Introduction

Top Abstract Introduction Methods Results Discussion References

Molecular, biochemical and physiological studies demonstrate the presence of multiple nicotinic receptor (AChR) subunits in brain and other tissues (Alkondon and Albuquerque, 1993; Deneris et al., 1991; Papke, 1993). One of the predominant nicotinic receptor subtypes in the brain contains the alpha-7 subunit, especially in telencephalic regions such as hippocampus and neocortex, based on high affinity to BTX binding studies (Clarke et al., 1985; Marks et al., 1986). These receptors function as homo-oligomers when expressed in oocytes, where they demonstrate characteristic high affinity binding to -BTX, high calcium-permeability and rapid desensitization (Couturier et al., 1990; Seguela et al., 1993; de Fiebre et al., 1995).

The identification of agents selective for the alpha-7 AChR subtype has drawn attention to their roles in neuronal survival and memory-related behaviors (de Fiebre et al., 1995). One of these compounds, GTS-21 (also known as DMXB), is presently in a clinical trial for Alzheimer's disease, based on its behavioral actions and ability to protect rat neurons from various apoptotic and necrotic insults, both in vivo and in vitro (Martin et al., 1994; Meyer et al., 1998a, Shimohama et al., 1998). These insults include amyloid-exposure, removal of nerve growth factor from differentiated PC12 cells, NMDA-induced toxicity and axotomy of septal-hippocampal cholinergic neurons. Recently, several selective alpha-7 agonists with differing residual alpha-7 receptor-antagonist activities were compared relative to their ability to protect differentiated PC12 cells from trophic factor deprivation (Meyer et al., 1998b). This apparent residual inhibition of alpha-7 receptors most likely represents some form of channel block by agonist or a form of desensitization that is unique from the rapidly reversible desensitization that occurs with the application of high concentrations of ACh (de Fiebre et al., 1995; Papke et al., 1997). Agonists with significant residual antagonism were not cytoprotective, suggesting that an analysis of both agonist and antagonist activities may be necessary for predicting the cytoprotective efficacy of alpha-7 AChRs.

GTS-21 has also been shown to improve memory-related behaviors in nonhuman primates (Briggs et al., 1997), aged rabbits (Woodruff-Pak et al., 1994), nucleus-basalis lesioned rats (Meyer et al., 1994) and aged rats (Arendash et al., 1995). The improvement in delayed pair-matching behavior in aged primates observed after GTS-21 administration is particularly interesting because this drug reportedly is only very weakly efficacious at human alpha-7 receptors. Although this result may indicate that nonnicotinic receptors underlie at least some actions of GTS-21 in vivo, it is also conceivable that one or more metabolites of GTS-21 possesses nicotinic agonist activity in humans.

An initial pharmacokinetic study of GTS-21 in the rat indicated that only minor amounts of this compound were excreted in an unaltered form (Mahnir et al., 1998). Both methoxy substituents on the benzylidene ring of GTS-21 were potential sites of primary hepatic metabolism by O-demethylation (Kem et al., 1996). The principal human primary metabolite is 3-(4-hydroxy,2-methoxybenzylidene)- anabaseine, or 4OH-GTS-21 (Azuma et al., 1996). Because little is known about the properties of this metabolite relative to nicotinic receptors or their biological functions, we investigated these properties in several systems. First, to determine whether 4OH-GTS-21 was an agonist or antagonist at human and rat nicotinic receptor subtypes, its effects were measured on different AChR subunit combinations in the Xenopus oocyte system. Its binding to alpha-7 receptors was characterized by displacement of high affinity -BTX binding in human (SK-N-SH) and rat (PC12) cell lines. Finally, its biological activity was compared to GTS-21 relative to cytoprotection against amyloid-induced toxicity in both cell lines. An amyloid model for neurotoxicity was chosen because of its sensitivity to nicotinic receptor-mediated cytoprotection (Kihara et al., 1997; Zamani et al., 1997) and its ability to affect PC12 cells (Fagarasan and Efthimiopoulos, 1996); in addition to the well established involvement of amyloid-deposition in Alzheimer's disease. Our results indicate that 4OH-GTS-21 is an alpha-7-selective partial agonist with at least 10-fold greater efficacy for both human and rat 7 AChRs than for any -subunit containing AChR. It is also cytoprotective in human and rat cells, unlike GTS-21, which was protective only in the rat cell line.

	Methods

Top Abstract Introduction Methods Results Discussion References

Animals. Female Xenopus laevis were housed in aquarium tanks maintained precisely at 18°C to reduce potential problems with seasonal variability in oocyte viability. Frogs were fed frog brittle (Nasco) and also kept on a 12 hr light:dark cycle.

Cell cultures. Rat PC12 cells and human SK-N-SH cells were purchased from American Tissue Culture Co. (Rockville, MD). PC12 cells were cultured by modifications of Greene and Tischler (1976). They were plated at 20 to 30% confluence on 35-mm plates precoated with poly-L-lysine (10 g/liter), then grown in Dulbecco's modified Eagle medium containing 10% heat-inactivated horse serum, 5% fetal bovine serum, 100 U/ml penicillin, 100 µg/ml streptomycin and 0.5 mM L-glutamine. SK-N-SH cells were cultured and grown similarly, except without poly-L-lysine precoated plates. Cultures were maintained at 37°C, 94% O₂/6% CO₂ and 90 to 92% humidity. Nearly 100% confluent cultures were used for binding studies, although cytoprotection studies used 40 to 50% confluent plates.

Xenopus oocyte expression and recording. Preparation of in vitro synthesized cRNA transcripts and oocyte injection were described previously (de Fiebre et al., 1995, Papke et al., 1997). The measurement of human 42 responses was as described (Gerzanich et al., 1995), although all other conventional recordings were as in Papke et al. (1997). Recordings were made 2 to 7 days after mRNA injections. Current responses to drug administration were studied under two electrode voltage clamp at a holding potential of 50 mV. Recordings were made using a Warner Instruments oocyte amplifier interfaced with National Instruments' (Dallas, Texas) LabView software. Current electrodes were filled with 250 mM CsCl, 250 mM CsF and 100 mM EGTA, pH 7.3 and had resistances of 0.5 to 2.0 M. Voltage electrodes were filled with 3 M KCl and had resistances of 1 to 3 M. Because the general health of the oocytes, their ability to express the heterologous receptors and to maintain stable holding potential were all correlated with the cell's initial resting potential, a criterion for minimal initial resting potential was set at 30 mV. For analysis of concentration-responses relationships, oocytes were placed in a Lucite recording chamber with a total volume of 0.6 ml and were perfused at room temperature with frog Ringer's (115 mM NaCl, 2.5 mM KCl, 10 mM HEPES pH 7.3, 1.8 mM CaCl₂) plus 1 µM atropine to block potential muscarinic responses. The bath solution maintains calcium at a physiologically relevant concentration because extracellular calcium is an important modulator of neuronal nAChR function (Mulle et al., 1992, Vernino et al., 1992); calcium will also have indirect effects on agonist-evoked responses via calcium-dependent chloride channels at this concentration. However, we have previously shown that calcium-dependent effects produce only linear amplification of peak responses and do not distort the concentration-responses relationships over a wide range of agonist concentrations (Papke et al., 1997).

-BTX binding measurements. Cultured cells were homogenized in 10 volumes of iced cold Krebs-Ringer's HEPES (KRH) buffer (NaCl, 118 mM; KCl, 4.8 mM; MgSO₄, 1.2 mM; CaCl₂, 2.5 mM; and HEPES, 20 mM; pH adjusted to 7.5 with NaOH). These homogenates were centrifuged at 20,000 × g for 20 min at 4°C, and the resulting pellets resuspended in Krebs-Ringer's HEPES. Binding of [¹²⁵I]-BTX was measured using the method of Marks et al. (1986). The final incubation contained 200 to 400 µg protein/250 µl with 2 nM [¹²⁵I]-BTX; this 2-hr incubation was at 37°C. Binding was terminated by diluting with 3 ml of ice-cold KRH buffer, followed immediately by filtration through glass fiber filters (Whatman GFB; Springfield Mill, UK) soaked in buffer containing 0.5% polyethylenimine for 30 min at room temperature. Nonspecific binding was determined with 1 mM unlabeled nicotine. Each condition was measured in triplicate, and each assay conducted with at least three separate preparations of tissue. K_i values were calculated by the equation of Cheng and Prusoff (1973) using K_d values calculated for each cell type.

Neuroprotection assay. PC12 or SK-N-SH cells were exposed to 20 µM A 25 to 35 in fresh DMEM for 24 hr in the presence of specified drug treatments. Medium was replaced 10 min before addition of amyloid, with drugs added with the new medium. Cell density was estimated using the NIH Image program version 1.55 (Martin et al., 1994). A Nikon inverted microscope (100× magnification) was attached to a Mac II computer via a monochrome video camera (Cohu, Inc., San Diego, CA). Four random areas were counted per plate and there were four plates per treatment group, unless otherwise indicated. Values were compared by one-way analysis of variance using the Statview II program.

Chemicals. GTS-21 was provided by Taiho Pharmaceuticals (Tokushima, Japan), and 4OH-GTS-21 by Dr. Bill Kem at the University of Florida. All other chemical reagents were purchased from Sigma Chemical Co. (St. Louis, MO).

	Results

Top Abstract Introduction Methods Results Discussion References

GTS-21 and 4OH-GTS-21 concentration-response relationships for human and rat alpha-7 AChR. Data on the full concentration responses relationships for ACh, GTS-21 and 4OH-GTS-21 were generated for human and rat alpha-7 AChR with regard to both activation and the residual inhibition produced to subsequent ACh applications (fig. 1). To compare these results to previous studies (Briggs et al., 1997; de Fiebre et al., 1995; Gopalakrishnan et al., 1995; Hunter et al., 1994; Papke et al., 1997, Sands et al., 1993; Séguéla et al., 1991), responses were plotted as a function of applied concentration. The resulting values are in reasonable agreement with those of previous studies and represent the relative scaling of potencies and efficacies.

View larger version (25K): [in this window] [in a new window]   Fig. 1.   Rat and human alpha-7 AChR responses to ACh, GTS-21 and 4OH-GTS-21. A, The concentration-response relationships for the peak agonist activated currents of oocytes injected with RNA coding for the rat alpha-7 subunit to ACh, GTS-21 and 4OH-GTS-21. All responses were initially measured relative to the individual oocyte's response to 300 µM ACh, applied 5 min before the experimental application. B, The concentration-response relationships for the peak agonist activated currents of oocytes injected with RNA coding for the human alpha-7 subunit to ACh, GTS-21 and 4OH-GTS-21. All responses were initially measured relative to the individual oocyte's response to 300 µM ACh, applied 5 min before the experimental application. To obtain fits that were satisfactory by eye, the maximum responses for ACh and GTS-21 were constrained to the maximum responses achieved experimentally. C, The residual inhibition of control (300 µM) ACh responses of rat alpha-7 injected oocytes after the application of ACh, GTS-21 or 4OH-GTS-21 at the indicated concentrations. All responses are expressed relative to the individual oocyte's response to 300 µM ACh, applied 5 min before the application of the experimental agonist. D, The residual inhibition of control (300 µM) ACh responses of human alpha-7 injected oocytes after the application of ACh, GTS-21 or 4OH-GTS-21 at the indicated concentrations. All responses are expressed relative to the individual oocyte's response to 300 µM ACh, applied 5 min before the application of the experimental agonist.

View this table: [in this window] [in a new window]   TABLE 1 Efficacy and potency of GTS-21 and 4OH-GTS-21 on rat and human 7 AChR

View this table: [in this window] [in a new window]   TABLE 2 Residual inhibition of 7 AChRs (from fig. 1)

-BTX binding in rat and human cell lines. Binding studies with [¹²⁵I]-BTX indicated that both rat PC12 cells and human SK-N-SH cells had similar K_ds for the toxin, 1.7 and 2.2 nM, respectively. Based on an -BTX binding concentration of 2 nM, K_i values for the inhibition of -BTX binding to PC12 cells were 0.31 ± 0.02 µM and 0.17 ± 0.02 µM (mean ± S.E.M. of three experiments) for GTS-21 and 4OH-GTS-21, respectively (fig. 2). For SK-N-SH cells, the K_i values for the inhibition of [¹²⁵I]-BTX binding were 0.45 ± 0.03 µM and 23 ± 2 µM for 4OH-GTS-21 and GTS-21, respectively.

View larger version (16K): [in this window] [in a new window]   Fig. 2.   Effects of GTS-21 and 4OH-GTS-21 on high affinity -BTX binding in membranes from rat PC12 cells and human SK-N-SH cells. Membranes from PC12 cells (top) or SK-N-SH cells (bottom) were assayed for nicotine-displaceable, high affinity binding of -BTX (2 nM) in the presence of specified concentrations of GTS-21 or 4OH-GTS-21. Each value is the mean ± S.E.M. of three plates per group, each assayed in triplicate.

Selectivity of 4OH-GTS-21 for 7 AChRs. The low relative efficacy of GTS-21 for a variety of rat neuronal nicotinic AChRs (i.e., 22, 24, 32, 34, 42 and 44) has previously been reported (Meyer et al., 1997). Specifically, 100 µM GTS-21 produced responses in these AChRs that were less than 1% of the maximum response obtainable with ACh on all these AChR subtypes. To test the hypothesis that the 4OH-GTS-21 derivative would show a similar selectivity for alpha-7-type receptors compared to other nAChRs, the activity of 100 µM 4-OH-GTS-21 was measured on mouse muscle AChR and rat neuronal AChR. Responses to 4OH-GTS-21 were less than 1% of the ACh maximum for 11, 22, 34 and 42 AChR, and 4 ± 1% the ACh maximum for 32 AChR (data not shown).

Effects of GTS-21 and 4OH-GTS-21 on the human 42 AChRs. Because the 42 subunit combination represents the predominant high affinity nicotine receptor in the brain (Flores et al., 1992; Nakayama et al., 1991; Whiting and Lindstrom, 1988), it was also of interest to evaluate the activity of the two anabaseine compounds on the human form of this AChR. Consistent with the results obtained with rodent AChRs, both GTS-21 and 4OH-GTS-21 appeared to be only very weak partial agonists on the human 42 AChRs, with relative efficacies compared to ACh of no more than 5 or 1%, respectively (fig. 3).

View larger version (18K): [in this window] [in a new window]   Fig. 3.   The effects of GTS-21 and 4OH-GTS-21 on human 42 AChRs. A, Responses of an oocyte expressing human 42 AChR to either 10 µM GTS-21 (dotted line) or 10 µM ACh (solid line). The duration of drug application is indicated by the bar above the trace. B, Concentration response relationship for the inhibition of human 42 responses to 10 µM ACh by increasing concentration of GTS-21. C, Responses of an oocyte expressing human 42 AChR to either 100 µM 4OH-GTS-21 (dotted line) or 10 µM ACh (solid line). The maximum response to 4OH-GTS-21 was less than 1% of the ACh maximum. The duration of drug application is indicated by the bar above the trace. D, Concentration response relationship for the inhibition of human 42 responses to 10 µM ACh by increasing concentrations of 4OH-GTS-21.

Cytoprotective effects of GTS-21 and 4OH-GTS-21 on PC12 and SK-N-SH cells. A 25-35 reduced the viability of both PC12 cell and SK-N-SH cells to a similar extent at a 20 µM concentration over a 24-hr interval (fig. 4). GTS-21 and 4OH-GTS-21 both protected against this A 25-35 induced toxicity in PC12 cells at a 10 µM concentration; however, only 4OH-GTS-21 protected SK-N-SH cells at this concentration. The cytoprotective activity of 4OH-GTS-21 in SK-N-SH cells was blocked by mecamylamine, a nicotinic antagonist, indicating the role of nicotinic receptors in its cytoprotective action (fig. 5). Mecamylamine alone had no effect on cell viability, indicating that the endogenous agonist choline, which is generated by cell plasma membranes, had no apparent effect on cell viability.

View larger version (28K): [in this window] [in a new window]   Fig. 4.   Effects of GTS-21 and 4OH-GTS-21 on PC12 and SK-N-SH cell viability in the presence of A 25-35. Cell cultures were incubated for 10 min with the specified drug (10 µM each) before the addition of 20 µM A 25-35. Cell viability was then measured 24 hr later using an NIH image system, with four random areas counted/plate. Each value is the mean ± S.E.M. of cell density (cells/6 mm2 counting area) for at least four plates per group. *P < .05 compared to untreated control group (one-way analysis of variance).

View larger version (25K): [in this window] [in a new window]   Fig. 5.   Effects of mecamylamine on 4OH-GTS-21-induced neuroprotection against A 25-35 in SK-N-SH cells. SK-N-SH cells were treated as described in figure 4, with mecamylamine (1 mM) added to the medium prior to the 4OH-GTS-21 when specified. Each value is the mean ± S.E.M. of at least four plates per group of cell density (cells/6 mm2 counting area). *P < .05 compared to corresponding control value (with or without mecamylamine), one-way analysis of variance.

	Discussion

Top Abstract Introduction Methods Results Discussion References

Nicotine has been studied as a potential treatment for Alzheimer's disease because of its ability to improve a variety of spatial memory and nonspatial avoidance tasks in animals and to enhance delayed recall, attention and other memory-related behaviors in humans. Unfortunately, the toxic side effects associated with this drug reduce its practicality as a therapeutic agent, which has led to the search for novel agents with more selectivity for the nicotinic receptors underlying these behaviors. The observations that GTS-21 and related 3-substituted anabaseine compounds were highly selective alpha-7 agonists, and that these agents improved a variety of memory related behaviors in primates as well as several rodent species, suggested that they may be useful along these lines (Briggs et al., 1997; Woodruff-Pak et al., 1994; Meyer et al., 1994; Arendash et al., 1995). However, GTS-21 was also found to be a very weak agonist at human alpha-7 receptors, arguing that its behavioral activity in primates may depend on some other mechanism of action (Briggs et al., 1997).

Our results demonstrate that the principal metabolite of GTS-21 in humans, 4OH-GTS-21, is a selective partial agonist for both human and rat alpha-7 receptors. It retains the cytoprotective activity of GTS-21 previously reported in rat cell lines, but is the first alpha-7 AChR selective agonist demonstrated to be cytoprotective in human cells. Our results also extend the concept that benzylidine-anabaseine derivatives are selective for alpha-7 AChRs (de Fiebre et al., 1995). The specific substitutions on the benzylidine ring influence this selectivity in terms of regulating the potency, efficacy and the balance between agonist and antagonist properties. The 4OH-metabolite of GTS-21 (Azuma et al., 1996) appears to have a favorable activity profile for the human alpha-7 AChR. The selective retention of agonist activity by 4OH-GTS-21, compared to GTS-21, is consistent with binding experiments on human neuroblastoma SK-N-SH cells and rat PC12 cells. In human cells, GTS-21 produced no consistent concentration-dependent displacement of [¹²⁵I]-BTX at concentrations of up to 10 µM, although similar concentrations of 4OH-GTS-21 produced nearly total displacement of [¹²⁵I]-BTX binding. These observations suggest that GTS-21 itself may function as a pro-drug in therapeutic applications, gaining activity once it is demethylated to 4OH-GTS-21. The activity of the metabolite may therefore have contributed to the efficacy demonstrated for GTS-21 in delayed match to sample experiments with monkeys (Briggs et al., 1997).

Studies with a variety of alpha-7 agonists indicate that agonist and residual inhibitory activities are both important for predicting receptor function (Meyer et al., 1998b; de Fiebre et al., 1995). Residual inhibitory activity appears to interfere particularly with the neuroprotective effects of alpha-7 agonists, and has less effect on memory-related behavioral improvements (Meyer et al., 1998b). 4OH-GTS-21 is only a 50% partial agonist for human and rat alpha-7 AChRs, but has little inhibitory activity these receptors. This profile might permit 4OH-GTS-21 to cause more receptor-activation over extended intervals than seen with full agonists that induce greater inhibition, e.g., DMAC (de Fiebre et al., 1995). The lack of inhibitory activity would also appear to be important for the neuroprotective activity of 4OH-GTS-21.

Amyloid-induced neurotoxicity has been characterized extensively because of the accumulation of this protein in Alzheimer's disease. Mutations in this gene product that increase expression of aggregatable forms of amyloid predispose individuals to the disease, presumably because of aggregating capacity of this A 1-41 form (Hardy, 1997). The peptide fragment A 25-35 also aggregates to form fiber-like structures similar to those formed by A 1-41; however, this peptide does not require several hours of preincubation for the aggregation to occur as A 1-41 does (Cafe et al., 1996; Cotman, 1997). Exposure to either A 25-35 or A 1-41 causes cell death through a process that appears to involve membrane peroxidation and increased intracellular calcium accumulation (Fukuyama et al., 1994; Richardson et al., 1996). Nonselective nicotinic receptor activation as well as GTS-21 have been shown previously to reduce this amyloid-induced toxicity in rodent cells (Kihara et al., 1997; Zamani et al., 1997). Our results extend this observation to human cells for the first time for 4OH-GTS-21. It therefore seems very likely that the neuroprotective activity of this class of compound indeed derives from the ability to activate of this class of receptor, because the former was much more potent at these receptors in oocyte and binding studies.

Nicotinic receptor activation was found previously to exert a neuroprotective action in other model systems for cell viability as well, including trophic factor deprivation in sympathetic ganglia (Koike et al., 1989) or differentiated PC12 cells (Martin et al., 1994), lesioned brain dopamine neurons (Janson, 1988), NMDA-induced toxicity in primary neocortical neurons (Akaike et al., 1994), and ischemia in vivo (Shimohama et al., 1998). Each of these neuroprotective actions was found to be sensitive to antagonists of alpha-7 receptors, including mecamylamine (Martin et al., 1994), which was recently confirmed to inhibit these receptors (Meyer et al., 1997). The antiapoptotic activity of nicotinic receptor activation seems to occur through a protein kinase C-dependent pathway, at least in part (Wright et al., 1993). Nicotinic receptors, particularly highly calcium permeant alpha-7 containing subtypes, may therefore trigger this calcium-sensitive transduction pathway to elicit a neuroprotective action. Studies investigating this neuroprotective transduction process are currently in progress, focusing on the apparent role of protein kinase C activity.

In addition to providing evidence that GTS-21 may serve as a pro-drug for clinical conditions sensitive to alpha-7 activation, our study indicates that species differences in alpha-7 receptor structure may provide important clues for elucidating ligand-receptor interactions among potential agonists. The sequence for alpha-7 subunits is well conserved between rat and human, with 94% sequence identity. Of the 30 nonidentical residues, one site, f222, a tyrosine to phenylalanine difference in the human, stands out as a likely element for species specific effects. This site is nine amino acids downstream from cysteines known to be essential for agonist binding and is conserved as a tyrosine in other putative vertebrate nicotinic alpha-7 subunits cloned to date. Our results suggest that an investigation of the influence of the tyrosine to phenylalanine conversion, along with other subtle differences between these species, is likely to lead to the design of better therapeutic agents for nicotinic receptors.