Pharmacological profiles and anti-inflammatory activity of pCN-diEPP and mCN-diEPP, new alpha9alpha10 nicotinic receptor ligands

Abstract

Pain due to inflammation can be reduced by targeting the noncanonical nicotinic receptors (NCNR) in cells of the immune system that regulate the synthesis and release of pro- and anti-inflammatory cytokines. Although NCNR do not generate ion channel currents, the pharmacology of ion-channel forms of the receptors can predict drugs which may be effective regulators of the cholinergic anti-inflammatory system (CAS). Agonists of a7 type receptors have been definitively associated with CAS. Receptors containing a9 and a10 subunits have also been implicated, and selective a9a10 antagonist conopeptides have been shown to effectively reduce inflammatory pain. We have recently characterized two small molecules, pCN-diEPP and mCN-diEPP, as selective a9a10 agonists and antagonists, respectively. We used these drugs, along with nicotine, an a7 agonist and a9a10 antagonist, to probe the mixed populations of receptors that are formed when a7, a9, and a10 are all expressed together in Xenopus oocytes. We also evaluated the effects of the CN-diEPP compounds on regulating the ATP-induced release of interleukin-1β from monocytic THP-1 cells, which express noncanonical forms of these receptors. The compounds successfully identified separate populations of receptors when all three subunits were co-expressed, including a potential population of homomeric a10 receptors. The a9a10 agonist pCN-diEPP was the more effective regulator of interleukin-1β release in THP-1 cells. pCN-diEPP was also fully effective in a mouse model of inflammatory pain, while mCN-diEPP had only partial effects, requiring a higher dosage. The analgetic effects of pCN-diEPP and mCN-diEPP were retained in a7 knockout mice.

Introduction

In light of the present opiate epidemic, which has been precipitated by the widespread use of prescription opiates as pain killers, there is an acute need to develop alternative non-addicting analgetic therapies. The discovery that cells of the immune system have nicotinic acetylcholine receptors (nAChRs) that can be stimulated to decrease the release of pro-inflammatory cytokines (Borovikova et al., 2000) has generated a great deal of interest in the cholinergic anti-inflammatory systems (CAS) as new avenues for treating pain associated with inflammation (Hone and McIntosh, 2023). Many studies have demonstrated the essential role played by homomeric a7 nAChR in CAS (Bencherif et al., 2011; Fujii et al., 2017; Piovesana et al., 2021; Treinin et al., 2017) and, more recently, by a9* nAChR (a9 homomeric receptors or, more often, a9 co-assembled with a10 subunits), have been have identified as alternative targets in CAS (AlSharari et al., 2020; Hone et al., 2018). Although only partially understood, numerous downstream signaling mechanisms are activated in response to nAChR stimulation of immune cells (for review: (Richter and Grau, 2023)). On the one hand, stimulation of mainly a7 nAChRs down-regulates the expression of pro-inflammatory mediators on the transcriptional and translational level (for review: (Richter and Grau, 2023)). On the other hand, stimulation of nAChRs containing subunits a9, a7, and/or a10 inhibits the function of the ATP-sensitive P2X7 receptor and down-regulates the maturation and release of inflammasome-dependent cytokines including interleukin (IL)-1β (Hecker et al., 2015; Richter et al., 2016; Zakrzewicz et al., 2017). Of note, the P2X7 receptor and IL-1β play an important role in pain (Ren and Illes, 2022).

Although rapidly desensitizing and with low probability of channel opening (Papke and Horenstein, 2021), a-bungarotoxin-sensitive a7 nAChRs (Amar et al., 1993) have been widely studied as modulators of brain function and considered as potential targets for disorders of the central nervous system such as Alzheimer’s disease (Bouzat et al., 2018; Russo et al., 2014) and schizophrenia (Terry and Callahan, 2020). The first functional roles identified for a9* nAChRs were as regulators of auditory function (Elgoyhen et al., 2001; Vetter et al., 1999). Both a7 and a9* nAChRs form functional ligand-gated ion channels in heterologous expression systems such as Xenopus oocytes, which has allowed for extensive characterization of their pharmacological properties, especially in the case of a7 nAChRs (Papke and Horenstein, 2021). Early studies of a-bungarotoxin-sensitive a9* nAChRs highlighted several unusual pharmacological features, including the fact that nicotine, the eponymous agonist for the receptor family, was an antagonist rather than an agonist (Verbitsky et al., 2000). One of the most selective ligands of a9* nAChRs is the conopeptide antagonist alpha-RgIA (Azam and McIntosh, 2012), although recently other small molecule antagonists have been proposed to be a9 selective (Bavo et al., 2022).

Consistent with their characterization in heterologous expression systems, a7 and a9* nAChR have been shown to mediate electrical signaling in the brain (Frazier et al., 1998) and inner ear (Fuchs, 1996), respectively. Additionally, the expression of a9 and a10 has been reported in dorsal root sensory ganglia (Lips et al., 2002). However, whether the expression of either a7 or a9* in cells of the immune system that mediate CAS is associated with ion channel activation is highly questioned. These receptors have, rather, been proposed to function as metabotropic receptors (Richter and Grau, 2023) and therefore may be considered “noncanonical nicotinic receptors” (NCNR). The mechanism accounting for this different form of receptor function is unknown, and while there is, in general, a concurrence between the pharmacology established with heterologous expression systems and ligands modulating CAS, there are also remarkable differences, such that ligands that produce little or no channel activation have been shown to be able to mediate CAS signaling (Papke et al., 2023; Piovesana et al., 2021): NS6740 in the case of a7 nAChRs (Papke et al., 2015; Thomsen and Mikkelsen, 2012) and phosphocholine in the case of a9* nAChRs (Richter et al., 2016). The conformational dynamics of ligand-gated ion channels involves the induction of both conducting and non-conducting (i.e. desensitized) states. Even the most efficacious of a7 agonists more effectively induce desensitization than channel activation (Papke and Lindstrom, 2020), and some ligands promote desensitization without any channel activation. Such ligands have been referred to as “silent agonists”, and their ability to induce desensitized conformations can be confirmed through the use of type II positive allosteric modulators such as PNU-120596 that can reactivate the desensitized receptors (Papke et al., 2023). Numerous silent agonists have been shown to be effective modulators of CAS, and it has been proposed that for the non-conducting forms of the receptors in immune cells, the activation of a metabotropic pathway is associated with non-conducting conformations corresponding to the desensitized states of channel-forming receptors (Horenstein and Papke, 2017; Papke et al., 2015). Based on the metabotropic activity of ligands like phosphocholine, it has been suggested that similar mechanisms account for CAS function of a9* NCNR (Zakrzewicz et al., 2017).

A recent study compared the metabotropic activity of the weak a7 partial agonist pCF3-diEPP (Quadri et al., 2018) and phosphocholine in various immune cell preparations (Richter et al., 2022). It was found that the activity of pCF3-diEPP could be blocked by either the a7-specific conopeptide [V11L,V16D]ArIB (Whiteaker et al., 2008) or by the a9a10 nAChR-specific conopeptide RgIA4 (Christensen et al., 2017). Consistent with this, it was shown that, in addition to being an weak a7 partial agonist, pCF3-diEPP was a more efficacious partial agonist of a9 nAChRs. This inspired a follow-up study of a family of 1,1-diethyl-4-phenylpiperazin-1-ium compounds (diEPPs) that were previously studied for their a7 silent agonist activity (Quadri et al., 2016), which identified several a9-selective agonists and antagonists (Papke et al., 2022).

While a7-mediated CAS activity is associated with receptor activation and/or desensitization (Papke et al., 2023), studies with a9 conopeptide antagonists (AlSharari et al., 2020; Christensen et al., 2017; Huynh et al., 2019; Pacini et al., 2016; Romero et al., 2017) suggested that a9 CAS activity would arise from a9 inhibition in vivo (Toma et al., 2020). However, this was inconsistent with the effect of the conopeptide in the cell-based assays, which suggested a9 nAChR agonism was the basis for the CAS effects.

In this paper, we address this controversy by using two closely related compounds, the a9 nAChR agonist pCN-diEPP and the a9 nAChR antagonist mCN-diEPP (Papke et al., 2022), characterizing their activity with mixed populations of a7 and a9a10 nAChRs in Xenopus oocytes, with human monocytic THP-1 cells, and with an animal model of inflammatory pain using both wild-type (WT) and a7-knockout (KO) mice.

Note that, in general, heterologous expression of nAChR takes advantage of the fact that each receptor subtype can be studied in isolation. However, in the in vivo and cell-based assays, both a7 and a9a10 nAChRs may be present in the same cells. Therefore, some experiments were conducted in oocytes expressing a7 along with a9 and a10 to validate the usefulness of our probe compounds with mixed receptor populations.

Methods

pCN-diEPP and mCN-diEPP

4-(4-cyanophenyl)-1,1-diethylpiperazin-1-ium (pCN-diEPP) and 4-(3-cyanophenyl)-1,1-diethylpiperazin-1-ium (mCN-diEPP) were synthesized as previously documented (Quadri et al., 2016). For oocyte experiments 100 mM stock solutions were prepared in dimethyl sulfoxide and stored in −20°, then freshly diluted in Ringer’s solution.

Expression of human nAChR subunits in Xenopus laevis oocytes

Plasmid DNA encoding the human a7 nAChR was obtained from Jon Lindstrom (University of Pennsylvania, Philadelphia, PA). The human resistance-to-cholinesterase 3 (RIC3) clone was obtained from Millet Treinin (Hebrew University, Jerusalem, Israel) and RNA co-injected with a7 to improve the level and speed of receptor expression without affecting the pharmacological properties (Halevi et al., 2003). Plasmid DNA encoding the human a10 nAChR was obtained from J. Michael McIntosh (University of Utah, Salt Lake City, UT). Plasmid DNA encoding the human a9 nAChR and the human receptor-associated protein of the synapse (RAPSYN) with codon optimization for expression in Xenopus laevis were obtained from Eurofins Genomics (Ebersberg, Germany). RAPSYN RNA was co-injected with the a9 and a10 to improve expression (Richter et al., 2016). After linearization and purification of the plasmid DNAs, RNAs were prepared using the mMessage mMachine in vitro RNA transcription kit (Ambion, Austin, TX).

Frogs were maintained in the Animal Care Service facility of the University of Florida, and all procedures were approved by the University of Florida Institutional Animal Care and Use Committee (approval number 202002669). In brief, the animals were first anesthetized for 15–20 min in 1.5 l frog tank water containing 1 g of MS-222 buffered with sodium bicarbonate. Oocytes were obtained surgically from mature female Xenopus laevis (Nasco, Ft. Atkinson, WI) and treated with 1.4 mg/ml type 1 collagenase (Worthington Biochemicals, Freehold, NJ, USA) for 2–4 h at room temperature in calcium-free Barth’s solution (88 mM NaCl, 1 mM KCl, 2.38 mM NaHCO₃, 0.82 mM MgSO₄, 15 mM HEPES, and 12 mg/l tetracycline, pH 7.6) to remove the ovarian tissue and the follicular layers. Stage V oocytes were injected with 4–6 ng a7 RNA and 2–3 ng RIC3 RNA (2:1 ratio) in 50 nl water, or with 12 ng a9 RNA, or along with 12 ng a10 RNA, and 3 ng RAPSYN RNA in 50 nl water. Oocytes were maintained in Barth’s solution containing additionally 0.32 mM Ca(NO₃)₂ and 0.41 mM CaCl₂, and recordings were carried out 2–20 days after injection.

In order to obtain mixed populations of a7 and a9a10 receptors, RNA for each nAChR subunit and their supporting chaperone proteins (RIC3 for a7 and RAPSYN for a9* receptors) were injected at varying ratios. Since a7 expression is consistently more robust, it was necessary to inject a9 and a10 at higher levels. The injections were as indicated in Table 2.

Table 2.

RNA values in ratio experiments

	α7	RIC3	α9	α10	RAPSYN
α7 alone	6 ng	3 ng
α7, α9α10 1:2	4 ng	2 ng	8 ng	8 ng	2 ng
α7, α9α10 1:3	3.6 ng	18 ng	10.7 ng	10.7 ng	2.5 ng
α7, α9α10 1:4	2.48 ng	1.24 ng	12 ng	12 ng	2.7 ng
α9α10 alone			12 ng	12 ng	3.0 ng

Two-electrode voltage-clamp electrophysiology

Two-electrode voltage-clamp experiments were conducted using OpusXpress 6000A (Molecular Devices, Union City CA, USA) (Papke and Smith-Maxwell, 2009). Both the voltage and current electrodes were filled with 3 M KCl. Oocytes were voltage-clamped at −60 mV at RT. The oocytes were perfused with Ringer’s solution (115 mM NaCl, 2.5 mM KCl, 1.8 mM CaCl₂, 10 mM HEPES, 1 μM atropine, pH 7.2) at 2 ml/min. NaCl, KCl, CaCl₂, and HEPES came from Fisher Scientific (Waltham MA, USA), and unless otherwise noted, all other salts and chemicals were obtained from Sigma-Aldrich (St. Louis MO, USA). To evaluate the effects of experimental compounds, responses were compared to control ACh-evoked responses, defined as the average of two initial applications of 60 μM ACh made before test applications. Drug applications were 12 s in duration, followed by 181 s washout periods.

Each episode of data acquisition was a total of 210 s and included an initial 30 s period used to define the baseline for the drug-evoked responses. After 30 s, drugs were applied, and the following 120 s were defined as the drug response period for analysis. The responses were calculated as both peak-current amplitudes and net charge, as previously described (Papke and Papke, 2002). Data were collected at 50 Hz, filtered at 20 Hz, and analyzed by Clampfit (Molecular Devices) and Excel (Microsoft, Redmond WA, USA). Data for concentration-response studies were expressed as averages ± SEM from at least five oocytes for each experiment and plotted with Kaleidagraph 4.5.2 (Synergy Software, Reading PA, USA). The values for the curve fits were generated using the Levenberg-Marquardt algorithm to obtain the best Chi-Square fit to the Hill equation using the Kaleidagraph 4.5.2 plotting program. The errors in Table 1 are the calculated standard errors of the fit parameters based on the goodness of fit.

Table 1.

Curve fit parameters from Figure 1

pCN-diEPP activation of α9α10
Imax	n	EC50	ChiSq	R
0.94 ± 0.06	1.07 ± 0.2	2.57 ± 0.57 μM	0.012	0.993
mCN-diEPP inhibition of α9α10
	n	IC50	ChiSq	R
	−1.41 ± 0.16	17.1 ± 1.54 μM	0.0083	0.996
Nicotine activation of α7
Imax	n	EC50	ChiSq	R
0.85 ± 0.04	1.45 ± 0.34	14.9 ± 2.8 μM	0.018	0.991
Nicotine inhibition of α9α10
	n	IC50	ChiSq	R
	−1.98 ± 0.53	437 ± 73 μM	0.044	0.982

Multi-cell averages were calculated for comparisons of complex responses. Averages of the normalized data were calculated for each of the 10,322 points in each of the 206.44 s traces (acquired at 50 Hz), as well as the standard errors for those averages. All other data are plotted as the individual replicates normalized to the ACh control responses obtained from the same cells. ANOVA was conducted in Kaleidagraph 4.5.2.

Monocytic THP-1 cells

Monocytic THP-1 cells were obtained from the German Collection of Microorganisms and Cell Cultures (Braunschweig, Germany). Cells were cultured in RPMI 1640 medium (Capricorn Ebsdorfergrund, Germany, Cat# RPMI-A) supplemented with 10% FBS from Capricorn (Cat# FBS-16A) under 5% CO₂ atmosphere at 37°C. To investigate IL-1β release, monocytic cells were resuspended in FBS-free RPMI medium. 0.5 × 10⁶ cells/0.5 ml per well were seeded in 48-well plates (Greiner Bio-One, Frickenhausen, Germany;). Cells were primed for 5 h with 1 μg/ml LPS (E. coli O26:B6, Merck, Darmstadt, Germany, Cat# L2654) as described previously (Richter et al., 2022). Then the P2X7 receptor agonist BzATP (2’(3’)-O-(4-benzoyl-benzoyl)ATP trieethylammonium salt; Jena Bioscience, Jena, Germany, Cat# NU-1620–5) was added for 40 min in the absence or the presence of different concentrations of cholinergic agonists: ACh (Merck, Cat# A6625), phosphocholine (Merck, Cat# P0378), pCN-diEPP, or mCN-diEPP. After treatment, cells were spun down (500 g, 8 min, 4°C) and the supernatants were collected and stored at −20°C for later IL-1β and lactate dehydrogenase (LDH) measurements.

IL-1β concentrations in cell-free supernatants obtained from experiments on THP-1 cells were measured using the Human IL-1β/IL-1F2 DuoSet enzyme-linked immunosorbent assay (ELISA) from R&D Systems (Minneapolis MN, USA, Cat# DY201) according to the supplier’s instructions. To test for cell viability at the end of the cell culture experiments, LDH activity was measured using the CytoTox 96^® Non-Radioactive Cytotoxicity Assay (Promega, Madison WI, USA; Cat# G1780) according to the supplier’s instructions. LDH activities determined in cell-free supernatants are given as percentage of the total LDH activity of lysed control cells.

Statistical analysis of cytokine data

Statistical analysis was performed using GraphPad Prism 9 (GraphPad Software, La Jolla CA, USA). Results obtained in the cell-based assays on BzATP-induced cytokine release were analyzed using SPSS (Version 23, IBM, Armonk, NY, United States) visualized using Inkscape version 0.48.5 r10040 (Free and Open Source Software licensed under the GPL).

The n-numbers (≤ 6) in the experiments with were too small to determine a normal distribution with certainty. Therefore, non-paramateric test were used. Dependent data sets were analyzed first by the Friedman test. When the P value was below 0.05, the two-tailed Wilcoxon signed-rank test was performed for pairwise comparison.

All numbers (n) of the individual experiments and details on the type of statistical test used for each experiment are indicated in the Results section and Figure legends. No outliers were excluded from the analyses.

Measurements of inflammatory pain

Animals

Experiments were conducted using adult (10–15 weeks) male and female C57BL/6J mice from Jackson Laboratory (Bar Harbor ME, USA): a7 WT and KO mice on a C57BL/6J background. Mice null for the a7 subunit along with their WT littermates were initially procured from Jackson Laboratory and later bred in an approved animal care facility at Virginia Commonwealth University. The breeding scheme involved crossing heterozygous mice and backcrossing progeny for at least 12 to 15 generations, to control for irregularities that might occur crossing solely mutant animals, to generate both mutant and WT animals. Then mice were weaned at 21 days of age and subsequently housed in groups of two to five with Teklad corn cob bedding (#7097, Envigo Teklad, Madison WI, USA). Initially, they were maintained in a temperature- and humidity-controlled vivarium space (21 ± 3◦C, 55 ± 10%) on a 12-h light/dark cycle (lights on at 7:00 AM) with free access to food (Teklad LM-485 mouse sterilized diet, Harlan Laboratories Inc., Indianapolis IN, USA) and water until needed. Then mice were retrieved from the vivarium and housed (4–5 mice per cage) for the duration of the study in a temperature- and humidity-controlled out-of-vivarium space on the same light/dark cycle. They were given ad libitum food and water. All experiments were performed during the light cycle. This study was approved by the Institutional Animal Care and Use Committee of Virginia Commonwealth University (approval # AM10142) and carried out in accordance with the National Institutes of Health’s Guide for the Care and Use of Laboratory Animals. All experimental animals were included in further behavioral testing and none of them showed behavioral disturbances unrelated to the pain induction procedure.

Induction of inflammatory pain by complete Freund’s adjuvant (CFA)

We explored the effects of pCN-diEPP and mCN-diEPP in the CFA test, composed of inactivated and dried Mycobacterium tuberculosis and adjuvant, a widely used model of persistent inflammatory pain (Kaliyaperumal et al., 2020). CFA was purchased from Sigma-Aldrich (St. Louis MO, USA). The CFA model is based on hypersensitivity, paw swelling, and nuclear factor-κB–mediated transcription of tumor necrosis factor α involved in the formation of the principal mediators of inflammation (Hartung et al., 2015). Mice were injected intraplantarly with 20 μl of CFA (50%, diluted in mineral oil; Sigma-Aldrich). Mechanical sensitivity (see the measurement of the von Frey test) was measured before and 3 days after CFA injection. pCN-diEPP (0.1, 0.3, and 1 mg/kg) and mCN-diEPP (0.6, 1, and 3 mg/kg), dissolved in a mixture of 1:1:18 [1 volume ethanol/1 volume Emulphor-620 (Rhone-Poulenc, Inc., Princeton NJ, USA)/18 volumes distilled water] or vehicle was injected intraperitoneally (i.p.) on day 3 after CFA injection, and mice were tested for mechanical sensitivity at different time points (1, 3, and 6 h) after drug injection.

Evaluation of mechanical sensitivity

Mechanical sensitivity thresholds were determined according to the method of Chaplan et al, (Chaplan et al., 1994) and as adapted in Bagdas et al., 2015 (Bagdas et al., 2015). A series of calibrated von Frey filaments (Stoelting, Wood Dale IL, USA) with logarithmically incremental stiffness ranging from 2.83 to 5.07 expressed as diameter sensitivity (ds) log 10 of 10 x force (in milligrams) was applied to the paw with a modified up-down method (Dixon, 1965). The mechanical threshold was expressed as log 10 of 10 x force (in milligrams), indicating the force of the von Frey hair to which the animal reacted (paw withdrawn, licking, or shaking). All behavioral testing on animals was performed in a blinded manner.

The data obtained were discrete values that were not normally distributed and, hence, not suitable for parametric analysis. The data were therefore first evaluated with the non-parametric Kruskal–Wallis test, a one-way analysis of variance by ranks. A significant Kruskal–Wallis test indicated that at least one sample stochastically dominated the other samples, in this case, the baseline data for all groups. For analyzing the specific sample pairs at the different time points for stochastic dominance, as a second step the groups were tested pairwise with the Mann–Whitney test. The P values calculated by the Mann-Whitney test are provided in the tables or shown in the figures.

Locomotor activity

Mice were placed into individual Omnitech photocell activity cages (28 x 16.5 cm) (Columbus OH, USA) 1 h after administration of either vehicle, pCN-diEPP (1 mg/kg, i.p.), or mCN-diEPP (3 mg/kg, i.p.). Interruptions of the photocell beams (two banks of eight cells each) were then recorded for the next 60 min. Data are expressed as number of photocell interruptions.

Body temperature

Rectal temperature was measured by a thermistor probe (inserted 24 mm) and digital thermometer (Yellow Springs Instrument Co., Yellow Springs OH, USA). Readings were taken just before and at 60 min after the injection of either vehicle, pCN-diEPP (1 mg/kg, i.p.), or mCN-diEPP (3 mg/kg, i.p.) (n = 8/group). The difference in rectal temperature before and after treatment was calculated for each mouse. The ambient temperature of the laboratory varied from 21–24°C from day to day.

Results and Discussion

Characterization of probe compounds for a7 and a9a10 receptors

We evaluated several compounds for their ability to distinguish a7 from a9a10 nAChR in anticipation of doing experiments with mixed populations of receptors. We used agonists and antagonists that were selective for one nAChR population or another.

Position of the cyano group determines the activity of CN-diEPPcompounds

pCN-diEPP is a partial agonist for a7 nAChRs that produced substantial PNU-120596-sensitive desensitization (Quadri et al., 2016), while mCN-diEPP produced no activation and only modest inhibition of a7 nAChR (for structures, see Figure 1). When characterized on a9 homomeric nAChRs, pCN-diEPP was shown to be a full agonist with an EC₅₀ of 0.368 ± 0.10 μM, while mCN-diEPP was an antagonist with an IC₅₀ of 21 ± 2.71 μM (Papke et al., 2022). Similar results were obtained with heteromeric a9a10 receptors (Figure 1A). pCN-diEPP was a very efficiacious agonist with an I_max 0.94 ± 0.06 that of ACh and an EC₅₀ of 2.57 ± 0.57 μM, while mCN-diEPP antagonized the ionotropic responses of a9a10 nAChR to ACh (60 μM) with an IC₅₀ of 17.1 ± 1.57 μM (Papke et al., 2022). Curve fit parameters and error estimates are provided in Table 1, and plots of the individual replicates are provided in Supplemental Figure 1A.

Figure 1.

Concentration-response studies with probe compounds. A) Cells co-expressing human a9, a10, and RAPSYN were evaluated for their sensitivity to pCN-diEPP as an agonist or mCN-diEPP as an antagonist of 60 μM ACh-evoked control responses. Structures are shown as inserts. Plotted are the average peak current responses of at least 6 cells at each concentration (± SEM). Data were fit to the Hill equation with either a positive Hill coefficient for agonist activity or a negative Hill coefficient for antagonist activity. Responses to pCN-diEPP are relative to ACh maximum. Individual replicates are plotted in Supplemental Figure 1A. Responses to co-applications of mCN-diEPP are relative to ACh alone. Curve fit parameters are provided in Table 1. B) Nicotine as an agonist of a7 and an antagonist for a9a10 receptors. Plotted are the average peak current responses of at least 6 cells at each concentration (± SEM). Data were fit to the Hill equation with either a positive Hill coefficient for agonist activity or a negative Hill coefficient for antagonist activity. Responses of a7 nAChR are relative to ACh maximum. Responses of a9a10 nAChR to co-applications of nicotine are relative to responses to ACh alone. Curve fit parameters are provided in Table 1. Individual replicates are plotted in Supplemental Figure 1B.

Nicotine as a probe to distinguish a7 from a9a10 receptors

Nicotine is an efficacious agonist of a7 nAChR (Papke et al., 2007), yet it has been reported to be an antagonist of a9 nAChR ion channel function (Verbitsky et al., 2000), so that nicotine responses may be useful to separate components of mixed receptor populations (Figure 1B). The I_max of nicotine for a7 net-charge response was 85 ± 4% that of ACh, with an EC₅₀ of 14.9 ± 2.8 μM. The IC₅₀ for the inhibition of a9a10 peak current responses was 437 ± 73 μM (Figure 1B). The IC₅₀ of nicotine at a9a10 nAChR heteromers seems to be high compared to the published IC₅₀ of 32 μM at a9 homomers (Verbitsky et al., 2000). However, the previous study used ACh at a concentration of 10 μM, while we used 60 μM ACh for stimulation, which may account at least in part for the different IC₅₀ values. Curve fit parameters and error estimates are provided in Table 1, and plots of the individual replicates are provided in Supplemental Figure 1B.

Probing mixed populations of receptors

While there are likely to be factors in vivo that regulate the subunit composition of nAChR formed in specific cells, expression in Xenopus oocytes may be more permissive and able to be regulated by mass action. For example, it is a common practice to bias the subunit composition of a4b2-containing nAChR by altering the ratio of a4 and b2 RNA injected (Zwart et al., 2006). Therefore, Xenopus oocytes were injected with RNA for a7 alone, a9 together with a10 alone, or a7 in combination with a9 and a10 at various ratios (see Table 2) to determine whether receptors expressed would be mixed populations of typical a7 and a9a10 nAChRs or perhaps contain receptors with mixed a7 and a9* properties.

In addition to having differing responses to the pharmacological probes illustrated in Figure 1, due to the unique rapid concentration-dependent desensitization of a7 receptors (Papke and Horenstein, 2021), ACh control responses are also distinctly different between a7 and a9* nAChRs (Papke et al., 2022) (Figure 2A), with a9a10 responses having an initial peak followed by later current, resulting in an increased ratio of net charge-to-peak current. With the higher levels of a9a10 expression, the response wave forms took on the characteristic shape of the a9a10 ACh responses, having the higher ratio of charge-to-peak current. At the lower two ratios of a9a10 to a7, the co-expression a9 and a10 had the effect of decreasing ACh responses compared to a7 expressed alone in both peak and net charge (Figure 2B, see Table 3 for ANOVA). We should be very circumspect in comparing one set of injected Xenopus oocytes to another, since both a7 and a9a10 RNA amounts were varied. However, the observation that at the 1:2 and 1:3 ratios, mixed a7 and a9a10 expression seemed to reduce both peak currents and net charge in response to ACh is interesting. It might suggest that a7a9a10 nAChR heteromers could be forming in Xenopus oocytes that do not function as ligand-gated ion channels.

Figure 2.

ACh control responses of cells expressing a7, a9 plus a10, or all three nAChR subunits at the various ratios indicated (see Methods and Table 2 for RNA ratios). A) Averaged first 60 μM ACh control responses (dark line) and SEM (tan shaded areas) of 6–7 cells, calculated for each of 10,000 points over 210-second acquisition periods. B) Data on individual responses (two responses from each of 6–7 cells). Both peak current amplitude and net charge (Papke and Papke, 2002) over a 120 s interval following the start of the drug application were measured. Additionally, the ratio of peak currents (μA) and net charge (millicoulombs) were calculated and plotted (see Table 3 for statistical analyses).

Table 3.

Statistical analysis of Control responses in Figure 2

Table 3A Peak currents
One Way ANOVA
One Way ANOVA
Data Table: Three ratios controls
Factor A: 5 Groups
a7 peak, 1 to 2 peak, 1 to 3 peak, 1 to 4 peak, a9a10 peak
Analysis of Variance Results
Source	DF	SS	MS	F	P
Total	69	1196.6266	17.342414
A	4	411.19963	102.79991	8.5074672	< .0001
Error	65	785.42695	12.083491

Bonferroni’s All Pairs Comparison
Comparison	Mean Difference	|t|	P	95% CL
α7 peak vs 1 to 2 peak	6.04695	4.6025	0.0002	2.2289 to 9.865
α7 peak vs 1 to 3 peak	6.95562	5.2941	< .0001	3.1375 to 10.774
α7 peak vs 1 to 4 peak	3.8897	2.9605	0.0428	0.071615 to 7.7078
α7 peak vs a9a10 peak	3.33386	2.5375	0.1357	−0.48422 to 7.1519
1 to 2 peak vs 1 to 3 peak	0.90867	0.6916	1	−2.9094 to 4.7268
1 to 2 peak vs 1 to 4 peak	−2.15726	1.6419	1	−5.9753 to 1.6608
1 to 2 peak vs α9α10 peak	−2.71309	2.065	0.4292	−6.5312 to 1.105
1 to 3 peak vs 1 to 4 peak	−3.06593	2.3335	0.2272	−6.884 to 0.75215
1 to 3 peak vs α9α10 peak	−3.62176	2.7566	0.0757	−7.4398 to 0.19632
1 to 4 peak vs α9α10 peak	−0.555837	0.4231	1	−4.3739 to 3.2622

Table 3B. Net Charge
One Way ANOVA
Data Table: Three ratios controls
Factor A: 5 Groups
α7 charge, 1 to 2 charge, 1 to 3 ratio, 1 to 4 charge, α9α10 charge
Analysis of Variance Results
Source	DF	SS	MS	F	P
Total	67	18924.291	282.45211
A	4	7757.1161	1939.279	10.940509	< .0001
Error	63	11167.175	177.25675

Bonferroni’s All Pairs Comparison
Comparison	Mean Difference	|t|	P	95% CL
α7 charge vs 1 to 2 charge	20.1261	3.8426	0.0029	4.8885 to 35.364
α7 charge vs 1 to 3 ratio	20.1255	3.8425	0.0029	4.8879 to 35.363
α7 charge vs 1 to 4 charge	−3.85703	0.7364	1	−19.095 to 11.381
α7 charge vs α9α10 charge	−0.12008	0.0229	1	−15.358 to 15.118
1 to 2 charge vs 1 to 3 ratio	−0.000564022	0.0001	1	−14.64 to 14.639
1 to 2 charge vs 1 to 4 charge	−23.9831	4.766	0.0001	−38.623 to −9.3433
1 to 2 charge vs α9α10 charge	−20.2462	4.0234	0.0016	−34.886 to −5.6064
1 to 3 ratio vs 1 to 3 charge	−23.9826	4.7659	0.0001	−38.622 to −9.3428
1 to 3 ratio vs α9α10 charge	−20.2456	4.0233	0.0016	−34.885 to −5.6058
1 to 4 charge vs α9α10 charge	3.73695	0.7426	1	−10.903 to 18.377

Table 3C. Charge:Peak
One Way ANOVA
Data Table: Three ratios controls
Factor A: 5 Groups
α7 ratio, 1 to 2 ratio, 1 to 3 ratio, 1 to 4 ratio, α9α10 ratio
Analysis of Variance Results
Source	DF	SS	MS	F	P
Total	67	2736.7877	40.847578
A	4	801.65783	200.41446	6.5246839	0.00019
Error	63	1935.1299	30.716348

Bonferroni’s All Pairs Comparison
Comparison	Mean Difference	|t|	P	95% CL
α7 ratio vs 1 to 2 ratio	−0.247095	0.1133	1	−6.5902 to 6.096
α7 ratio vs 1 to 3 ratio	−8.29304	3.8036	0.0032	−14.636 to −1.95
α7 ratio vs 1 to 4 ratio	−7.26774	3.3334	0.0144	−13.611 to −0.92467
α7 ratio vs α9α10 ratio	−3.39605	1.5576	1	−9.7391 to 2.947
1 to 2 ratio vs 1 to 3 ratio	−8.04594	3.841	0.0029	−14.14 to −1.9517
1 to 2 ratio vs 1 to 4 ratio	−7.02065	3.3515	0.0136	−13.115 to −0.92642
1 to 2 ratio vs α9α10 ratio	−3.14896	1.5032	1	−9.2432 to 2.9453
1 to 3 ratio vs 1 to 4 ratio	1.0253	0.4895	1	−5.0689 to 7.1195
1 to 3 ratio vs α9α10 ratio	4.89699	2.3377	0.2259	−1.1972 to 10.991
1 to 4 ratio vs α9α10 ratio	3.87169	1.8483	0.6926	−2.2225 to 9.9659

Nicotine allowed for the separation of sensitive a7 responses from insensitive a9a10 nAChR (Figure 3A). Note that these data were normalized to their respective ACh controls, which factored out the overall lower responses associated with a9a10 co-expression. The two lower ratios of a9a10 co-expression produced a mixed population of nicotine-sensitive and -insensitive receptors, while at the highest ratio of a9a10 co-injection, the cells were insensitive to nicotine, similar to those injected with just a9 and a10 (see Table 4A for ANOVA).

Figure 3.

The pharmacological separation of receptors in mixed populations. A) Detection of (30 μM) nicotine-sensitive a7 responses. B) Detection of (30 μM) pCN-diEPP-sensitive responses, primarily representing a9a10 receptors. C) Inhibition of 60 μM ACh-evoked responses by co-application of 30 μM mCN-diEPP. In all panels, data are shown for both peak currents and net charge (n = 6–7 for each condition, see Table 4 for statistical analyses).

Table 4.

analysis of Figure 3 Pharmacology

Table 4A Nicotine peak currents
One Way ANOVA
Data Table: Data with three ratios
Factor A: 5 Groups
α7 peak, α7(α9α10)2 peak, α7(α9α10)3 peak, α7(α9α10)4 peak, α9α10 peak
Analysis of Variance Results
Source	DF	SS	MS	F	P
Total	33	47.725917	1.4462399
A	4	31.079279	7.7698197	13.535752	< .0001
Error	29	16.646639	0.57402202

Bonferroni’s All Pairs Comparison
Comparison	Mean Difference	|t|	P	95% CL
α7 peak vs α7(α9α10)2 peak	0.977414	2.4135	0.2234	−0.25293 to 2.2078
α7 peak vs α7(α9α10)3 peak	1.69071	4.1748	0.0025	0.46037 to 2.921
α7 peak vs α7(α9α10)4 peak	2.51034	6.1987	< .0001	1.28 to 3.7407
α7 peak vs α7(α9α10)2 peak	2.50163	5.9349	< .0001	1.2211 to 3.7822
α7(α9α10)2 peak vs α7(α9α10)3 peak	0.713293	1.7613	0.8873	−0.51705 to 1.9436
α7(α9α10)2 peak vs α7(α9α10)4 peak	1.53292	3.7852	0.0071	0.30258 to 2.7633
α7(α9α10)2 peak vs α9α10 peak	1.52421	3.6161	0.0112	0.24364 to 2.8048
α7(α9α10)3 peak vs α7(α9α10)4 peak	0.819629	2.0239	0.5228	−0.41071 to 2.05
α7(α9α10)3 peak vs α9α10 peak	0.810921	1.9238	0.6424	−0.46966 to 2.0915
α7(α9α10)4 peak vs α9α10 peak	−0.00870844	0.0207	1	−1.2893 to 1.2719

Nicotine Net charge
One Way ANOVA
Data Table: Data with three ratios
Factor A: 5 Groups
α7 charge, α7(α9α10)2 charge, a7(a9a10)3 charge, a7(a9a10) 4 charge, a9a10 charge
Analysis of Variance Results
Source	DF	SS	MS	F	P
Total	33	6.6619212	0.2018764
A	4	6.2062219	1.5515555	98.738614	< .0001
Error	29	0.45569922	0.015713766

Bonferroni’s All Pairs Comparison
Comparison	Mean Difference	|t|	P	95% CL
α7 charge vs α7(α9α10)2 charge	0.371516	5.5446	< .0001	0.16795 to 0.57508
α7 charge vs α7(α9α10)3 charge	0.844269	12.6001	< .0001	0.64071 to 1.0478
α7 charge vs α7(α9α10)4 charge	1.08297	16.1626	< .0001	0.87941 to 1.2865
α7 charge vs α9α10 charge	1.08529	15.5618	< .0001	0.87341 to 1.2972
α7(α9α10)2 vs α7(α9α10)3 charge	0.472753	7.0555	< .0001	0.26919 to 0.67632
α7(α9α10)2 vs α7(α9α10)4 charge	0.711454	10.6179	< .0001	0.50789 to 0.91502
α7(α9α10)2 charge vs α9α10 charge	0.713775	10.2347	< .0001	0.5019 to 0.92565
α7(α9α10)3 vs α7(α9α10)4 charge	0.2387	3.5624	0.0129	0.035136 to 0.44226
α7(α9α10)3 vs α9α10 charge	0.241021	3.456	0.0171	0.029145 to 0.4529
α7(α9α10)4 vs α9α10 charge	0.00232117	0.0333	1	−0.20955 to 0.2142

Table 4B pCN-diEPP peak currents
One Way ANOVA
Data Table: Data with three ratios
Factor A: 5 Groups
α7 peak, α7(α9α10)2 peak, α7(α9α10)3 peak, α7(α9α10)4 peak, α9α10 peak
Analysis of Variance Results
Source	DF	SS	MS	F	P
Total	34	25.446475	0.74842572
A	4	18.238626	4.5596566	18.977882	< .0001
Error	30	7.2078483	0.24026161

Bonferroni’s All Pairs Comparison
Comparison	Mean Difference	|t|	P	95% CL
a7 peak vs α7(α9α10)2 peak	−0.0109723	0.0419	1	−0.80479 to 0.78285
a7 peak vs a7(a9a10)3 peak	−0.917661	3.5025	0.0147	−1.7115 to −0.12384
a7 peak vs a7(a9a10)4 peak	−1.83001	6.9847	< .0001	−2.6238 to −1.0362
a7 peak vs a9a10 peak	−1.31903	5.0344	0.0002	−2.1129 to −0.52522
α7(α9α10)2 peak vs a7(a9a10) 3 peak	−0.906689	3.4606	0.0164	−1.7005 to −0.11287
α7(α9α10)2 peak vs a7(a9a10) 4 peak	−1.81904	6.9428	< .0001	−2.6129 to −1.0252
α7(α9α10)2 peak vs a9a10 peak	−1.30806	4.9925	0.0002	−2.1019 to −0.51424
a7(a9a10)3 peak vs a7(a9a10)4 peak	−0.912348	3.4822	0.0155	−1.7062 to −0.11853
a7(a9a10)3 peak vs a9a10 peak	−0.401373	1.5319	1	−1.1952 to 0.39245
a7(a9a10)4 peak vs a9a10 peak	0.510974	1.9503	0.6054	−0.28285 to 1.3048

pCN-diEPP Net Charge
One Way ANOVA
Data Table: Data with three ratios
Factor A: 5 Groups
a7 charge, α7(α9α10)2 charge, a7(a9a10)3 charge, a7(a9a10)4 charge, a9a10 charge
Analysis of Variance Results
Source	DF	SS	MS	F	P
Total	34	9.0405759	0.26589929
A	4	7.8405214	1.9601304	49.001034	< .0001
Error	30	1.2000545	0.040001816

Bonferroni’s All Pairs Comparison
Comparison	Mean Difference	|t|	P	95% CL
α7 charge vs α7(α9α10)2 charge	0.008322	0.0778	1	−0.31558 to 0.33223
α7 charge vs α7(α9α10)3 ch.	−0.878618	8.2185	< .0001	−1.2025 to −0.55471
α7 charge vs α7(α9α10)4 ch.	−0.875816	8.1923	< .0001	−1.1997 to −0.55191
α7 charge vs α9α10 charge	−1.09152	10.21	< .0001	−1.4154 to −0.76761
α7(α9α10)2 charge vs α7(α9α10)3 ch.	−0.88694	8.2964	< .0001	−1.2108 to −0.56303
α7(α9α10)2 charge vs α7(α9α10)4 ch.	−0.884138	8.2702	< .0001	−1.208 to −0.56023
α7(α9α10)2 charge vs α9α10 charge	−1.09984	10.2878	< .0001	−1.4237 to −0.77593
α7(α9α10)3 ch. vs α7(α9α10)4 char	0.00280213	0.0262	1	−0.3211 to 0.32671
α7(α9α10)3 ch. vs α9α10 charge	−0.2129	1.9914	0.556	−0.53681 to 0.11101
α7(α9α10)4 ch. vs α9α10 charge	−0.215702	2.0177	0.5265	−0.53961 to 0.1082

Table 4C Inhibition of ACh responses by mCN-diEPP
Peak currents
One Way ANOVA
Data Table: Data with three ratios
Factor A: 5 Groups
a7 peak, α7(α9α10)2 peak, a7(a9a10) 3 peak, a7(a9a10) 4 peak, a9a10 peak
Analysis of Variance Results
Source	DF	SS	MS	F	P
Total	32	1.9254888	0.060171524
A	4	1.3377141	0.33442853	15.931274	< .0001
Error	28	0.58777464	0.020991952

Bonferroni’s All Pairs Comparison
Comparison	Mean Difference	|t|	P	95% CL
α7 peak vs α7(α9α10)2 peak	0.156714	1.9442	0.6199	−0.08889 to 0.40232
α7 peak vs α7(α9α10)3 peak	0.126972	1.5752	1	−0.11863 to 0.37258
α7 peak vs α7(α9α10)4 peak	−0.162472	2.0156	0.5353	−0.40808 to 0.083132
α7 peak vs α9α10 peak	0.458369	5.4796	< .0001	0.20349 to 0.71324
α7(α9α10)2 peak vs α7(α9α10)3 peak	−0.0297416	0.384	1	−0.26571 to 0.20623
α7(α9α10)2 peak vs α7(α9α10)4 peak	−0.319186	4.1215	0.003	−0.55515 to −0.083217
α7(α9α10)2 peak vs α9α10 peak	0.301655	3.7423	0.0083	0.056051 to 0.54726
α7(α9α10)3 peak vs α7(α9α10)4 peak	−0.289445	3.7374	0.0085	−0.52541 to −0.053476
α7(α9α10)3 peak vs α9α10 peak	0.331396	4.1113	0.0031	0.085792 to 0.577
α7(α9α10)4 peak vs α9α10 peak	0.620841	7.7021	< .0001	0.37524 to 0.86645

Net Charge
One Way ANOVA
Data Table: Data with three ratios
Factor A: 5 Groups
a7 charge, α7(α9α10)2 charge, a7(a9a10)3 charge, a7(a9a10) 4 charge, a9a10 charge
Analysis of Variance Results
Source	DF	SS	MS	F	P
Total	32	1.9115891	0.05973716
A	4	1.526624	0.381656	27.759314	< .0001
Error	28	0.38496513	0.013748755

Bonferroni’s All Pairs Comparison
Comparison	Mean Difference	|t|	P	95% CL
α7 charge vs α7(α9α10)2 charge	0.0839687	1.2872	1	−0.1148 to 0.28273
α7 charge vs α7(α9α10)3 charge	−0.198647	3.0451	0.0502	−0.39741 to 0.00011879
α7 charge vs α7(α9α10)4 charge	−0.114133	1.7496	0.9114	−0.3129 to 0.084632
α7 charge vs α9α10 charge	0.436095	6.4418	<.0001	0.22983 to 0.64236
α7(α9α10)2 charge vs α7(α9α10)3 ch.	−0.282615	4.5092	0.0011	−0.47358 to −0.091648
α7(α9α10)2 charge vs α7(α9α10)4 char	−0.198102	3.1608	0.0376	−0.38907 to −0.0071345
α7(α9α10)2 charge vs α9α10 charge	0.352126	5.3978	< .0001	0.15336 to 0.55089
α7(α9α10)3 ch vs α7(α9α10)4 char	0.0845132	1.3484	1	−0.10645 to 0.27548
α7(α9α10)3 ch vs α9α10 charge	0.634742	9.7301	<.0001	0.43598 to 0.83351
α7(α9α10)4 char vs α9α10 charge	0.550228	8.4346	<.0001	0.35146 to 0.74899

As expected, pCN-diEPP produced minimal activation of a7 nAChRs and cells injected with the lowest ratio of a9a10, while substantial pCN-diEPP responses were observed at the higher ratios of a9a10 expression (Figure 3B, see Table 4B for ANOVA). In contrast to the data with nicotine and pCN-diEPP, where intermediate ratios of expression suggested intermediate responses, presumably due to mixed populations of a7 and a9a10 nAChR, it appeared that any level of a7 co-expression provided protection from the profound inhibition by mCN-diEPP that was seen in oocytes in which only a9a10 nAChR subunits were expressed (Figure 3C, see Table 4C for ANOVA). These data might indicate that hybrid a7a9a10 nAChR are being formed and that their ionotropic function is somehow protected from the antagonism by mCN-diEPP. This is, however, in contrast to the above mentioned suggestion that a7a9a10 nAChR heteromers may lack ionotropic function. This difference may relate to the specific subunit stoichiometries of the heteromeric receptors.

Possible activation of a10 homomeric receptors

While our co-injections of a7, a9, and a10 largely seemed to indicate the formation of separate functional a7 and a9* nAChR (Figures 2 & 3), it may be the case that co-expression of a9 and a10 will naturally result in multiple populations of receptors: a9 homomers, a9a10 heteromers, and possibly even some a10 homomers, which may be assembled but not readily detected as functional. The possibility that quiescent homomeric a10 nAChR may be converted, at least transiently, into functional receptors is supported by a recent report (Hone and McIntosh, 2022) that quiescent homomeric a10 nAChR can be functionally activated by exposure to certain alkaloids, for example methyllycaconitine (MLA) or strychnine. MLA is often purported to be an a7-selective antagonist, although it has been argued that it behaves more like an inverse agonist than a simple competitive antagonist (Papke et al., 2018; Williams et al., 2011). MLA has been used to implicate an a7-dependent mechanism in some studies of CAS activity (Quadri et al., 2018), where an a9 mechanism has subsequently also been implicated (Richter et al., 2022). We therefore evaluated the ability of MLA co-applications to inhibit the ACh-evoked responses of cells expressing a7, or a9 alone, or a9 co-expressed with a10 (Figure 4). We measured inhibition of both peak currents and net charge, with net charge probably being the better measurement (Papke and Papke, 2002). All three populations of receptors were inhibited by sub-micromolar concentrations of MLA (net charge IC₅₀s were 0.12 ± 0.05, 0.19 ± 0.02, and 0.24 ± 0.13 μM for a7, a9, and a9a10 expressing cells, respectively, Figure 4A). We noticed, however, that the MLA applications had a potentiating effect on successive ACh responses of the cells expressing both a9 and a10 nAChR subunits, but not for cells expressing a7 or a9 alone (Figure 4B & C, see Table 5 for ANOVA). This would be consistent with an activation of a previously quiescent population of homomeric a10 nAChRs by MLA. Interestingly, we saw that, although 30 μM nicotine did not activate currents in a9a10-expressing cells, it did have a similar potentiating effect on subsequent responses to ACh (Figure 4D). This is consistent with the hypothesis that if there were homomeric a10 receptors present, ion currents in response to ACh would be facilitated if a certain number of the five potential ligand binding sites of homomeric a10 nAChRs remained occupied by an antagonist. An alternative explanation might be a MLA- or nicotine-mediated induction of the a10 receptor expression on the cell surface. In the light of the short time between pretreatment and the application of the nAChR agonist, the latter explanation seems to be less probable. More in-depth analyses are warranted to clarify this interesting aspect and to test if it is relevant to the in vivo situation.

Figure 4.

Probing a9a10-expressing cells for the possible activation of a10 homomeric receptors with methyllycaconitine (MLA). A) Evaluation of MLA antagonist activity. MLA was co-applied with 60 μM ACh to cells expressing a7, a9, or both a9 and a10. Normalized responses were measured relative to the average of two initial ACh controls on both peak currents (left panel) and net charge (right panel). Each cell was tested with a single concentration of MLA, and the individual cell data were used to obtain fits to the Hill equation assuming a negative Hill coefficient. The IC₅₀ values for the inhibition of peak currents were 0.62 ± 0.05, 2.64 ± 1.75, and 6.78 ± 2.18 μM for a7, a9, and a9a10, respectively. The IC₅₀ values for the inhibition of net charge were 0.12 ± 0.05, 0.19 ± 0.02, and 0.24 ± 0.13 μM for a7, a9, and a9a10, respectively. B) Effects of MLA co-application on subsequent responses to ACh. Following the co-application of MLA and ACh (represented in panel A), an additional application of 60 μM ACh was made, and those responses (peak currents, left panel and net charge in right panel) were normalized to the average of the two initial ACh control responses. While the subsequent responses of a7- and a9-expressing cells tended to decrease following MLA applications in a concentration-dependent manner, responses of a9a10-expressing cells became progressively larger. We conducted ANOVA on the net-charge responses (Figure 4C, Table 5). At all concentrations greater than 300 nM, the ACh responses of a9a10-expressing cells were greater than those expressing either a7 or a9 alone and greater for the highest concentration of MLA than for the lowest. C) Averaged traces of the ACh responses of cells expressing a9 and a10 (n = 6) before (black line with tan background, representing the SEM) and after (dark blue line with light blue SEM) the application of 30 μM MLA. D) Averaged responses of cells expressing a9 and a10 (n = 7) before, during, and after the application of 30 μM nicotine (Figure 3A). The insert shows the overlay of the before and after responses, with the currents after nicotine in blue.

Table 5.

ANOVA of Figure 4 net-charge responses after MLA

One Way ANOVA
Data Table: ANOVA
Factor A: 24 Groups
α7 .03, α7 0.1, α7 0.3, α7 1, α7 3, α7 10, α7 30, α7 100, α9 .03, α9 0.1, α9 0.3, α9 1, α9 3, α9 10, α9 30, α9 100, α9α10 .03, α9α10 0.1, α9α10 0.3, α9α10 1, α9α10 3, α9α10 10, α9α10 30, α9α10 100
Analysis of Variance Results
Source	DF	SS	MS	F	P
Total	159	279.61159	1.7585634
A	23	226.03708	9.8276991	24.947818	< .0001
Error	136	53.574508	0.39393021

Bonferroni’s Pairs Comparison, Focus on specific concentrations compared to a9a10
Comparison	Mean Difference	|t|	P	95% CL
α7 .03 vs α9 .03	0.115355	0.3304	1	−1.229 to 1.4597
α7 .03 vs α9α10 .03	−0.547576	1.5111	1	−1.9426 to 0.8475
α7 0.1 vs α9 0.3	−0.108643	0.3111	1	−1.453 to 1.2357
α7 0.1 vs α9α10 0.1	−0.987013	2.7238	1	−2.3821 to 0.40806
α7 0.3 vs α9 .03	−0.545087	1.6248	1	−1.8367 to 0.7465
α7 0.3 vs α9α10 .03	−1.20802	3.4595	0.1996	−2.5523 to 0.13631
α7 1 vs α9 1	−0.245152	0.7307	1	−1.5367 to 1.0464
α7 1 vs α9α10 1	−1.95643	5.6028	< .0001	−3.3008 to −0.6121
a7 3 vs α9 3	−0.346432	1.0326	1	−1.638 to 0.94515
α7 3 vs α9α10 3	−2.43049	7.2447	< .0001	−3.7221 to −1.1389
α7 10 vs α9 10	−0.18647	0.5558	1	−1.4781 to 1.1051
α7 10 vs α9α10 10	−2.48328	7.1116	< .0001	−3.8276 to −1.139
α7 30 vs α9 30	−0.125675	0.3746	1	−1.4173 to 1.1659
α7 30 vs α9α10 30	−3.03277	9.0399	< .0001	−4.3244 to −1.7412
α7 100 vs a9 100	−0.0652598	0.1801	1	−1.4603 to 1.3298
α7 100 vs a9a10 100	−4.66831	13.3691	< .0001	−6.0126 to −3.324
α9 .03 vs a9a10 .03	−0.662931	1.8985	1	−2.0073 to 0.68139
α9 0.1 vs a9a10 0.1	−0.762414	2.1834	1	−2.1067 to 0.58191
α9 0.1 vs a9a10 0.3	−1.06231	3.1665	0.5258	−2.3539 to 0.22928
α9 0.3 vs a9a10 0.3	−1.17827	3.5121	0.1667	−2.4699 to 0.11332
α9 1 vs α9α10 1	−1.71127	4.9008	0.0007	−3.0556 to −0.36695
α9 3 vs α9α10 3	−2.08406	6.212	< .0001	−3.3756 to −0.79247
α9 10 vs α9α10 10	−2.29681	6.5776	< .0001	−3.6411 to −0.95249
α9 10 vs α9α10 30	−2.83536	8.4515	< .0001	−4.127 to −1.5438
α9 10 vs α9α10 100	−4.42618	13.1933	< .0001	−5.7178 to −3.1346
α9 30 vs α9α10 30	−2.90709	8.6653	< .0001	−4.1987 to −1.6155
α9 100 vs α9α10 100	−4.60305	13.1822	< .0001	−5.9474 to −3.2587

The impact of pCN-diEPP and mCN-diEPP on the ATP-mediated release of IL-1β by monocytic cells

It was shown before that ACh and phosphocholine function as unconventional agonists at monocytic nAChRs containing a7 and/or a9/a10 subunits and as potent inhibitors of the BzATP-induced maturation and release of IL-1β in monocytic cells (Richter et al., 2016; Zakrzewicz et al., 2017). To test if pCN-diEPP and mCN-diEPP exert similar effects on monocytic cells, human monocytic THP-1 cells were primed with LPS (1 μg/ml) for 5 h, followed by stimulation with the P2X7 receptor agonist BzATP (100 μM) for another 40 min in the presence or the absence of ACh (10 μM), phosphocholine (200 μM), pCN-diEPP (100 μM), or mCN-diEPP (100 μM) (Figure 5). Thereafter, the concentration of released IL-1β was measured in cell culture supernatants by ELISA. As expected, untreated cells and cells primed with LPS did not release relevant amounts of IL-1β, whereas stimulation with BzATP resulted in elevated IL-1β levels released by monocytic THP-1 cells, in the range of 84 pg/ml to 236 pg/ml (Figure 5A). The BzATP-induced release of IL-1β was significantly inhibited by phosphocholine, ACh, and pCN-diEPP, while for mCN-diEPP only a tendency (P = 0.08, Friedman test followed by the two-tailed Wilcoxon signed-rank test) towards a mild inhibition was seen. The effects of pCN-diEPP and mCN-diEPP on the IL-1β release were repeated and confirmed in a second independent set of experiments (Figure 5B). When both experiments were taken together (Figure 5A & B), the weak inhibitory effects of mCN-diEPP on the BzATP-induced release of IL-1β were statistically significant (P = 0.014, n = 12 each, Friedman test followed by the two-tailed Wilcoxon signed-rank test). The biological relevance of such subtle changes in IL-1β release are, however, questionable.

Figure 5.

The impact of pCN-diEPP (pCN) and mCN-diEPP (mCN) on the ATP-mediated release of interleukin-1β (IL-1β) by human monocytic THP-1 cells. THP-1 cells were primed with lipopolysaccharide (LPS; 1 μg/ml, 5 h). Then the P2X7 receptor agonist BzATP ((2’/3’-O-(4-benzoylbenzoyl)adenosine-5’-triphosphate, tri(triethylammonium) salt) was added for another 40 min to trigger IL-1β release, which was measured by ELISA. A) The BzATP (100 μM) -induced release of IL-1β was investigated in the presence and absence of the compounds pCN (100 μM) or mCN (100 μM), as well as phosphocholine (PC, 200 μM) or acetylcholine (ACh, 10 μM). PC, ACh, and pCN inhibited the BzATP-induced release of IL-1β, while mCN alone had no effect. B) The inhibitory effect of pCN on the BzATP-induced release of IL-1β was partially reversed in the presence of mCN. In each experiment, the IL-1β concentrations obtained after stimulation with BzATP + solvent were set to 100%, and all other values were calculated accordingly. Data are presented as individual data points; the bar represents the median, and whiskers encompass the 25^th to 75^th percentile. *P ≤ 0.05, different from LPS-primed cells stimulated with BzATP alone. #P ≤ 0.05 significantly different from samples where BzATP plus corresponding agonist were given. Statistics are based on the Friedman test followed by the two-tailed Wilcoxon signed-rank test.

These results are in line with our expectations that stimulation of a9* nAChRs in monocytic cells metabotropically inhibits the ionotropic function of the ATP-sensitive P2X7R and, hence, the release of IL-1β (Richter et al., 2016; Zakrzewicz et al., 2017). However, we cannot predict which nAChR subunits are involved in the effects of pCN-diEPP. They might indeed be mediated via a9a10 nAChR, similar to the unconventional nAChR agonists glycerophosphocholine or lysophosphatidylcholine (Zakrzewicz et al., 2017). In addition, a7 nAChR subunits might be involved, as we have shown before that a7 nAChR silent agonists can inhibit the BzATP-induced release of IL-1β as well (Richter et al., 2022). As pCN-diEPP was characterized as a very weak but strongly desensitizing agonist at a7 nAChRs (Quadri et al., 2016), properties that are close to silent agonists, it might also induce metabotropic functions of this receptor. More research including the use of a7- and a9-specific nAChR antagonists and gene silencing experiments might address these questions.

As expected, the a9a10 nAChR antagonist mCN-diEPP reversed the inhibitory effect of ACh, phosphocholine (Figure 5A), and pCN-diEPP (Figure 5A & B). Surprisingly, pCN-diEPP also antagonized the effects of ACh and phosphocholine (Figure 5A). We conclude that both pCN-diEPP and mCN-diEPP can exert agonistic and antagonistic functions in this experimental setting. This observation adds another conundrum to the specific characteristics of NCNRs, and suggests that the ionotropic effects of a certain ligand at conventional nAChRs do not fully predict the effects induced at NCNRs.

As shown in Supplementary Figure 2A & B, LDH values remained below 14% of the total release, irrespective of the experiment performed, which suggests that, at least in the short term, pCN-diEPP and mCN-diEPP are not overtly toxic to mononuclear phagocytes.

pCN-diEPP fully attenuates CFA-induced inflammatory pain in an α7 nAChR-independent manner.

Animals treated with CFA showed greatly reduced mechanical pain threshold compared to vehicle-treated controls (P = 0.001, Table 6). Pain thresholds were elevated by treatments with pCN-diEPP (Figure 6A and Table 6). Note that these data were not well fit to nomal distributions, especially under control conditions where the majority of responses were the same high threshold value of 3.63 g. Therefore, the data in Figure 6A are the median scores for each condition (along with the 25–75% ranges). The individual responses are shown in Supplemental Figure 3. Statistical analysis was conducted using a Kruskal-Wallis test followed by two-tailed Mann–Whitney U tests at each time point following pCN-diEPP treatment. There were significant effects for all the doses tested at the 1 hour time point (P = 0.001, Table 6), and for the two highest dosages at the 3 hour time point (P = 0.014 and 0.006, Table 6), with no significant effects at the 6 hour time point.

Table 6.

Statistical analysis of Figure 6A

Kruskal-Wallis test followed by two-tailed Mann–Whitney U test N = 8, each
Kuskal-Wallis test
Time points	P
BL	P = 0.664
0 h	P = 0.000
1 h	P = 0.000
3 h	P = 0.000
6 h	P = 0.000
Mann–Whitney U test, BL
Compared groups	P
Veh/Veh vs. CFA/Veh	n.a.
Veh/Veh vs. Veh/pCN-diEPP (1.0 mg/kg)	n.a.
CFA/Veh vs. CFA/pCN-diEPP (0.1 mg/kg)	n.a.
CFA/Veh vs. CFA/pCN-diEPP (0.3 mg/kg)	n.a.
CFA/Veh vs. CFA/pCN-diEPP (1.0 mg/kg)	n.a.
Mann–Whitney U test, 0 h
Compared groups	P
Veh/Veh vs. CFA/Veh	0.001
Veh/Veh vs. Veh/pCN-diEPP (1.0 mg/kg)	0.643
CFA/Veh vs. CFA/pCN-diEPP (0.1 mg/kg)	0.632
CFA/Veh vs. CFA/pCN-diEPP (0.3 mg/kg)	0.105
CFA/Veh vs. CFA/pCN-diEPP (1.0 mg/kg)	0.874
Mann–Whitney U test, 1 h
Compared groups	P
Veh/Veh vs. CFA/Veh	0.001
Veh/Veh vs. Veh/pCN-diEPP (1.0 mg/kg)	0.353
CFA/Veh vs. CFA/pCN-diEPP (0.1 mg/kg)	0.001
CFA/Veh vs. CFA/pCN-diEPP (0.3 mg/kg)	0.001
CFA/Veh vs. CFA/pCN-diEPP (1.0 mg/kg)	0.001
Mann–Whitney U test, 3 h
Compared groups	P
Veh/Veh vs. CFA/Veh	0.001
Veh/Veh vs. Veh/pCN-diEPP (1.0 mg/kg)	0.822
CFA/Veh vs. CFA/pCN-diEPP (0.1 mg/kg)	0.277
CFA/Veh vs. CFA/pCN-diEPP (0.3 mg/kg)	0.014
CFA/Veh vs. CFA/pCN-diEPP (1.0 mg/kg)	0.006
Mann–Whitney U test, 6 h
Compared groups	P
Veh/Veh vs. CFA/Veh	0.001
Veh/Veh vs. Veh/pCN-diEPP (1.0 mg/kg)	0.952
CFA/Veh vs. CFA/pCN-diEPP (0.1 mg/kg)	0.704
CFA/Veh vs. CFA/pCN-diEPP (0.3 mg/kg)	0.194
CFA/Veh vs. CFA/pCN-diEPP (1.0 mg/kg)	0.196

Figure 6.

The effects of systemic pCN-diEPP in the CFA-induced chronic inflammatory pain model. A) Antiallodynic effects after intraperitoneal administration of various doses of pCN-diEPP (0.1, 0.3, and 1 mg/kg) in WT mice. The mechanical paw withdrawal thresholds were determined 3 days after intraplantar injection of CFA (100%) at 1, 3, and 6 h after the drug administration (n = 8/group, 50% male and 50% female). pCN-diEPP fully reversed the mechanical hypersensitivity in a dose-dependent manner. Replicate data are provided in Supplementary Figure 3. Statistical results based on the Friedman test followed by the two-tailed Wilcoxon signed-rank test are provided in Table 6 B) The antiallodynic effects of systemic pCN-diEPP in the CFA model was not lost in the a7 KO mice. Antiallodynic effects after intraperitoneal administration of pCN-diEPP (1 mg/kg) in WT and in alpha7 KO mice. The mechanical paw withdrawal thresholds were determined 3 days after intraplantar injection of CFA (100%) at 1 h after the drug administration (n = 6–8/group, 50% male and 50% female). Plotted are the median scores under each condition, whiskers encompass the 25^th to 75^th percentile. P values (Table 6) as determined by the Friedman test followed by the two-tailed Wilcoxon signed-rank test are indicated in the figure.

These data are in line with the common notion that full and silent agonists at a7 nAChRs exert anti-inflammatory functions in vivo, and hence are also expected to reduce inflammatory pain. Therefore, the next logical step was to investigate if the a7 nAChR mediates the effect of pCN-diEPP.

In a separate experiment (Figure 6B), we tested CFA-treated a7 WT and KO mice with pCN-diEPP (1 mg/kg) or vehicle and evaluated their mechanical hypersensitivity 3 days after CFA. Our experiments indicated that the antinociceptive effect of pCN-diEPP was independent of a7 nAChRs since the effects of 1mg/kg pCN-diEPP were essentially the same in both WT and a7 nAChR KO animals.

As pCN-diEPP functioned as a full agonist at a9a10 nAChRs, our data strongly suggest that the reduction of CFA-induced pain was mediated via the activation of a9* nAChRs. However, for a firm conclusion, experiments on a9a10 nAChR KO mice are warranted. It would be of high clinical relevance to test pCN-diEPP in experimental neuropathic pain and to treat animals suffering from CFA-induced inflammatory pain with the conopeptide RgIA4. In contrast to a conopeptide, small molecules such as pCN-diEPP generally have a higher probability of entering the clinical arena.

mCN-diEPP only partially attenuates CFA-induced inflammatory pain.

The highest concentrations of mCN-diEPP produced a modest reduction of CFA-induced mechanical hypersensitivity 1 and 3 h after injection (Figure 7A, Table 7). The effect of that dose dissipated 6 h after injection. In sham-treated mice 3 mg/kg mCN-diEPP did not alter von Frey responses. As with Figure 6A, data presented are the median scores for each condition (along with the 25–75% ranges). The individual responses are shown in Supplemental Figure 4.

Figure 7.

The effects of systemic mCN-diEPP in the CFA-induced chronic inflammatory pain model. A) Antiallodynic effects after intraperitoneal administration of various doses of mCN-diEPP (0.6, 1, and 3 mg/kg) in WT mice. The mechanical paw withdrawal thresholds were determined 3 days after intraplantar injection of CFA (100%) at 1, 3, and 6 h after the drug administration (n = 8/group, 50% male and 50% female). Plotted are the median scores under each condition bracketed by the 25 – 75% ranges. Replicate data are provided in Supplementary Figure 4. Statistical results based on Friedman test followed by the two-tailed Wilcoxon signed-rank test are provided in Table 7. B). Antiallodynic effects of systemic mCN-diEPP in the CFA model were observed in the alpha7 KO mice. Antiallodynic effects after intraperitoneal administration of mCN diEPP (3 mg/kg) in WT and in alpha7 KO mice. The mechanical paw withdrawal thresholds were determined 3 days after intraplantar injection of CFA (100%) at 1 h after the drug administration (n = 8/group, 50% male and 50% female). Plotted are the median scores under each condition, whiskers encompass the 25^th to 75^th percentile. P values (Table 7) as determined by the Friedman test followed by the two-tailed Wilcoxon signed-rank test are indicated in the figure.

Table 7.

Statistical analysis of withdrawal thresholds in Figure 7A

Kruskal-Wallis test followed by two-tailed Mann–Whitney U test N = 8, each
Kruskal-Wallis test
Time points	P
BL	P = 0.943
0 h	P = 0.000
1 h	P = 0.000
3 h	P = 0.000
6 h	P = 0.000
Mann–Whitney U test, BL
Compared groups	P
Veh/Veh vs. CFA/Veh	n.a.
Veh/Veh vs. Veh/mCN-diEPP (3 mg/kg)	n.a.
CFA/Veh vs. CFA/mCN-diEPP (0.6 mg/kg)	n.a.
CFA/Veh vs. CFA/mCN-diEPP (1 mg/kg)	n.a.
CFA/Veh vs. CFA/mCN-diEPP (3 mg/kg)	n.a.
Mann–Whitney U test, 0 h
Compared groups	P
Veh/Veh vs. CFA/Veh	0.001
Veh/Veh vs. Veh/mCN-diEPP (3 mg/kg)	0.601
CFA/Veh vs. CFA/mCN-diEPP (0.6 mg/kg)	0.817
CFA/Veh vs. CFA/mCN-diEPP (1 mg/kg)	0.134
CFA/Veh vs. CFA/mCN-diEPP (3 mg/kg)	0.134
Mann–Whitney U test, 1 h
Compared groups	P
Veh/Veh vs. CFA/Veh	0.001
Veh/Veh vs. Veh/mCN-diEPP (3 mg/kg)	0.954
CFA/Veh vs. CFA/mCN-diEPP (0.6 mg/kg)	1.000
CFA/Veh vs. CFA/mCN-diEPP (1 mg/kg)	0.134
CFA/Veh vs. CFA/mCN-diEPP (3 mg/kg)	1.000
Mann–Whitney U test, 3 h
Compared groups	P
Veh/Veh vs. CFA/Veh	0.001
Veh/Veh vs. Veh/mCN-diEPP (3 mg/kg)	0.865
CFA/Veh vs. CFA/mCN-diEPP (0.6 mg/kg)	0.284
CFA/Veh vs. CFA/mCN-diEPP (1 mg/kg)	0.100
CFA/Veh vs. CFA/mCN-diEPP (3 mg/kg)	0.004
Mann–Whitney U test, 6 h
Compared groups	P
Veh/Veh vs. CFA/Veh	0.001
Veh/Veh vs. Veh/mCN-diEPP (3 mg/kg)	0.484
CFA/Veh vs. CFA/mCN-diEPP (0.6 mg/kg)	0.913
CFA/Veh vs. CFA/mCN-diEPP (1 mg/kg)	0.643
CFA/Veh vs. CFA/mCN-diEPP (3 mg/kg)	0.153

In a separate experiment, we tested CFA-treated a7 nAChR WT and KO mice with mCN-diEPP (3 mg/kg) or vehicle and evaluated their mechanical hypersensitivity 3 days after CFA (Figure 7B). At the 1 hour time point, mCN-diEPP showed a significant effect of treatment in both WT and the a7 KO mice, although in both cases the treatments did not fully reverse the effects of CFA treatment. This result suggests that a7 nAChR subunits are not involved in the weak antinociceptive effects of mCN-diEPP. It rather suggests that for some cellular mediators of CAS, a9 nAChR antagonism may have some limited efficacy. However, compared to pCN-diEPP, mCN-diEPP seems to be a far less promising candidate for clinical application.

We also confirmed that neither the highest dose tested of pCN-diEPP (1 mg/kg) nor the highest dose tested of mCN-diEPP (3 mg/kg) had any significant effects on locomotor activity or body temperature of mice compared to vehicle-treated animals (see Supplemental Figure 5 and associated statistical anlysis in supplemental data).

Conclusions

Numerous cell types have been implicated to be engaged in the NCNR-mediated modulation of CAS, including various leukocyte lineages, macrophages, and microglia. It seems likely that in whole-animal studies, multiple different cell types may contribute to the anti-inflammatory activity of any particular drug. Much of the work characterizing the activity of the NCNR that regulate the CAS has focused on a7 due to a partial correspondence between the pharmacology of a7 ion channel pharmacology and efficacy in models of CAS, with the large caveat that a7 channel desensitizers may have greater efficacy in CAS than ion-channel activators (Papke and Lindstrom, 2020; Papke et al., 2023; Piovesana et al., 2021). Several in vivo studies using a9*-selective conopeptides have also implicated a9 nAChRs as potential modulators of CAS (AlSharari et al., 2020; Christensen et al., 2017; Huynh et al., 2019; Pacini et al., 2016). Cell-based assays like those reported in this study implicate both a7 and a9* nAChRs as regulators of CAS (Richter et al., 2016; Zakrzewicz et al., 2017), but with the caveat that the a9*-selective conopeptide RgIA4 blocks the effects of other ligands that activate a9* receptors (Richter et al., 2022). Therefore, it remains a challenge to sort out exactly which receptor subtypes contribute to the NCNR that affect inflammation, particularly in whole-animal experiments, where all the nAChR subtypes will be present and expressed at different levels in various cell types.

For some ligands (PNU-282987, GAT-107, and NS6740) that were effective with in vivo models of inflammatory pain, an a7 nAChR-dependent mechanism was confirmed by demonstrating that effects were lost in a7 nAChR KO animals (Bagdas et al., 2016; Donvito et al., 2017; Papke et al., 2015). However, such data do not support the hypothesis that a7 nAChR expression alone is sufficient for CAS activity. In the present study, we demonstrate that for the a9 nAChR agonist pCN-diEPP, a7 expression is not necessary for CAS activity in the CFA model of inflammatory pain. The retention of full activity of pCN-diEPP in a7 nAChR KO animals strongly implicates an a9* nAChR-dependent mechanism. Likewise, the milder anti-nociceptive effects of mCN-diEPP also seemed to be independent of a7 nAChR subunits.

In some studies (Costa et al., 2012; Garai et al., 2018; Patel et al., 2017; Pinheiro et al., 2020; Quadri et al., 2018; Toma et al., 2019), block by MLA has been proposed to demonstrate an a7-dependent mechanism. However, due to the a9* nAChR sensitivity to MLA (Figure 4), the interpretation of those results should be called into question.

The mystery remains concerning the degree to which agonist and/or antagonist properties of canonical ion channel function of the a-bungarotoxin-sensitive nAChR can really be predictive of the ability of the same ligands to regulate the CAS function of NCNR. On this point, it should be kept in mind that even the most efficacious of agonists do not just promote ion channel activation, which in any case is usually just transient, particularly in the case of a7 nAChR. Agonists stabilize ligand-dependent non-conducting states as well. This has led to the hypothesis that signal transduction by NCNR is associated with the induction of non-conducting conformations (Papke and Lindstrom, 2020; Papke et al., 2023), consistent with the CAS activity of the strongly desensitizing a7 nAChR silent agonist NS6740 (Papke et al., 2015; Thomsen and Mikkelsen, 2012) and the nonconventional a9 nAChR agonist phosphocholine (Richter et al., 2016).

The ultimate development of new therapeutic agents will require study of basic ADMET (absorption, distribution, metabolism, excretion, and toxicity) properties. A further consideration is brain permeation if microglia, which are potential mediators of inflammation in neurodegenerative diseases are to be regarded as targets. Due to their quaternary ammonium, the diEPP compounds are unlikely to have good brain penetration, a limitation that might be overcome with tertiary amine analogs. However, in general, peripheral sensitization in nociceptors is essential for the development of inflammatory pain, so our finding of the a9-dependent CAS activity of pCN-diEPP highlights the utility of targeting of a9* nAChR for the development of non-opiate analgetics.