throughout the brain and play roles in synapse formation and activity, but the way they work remains unclear. An obstacle to understanding how GluD receptors work is the lack of molecules that can specifically block these receptors’ ion channel activity.

Lemoine et al. have developed a tool that enables control of the ion channel in GluD receptors using light. Human cells grown in the lab were genetically modified to produce a version of GluD2 (a member of the GluD family) with a light-sensitive molecule attached. In darkness or under green light, the light-sensitive molecule blocks the channel and prevents ions from passing through. Under violet light, the molecule twists, and ions can flow through the channel.

With this control over the GluD2 ion channel activity, Lemoine et al. were able to validate previous research showing that the activation of metabotropic glutamate receptors can trigger GluD2 to open. The next step will be to test this approach in neurons. This will help researchers to understand what role GluD ion channels play in neuron to neuron communication.

## Introduction

The delta glutamate receptors, GluD1 and GluD2, belong to the ionotropic glutamate receptor (iGluR) family, yet they don’t bind glutamate (Yuzaki and Aricescu, 2017). They are considered as glutamate receptors solely based on their strong sequence and structure homology with AMPA, NMDA and kainate receptors (Lomeli et al., 1993; Araki et al., 1993; Schmid and Hollmann, 2008; Elegheert et al., 2016; Burada et al., 2020a; Burada et al., 2020b). GluD receptors are widely expressed throughout the brain, GluD1 predominantly in the forebrain, while GluD2 is highly enriched in cerebellar Purkinje cells (PCs) (Konno et al., 2014; Hepp et al., 2015; Nakamoto et al., 2020b). Both GluD1 and GluD2 play a role in the formation, stabilization, function and plasticity of synapses through their interaction with members of the cerebellin (Cbln) family (Fossati et al., 2019; Tao et al., 2018; Matsuda et al., 2010; Kakegawa et al., 2008). Cbln1 notably binds both the N-terminal domain of postsynaptic GluD2 and presynaptic neurexins, leading to a trans-synaptic bridge that promotes synaptogenesis and is essential for GluD2 signaling in vivo (Elegheert et al., 2016; Suzuki et al., 2020). Deletion of genes coding for GluD1 or GluD2 in mouse results in marked behavioral alterations (Yadav et al., 2012; Lalouette et al., 2001; Yadav et al., 2013; Nakamoto et al., 2020a), and mutations in these genes in humans have been associated with neurodevelopmental and psychiatric diseases (Griswold et al., 2012; Treutlein et al., 2009; Greenwood et al., 2011; Cristino, 2019), attesting to their functional importance in brain circuits.

Despite their structural similarity with other iGluRs, and notably the presence of a ligand-binding domain (LBD), GluDs stand out because they are not activated by glutamate (Araki et al., 1993; Lomeli et al., 1993). Nonetheless, recent studies revealed that GluD pore opening could be triggered indirectly, through the activation of Gq-coupled metabotropic receptors, and contributes to neurotransmission and neuronal excitability (Ady et al., 2014; Dadak et al., 2017; Benamer et al., 2018; Gantz et al., 2020). Indeed, the activation of GluD channels triggered by metabotropic glutamate receptors (mGlu1/5) underlies the slow excitatory postsynaptic current in cerebellar PCs (GluD2, Ady et al., 2014) and in midbrain dopaminergic neurons (GluD1, Benamer et al., 2018). Moreover, the demonstration that GluD1 channels carry the noradrenergic slow excitatory current in dorsal raphe neurons (Gantz et al., 2020), suggests that the contribution of GluD channels to neuronal excitability and synaptic physiology may be widespread. The above studies relied largely on genetic tools, such as dead-pore mutants or targeted deletions, to assess the ion channel function of GluDs. Yet, due to the absence of specific pharmacological tools to block their ion permeation, the role of GluD1/2 channels in the regulation of neural activity remains largely elusive.

Pore blockers for GluDs, such as pentamidine and 1-Naphthyl acetyl spermine (NASPM), were previously identified using a point mutation (A654T) in GluD2 that confers constitutive ion flow and causes the degeneration of cerebellar PCs in Lurcher (GluD2Lc) mice (Wollmuth et al., 2000; Zuo et al., 1997). These molecules are however also pore blockers of NMDA and AMPA receptors, respectively. Other ligands such as D-serine and glycine bind to the LBD and reduce spontaneous currents in GluD2Lc, which suggests a coupling between the LBD and the channel (Naur et al., 2007; Hansen et al., 2009), but these molecules have broad spectrum activity. Finally, 7-chlorokynurenic acid has been identified to modulate GluD2Lc current by binding to the D-serine site but it is also a GluN1 competitive antagonist (Kristensen et al., 2016).

To fill this gap, we bestowed light-sensitivity to the GluD ion channel pore using a photoswitchable tethered ligand (PTL) approach (Paoletti et al., 2019; Mondoloni et al., 2019). Using structure-based design, we incorporated a cysteine point mutation at the surface of GluD2, right above the hypothetical channel lumen, onto which can be anchored a photoswitchable pore blocker. Different wavelengths of light are then used to modify the geometry of the PTL, thereby presenting/removing the blocker to/from the channel, resulting in optical control of ionotropic activity. Here we demonstrate rapid and reversible, optical control of ion current through a cysteine-substituted GluD2 receptor. This novel tool, called light-controllable GluD2 (LiGluD2), allows rapid, reversible and pharmacologically-specific control of ionic current through GluD2, and may help provide a mechanistic understanding of how this receptor contributes to brain circuit function and behaviors.

## Results

### Designing a light-controllable GluD receptor

Our approach to probing the functionality of the ion channel in GluD is to install a photo-isomerizable pore blocker at the extracellular entrance to the channel lumen (Figure 1A). The tethered ligand is site-specifically attached to a cysteine-substituted residue. In darkness or under green light (500–535 nm), the PTL adopts an elongated shape and reaches the lumen, resulting in ion channel blockade, while under violet light (380–390 nm), it switches to a twisted, shorter configuration, relieving blockade. Our design of the PTL was based on the chemical structure of pentamidine (Figure 1B), a pore blocker that efficiently blocks current through GluD2Lc receptors (Williams et al., 2003). The PTL, called MAGu, contains a thiol-reactive maleimide (M) moiety, a central photo-isomerizable azobenzene (A) chromophore, and a guanidinium (Gu) head group that resembles the amidinium groups of pentamidine (Figure 1C). MAGu was selected notably because its synthesis route has been described (referred to as PAG1c in the original article) and because it was shown to have no adverse effect on native brain tissue (Lin et al., 2015). In aqueous solution, MAGu could be converted to its cis form using 380 nm light, and converted back to trans either slowly in darkness (t1/2 ~ 20 min) or rapidly upon illumination with 525 nm light (Figure 1—figure supplement 1A–B), in agreement with previous reports (Lin et al., 2018). To find the best attachment site for MAGu on GluD, we developed a homology model of the GluD2 receptor, based on the structure of the recently crystallized GluA2 receptor (Twomey et al., 2017) (see methods). Using this model, we selected a series of 15 residues, located on the peptide that links the LBD to the third transmembrane domain (M3) that lines the channel lumen, for mutation to cysteine (Figure 1D–E).

Figure 1 with 1 supplement see all

### Cysteine screening

Since no known ligand directly gates the ion channel of GluD2, we used a Lc mutant, A654T, which displays a constitutively open channel (Wollmuth et al., 2000; Zuo et al., 1997), for screening the 15 single-cysteine mutations. Accordingly, we found that heterologous expression of GluD2-A654T, but not of the wild-type (WT) protein, in HEK cells produces large currents that reverse at membrane potential close to 0 mV and are reduced by externally-applied pentamidine (Figure 2A). Subtracted Lc current showed clear rectification at positive potentials, as reported with the blockade by NASP, another GluD blocker (Kohda et al., 2000). Therefore, the A654T Lc mutant was subsequently used as a screening platform to find the best attachment site for MAGu on GluD2. Each of the 15 residues identified in Figure 1D were mutated individually to cysteine on the A654T background, and tested using patch-clamp electrophysiology. Cells were treated with MAGu (20 μM, 20 min) and Lc currents were measured in voltage-clamp mode (?60 mV) under different illumination conditions to toggle MAGu between its cis and trans states. As expected, current through A654T was not affected by light, indicating that in the absence of a properly-positioned cysteine, MAGu has no effect on this Lc channel. In contrast, we found several cysteine mutants for which current was significantly larger under 380 than under 535 nm light, and one mutant (Q669C) for which there was a tendency for ‘reverse photoswitching’, that?is larger currents under 535 than under 380 nm light (Figure 2B). We then quantified the degree of photoswitching by comparing the block in darkness (trans state) to the block evoked by a saturating concentration of pentamidine (100 μM). We excluded from the analysis mutants that displayed no pentamidine-decreased leak current (i.e. mutants for which pentamidine block was significantly smaller than that observed on A654T, Figure 2—figure supplement 1A–B), because they were likely either not expressed or not functional. Photoswitching was significant for Q666C, Q669C, D670C, Q674C, and I677C, suggesting that MAGu covalently reacted with these cysteine mutants and that, once tethered, it could modulate current in one of its conformer (Figure 2C). Importantly, photomodulation was absent in the control A654T and the other cysteine mutants, indicating that the effect of light is specific to the attachment of MAGu to the above-mentioned cysteine mutants. From a structural point of view, the photocontrollable mutants are all located at the very top of the linker, that?is further away from the membrane domain compared to other tested residues (Figure 2D).

Figure 2 with 1 supplement see all

### Photocontrol of GluD2-A654T-I677C tethered with MAGu

We then selected the best mutant GluD2-A654T-I677C for further characterization. Because pentamidine does not fully block GluD2, even at saturating concentrations (Williams et al., 2003), we quantified the extent of photoswitching by blocking leak current completely, using impermeant N-Methyl-D-glucamine (NMDG). We found that MAGu blocked about 33% of the leak current in its trans form (Figure 3A). Photoregulation was fully reversible over many cycles of 380 and 535 nm light (Figure 3B), in agreement with the fact that azobenzenes photobleach minimally (Beharry and Woolley, 2011). Under our illumination conditions, light pulses of 200 ms were sufficient to fully unblock the current, while shorter illumination times could be used to finely tune the degree of blockade (Figure 3C). Once in the cis configuration, MAGu relaxes back to its thermodynamically stable trans state slowly, with a half-life of about 20 min in solution (Figure 1—figure supplement 1B). Accordingly, relief of blockade persisted for many seconds in darkness after a brief flash of 380 nm light (Figure 3D), eliminating the need for constant illumination, an important feature for future neurophysiology experiments.

Figure 3

From a pharmacological point of view, current blockade occurred in the trans state (535 nm) and was relieved in the cis configuration (380 nm) for all membrane potential tested, with very little voltage-dependence (Figure 4A), which contrasts with the profound voltage-dependence of block observed with pentamidine. This suggested to us that the positive charge of MAGu may not sense the electrical field of the membrane as much as pentamidine does, and thus that the two molecules may bind to different sites. To investigate whether MAGu and pentamidine compete for the same binding site, we evaluated the dose-response relationship of pentamidine block on GluD2-A654T-I677C conjugated with MAGu, under both 380 and 535 nm light (Figure 4B). We found the IC50s under both wavelengths to be virtually indistinguishable, favoring the idea that MAGu and pentamidine have distinct, non-overlapping binding sites. To get further molecular insight into trans MAGu-induced reduction of current, we performed molecular modeling experiments. After inserting the cysteine mutation, trans MAGu was docked by covalent docking, that?is the reactive maleimide moiety was forced to be in contact with the cysteine while the rest of the molecule was free to move. We found that the guanidinium headgroup of trans MAGu couldn’t reach the membrane-embedded lumen of GluD2 (Figure 4C), in agreement with our electrophysiology data. The effect of trans MAGu on the ion current was tested with the MOLEonline webserver (Pravda et al., 2018), which allowed to compute the geometry of the ion channel. We found that the photoswitch has a direct steric effect on the size of the cavity above the channel, as shown by the comparison of the computed channel in presence or absence of the photoswitch (Figure 4D). In addition, the charge of the photoswitch could modify the electrostatic potential in the cavity and thereby affect ion transfer.

Figure 4

### Optical control of the native GluD2 channel

We next sought to determine whether the non-Lc, native channel could be photocontrolled after installation of MAGu on the cysteine-substituted GluD2-I677C receptor. In heterologous expression system, activation of mGlu1 using the selective agonist 3,5-Dihydroxyphenylglycine (DHPG) was reported to trigger opening of GluD2 receptors (Ady et al., 2014; Dadak et al., 2017). Therefore, we co-expressed the b isoform of mGlu1, which displays low basal activity (Prézeau et al., 1996), together with GluD2 in HEK cells. Cells were labeled with MAGu and DHPG currents were recorded while alternating between 380 and 535 nm light. We found that DHPG-induced currents were reversibly reduced by about 23% under 535 nm compared to 380 nm light for I677C, indicating that optical blockade with MAGu could be transposed to the native, non-Lc GluD2 (Figure 5A). Importantly, DHPG-induced currents were identical in both wavelengths of light for the WT receptor (Figure 5B), confirming that the effect of light is specific to the attachment of MAGu to I677C (Figure 5C). In addition, we observed that the holding current increased when switching from darkness to 380 nm light for I677C, and decreased when switching back to 535 nm light, but remained constant in both wavelengths of light for WT (Figure 5D). This suggests that a fraction GluD2 receptors are constitutively open prior to DHPG application, likely due to some basal mGlu1 activity in these cells. Altogether, these results show that the GluD2 I677C mutant labeled with MAGu (a.k.a. LiGluD2) possesses a functional ion channel, which can be gated through the mGlu signaling pathway, and which can be reversibly blocked and unblocked with green and purple light, respectively.

Figure 5 with 1 supplement see all

In order to evaluate the usefulness of LiGluD2 in neurons, we verified that MAGu treatment does not lead to photosensitization of native glutamate currents in PCs of WT mice. Since GluD2 is enriched at the parallel fiber-PC synapse (Landsend et al., 1997), we used local application of DHPG (200 μM) to induce inward current in both MAGu- and vehicle-treated slices. We found that the DHPG current amplitude remained unchanged under 535 nm compared to 380 nm in the two treatment conditions, resulting in a ratio of current amplitude I380/I535 measured for each cell not significantly different from 1 (Figure 5—figure supplement 1A). We then recorded AMPA-mediated excitatory post-synaptic currents (EPSCs) in both vehicle- and MAGu-treated PCs. We found that the amplitude of electrically evoked EPSCs was stable under 535 nm compared to 380 nm, and that the ratio of EPSC amplitudes I380/I535 was not significantly different from one in both conditions (Figure 5—figure supplement 1B). These control experiments demonstrate that wild-type GluD and GluA receptors, which lack a properly-positioned cysteine residue near the pore lumen, remain insensitive to light after MAGu treatment.

## Discussion

The PTL strategy has been successfully applied to several members of the iGluR family, including kainate (Volgraf et al., 2006) and NMDA (Berlin et al., 2016) receptors. In these former studies, the photoswitches were made with a glutamate head group, and were tethered to the LBD in proximity to the glutamate binding pocket, providing photocontrol of channel gating (Reiner et al., 2015). Because their activation mechanism is still unknown, we adopted a different strategy for photocontrolling GluD receptors. We installed the photoswitchable ligand MAGu in proximity to the pore lumen, in hope to alter ion conduction through non-competitive antagonism. We found several cysteine mutants for which current was specifically modulated by light after attachment of MAGu, notably I677C (a.k.a. LiGluD2). In LiGluD2, trans MAGu likely does not reach the pore lumen as originally designed. Nevertheless, it reversibly modulates current through the open GluD2 channel with high temporal and pharmacological precision.

The compounds traditionally used to probe the ionic function of GluDs, such as pentamidine and NASPM (Kohda et al., 2000; Williams et al., 2003), are not specific of GluD and also block NMDA and AMPA receptors. As to D-serine and glycine, they partially inhibit GluD2Lc and mGlu1-gated GluD currents (Naur et al., 2007; Ady et al., 2014; Benamer et al., 2018), but they are also co-agonists of NMDA receptors. Here, the pharmacological specificity of LiGluD2 is exquisite: after MAGu treatment, only the I677C mutant, and not the WT receptor, became sensitive to light. Likewise, MAGu did not photosensitize other WT glutamate receptors expressed in native brain tissue, in agreement with previous report demonstrating that MAGu has no off-target effects on WT GABA receptors, glutamate receptors and voltage-gated ion channels (Lin et al., 2015). Indeed, even though the maleimide group of MAGu reacts in principle with any extracellular cysteine freely accessible on the cell surface, the ability of the tethered ligand (here guanidinium) to reach a particular site of action on any given protein in one configuration, but not the other (e.g. trans but not cis), is highly improbable. In fact, the PTL approach has already demonstrated exquisite pharmacological specificity for a large variety of cysteine-substituted ion channels and receptors (Paoletti et al., 2019; Mondoloni et al., 2019), even in complex biological settings such as brain slices (Berlin et al., 2016) or intact neuronal circuits in vivo (Lin et al., 2015; Durand-de Cuttoli et al., 2018).

Attachment of MAGu to GluD2 requires a single amino acid substitution, which is unlikely to disrupt the function of the receptor. In line with this, we found that the functional coupling of GluD2 with mGlu1 signaling (Ady et al., 2014; Dadak et al., 2017) was intact in LiGluD2. This enabled us to validate that activation of mGlu1 triggers the opening of the GluD2 channel in heterologous expression system, in support of earlier evidence that opening of the ion channel of GluD receptors can be triggered in response to metabotropic signaling mechanisms (Ady et al., 2014; Dadak et al., 2017; Benamer et al., 2018; Gantz et al., 2020). Even though light-induced blockade in LiGluD2 is partial, the rapid kinetics of block/unblock, coupled to the genetic specificity of the methodology, provide a unique opportunity to detect even small variations in GluD2 current, such as the tonic current we observed in heterologous expression system. LiGluD2 remains to be deployed in neuronal setting, yet we believe it will be a crucial tool for probing the ionotropic contribution of this orphan receptor to synaptic physiology.

## Materials and methods

Key resources table

### Chemicals

Bio-grade Chemicals products was provided by Sigma-Aldrich from Merck. MAGu was synthesized as previously described (Lin et al., 2015) and provided by Enamine Ltd., Kyiv, Ukraine (www.enamine.net). MAGu was stored at ?80°C as stock solutions in anhydrous DMSO.

### Spectrophotometry

Request a detailed protocol

UV-visible spectra were recorded on a Nanodrop 2000 (Thermo Scientific, 1 mm path) with 100 μM MAGu in PBS pH 7.4 (10% final DMSO). The sample was illuminated for 1 min using ultra high-power LEDs (Prizmatix) connected to an optical fiber (URT, 1 mm core, Thorlabs), followed by an immediate measurement of absorbance. Light intensity at the tip of the 1 mm fiber was 100 mW for the 390 nm LED, and 150 mW for the 520 nm LED.

### Molecular biology

Request a detailed protocol

The single-cysteine mutations of GuD2 were generated by site-directed mutagenesis using the Quick Change II kit (Agilent technology) performed on pcDNA3-GluD2 (Ady et al., 2014). All mutants were verified by sequencing.

### Cell line

Request a detailed protocol

We used human Embryonic Kidney cells (HEK tsA201, Sigma-Aldrich # 96121229). Cells were certified by Sigma-Aldrich. Mycoplasma contamination status were negative.

### Cell culture

Request a detailed protocol

Cells were cultured in 25 cm2 tissue culture flask (Falcon, Vented Cap, 353109) with a culture medium composed of Dulbeco’s Modified Eagle Medium (Gibco life technologies, 31966047) containing Glutamax and supplemented with Fetal Bovine Serum (10%, Gibco life technologies, 10500064), Nonessential Amino-Acids (1%, Life Technologies, 11140–035), ampicillin, streptomycin (50,000 U, Gibco, life technologies, 15140–122) and mycoplasma prophylactic (2.5 mg, InvivoGen) antibiotics.

### Transfection

Request a detailed protocol

HEK tsA201 cells were freshly seeded and plated out in a 6-well plate, on coverslips (10 mm) treated with poly-L-lysine hydrobromide (Sigma, P6282-5MG). Cells were transiently transfected using calcium-phosphate precipitation, as described in Lemoine et al., 2016, using 1 μg of cDNA of GluD2 cysteine mutant per well. For co-transfection experiments, we used mGlu1b/GluD2 ratio from 0.7 to 1, with a maximum of 2 μg of total DNA. The plasmid pRK5-mGlu1b used in this study is a generous gift of L. Prezeau (IGF, Montpellier).

### In vitro electrophysiology

Request a detailed protocol

Electrophysiological currents were recorded on HEK tsA201 cells at room temperature (21–25°C), 24–48 hr after transfection. Prior to whole-cell patch-clamp experiments, cells were incubated for 20 min with an extracellular solution containing 20 μM MAGu, and then washed for at least 5 min with a fresh external solution. Cells were perfused with an external solution containing (in mM): 140 NaCl, 2.8 KCl, 2 CaCl2, 2 MgCl2, 12 glucose, 10 HEPES and NaOH-buffered at pH 7.32. The external NMDG solution contained (in mM): 140 NMDG, 2.8 KCl, 2 CaCl2, 2 MgCl2, 12 glucose, 10 HEPES and was KOH-buffered at pH 7.32. Cells were patched with a borosilicate pipette (4–5 MΩ) containing an intracellular solution containing (in mM): 140 KCl, 5 MgCl2, 5 EGTA, 10 HEPES, and pH-adjusted to 7.32 with KOH. For recording metabotropic activation of GluD2 by mGlu1, the internal solution contained (in mM): 140 K-gluconate, 6 KCl, 12.6 NaCl, 0.1 CaCl2, 5 Mg-ATP, 0.4 Na-GTP, 1 EGTA, 10 HEPES, and was adjusted to pH 7.32 with KOH. Pentamidine and NMDG solutions were applied using a fast-step perfusion system equipped with three square tubes (SF77B, warning instruments), as described in Lemoine et al., 2016. Illumination was carried out using a high-power LED system (pE-2, Cooled) mounted directly on the epifluorescence port of a vertical microscope (SliceScope Pro 6000, Scientifica). Light output at the focal plane was 5 and 11.7 mW/mm2 for the 380 and 535 nm LEDs, respectively. Currents were recorded with an axopatch 200B and digitized with a digidata 1440 (Molecular devices). Signals were low-pass filtered (Bessel, 2 kHz) and collected at 10 kHz using the data acquisition software pClamp 10.5 (Molecular Devices). Electrophysiological recordings were extracted using Clampfit (Molecular Devices) and analyzed with R.

### Slice electrophysiology

Request a detailed protocol

Animal breeding and euthanasia were performed in accordance with European Commission guidelines and French legislation (2010/63/UE) and procedures were approved by the French Ministry of Research (Agreement APAFIS#16198–2018071921137716 v3). Mice at age P30-40 were anesthetized with isoflurane and decapitated. Cerebella were rapidly extracted and transferred into ice-cold ACSF supplemented with 50 mM sucrose and 1 mM kynurenic acid. The composition of ACSF in mM was as follows: 126 NaCl, 26 NaHCO3, 2.5 KCl, 1.25 NaH2PO4, 1 MgCl2, 2 CaCl2, 20 glucose. pH was adjusted to 7.4 by continuous gassing with carbogen. Sagittal slices (250 μm) were sectioned from the vermis on a vibratome (Leica VT 1200S) and transferred to oxygenated ACSF. Slices were incubated for 15 min at 30°C then transferred at room temperature before recording. Unless stated otherwise, all the steps were performed at room temperature. Purkinje cells were visually identified using infrared Dodt contrast imaging with a 60?×?water immersion objective. Whole-cell recordings from Purkinje cells in cerebellar lobules IV-VI (voltage-clamped at ?70 mV, liquid junction potentials not corrected) were performed with borosilicate glass pipettes (WPI, 2–4 MΩ) pulled with a horizontal micropipette puller (Sutter instruments). Internal pipette solutions contained (in mM): 140 Cs-gluconate, 5 CsCl, 2 MgCl2, 0.5 EGTA, 2 Na-ATP (pH 7.3, adjusted with CsOH). Whole-cell currents were recorded at 20 kHz and filtered with a Bessel low-pass filter at 4 kHz using a patch-clamp amplifier (Multiclamp 700B, Molecular Devices) connected to a Digidata 1440A interface board (Molecular Devices). Only Purkinje cells with a series resistance?<12 MΩ (not compensated; monitored during experiments by applying 200 ms, ?5 mV voltage pulses) were used for the analyses.

Slices were incubated for 20 min under continuous oxygenation with MAGu 20 μM or vehicle (DMSO 0.4%) dissolved in 750 μl ACSF in a well of 24-well plate. The slices were then washed in ACSF for at least 20 min, and transferred in the recording chamber. Photoswitching was achieved by illuminating the slice as described above alternatively at 380 nm and 535 nm for 1 s.

(RS)-DHPG (200 μM) was diluted in ACSF and locally pressure-applied using a patch pipette placed in the dendrites of the recorded Purkinje-cell. A pneumatic microinjector (Picopump, WPI) was used to deliver 0.1–0.2 ms air pressure pulses (4–10 PSI) every minute. DHPG-mediated currents were recorded at room temperature in presence of CNQX 10 μM, D-APV 25 μM, gabazine 10 μM and CGP 55845 0.5 μM. DHPG was applied immediately after the light stimulation.

Parallel fiber stimulation was achieved every 10 s with a glass pipette filled with ACSF and placed in the outer half part of molecular layer. A constant voltage isolation unit (DS3, Digitimer Ltd) was used to deliver 10 μs rectangular pulses (50–200 μA) for extracellular stimulation. Parallel fibers inputs were identified by paired-pulse facilitation at 50 ms inter stimulus interval. Amplitudes of evoked EPSCs were averaged from 6 to 12 traces. EPSCs were recorded at 30°C in presence of gabazine 10 μM. Electrical stimulations were performed immediately after the light stimulation.

### Molecular modeling

Request a detailed protocol

The model of the GluD receptor has been obtained by homology modeling using the software modeller version 9.19 (Webb and Sali, 2016). The template was that of the glutamate receptor GluA2 (PDB code 5weo) (Twomey et al., 2017). The automodel class has been used with slow level of MD refinement and the optimization has been repeated three times for each model. 500 models were prepared and the best, as assessed by the DOPE score, was retained for further studies.

The structure of the protein and ligand were converted to pdbqt files with the software open babel 2.4.1. Covalent docking was then performed with the software smina (Koes et al., 2013). The box of 25*25*25 angstrom was defined manually to encompass the mutated residue and extend to the?axis of symmetry. Covalent docking forced the maleimide to be in direct contact with the SG atom of the cysteine with which it is shown experimentally to form a covalent bond. The geometry of the ion channel has been computed with MOLEonline webserver, with the 'pore' mode. The resulting ion channel was color-coded as a function of the diameter of the channel allowing to illustrate the reduction of the diameter from a large (green) to a small (red) diameter.

### Data analysis

Request a detailed protocol

Data are plotted as mean?±?SEM. Total number (n) of cells in each group and statistics used are indicated in figure and/or figure legend. Comparisons between means were performed using parametric tests (two-sample t-test, Normality always verified, Shapiro-Wilk test of normality). Homogeneity of variances was tested preliminarily and the t-tests were Welch-corrected accordingly. For comparison with theoretical values of 0 or 1, we performed either one-sample t-tests when Normality was verified, or a non-parametric test (one-sample Wilcoxon tests) when Normality was not verified. #p<0.1, *p<0.05, **p<0.01, ***p<0.001.

Time-course of current unblock and of thermal relaxation were fitted with the following mono-exponential function:

(1) $y=100×\left(1-{e}^{-kx}\right)$

with k the decay constant, and ln2/k the half-life.

Dose-response relationships were fitted with the following equation:

(2) $y=MIN+\frac{MAX-MIN}{1+{\left(\frac{x}{IC50}\right)}^{nH}}$

with MAX the maximal current, MIN the minimal current, IC50 the pentamidine concentration yielding half block, and nH the Hill number.

## References

1. 1
2. 2
3. 3
4. 4
5. 5
6. 6
7. 7
8. 8
9. 9
10. 10
11. 11
12. 12
13. 13
14. 14
15. 15
16. 16
17. 17
18. 18
19. 19
20. 20
21. 21
22. 22
23. 23
24. 24
25. 25
26. 26
27. 27
28. 28
29. 29
30. 30
Cell-Specific neuropharmacology (2019)
Trends in Pharmacological Sciences 40:696–710.
https://doi.org/10.1016/j.tips.2019.07.007
31. 31
32. 32
33. 33
34. 34
35. 35
36. 36
Changes in the carboxyl-terminal domain of metabotropic glutamate receptor 1 by alternative splicing generate receptors with differing agonist-independent activity
(1996)
Molecular Pharmacology 49:422–429.
37. 37
38. 38
39. 39
40. 40
41. 41
42. 42
43. 43
44. 44
Comparative protein structure modeling using MODELLER (2016)
Current Protocols in Bioinformatics 54:cpbi.3.
https://doi.org/10.1002/cpbi.3
45. 45
Pharmacology of delta2 glutamate receptors: effects of pentamidine and protons (2003)
Journal of Pharmacology and Experimental Therapeutics 305:740–748.
https://doi.org/10.1124/jpet.102.045799
46. 46
47. 47
48. 48
49. 49
A GluD Coming-Of-Age story (2017)
Trends in Neurosciences 40:138–150.
https://doi.org/10.1016/j.tins.2016.12.004
50. 50

## Decision letter

Reviewing Editor; Stanford University School of Medicine, United States
2. Kenton J Swartz
Senior Editor; National Institute of Neurological Disorders and Stroke, National Institutes of Health, United States
3. Stephanie C Gantz
Reviewer; University of Iowa, United States
4. Lonnie Wollmuth
Reviewer; Stony Brook University, United States
5. Anna Koster
Reviewer; Stanford University, United States

In the interests of transparency, eLife publishes the most substantive revision requests and the accompanying author responses.

Acceptance summary:

GluD2 is a member of the ionotropic glutamate receptor family, but its specific role in the central nervous system is not yet understood. This study demonstrates a novel chemogenetic tool to probe function of GluD2 in HEK cells. This tool has the potential to be very powerful to advance the understanding of GluD2 channel function in neurons since it appears highly selective.

Decision letter after peer review:

Thank you for submitting your article "Probing the ionotropic activity of the orphan glutamate delta 2 receptor with genetically-engineered photopharmacology" for consideration by eLife. Your article has been reviewed by three peer reviewers, and the evaluation has been overseen by a Reviewing Editor (Merritt Maduke) and Kenton Swartz as the Senior Editor. The following individuals involved in review of your submission have agreed to reveal their identity: Stephanie C Gantz (Reviewer #1); Lonnie Wollmuth (Reviewer #2); Anna Koster (Reviewer #3).

The reviewers have discussed the reviews with one another, and the Reviewing Editor has drafted this decision to help you prepare a revised submission.

The editors have judged that your manuscript is of interest; however, as described below, additional experiments are required to support the conclusions. Therefore, we would like to draw your attention to changes in our revision policy that we have made in response to COVID-19 (http://www.asadoresteatre.com/articles/57162). In recognition that many researchers have temporarily lost access to the labs, we will give authors as much time as they need to submit revised manuscripts.

Summary:

Unlike other ionotropic glutamate receptors, GluD2 is not gated by glutamate. No specific or high-affinity chemical modulators that induce channel activity exist for this receptor. To address this challenge, the authors used a previously characterized photoswitchable tethered ligand (PTL) called MAGu to target a very non-specific blocker (pentamidine) to a new ion channel target (the GluD2 receptor). This approach (using this exact PTL) has been used to target knock-in cysteine mutants of the GABAA receptor in mouse brain slices and in vivo in an awake, behaving mouse. Based on this precedent, it is not unreasonable to believe that this tool could similarly be used for the GluD2 receptor, which would be a significant advance in the field for understanding the physiological role of this protein in disease. However, there are concerns about signal-to-noise, since the pore block by trans-MAGu is only a fraction of total presumed current through GluD; therefore, it has not yet been demonstrated that the MAGu response will be sufficient for physiological studies.

Essential revisions:

1) In order to evaluate the potential of this new tool, it is essential that the authors provide a more thorough quantification of its effects, as described below, which will be relevant to the signal:noise in physiological experiments.

The photo-switchable capability of MAGu is convincing, but what is not clear is how good of a blocker trans-MAGu is. It is clear that trans-MAGu is not producing full block of leak current, but it is not quantified. These data need to be added to have a clear understanding of the strategy and mechanism. cis-MAGu unblocks a portion of leak current, but it is not clear how much of the total leak current is blocked by trans-MAGu. Knowing how much leak current is blocked by trans-MAGu is critical to interpreting the effect on mGluR-activated GluD2 current.

In Figure 2C I677C, there is still substantial leak in 535 nm. The quantification in Figure 2C (% photoswitching) shows the % of I-Blockphoto over I-Blockpenta, but the arrows in the righthand trace, it would appear I-Blockphoto is actually the current unblocked. It would be helpful to quantify the amount of leak current blocked by trans-MAGu. In Figure 2C, the extent of block for photoswitching is being quantified relative to that for pentamidine, which is reasonable. However, for pentamidine, what is the concentration used for the experiments? Where is it at on the concentration-block curve for pentamidine? Presumably, if complete block the leak current should go to zero and hence the efficacy of block of photoswitching blocker would be less (e.g., Figure 4B). Please clarify. Additional discussion as the structural basis for incomplete block may also be helpful.

How does MAGu work on the cysteine-engineered receptor that would presumably be used for future in vivo studies? Because the GluD2-I677C point mutant (lacking the L654T background) does not show current, the authors use the known effect of mGlu1 receptor agonism as a readout of GluD2-I677C activity in response to light and only see a 23% decrease in mGlu1 current--is this very small effect physiologically significant or to be expected? It seems like MAGu might be a useful tool to modulate GluD2 in Lurcher mice (which harbor the L654T mutation), but it is hard to know what the probe efficacy and usefulness is for evaluating the physiology of the WT GluD2 receptor in the absence of a way to measure a direct functional effect on the channel. How else might this be addressed?

Discussion paragraph two states that the WT receptor is insensitive to MAGu, but it is not clear where those data are presented. It would be beneficial to show the magnitude of the DHPG-induced current in WT GluD2-expressing cells before and after addition of MAGu to address the possibility that MAGu affects the current irrespective of trans- or cis- conformation.

It is also not clear how MAGu will be selective for site-specific conjugation when introduced in a neuronal setting. Is it expected MAGu will react with any available cysteine? It would be helpful to discuss possible limitations going forward towards use in neurons.

2) The Introduction and Abstract are rather general and antiquated, to the disservice of the readers. It may be time to move away from the notion that ion channel function of GluD is debated. The authors have published many elegant studies demonstrating ion channel function. By appearances of the literature, the interpretation of these studies are not contested. In addition to pharmacology, ion channel function of GluD has been demonstrated using selective genetic strategies (e.g. Ady et al., 2013; Benamer et al., 2018; Gantz et al., 2020). To this end, the Abstract and Introduction should be changed. It does not seem fitting to state "direct evidence for ionotropic activity of GluD in neuronal setting [sic] is lacking" provided the studies referenced above. Broadly, the readers would benefit from restructuring of the Introduction and Abstract to state the specific issue addressed by the present study (i.e. the lack of specific antagonists/pore blockers to study GluD without affecting other iGluRs) and highlight the potential application of the ligand.

3) It would be helpful to define early and explicitly what the photoswitchable functional strategy is – that it is working via a pore block mechanism. In the Abstract, for example, instead of calling it “…a photoswitchable ligand.” how about just “…a photoswitchable pore blocker." The functional strategy – that you are generating a photoswitchable pore blocker – should also be explicitly stated in the Introduction, where right now it is touched on but not explicitly stated.

4) PTLs have been shown to generate a high local concentration of ligand to accelerate pharmacological response (and in this case, provide some level of specificity for a very non-specific, greasy cation), but it is hard to rationalize "absolute" pharmacological specificity claimed by the authors (Abstract, Discussion paragraph two). At the mid-μM concentrations required to elicit response, it seems unlikely that MAGu will not react with any other extracellular cysteines present in cells. Further, the guanidinium group by itself will certainly not direct the maleimide reactivity towards GluD2 over any other cation channel or electronegative protein surface. The language of this claim should be modified in the absence of other types of specificity assays.

5) Figure 4A. Please also show the difference current and contrast/compare to what is shown in Figure 2A. This would clarify the “voltage-independence” of action for those unfamiliar.

6) Figure 4D. It is not clear how the “ion channel” or red/green pore was generated. Is this from the structure or from some modeling? Please add details. This is an interesting figure, but it appears somewhat speculative and requires more details for the reader to understand its basis. What is driving the positioning of the trans MAGu? Is it being fixed? And what is driving the change in the coloration – presumed pore blocking by trans MAGu?

[Editors' note: further revisions were suggested prior to acceptance, as described below.]

Thank you for submitting your article "Probing the ionotropic activity of glutamate GluD2 receptor in HEK cells with genetically-engineered photopharmacology" for consideration by eLife. Your article has been reviewed by the three original peer reviewers, and the evaluation has been overseen by Merritt Maduke as the Reviewing Editor and Kenton Swartz as the Senior Editor. The following individuals involved in review of your submission have agreed to reveal their identity: Stephanie C Gantz (Reviewer #1); Lonnie Wollmuth (Reviewer #2); Anna Koster (Reviewer #3).

The reviewers have discussed the reviews with one another. They agree the revisions have greatly strengthened the manuscript but that a few additional relatively minor revisions are essential. The Reviewing Editor has drafted this decision to help you prepare a revised submission.

We would like to draw your attention to changes in our revision policy that we have made in response to COVID-19 (http://www.asadoresteatre.com/articles/57162). Specifically, we are asking editors to accept without delay manuscripts, like yours, that they judge can stand as eLife papers without additional data, even if they feel that they would make the manuscript stronger. Thus the revisions requested below only address clarity and presentation.

Summary:

Unlike other ionotropic glutamate receptors, GluD2 is not gated by glutamate. No specific or high-affinity chemical modulators that induce channel activity exist for this receptor. To address this challenge, the authors used a previously characterized photoswitchable tethered ligand (PTL) called MAGu to target a very non-specific blocker (pentamidine) to a new ion channel target (the GluD2 receptor). This approach (using this exact PTL) has been used to target knock-in cysteine mutants of the GABAA receptor in mouse brain slices and in vivo in an awake, behaving mouse. Based on this precedent, it is not unreasonable to believe that this tool could similarly be used for the GluD2 receptor, which would be a significant advance in the field for understanding the physiological role of this protein in disease. However, there are concerns about signal-to-noise, since the pore block by trans-MAGu is only a fraction of total presumed current through GluD; therefore, it has not yet been demonstrated that the MAGu response will be sufficient for physiological studies.

Revisions:

It is unclear how the block of DHPG-induced current was measured, especially in regards to the tonic current. Figure 5A top-left shows the tonic current and then the additional DHPG-induced current. The dashed line makes it seem that this is where the magnitude of the DHPG-induced current was measured from. But it would be best to measure DHPG-induced current as a change from the tonic inward current, which is perhaps what the red arrow is indicating? Including details in the figure legend would be helpful.

If the inhibition of DHPG-induced current included the tonic current, it may be substantially less than 22% block. If this is the case, we would suggest paired statistics to be sure the decrease in DHPG-current is significant, especially since there is some variability in both the mutant and WT conditions (seen in source data).

https://doi.org/10.7554/eLife.59026.sa1

## Author response

Essential revisions:

1) In order to evaluate the potential of this new tool, it is essential that the authors provide a more thorough quantification of its effects, as described below, which will be relevant to the signal:noise in physiological experiments.

The photo-switchable capability of MAGu is convincing, but what is not clear is how good of a blocker trans-MAGu is. It is clear that trans-MAGu is not producing full block of leak current, but it is not quantified. These data need to be added to have a clear understanding of the strategy and mechanism. cis-MAGu unblocks a portion of leak current, but it is not clear how much of the total leak current is blocked by trans-MAGu. Knowing how how much leak current is blocked by trans-MAGu is critical to interpreting the effect on mGluR-activated GluD2 current.

We understand the reviewers concern. To address this issue, we now have compared, for the best mutant (A654T-I677C), the degree of block with pentamidine (100 uM) and with transMAGu to that induced by NMDG, a large impermeant organic cation that blocks virtually all leak current (see new Figure 3A). We found that pentamidine blocks about 75% of the leak current, while trans MAGu blocks about 33% of the leak current.

In Figure 2C I677C, there is still substantial leak in 535 nm. The quantification in Figure 2C (% photoswitching) shows the % of I-Blockphoto over I-Blockpenta, but the arrows in the righthand trace, it would appear I-Blockphoto is actually the current unblocked. It would be helpful to quantify the amount of leak current blocked by trans-MAGu. In Figure 2C, the extent of block for photoswitching is being quantified relative to that for pentamidine, which is reasonable. However, for pentamidine, what is the concentration used for the experiments? Where is it at on the concentration-block curve for pentamidine? Presumably, if complete block the leak current should go to zero and hence the efficacy of block of photoswitching blocker would be less (e.g., Figure 4B). Please clarify. Additional discussion as the structural basis for incomplete block may also be helpful.

Blockphoto was renamed BlockMAGu for clarity, because block occurs also in darkness, and not just under green light. In Figure 2C, current is first recorded in darkness, when MAGu is in its trans, blocking state. Illuminating with 380 nm light converts MAGu to cis and relieves blockade. The arrow in the figure has now been moved to the dark period for clarity on how we quantify BlockMAGu.

For pentamidine, we used a saturating concentration (100 uM, see Figure 4B). This is now clearly stated both in the text and in the figure legend. Yet, even at saturating concentration, pentamidine does not block GluD fully (Figure 4B), as previously reported (Williams et al., 2003). Therefore, it makes sense indeed to quantify the degree of photoswitching by comparing the block under 535 nm light to that induced by NMDG. This is now done in Figure 3C, for the “best” mutant I677C (see response to previous comment).

In agreement with the study by Williams et al., 2003, we observed partial block of GluD2Lc current, even for saturating concentrations of pentamidine. It was suggested that pentamidine may easily permeate GluD channels. Our current manuscript does not add any new information regarding the structural basis for pentamidine block, and thus we believe adding such discussion would be too speculative.

How does MAGu work on the cysteine-engineered receptor that would presumably be used for future in vivo studies? Because the GluD2-I677C point mutant (lacking the L654T background) does not show current, the authors use the known effect of mGlu1 receptor agonism as a readout of GluD2-I677C activity in response to light and only see a 23% decrease in mGlu1 current--is this very small effect physiologically significant or to be expected? It seems like MAGu might be a useful tool to modulate GluD2 in Lurcher mice (which harbor the L654T mutation), but it is hard to know what the probe efficacy and usefulness is for evaluating the physiology of the WT GluD2 receptor in the absence of a way to measure a direct functional effect on the channel. How else might this be addressed?

The 23% decrease in mGlu1-induced current is not very different from the 33% photoswitching we are now reporting for MAGu blockade when compared to NMDG block (Figure 3A). We understand that such amount of photoswitching may appear small. However, the PTL technology has two important features that should relief the reviewers’ concerns.

First, the PTL technology is highly specific to the engineered cysteine mutation, as already demonstrated for a large variety of cysteine-substituted ion channels and receptors, including voltage-gated potassium channels, GABA receptors, metabotropic and ionotropic glutamate receptors and nicotinic acetylcholine receptors (see our recent review Paoletti et al., 2019 for a detailed list). This pharmacological specificity lies in the strong geometrical constrains between the tethered ligand and the engineered receptor, notably the strict distance between the attachment site (the cysteine mutation) and the ligand binding pocket. As a result, during the engineering step, we usually find that only one (or a handful at most) of the screened cysteine mutants can be made photocontrollable using this strategy. The PTL technology has now been deployed in complex neuronal setting, including mice and fish in vivo, with absolutely no adverse effects (see our review). Most importantly, the photoswitch MAGu we have used here has already been deployed in mice in vivo to control cysteine-substituted GABAA receptors (Lin et al., 2015), and showed absolutely no off-target effect on wild-type GABA receptors, glutamate receptors, or voltage-gated ion channels (see Figure S3 in Lin et al., 2015). We now have conducted similar controls (new Figure 5—figure supplement 1) and demonstrate in this article the specificity of our approach. We notably show that after MAGu treatment, GluD and GluA currents remain insensitive to the photostimulations, most likely because they lack a properly-positioned cysteine residue near the pore lumen.

Second, with PTLs, antagonism (or agonism) can be switched on and off within (milli)seconds, owing to the rapid kinetics of azobenzene photoswitching (see Figure 3B). This contrasts with traditional pharmacological compounds: the onset of their effect upon exposure takes a long time to develop and they can be difficult to remove, especially when they have high affinity, highly selectivity for their receptor. Hence, with PTLs, even a slight change in receptor activation efficacy can be reported rapidly with high fidelity and accuracy. This is for instance illustrated in Figure 5A: switching between 380 and 525 nm light rapidly and reversibly changes the holding current in the absence of DHPG. The light-induced current is small in amplitude, but because it can be reversibly switched on and off with high reproducibility, it unveils the existence of a tonically-activated GluD2 receptor in these cells.

For these reasons, we are confident that LiGLuD2 will be useful, in combination with a sensitive technique such as electrophysiology, to measure the direct functional effect of GluD2 photocontrol on neuronal physiology. Additionally, antagonism does not have to be complete to observe a strong physiological and/or phenotypical impact (see for instance Durand-de Cuttoli et al., 2018). The I677C mutant could be virally transduced in neuronal cells, either of the WT or the GluD2 knock-out mouse, with the idea of understanding when and where GluD2 is active. We don’t think Lurcher mice would help address this question.

These aspects of the PTL technology are now discussed in greater details in the Discussion section of the manuscript.

Discussion paragraph two states that the WT receptor is insensitive to MAGu, but it is not clear where those data are presented. It would be beneficial to show the magnitude of the DHPG-induced current in WT GluD2-expressing cells before and after addition of MAGu to address the possibility that MAGu affects the current irrespective of trans- or cis- conformation.

Our intention was to point that after MAGu treatment, the WT receptor remains insensitive to light, indicating that photocontrol is specific to the cysteine mutant. We agree with the reviewers that our sentence was misleading. We did not show that MAGu treatment per se has no effect on the WT receptor. Sentence was changed accordingly.

It is also not clear how MAGu will be selective for site-specific conjugation when introduced in a neuronal setting. Is it expected MAGu will react with any available cysteine? It would be helpful to discuss possible limitations going forward towards use in neurons.

The maleimide group of MAGu will indeed react with freely available cysteines, yet just on cell surface proteins (the highly reductive environment of the cytoplasm precludes maleimide conjugation to intracellular proteins). Nevertheless, as discussed above, this approach to photosensitizing receptors has proved to be remarkably specific in both fish and mice, with wild-type neurons or animals being unaffected by light after PTL treatment. Indeed, cysteine is a relatively infrequent amino acid (about 3%) often engaged in disulfide bonds (-S-S-) within or between polypeptide chains. Only free cysteines are available for PTL conjugation. In addition, there is a strong geometrical constrain for photocontrol: the cysteine residue has to be located at an ideal distance to the ligand binding pocket to result in photocontrol of protein activity. These limitations are now included in the Discussion.

2) The Introduction and Abstract are rather general and antiquated, to the disservice of the readers. It may be time to move away from the notion that ion channel function of GluD is debated. The authors have published many elegant studies demonstrating ion channel function. By appearances of the literature, the interpretation of these studies are not contested. In addition to pharmacology, ion channel function of GluD has been demonstrated using selective genetic strategies (e.g. Ady et al., 2013; Benamer et al., 2018; Gantz et al., 2020). To this end, the Abstract and Introduction should be changed. It does not seem fitting to state "direct evidence for ionotropic activity of GluD in neuronal setting [sic] is lacking" provided the studies referenced above. Broadly, the readers would benefit from restructuring of the Introduction and Abstract to state the specific issue addressed by the present study (i.e. the lack of specific antagonists/pore blockers to study GluD without affecting other iGluRs) and highlight the potential application of the ligand.

Even though some recent articles in the field still mention that GluD receptor do not generate ionic current, we agree that there is increasing evidence, from our group as well as from the lab of Stephanie Gantz, that GluD receptors can mediate slow excitatory current in neurons. Therefore, we modified the Introduction, and to lesser extent the Abstract, to better highlight the lack of (specific) pharmacology for GluDs, and the potential of our novel tool for the assessment of the ionotropic functions of GluDs in neurons.

3) It would be helpful to define early and explicitly what the photoswitchable functional strategy is – that it is working via a pore block mechanism. In the Abstract, for example, instead of calling it “…a photoswitchable ligand.” how about just “…a photoswitchable pore blocker." The functional strategy – that you are generating a photoswitchable pore blocker – should also be explicitly stated in the Introduction, where right now it is touched on but not explicitly stated.

We agree. We modified the text accordingly.

4) PTLs have been shown to generate a high local concentration of ligand to accelerate pharmacological response (and in this case, provide some level of specificity for a very non-specific, greasy cation), but it is hard to rationalize "absolute" pharmacological specificity claimed by the authors (Abstract, Discussion paragraph two). At the mid-μM concentrations required to elicit response, it seems unlikely that MAGu will not react with any other extracellular cysteines present in cells. Further, the guanidinium group by itself will certainly not direct the maleimide reactivity towards GluD2 over any other cation channel or electronegative protein surface. The language of this claim should be modified in the absence of other types of specificity assays.

MAGu will indeed react with other freely accessible cysteines on the cell surface, as discussed above. And because of the predictable low affinity of the guanidinium group for GluD2, we don’t expect any increase selectivity for I677C labeling. Yet, as discussed above, and as demonstrated with our supplementary figure, even though tethering is not specific to GluD2-I677C, photosensitization is. That said, we removed the word “absolute” which was somewhat excessive.

5) Figure 4A. Please also show the difference current and contrast/compare to what is shown in Figure 2A. This would clarify the “voltage-independence” of action for those unfamiliar.

We have modified Figure 4A for better comparison with Figure 2A. Instead of showing representative traces, we now show average traces + SEM for both wavelengths of light, together with the subtracted trace.

6) Figure 4D. It is not clear how the “ion channel” or red/green pore was generated. Is this from the structure or from some modeling? Please add details. This is an interesting figure, but it appears somewhat speculative and requires more details for the reader to understand its basis. What is driving the positioning of the trans MAGu? Is it being fixed? And what is driving the change in the coloration – presumed pore blocking by trans MAGu?

We added details to the Materials and methods and Results sections, to explain the modeling better. The geometry of the ion channel has been computed with MOLEonline webserver, with the “pore” mode. The resulting ion channel was color-coded as a function of the diameter of the channel allowing to illustrate the reduction of the diameter from a large (green) to a small (red) diameter. The maleimide moiety, i.e. the reactive part of MAGu, was forced to be in contact with the cysteine while the rest of the molecule was free to move.

[Editors' note: further revisions were suggested prior to acceptance, as described below.]

Revisions:

It is unclear how the block of DHPG-induced current was measured, especially in regards to the tonic current. Figure 5A top-left shows the tonic current and then the additional DHPG-induced current. The dashed line makes it seem that this is where the magnitude of the DHPG-induced current was measured from. But it would be best to measure DHPG-induced current as a change from the tonic inward current, which is perhaps what the red arrow is indicating? Including details in the figure legend would be helpful.

If the inhibition of DHPG-induced current included the tonic current, it may be substantially less than 22% block. If this is the case, we would suggest paired statistics to be sure the decrease in DHPG-current is significant, especially since there is some variability in both the mutant and WT conditions (seen in source data).

The magnitude of the DHPG current was not measured from the dashed line, but from the change in tonic inward current, as indicated by the red arrow in the figure indeed. Hence the block is 22%, not substantially less. This is now clearly stated in the figure legend.

https://doi.org/10.7554/eLife.59026.sa2

## Article and author information

### Author details

1. #### Damien Lemoine

Neuroscience Paris Seine - Institut de Biologie Paris Seine (NPS - IBPS), CNRS, INSERM, Sorbonne Université, Paris, France
##### Contribution
Data curation, Formal analysis, Writing - review and editing
##### Competing interests
No competing interests declared
2. #### Sarah Mondoloni

Neuroscience Paris Seine - Institut de Biologie Paris Seine (NPS - IBPS), CNRS, INSERM, Sorbonne Université, Paris, France
##### Contribution
Data curation, Visualization, Writing - review and editing
##### Competing interests
No competing interests declared
3. #### Jérome Tange

Neuroscience Paris Seine - Institut de Biologie Paris Seine (NPS - IBPS), CNRS, INSERM, Sorbonne Université, Paris, France
##### Contribution
Data curation
##### Competing interests
No competing interests declared
4. #### Bertrand Lambolez

Neuroscience Paris Seine - Institut de Biologie Paris Seine (NPS - IBPS), CNRS, INSERM, Sorbonne Université, Paris, France
##### Contribution
Supervision, Writing - review and editing
##### Competing interests
No competing interests declared
5. #### Philippe Faure

Neuroscience Paris Seine - Institut de Biologie Paris Seine (NPS - IBPS), CNRS, INSERM, Sorbonne Université, Paris, France
##### Contribution
Supervision, Funding acquisition, Writing - review and editing
##### Competing interests
No competing interests declared
6. #### Antoine Taly

1. CNRS, Université de Paris, UPR 9080, Laboratoire de Biochimie Théorique, Paris, France
2. Institut de Biologie Physico-Chimique-Fondation Edmond de Rothschild, PSL Research University, Paris, France
##### Contribution
Data curation, Funding acquisition, Visualization, Writing - review and editing
##### Contributed equally with
Ludovic Tricoire and Alexandre Mourot
##### Competing interests
No competing interests declared
7. #### Ludovic Tricoire

Neuroscience Paris Seine - Institut de Biologie Paris Seine (NPS - IBPS), CNRS, INSERM, Sorbonne Université, Paris, France
##### Contribution
Data curation, Supervision, Funding acquisition, Writing - review and editing
##### Contributed equally with
Antoine Taly and Alexandre Mourot
##### For correspondence
ludovic.tricoire@upmc.fr
##### Competing interests
No competing interests declared
8. #### Alexandre Mourot

Neuroscience Paris Seine - Institut de Biologie Paris Seine (NPS - IBPS), CNRS, INSERM, Sorbonne Université, Paris, France
##### Contribution
Conceptualization, Formal analysis, Supervision, Visualization, Writing - original draft
##### Contributed equally with
Antoine Taly and Ludovic Tricoire
##### For correspondence
alexandre.mourot@inserm.fr
##### Competing interests
No competing interests declared

### Funding

#### Agence Nationale de la Recherche (ANR-16-CE16-0014-01)

• Ludovic Tricoire

#### Fondation pour la Recherche Médicale (FRM EQU201903007961)

• Philippe Faure

• Antoine Taly

• Antoine Taly

#### LABEX Biopsy

• Damien Lemoine

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

### Acknowledgements

Authors would like to thank Nadine Mouttajagane and Manel Badsi for their help with molecular biology work. This work was supported by funding provided by the French Agency for Research (ANR-16-CE16-0014-01 to LT, ANR-11-LABX-0011 to AT), by the ‘Initiative d'Excellence’ (cluster of excellence LABEX Dynamo) to AT, by the Foundation for Medical Research (FRM, Equipe FRM EQU201903007961 to PF) and by a post-doctoral fellowship from the Labex BioPsy to DL.

### Senior Editor

1. Kenton J Swartz, National Institute of Neurological Disorders and Stroke, National Institutes of Health, United States

### Reviewing Editor

1. Merritt Maduke, Stanford University School of Medicine, United States

### Reviewers

1. Stephanie C Gantz, University of Iowa, United States
2. Lonnie Wollmuth, Stony Brook University, United States
3. Anna Koster, Stanford University, United States

### Publication history

2. Accepted: October 27, 2020
3. Accepted Manuscript published: October 28, 2020 (version 1)
4. Version of Record published: November 20, 2020 (version 2)

? 2020, Lemoine et al.

## Metrics

• 685
Page views
• 117
• 0
Citations

Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.

A two-part list of links to download the article, or parts of the article, in various formats.

### Open citations (links to open the citations from this article in various online reference manager services)

1. Biochemistry and Chemical Biology

# Bacterial OTU deubiquitinases regulate substrate ubiquitination upon Legionella infection

Donghyuk Shin et al.
Research Article Updated

Legionella pneumophila causes a severe pneumonia known as Legionnaires’ disease. During the infection, Legionella injects more than 300 effector proteins into host cells. Among them are enzymes involved in altering the host-ubiquitination system. Here, we identified two LegionellaOTU (ovarian tumor)-like deubiquitinases (LOT-DUBs; LotB [Lpg1621/Ceg23] and LotC [Lpg2529]). The crystal structure of the LotC catalytic core (LotC14-310) was determined at 2.4 ?. Unlike the classical OTU-family, the LOT-family shows an extended helical lobe between the Cys-loop and the variable loop, which defines them as a unique class of OTU-DUBs. LotB has an additional ubiquitin-binding site (S1’), which enables the specific cleavage of Lys63-linked polyubiquitin chains. By contrast, LotC only contains the S1 site and cleaves different species of ubiquitin chains. MS analysis of LotB and LotC identified different categories of host-interacting proteins and substrates. Together, our results provide new structural insights into bacterial OTU-DUBs and indicate distinct roles in host–pathogen interactions.

1. Biochemistry and Chemical Biology
2. Microbiology and Infectious Disease

# Inhibition of SARS-CoV-2 viral entry upon blocking N- and O-glycan elaboration

Qi Yang et al.
Research Article Updated

The Spike protein of SARS-CoV-2, its receptor-binding domain (RBD), and its primary receptor ACE2 are extensively glycosylated. The impact of this post-translational modification on viral entry is yet unestablished. We expressed different glycoforms of the Spike-protein and ACE2 in CRISPR-Cas9 glycoengineered cells, and developed corresponding SARS-CoV-2 pseudovirus. We observed that N- and O-glycans had only minor contribution to Spike-ACE2 binding. However, these carbohydrates played a major role in regulating viral entry. Blocking N-glycan biosynthesis at the oligomannose stage using both genetic approaches and the small molecule kifunensine dramatically reduced viral entry into ACE2 expressing HEK293T cells. Blocking O-glycan elaboration also partially blocked viral entry. Mechanistic studies suggest multiple roles for glycans during viral entry. Among them, inhibition of N-glycan biosynthesis enhanced Spike-protein proteolysis. This could reduce RBD presentation on virus, lowering binding to host ACE2 and decreasing viral entry. Overall, chemical inhibitors of glycosylation may be evaluated for COVID-19.