Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

LAs block nerve conduction in the peripheral nervous system (Arias, 1999). They also have a wide range of behavioral effects implicating actions in the central nervous system (CNS). Among these effects are restlessness, euphoria, muscle twitching, and tremor, which have been attributed to selective depression of inhibitory neurons by LAs. Other effects of LAs suggesting CNS actions include drowsiness, disorientation, slurred speech, respiratory depression, tinnitus, and sedation. At high concentrations, LAs may cause loss of consciousness or even death (Naguib et al., 1998; Hodgson et al., 2000).

LAs also differ from one another in several ways, and different bases have been used to classify LAs. LAs have been classified into three categories by Arias (1999). Agents from one category of LAs (group I; see Fig. 1), typified by tetracaine, procaine and lidocaine, possess only one aromatic ring. An amide (lidocaine) or an ester (tetracaine and procaine) linkage couples the ring to one aliphatic chain that typically ends in a ternary amino group (tetracaine, procaine, and lidocaine) or a quaternary ammonium ion (the charged lidocaine analog QX-314 or the dimethylammonio variant QX-222). The group I esters also have a second amino (procaine) or alkylamino (butylamino for tetracaine) chain para to the ester-linked alkylamino chain. The group I amides include compounds (lidocaine, QX-314, and QX-222) having two methyl groups ortho to the alkylamino or alkylammonio chain. Thus, group I compounds may be subdivided into two subgroups representing para-(alkyl)amino-phenyl-alkylesters and ortho-dimethyl-phenyl-alkylamides. A second category of LAs (group II; see Fig. 1) includes molecules with two aromatic rings separated and linked by a single -carbon chain and is typified by proadifen and adiphenine. Group II LAs also have an ester linking the ring region through the -carbon to an aliphatic chain that ends in a ternary amino group (proadifen and adephinine) or in a quaternary ammonium ion (charged proadifen derivative meproadifen; not shown). In some cases, group II LAs have an additional alkyl chain (propyl for proadifen) linked to the -carbon bridging the two aromatic rings. A third category (group III; see Fig. 1), represented by dimethisoquin, contains drugs having a fused, two-ring structure (i.e., an isoquinoline backbone). An aliphatic chain ending again in either a ternary amino group (dimethisoquin) or in a quaternary ammonium ion (charged dimethisoquin analog trimethisoquin; not shown) is coupled via an ester linkage to the isoquinoline ring, which is further derivatized with an alkyl chain (butyl for dimethisoquin) meta to the ester link.

View larger version (20K): [in this window] [in a new window]   Fig. 1.   Molecular structures of LAs used in this study. Names for each LA and their position according to the classification scheme of Arias (1999) in groups I, II, or III are shown for each structure [note however that dimethisoquin, proadifen, and adiphenine are ternary amino ligands rather than quaternary ammonium ions as Fig. 2 in Arias (1999) may implicate]. Structures are listed left top-to-bottom and then right top-to-bottom in the rank order, with some minor exceptions, observed for actions at nAChR. See the text, Figs. 2 through 9, and Table 1 for experimental details.

Whereas peripheral effects of LAs are attributed to actions on voltage-sensitive Na⁺ channels, mechanisms of LA action on presumably CNS-mediated behaviors are not well understood. Ionotropic neurotransmitter receptors such as cholinergic and serotonergic receptors have been investigated as potential targets of LA action in the brain (Katz and Miledi, 1975; Neher and Steinbach, 1978; Lukas and Bennett, 1979; Horn et al., 1980; Forman and Miller, 1989; Charnet et al., 1990; Revah et al., 1991; Barann et al., 1993; Dilger and Vidal, 1994; Fan and Weight, 1994).

Nicotinic acetylcholine receptors (nAChR) are neurotransmitter-gated ion channels. At least 16 distinct genes encode nAChR subunits that combine in a variety of ways to form diverse, pentameric nAChR ion channels (for review, see Lukas, 1998). Some physiological roles of diverse nAChR subtypes are known whereas others remain incompletely defined (Albuquerque et al., 1996, 2000). However, each nAChR subtype has a unique pharmacological profile, and this may also translate into differences in sensitivity to LAs (Dani, 1993; Eterovic et al., 1993; Cuevas and Adams, 1994). Conversely, nAChR may prove to contribute to clinically relevant therapeutic actions and/or side effects of LAs.

This study assessed effects of several members from the three categories of LAs on function of four different, human nAChR subtypes. A preliminary report of these findings has appeared (Gentry et al., 2000).

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Drug Dilutions. All drugs were prepared fresh the day of the assay as stock solution in as much as 50% ethanol depending on the drug and then diluted so that final concentrations of ethanol were no higher than 5% in assay buffer (130 mM NaCl, 5.4 mM KCl, 2 mM CaCl₂·2H₂O, 5 mM glucose, and 50 mM HEPES) at the highest concentrations of LA used in each experiment. Ethanol alone up to 50% in assay buffer had no effects on function of any of the examined nAChR subtypes during a 3-min exposure (data not shown). Ethanol, -phenylbenzeneacetic acid-2-(diethylamino)ethyl ester HCl (adiphenine), carbamylcholine chloride (carb), 2-diethylamino-N-(2,6-dimethylphenyl)acetamide HCl (lidocaine), -phenyl--propylbenzeneacetic acid-2-(diethylamino)ethyl ester HCl (proadifen; SKF-525A), 4-aminobenzoic acid-2-(diethylamino)ethyl ester HCl (procaine), and 4-(butylamino)benzoic acid 2-(dimethylamino)ethyl ester HCl (tetracaine) were purchased from Sigma (St. Louis, MO). 2-(Trimethylammonio)-N-(2,6-dimethylphenyl)acetamide chloride (QX-222) and 2-(triethylammonio)-N-(2,6-dimethylphenyl)acetamide Br or Cl were obtained from Alomone Labs (Jerusalem, Israel). 3-Butyl-1-[2-(dimethylamino)ethoxy] isoquinoline HCl (dimethisoquin) was purchased from ICN Biomedicals, Inc. (Plainview, NY) (discontinued; now available from Research Diagnostics, Inc., Flanders, NJ).

Model Cell Lines and Cell Culture. The present study used low passage (less than 50) human cell lines naturally or heterologously expressing different nAChR subtypes to examine inhibition potencies of LAs. The TE671/RD human cell line naturally expresses muscle-type nAChR (1*-nAChR according to suggested nomenclature) (Lukas et al., 1999) as an assembly of two 1-subunits and one each of 1-, -, and -subunits (Lukas, 1986, 1989; Schoepfer et al., 1988; Luther et al., 1989). Human neuroblastoma-derived SH-SY5Y cells naturally express two nAChR subtypes found in autonomic neurons. However, the subtype that contains 3- and 4-subunits (34*-nAChR) dominates ion flux assay measures of nAChR function in these cells (Lukas, 1993; Lukas et al., 1993; Conroy and Berg, 1995). nAChR containing 4- and 2-subunits (42-nAChR) are a major form of high affinity nicotine binding site in mammalian brain, and nAChR containing 4- and 4-subunits (44-nAChR) may also be expressed substantially in the CNS. Human 42- and 44-nAChR were expressed stably and heterologously in native nAChR-null SH-EP1 human epithelial cells to create SH-EP1-h42 and SH-EP1-h44 cell lines, respectively, using techniques that have been reported previously (Peng et al., 1999; Eaton et al., 2000; Ferchmin et al., 2001). Cells were maintained at 37°C, under 95% O₂/5% CO₂ in Dulbecco's modified Eagle's medium supplemented with 5% fetal bovine serum (Hyclone, Logan, UT), 10% horse serum (Invitrogen, Carlsbad, CA), 1% sodium pyruvate (Cellgro; AK by Mediatech Inc., Herndon, VA), 2% glutamine penicillin-streptomycin (Irvine Scientific, Santa Ana, CA) and 0.02% amphotericin B (Sigma).

Assays of nAChR Function. Rubidium-86 (⁸⁶Rb⁺) efflux assays were performed using TE671/RD, SH-SY5Y, SH-EP1-h42, or SH-EP1-h44 cells according to the procedure of Lukas and Cullen (1988). Cells were cultured (~2 × 10⁵ cells per 15.5-mm-diameter well; ~150 µg of total cell protein per well) on Falcon 24-well culture plates (BD Biosciences, Bedford, MA) precoated with poly-D-lysine M_r 70,000 to 150,000 (Sigma) according to Bencherif et al. (1995). Cells were allowed to grow (overnight) until a confluent monolayer was formed and cells adhered to culture plate. Confluence was monitored using light microscopy. ⁸⁶Rb⁺ was obtained from PerkinElmer Life Sciences (Boston, MA). In all cases, cells were loaded with ⁸⁶Rb⁺ for no less than 4 h and subsequently rinsed twice with 2 ml per well of drug-free assay buffer. Cells expressing each nAChR subtype of interest were exposed to efflux buffer alone or experimental concentrations of LA and/or carbamylcholine for 3 min. Radioactivity released into the extracellular medium was quantified by Cerenkov counting using a Wallac Trilux system (40% efficiency). Levels of nonspecific ion flux from each cell line were comparable whether defined with samples containing agonist plus antagonist 100 µM d-tubocurarine or with control samples containing buffer alone without agonist. Specific nAChR function was defined as total, experimentally determined ion flux evoked by agonist carb in the absence or presence of LA minus nonspecific ion flux. Typical values for specific and nonspecific ⁸⁶Rb⁺ efflux for cell samples are 9000 and 1200 cpm for TE671/RD cells, 2900 and 900 cpm for SH-SY5Y cells, 3500 and 600 cpm for SH-EP1-h42 cells, and 9000 and 800 cpm for SH-EP1-h44 cells.

Data Analysis. Dose-response curves were fit to data points by the general equation Y = b + [(a  b)/1 + 10^{((c  X)n)}] where Y is the observed specific ⁸⁶Rb⁺ efflux response (% of control), X is the experimental concentration of LA, b is nonspecific flux, a is total ion flux (1 mM carb control), c is the log IC₅₀ value for antagonist dose-response profiles at fixed agonist concentration or the log EC₅₀ value for agonist dose-response profiles at fixed antagonist concentration, and n is the Hill coefficient (<0 for antagonist dose-response profiles; >0 for agonist dose-response profiles). Best fit, nonlinear regression variable slope curves were determined by an iterative process using GraphPad Prism (San Diego, CA) software, from which values of a, b, c, and n were derived for each experiment. Values of a and b parameters were normalized to percentage of control flux. Because LA acted as noncompetitive inhibitors in nearly all cases, there was no dependence on concentration of agonist used to stimulate nAChR function for LA IC₅₀ values and there was no need to apply modifications, such as the Cheng-Prusoff correction, to the data. Results from three to 16 independent measurements were fit to the logistic equation and plotted as mean ± S.E.M.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Acute Effects of LAs on 1*-nAChR Function. ⁸⁶Rb⁺ efflux assays with the TE671/RD cell line expressing muscle-type 1*-nAChR were used to evaluate receptor function during acute exposure to selected LAs. Simultaneous exposure of cells to increasing concentrations of LA and 1 mM carb indicated that all LAs inhibited 1*-nAChR function in a concentration-dependent manner (Fig. 2). Concentrations at which half-maximal block of TE671/RD cell 1*-nAChR function were achieved rank ordered from high to low inhibition potency were (IC₅₀ value precedes drug's name): 0.34 µM proadifen > 1.9 µM adiphenine > 2.4 µM dimethisoquin > 13 µM tetracaine > 19 µM QX-314 > 52 µM lidocaine > 230 µM procaine 3.4 mM QX-222 (Table 1). Thus, representative group II LAs exhibited greatest inhibitory potency at 1*-nAChR followed by group III LAs, with group I LAs exhibiting the lowest inhibitory potency (Fig. 1). Hill coefficients for antagonist concentration-response profiles ranged between 0.74 proadifen to 1.6 for lidocaine (Table 1).

View larger version (35K): [in this window] [in a new window]   Fig. 2.   LA dose dependence for functional blockade of 1*-nAChR in TE671/RD cells. Specific 86Rb+ efflux (ordinate, percentage of control) measured as described under Materials and Methods was determined in the presence of the indicated concentrations (abscissa, log molar scale) of proadifen (), dimethisioquin (), QX-314 (), or procaine () (A) or adiphenine (), tetracaine (), lidocaine (), or QX-222 () (B). Smooth curves drawn through data points (means ± S.E.M. from at least three separate experiments) are from iterative fits to the logistic function setting a = 94 to 101% (see Materials and Methods). Results yield IC50 values, Hill coefficients, and r2 values indicated in the text and/or in Table 1. Values for the parameter b as a percentage of control 86Rb+ efflux are 3.0 ± 2.9% for proadifen, 11.2 ± 5.4% for dimethisoquin, 13.8 ± 7.0% for QX-314, 1.3 ± 13.5% for procaine, 2.4 ± 2.0% for adiphenine, 1.6 ± 2.9% for tetracaine, and 2.4 ± 3.6% for lidocaine.

View larger version (30K): [in this window] [in a new window]   Fig. 3.   Dose-response profiles for carbamylcholine-stimulated 86Rb+ efflux through 1*-nAChR at different concentrations of LA. Measurements of specific 86Rb+ efflux (ordinate, percentage of 1 mM carbamylcholine control) were made with TE671/RD cells assayed in the presence of carb alone () at the indicated doses (abscissa, molar log scale) or in the presence of carb with adiphenine, dimethisoquin, tetracaine, or procaine at indicated concentrations. Data points are from three or more independent experiments (mean ± S.E.M.). Smooth curves drawn through the data points are from iterative fits to the logistic function (b = 0.26 to 5.1%; see Materials and Methods). Values for the parameter a (as a percentage of specific 86Rb+ efflux) and r2 values (in parentheses) are 103.7 ± 4.0% (0.90) for untreated control, 76.3 ± 2.9% (0.97) for 1.3 µM adiphenine, 63.2 ± 3.3% (0.95) for 2.5 µM adiphenine, 41.6 ± 3.3% (0.90) for 5 µM adiphenine, 72.4 ± 0.9% (0.99) for 1 µM dimethisoquin, 48.7 ± 1.5% (0.99) for 3 µM dimethisioquin, 29.5 ± 1.2% (0.97) for 10 µM dimethisoquin, 78.4 ± 2.2% (0.99) for 3 µM tetracaine, 47.5 ± 2.2% (0.98) for 10 µM tetracaine, 21.8 ± 1.8% (0.94) for 32 µM tetracaine, 84.6 ± 0.9% (0.99) for 32 µM procaine, 64.5 ± 1.4% (0.99) for 100 µM procaine, and 36.1 ± 1.4% (0.98) for 316 µM procaine. None of the carb dose-response profiles displayed any significant shift from EC50 value of control (39 µM ×/ 0.05).

Acute Effects of LAs on 34*-nAChR Function. ⁸⁶Rb⁺ efflux assays using SH-SY5Y cells expressing 34*-nAChR showed concentration-dependent functional blockade by all LAs tested (Fig. 4). Concentrations at which half-maximal block (IC₅₀) of SH-SY5Y cell 34*-nAChR function was achieved were (by rank order: high to low inhibition potency): 0.6 µM proadifen > 1.8 µM adiphenine > 4.7 µM dimethisoquin > 8.3 µM tetracaine  9.2 µM QX-314 > 63 µM lidocaine > 87 µM procaine > 150 µM QX-222 (Table 1). Thus, the general rank order inhibition potency of LAs observed in the TE671/RD cell line (group II > group III > group I) held for 34*-nAChR expressed by the SH-SY5Y cell line (Figs. 2 and 4; Table 1). Hill slopes ranged between 0.78 for proadifen and 1.23 for lidocaine.

View larger version (34K): [in this window] [in a new window]   Fig. 4.   LA dose dependence for functional blockade of 34*-nAChR in SH-SY5Y cells. Specific 86Rb+ efflux (ordinate, percentage of control) measured as described under Materials and Methods was determined in the presence of the indicated concentrations (abscissa, log molar scale) of proadifen (), dimethisioquin (), QX-314 (), or procaine () (A) or adiphenine (), tetracaine (), lidocaine (), or QX-222 () (B). Smooth curves drawn through data points (means ± S.E.M. from at least three separate experiments) are from iterative fits to the logistic function (a = 91 to 103%; see Materials and Methods). Results yield IC50 values, Hill coefficients, and r2 values indicated in the text and/or in Table 1. Values for the parameter b as a percentage of control 86Rb+ efflux are 5.7 ± 3.8% for proadifen, 3.5 ± 2.5% for dimethisoquin, 1.1 ± 3.6% for QX-314, 1.4 ± 5.1% for procaine, 1.9 ± 4.2% for adiphenine, 5.6 ± 4.0% for tetracaine, 1.0 ± 10.0% for lidocaine, and 7.0 ± 9.6% for QX-222.

View larger version (33K): [in this window] [in a new window]   Fig. 5.   Dose-response profiles for carb-stimulated 86Rb+ efflux through 34*-nAChR at different concentrations of LA. Measurements of specific 86Rb+ efflux (ordinate, percentage of 1 mM carb control) were made with SH-SY5Y cells assayed in the presence of carb alone () at the indicated doses (abscissa, molar log scale) or in the presence of carb with adiphenine, dimethisoquin, tetracaine, or procaine at indicated concentrations. Data points are from three or more independent experiments (mean ± S.E.M.). Smooth curves drawn through the data points are from iterative fits to the logistic function (b = 5.4 to 7.4%; see Materials and Methods). Values for the parameter a (as a percentage of specific 86Rb+ efflux) and r2 values (in parentheses) are 113.3 ± 9.5% (0.85) for untreated control, 68.2 ± 6.0% (0.72) for 0.5 µM adiphenine, 48.6 ± 6.0% (0.61) for 1 µM adiphenine, 40.9 ± 4.8% (0.60) for 2 µM adiphenine, 75.6 ± 5.5% (0.94) for 1 µM dimethisoquin, 41.6 ± 6.4% (0.82) for 3 µM dimethisoquin, 25.6 ± 3.6% (0.75) for 10 µM dimethisoquin, 74.2 ± 4.2% (0.96) for 3 µM tetracaine, 37.4 ± 2.6% (0.92) for 10 µM tetracaine, 17.9 ± 6.1% (0.66) for 32 µM tetracaine, 72.3 ± 4.2% (0.89) for 32 µM procaine, 48.4 ± 4.0% (0.81) for 100 µM procaine, and 24.4 ± 4.2% (n/a) for 320 µM procaine. None of the carb dose-response profiles displayed any significant shift in EC50 value (155 µM ×/ 0.12).

Acute Effects of Drugs on 42- or 44- nAChR Function. Acute effects of selected LAs on function of nAChR containing 4-subunits were also evaluated by ⁸⁶Rb⁺ efflux assay. SH-EP1 cells in which 42- or 44- nAChR were expressed heterologously were exposed simultaneously to test concentrations of LA and 1 mM carb. All LAs inhibited 4*-nAChR function in a concentration-dependent manner with one exception (Figs. 6 and 7). The quarternary LA, QX-222, at a concentration of 2.5 mM only weakly inhibited 1 mM carb-evoked response mediated by 44-nAChR (Fig. 7). The rank order inhibition potency for LAs tested at 44-nAChR was identical with that for 1*- or 34*- nAChR and was (IC₅₀ value precedes drug name): 1.5 µM proadifen > 6.3 µM adiphenine  7.2 µM dimethisoquin > 30 µM tetracaine > 64 µM QX-314 > 250 µM lidocaine > 1.0 mM procaine 2.5 mM QX-222 (Table 1). Interestingly, the rank order inhibition potency for LAs at 42-nAChR was somewhat unique. Rank order and IC₅₀ values for LA inhibition of 42-nAChR function were: 2.0 µM proadifen > 3.7 µM adiphenine > 27 µM tetracaine > 61 µM dimethisoquin  68 µM QX-314 > 190 µM lidocaine > 390 µM QX-222 > 2.1 mM procaine (Fig. 6; Table 1). Thus, in general, group II LAs (adiphenine and proadifen) exhibited the greatest ability to inhibit carb-evoked 4*-nAChR function, whereas group I compounds exhibited weaker ability to inhibit. However, the group III compound has an inhibition potency for 42-nAChR within the range of potencies tested for group I compounds. Moreover, the diethylamino group I ligand, lidocaine, its triethylammonium derivative, QX-314, and the trimethylammonium analog, QX-222, had the most similar inhibitory capability for 42-nAChR compared with other nAChR subtypes (Figs. 1 and 6; Table 1). Hill slopes were shallowest for proadifen (0.73 to 0.94) and steepest for tetracaine (1.24 to 1.61) at 4*-nAChR (Table 1).

View larger version (34K): [in this window] [in a new window]   Fig. 6.   LA dose dependence for functional blockade of 42-nAChR in transfected SH-EP1-h42 cells. Specific 86Rb+ efflux (ordinate, percentage of control) measured as described under Materials and Methods was determined in the presence of the indicated concentrations (abscissa, log molar scale) of proadifen (), dimethisioquin (), QX-314 (), or procaine () (A) or adiphenine (), tetracaine (), lidocaine (), or QX-222 () (B). Smooth curves drawn through data points (means ± S.E.M. from at least three separate experiments) are from iterative fits to the logistic function (a = 92 to 101%; see Materials and Methods). Results yield IC50 values, Hill coefficients, and r2 values indicated in the text and/or in Table 1. Values for the parameter b as a percentage of control 86Rb+ efflux are 5.3 ± 6.8% for proadifen, 16.5 ± 5.1% for dimethisoquin, 4.1 ± 7.7% for QX-314, n/a for procaine, 1.2 ± 4.9% for adiphenine, 2.3 ± 2.5% for tetracaine, 3.1 ± 20% for lidocaine, and n/a for QX-222.

View larger version (34K): [in this window] [in a new window]   Fig. 7.   LA dose dependence for functional blockade of 44-nAChR in transfected SH-EP1-h44 cells. Specific 86Rb+ efflux (ordinate, percentage of control) measured as described under Materials and Methods was determined in the presence of the indicated concentrations (abscissa, log molar scale) of proadifen (), dimethisioquin (), QX-314 (), or procaine () (A), or adiphenine (), tetracaine (), lidocaine (), or QX-222 () (B). Smooth curves drawn through data points (means ± S.E.M. from at least three separate experiments) are from iterative fits to the logistic function (a = 85 to 97%; see Materials and Methods). Results yield IC50 values, Hill coefficients, and r2 values indicated in the text and/or in Table 1. Values for the parameter b as a percentage of control 86Rb+ efflux are 5.2 ± 2.3% for proadifen, 6.4 ± 2.8% for dimethisoquin, 7.0 ± 2.9% for QX-314, 12.3 ± 46% for procaine, 0.64 ± 4.4% for adiphenine, 1.4 ± 2.1% for tetracaine, 8.6 ± 7.6% for lidocaine, and n/a for QX-222.

View larger version (30K): [in this window] [in a new window]   Fig. 8.   Dose-response profiles for carb-stimulated 86Rb+ efflux through 42-nAChR at different concentrations of LA. Measurements of specific 86Rb+ efflux (ordinate, percentage of 1 mM carb control) were made with SH-EP1-h42 cells assayed in the presence of carb alone () at the indicated doses (abscissa, molar log scale) or in the presence of carb with adiphenine, dimethisoquin, tetracaine, or procaine at indicated concentrations. Data points are from three or more independent experiments (mean ± S.E.M.). Smooth curves drawn through the data points are from iterative fits to the logistic function (b = 3.7 to 1.2%; see Materials and Methods). Values for the parameter a (as a percentage of specific 86Rb+ efflux), and r2 values (in parentheses) are 100.8 ± 1.4% (0.98) for untreated control, 47.2 ± 3.0% (0.81) for 1 µM adiphenine, 36.9 ± 2.5% (0.83) for 3 µM adiphenine, 22.1 ± 2.6% (0.70) for 6 µM adiphenine, 54.1 ± 0.7% (0.99) for 32 µM dimethisoquin, 40.0 ± 3.2% (0.82) for 100 µM dimethisoquin, 13.2 ± 1.6% (0.95) for 320 µM dimethisoquin, 65.4 ± 1.5% (0.97) for 10 µM tetracaine, 44.0 ± 1.1% (0.98) for 32 µM tetracaine, 6.9 ± 0.3% (0.92) for 100 µM tetracaine, 85.4 ± 1.9% (0.98) for 100 µM procaine, 71.6 ± 2.9% (0.95) for 320 µM procaine, and 27.4 ± 1.3% (0.94) for 1000 µM procaine. All of the carb dose-response profiles produced a shift in EC50 value from (19 µM ×/ 0.03) for control to higher values (up to ~400 µM ×/ 0.08) in the presence of LAs.

View larger version (31K): [in this window] [in a new window]   Fig. 9.   Dose-response profiles for carb-stimulated 86Rb+ efflux through 44-nAChR at different concentrations of LA. Measurements of specific 86Rb+ efflux (ordinate, percentage of 1 mM carb control) were made with SH-EP1-h44 cells assayed in the presence of carbamylcholine alone () at the indicated doses (abscissa, molar log scale) or in the presence of carb with adiphenine, dimethisoquin, tetracaine, or procaine at indicated concentrations. Data points are from three or more independent experiments (mean ± S.E.M.). Smooth curves drawn through the data points are from iterative fits to the logistic function b = 1.5 to 1.6% (see Materials and Methods). Values for the parameter a (as a percentage of specific 86Rb+ efflux) and r2 values (in parentheses) are 97.9 ± 2.6% (0.99) for untreated control, 70.0 ± 3.1% (0.93) for 2 µM adiphenine, 48.7 ± 2.4% (0.91) for 4 µM adiphenine, 25.8 ± 3.0% (0.66) for 8 µM adiphenine, 68.9 ± 3.5% (0.97) for 4 µM dimethisoquin, 46.2 ± 2.9% (0.96) for 8 µM dimethisoquin, 52.0 ± 51.6% (0.89) for 16 µM dimethisoquin, 81.2 ± 1.2% (0.99) for 8 µM tetracaine, 72.2 ± 1.8% (0.97) for 16 µM tetracaine, 51.9 ± 1.6% (0.98) for 32 µM tetracaine, 85.0 ± 4.2% (0.95) for 250 µM procaine, 73.5 ± 8.6 (0.86) for 500 µM procaine, and 39.3 ± 7.4% (0.89) for 1000 µM procaine. The dose response profile for procaine produced a rightward shift in EC50 value (from 50 µM ×/ 0.03 control to 316 µM ×/ 0.25 in the presence of 1000 µM procaine) indicating that procaine may partially compete with carb for a binding site on the 44-nAChR. Adiphenine, dimethisoquin, tetracaine, and procaine dose-response profiles all produced functional blockade of the 42-nAChR insurmountable with increasing concentrations of carb suggesting noncompetitive binding properties.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The primary findings of this study are: 1) that LAs have reasonably strong ability to inhibit diverse, human nAChR subtypes (1*-, 34*-, 42-, and 44-nAChR), and 2) that, with a few exceptions, LAs act at these nAChR subtypes as noncompetitive functional inhibitors. Two exceptions to the noncompetitive inhibition rule are for LAs acting at 42-nAChR and for procaine acting at 44-nAChR. Although LAs can be classified in several ways, our data regarding LA function at nAChR are well segregated based on the structural classification for LAs as put forth by Arias (1999) (represented in Fig. 1). Although the sample size of LAs representing each structural category is limited, the data serve as a foundation for studies of LA-nAChR structure-activity relationships.

Group II LAs proadifen and adiphenine were the most potent functional inhibitors at all nAChR tested, with proadifen having 2- to 6-fold greater inhibition potency across nAChR subtypes than adiphenine (see Table 1). The only structural difference that may account for the greater inhibition potency of nAChR for proadifen is its additional propyl moiety linked to the -carbon (see Fig. 1).

In general, the group III ligand built on the butyl-isoquinoline backbone, dimethisoquin, had the next greatest ability to inhibit each nAChR subtype tested (see Table 1). Dimethisoquin, like the group II LAs, has a two carbon chain separating amino from ester moieties on one aliphatic chain (see Fig. 1). However, dimethisoquin is a dimethylamino rather than a diethylamino compound, suggesting that presence of the more compact amino group may contribute to lower inhibition potency for nAChR.

Group I drugs generally have the lowest inhibition potencies of the four nAChR subtypes examined. Tetracaine, which has a dimethylamino residue ending its ester-linked aliphatic chain but also has a markedly longer para-aminobutyl chain, consistently has 10-fold greater potency for the nAChR subtypes tested than the diethylamino, para-amino ligand, procaine (see Table 1 and Fig. 1). The ability of tetracaine to inhibit at all nAChR subtypes tested is only 4- to 8-fold greater than that of the diethylamino amide lidocaine. The quaternary, triethylammonium amide QX-314 generally has greater (up to 7-fold) inhibition potency for nAChR than does the ternary, diethylamino analog lidocaine, but QX-314 has much greater inhibition potency than does the more compact quaternary, trimethylammonium amide QX-222. Thus for group I esters, presence of a diethylamino (procaine) moiety rather than a more compact dimethylamino (tetracaine) group and perhaps the shortened length of the para-(alkyl)amino aliphatic chain correlates with diminished potency for nAChR (see Fig. 1). However, for group I amides, the diethylammonium (QX-314) has greater inhibition potency than does the more compact dimethylammonium (QX-222). The diethylammonium (QX-314) also has greater inhibition potency than does the diethylamino (lidocaine) compound. These findings indicate that structural determinants of LA ability to inhibit nAChR include bulk at the amino/ammonio moiety. However, these findings also indicate that these structural determinants differ between group I esters and amides, suggesting that ligands in group I should be further subclassified.

Of the nAChR subtypes tested, muscle-type 1*- and ganglionic 34*-nAChR naturally expressed in the periphery each were generally more sensitive to block by LAs than were CNS-type 42- or 44-nAChR. 1*- and 34*-nAChR had similar sensitivity to block by most of the LA tested. As an exception, 34*-nAChR showed nearly 20-fold higher sensitivity than 1*-nAChR to QX-222. Another notable exception was for 42-nAChR, which was ~9-fold more sensitive than 1*-nAChR to QX-222 block. Thus, across nAChR subtypes, the only rank order inhibition potency profile that deviated from proadifen > adiphenine > dimethisoquin > tetracaine > QX-314 > lidocaine > procaine QX-222 was for 42-nAChR, which had the profile tetracaine > dimethisoquin ~ QX-314 > lidocaine  QX-222 procaine. Comparisons between 42- and 44-nAChR indicated comparable inhibition potency or slightly higher potency at 42-nAChR for most of the LAs tested except for dimethisoquin (~9-fold greater inhibition potency at 44-nAChR) and QX-222 (>6-fold greater inhibition potency at 42-nAChR). Therefore, the small differences between 2- and 4-subunits influence potency for only a select few LAs. Curiously, 42-nAChR showed the highest level of discrimination between group I esters (tetracaine/procaine) but by far the lowest level of discrimination between group I amides (QX compounds and lidocaine).

There is substantial evidence that LAs bind to and inhibit nAChR via noncompetitive mechanisms implicating open channel block (for review, see Arias, 1999). Our findings are largely consistent with results and conclusions from these studies, but our studies extend to broader comparisons across LAs and across nAChR subtypes. For example, consistent with our results, QX-314 was found in electrophysiological studies to be a more effective blocker than QX-222 at frog extrajunctional (fetal or -subunit-containing) 1*-nAChR (Neher and Steinbach, 1978). The comparatively weak effects of these charged ligands, which we also observed, were attributed to inability to gain access to the cell's interior except via open channels. In another electrophysiological study, QX-314 and QX-222 were found to produce half-maximal inhibition of mouse muscle 1*-nAChR heterologously expressed in Xenopus oocytes at concentrations of 78 and 2780 µM, respectively (Pascual and Karlin, 1998), which compare with IC₅₀ values of 19 and 3400 µM, respectively, in the current studies. Electrophysiological analyses of effects on peak currents evoked by presumed 34*-nAChR naturally expressed in cultured rat parasympathetic cardiac neurons yielded IC₅₀ values of 28 µM for QX-222 and 2.8 µM for procaine (Cuevas and Adams, 1994), which have the same rank order but show considerably higher activity when compared with 150 and 87 µM IC₅₀ values for QX-222 and procaine, respectively, acting in the current study at human 34*-nAChR natively expressed in ganglionic SH-SY5Y cells. Another electrophysiological study reported voltage-dependent IC₅₀ values of 1 or 40 µM, respectively, for tetracaine acting at mouse muscle or Torpedo electroplax 1*-nAChR heterologously expressed in Xenopus oocytes (Eterovic et al., 1993); these values bracket the IC₅₀ value of 13 µM found in the current study for actions of tetracaine at human 1*-nAChR. Mouse 1*-nAChR expressed in Xenopus oocytes were blocked by procaine with an IC₅₀ of 66 µM (Yost and Dodson, 1993), showing higher sensitivity than human 1*-nAChR natively expressed in the TE671/RD cell line (IC₅₀ value of 230 µM). Differences in species, expression systems (cell environment), experimental approach (electrophysiological or ion flux assays), or perhaps degrees of nAChR desensitization (Ryan and Baenziger, 1999) could account for differences in IC₅₀ values observed in different studies.

Rank order inhibition potency for LA action at Torpedo marmorata electric organ 1*-nAChR measured using electron spin resonance and fluorescence techniques was procaine > tetracaine > QX-222 (Arias et al., 1990) and contrasts to our findings. This probably is because the biophysical measures taken may not relate to functional effects of LA at nAChR.

With regard to clinical relevance of LA inhibition of nAChR function, pharmacokinetic findings and ease of LA access to nAChR are of primary importance. Actions of charged ammonium LAs are likely restricted to the periphery. Also, peripheral routes of administration are less likely to yield LA concentrations high enough to induce neurotoxicity (Ritchie and Greengard, 1966). Roles played by blood-brain barrier-permeable LAs in neurological effects depend on perineural concentrations achieved rather than the dose of administered LA (Johnson, 2000), and plasma, cerebrospinal fluid (CSF), and brain concentrations of LAs are influenced by rates and routes of LA administration (Robinson et al., 1994; Xuecheng et al., 1997; Youngs, 1999).

For example, lidocaine is widely used for clinical procedures that include topical ointment application, epidural anesthesia, spinal administration, and systemic bolus administration. Peak plasma concentrations of lidocaine after topical application of 5% ointment to intact skin of healthy volunteers was 9 ng/ml (33.2 nM) and occurred 24 h after the treatment. Thus, absorption of lidocaine from intact skin is poor, and even if applied to damaged skin, effects of topical lidocaine at nAChR (IC₅₀ values of 52-250 µM) would be unlikely to occur. However, plasma and CSF concentrations of lidocaine after systemic bolus administration in humans peak at ~1.7 µg/ml (~6 µM) within 5 to 15 min rapidly declining thereafter (Tsai et al., 1998). Intravenous (i.v.) infusion of rabbits with lidocaine at a rate of 4 mg kg¹ min¹ produces whole brain  cortical brain > plasma > CSF concentrations of 373, 295, 166, and 77.5 µM, respectively, at the time of lidocaine-induced seizure onset (718 s) (Momota et al., 2000). These concentrations, especially those in cortical or whole brain, in which lidocaine accumulates, are higher than IC₅₀ values for blockade of nAChR. An independent investigation of adiphenine metabolism showed that concentration of ethyl-[¹⁴C]adiphenine 5 min after i.v. administration is 106.7 ± 17.2 ng/g brain tissue versus 7.1 ± 0.4 ng/g in blood, a difference of nearly 15 times more concentrated in brain tissue (Michelot et al., 1981). Spinal administration, despite dilution of drug in the subarachnoid space, of anesthetics provides perhaps the greatest opportunity for LA to access CNS receptors (Rigler et al., 1991). Spinally administered LAs are distributed nonhomogenously and can briefly reach very high local concentrations (Drasner et al., 1994; Robinson et al., 1994). For example, 5 min after routine spinal administration of a 5% solution of lidocaine, the mean CSF concentration can reach 14 mM (Van Zundert et al., 1996), which clearly would have effects at nAChR.

Collectively, these considerations and our results suggest therapeutic application of LA may produce peak plasma and/or brain concentrations capable of inducing significant functional inhibition of nAChR in the periphery or centrally. Enhanced deposition of LAs in the brain may actually make their concentrations there higher than in the periphery, meaning that inhibition of nAChR in the brain (e.g., 4*-nAChR and perhaps central 3*-nAChR) would occur even if their acute sensitivity to LAs is lower than LA sensitivity of nAChR in the periphery. Thus, nAChR need to be considered as potential targets for therapeutic actions of LAs and in mediation of side effects of LA action. Moreover, LAs present a broad and interesting family of pharmacophores, action of which at nAChR can be used to help define structural determinants on drugs and receptors that could be exploited to develop novel nAChR antagonists and safer and more effective LAs.