|Home | About | Journals | Submit | Contact Us | Français|
Xanomeline has been shown to bind in a unique manner at M1 and M3 muscarinic receptors, with interactions at both the orthosteric site and an allosteric site. We have previously shown that brief exposure of Chinese hamster ovary cells that express the M3 receptor to xanomeline followed by removal of free agonist results in a delayed decrease in radioligand binding and receptor response to agonists. In the current study, we were interested in determining the mechanisms of this effect.
Cells were treated with carbachol, pilocarpine or xanomeline for 1 h followed by washing and either used immediately or after waiting for 23 h. Control groups included cells that were not exposed to agonists and cells that were treated with agonists for 24 h. Radioligand binding and functional assays were conducted to determine the effects of agonist treatments.
The above treatment protocol with xanomeline resulted in similar effects of the binding of [3H]NMS and [3H]QNB. When receptor function is blocked using a variety of methods, the long-term effects of xanomeline binding were absent.
Our data indicate that xanomeline wash-resistant binding at the receptor allosteric site leads to receptor downregulation and that receptor activation is necessary for these effects.
There are five subtypes (M1–M5) of muscarinic receptors (mAChR) that are part of a large family of G-protein-coupled receptors. These receptors have seven transmembrane domains, in which interaction of the third cytoplasmic loop with the G-protein  and interactions within transmembrane domains 3, 5, 6, and 7 lead to receptor activation . Each of the subtypes has a unique pattern of expression. M3 mAChR are widely expressed throughout the brain [2, 3]. In addition, they are also expressed in salivary glands, the gastrointestinal tract and other peripheral organs. They are coupled to Gq and G11 G-proteins that in turn activate the phospholipase C signaling pathway . Activation of this receptor subtype may be responsible for some of the adverse effects of cholinesterase inhibitors in Alzheimer's patients, such as salivation and gastrointestinal tract problems [4, 5].
mAChR have a high degree of homology across subtypes in the orthosteric binding domain, which has made it difficult to develop subtype-selective agonists [2, 6]. However, mAChR have been shown to have allosteric binding domains that may differ between subtypes, thus offering the potential for the development of such agents that would be of great therapeutic use . Binding of a drug to an allosteric site on a receptor leads to changes in the conformation of the receptor's orthosteric domain that may result in receptor activation [7, 8]. Examples of such allosteric mAChR agonists include xanomeline , AC-42  and WIN51708 .
Xanomeline has been proposed to bind in a unique manner at the M1 receptor. Binding occurs in a reversible manner at the orthosteric (primary) binding site and in a wash-resistant manner at an allosteric (secondary) binding site . We have demonstrated that this unique binding profile also occurs at the M3 receptor . At both the M1 and M3 receptors, wash-resistant xanomeline binding leads to a long-term decrease in cell-surface receptor expression and attenuation of receptor function [13, 14]. This is similar to the well-documented changes in receptor expression and sensitivity induced by conventional reversible muscarinic agonists [15,16,17,18]. The aim of the current study was to determine the mechanisms of the long-term effects of xanomeline binding at the M3 receptor.
[3H]N-methylscopolamine ([3H]NMS) (81 Ci/mmol) was purchased from DuPont (Wilmington, Del., USA); [3H]quinuclidinyl benzylate ([3H]QNB) (52 Ci/mmol) was purchased from DuPont; myo-[3H]inositol (71 Ci/mmol) was purchased from GE Healthcare (Little Chalfont, UK); [14C]inositol-1-phosphate (300 mCi/mmol) was purchased from American Radiolabeled Chemicals (St. Louis, Mo., USA); Dulbecco's modified Eagle's medium was purchased from Invitrogen (Carlsbad, Calif., USA); geneticin was obtained from Calbiochem (San Diego, Calif., USA); siRNA was purchased from Ambion (Austin, Tex., USA), and bovine calf serum was supplied by Hyclone Laboratories (Logan, Utah, USA). Xanomeline tartrate was a generous gift from Eli Lilly & Co. (Indianapolis, Ind., USA); all other reagents were purchased from Sigma-Aldrich (St. Louis, Mo., USA).
Chinese hamster ovary (CHO) cells stably transfected with human M3 muscarinic acetylcholine receptor (hM3) (provided by Dr. M. Brann, University of Vermont Medical School) were grown in Dulbecco's modified Eagle's medium supplemented with 10% bovine calf serum and 50 μg/ml geneticin. Cells were grown to confluency in 24-well plates or 10-cm dishes at 37°C in a humidified atmosphere consisting of 5% CO2/95% air. All drug pretreatments and radioligand assays were performed on cells in monolayer unless otherwise noted.
CHO cells expressing hM3 receptor were incubated with agonist in monolayer at 37°C for 1 h. The cells were then washed 3 times with Hepes buffer (110 mmol/l NaCl, 5.4 mmol/l KCl, 1.8 mmol/l CaCl2, 1 mmol/l MgSO4, 25 mmol/l glucose, 20 mmol/l Hepes, 58 mmol/l sucrose; pH 7.4 ± 0.02; 340 ± 5 mosm) to remove any unbound drug. Cells were used immediately or after 23 h incubation in fresh culture medium. An additional group of cells was treated with agonist for 24 h prior to washing with Hepes buffer and used in the assay.
CHO cells expressing hM3 receptors were incubated with 50 nmol/l siRNA each for Gq and G11 G-proteins, with the transfection reagent Lipofectamine 2000, in Opti-Mem for 6–7 h. Media with no geneticin was added and the cells were allowed to incubate overnight. Subsequent pretreatment with xanomeline was the same as stated previously.
Two different protocols were used to assess agonist potency. CHO hM3 cells were incubated for 1 h concurrently with a fixed concentration of the muscarinic receptor ligand [3H]N-methylscopolamine ([3H]NMS) (0.2 nmol/l) and increasing concentrations of carbachol (100 nmol/l to 10 mmol/l) or pilocarpine (1 nmol/l to 100 μmol/l). In order to assess the effects of prolonged agonist treatments, cells were pretreated with increasing concentrations of carbachol or pilocarpine (1 pmol/l to 100 μmol/l) for 1 h followed by washing and used immediately or after incubation in ligand free media for 23 h. An additional group of cells was pretreated with agonists for 24 h prior to washing. In the binding assay, cells were incubated with [3H]NMS (0.2 nmol/l) for 1 h. In all instances, pretreatments and radioligand binding were performed in monolayer at 37°C. Non-specific binding was determined using 10 μmol/l atropine. Incubations were terminated by washing away free radioligand and then cells with bound radioligand were dissolved in 1 mol/l NaOH. Protein determinations were performed according to the method of Bradford . Bound radioactivity (disintegrations per minute (dpm)) was quantitated by liquid scintillation spectrometry.
The second protocol was the same as above except saturating concentrations of the muscarinic receptor ligands [3H]NMS (2.9 nmol/l) or [3H]QNB (1.4 nmol/l) were used and cells were pretreated with xanomeline (1 nmol/l to 100 μmol/l). The concentrations of the [3H]NMS and [3H]QNB were chosen so that the percent of receptor occupancy would be the same.
Two different protocols were used to determine the total cell-surface receptor density and radioligand-receptor equilibrium dissociation constant. CHO hM3 cells were incubated in monolayer for 1 h at 37°C in the absence or in the presence of xanomeline (10 μmol/l) alone or in conjunction with atropine (10 μmol/l) followed by three washes with Hepes buffer and 23 h incubation in control media. A second group of cells was untreated or treated with xanomeline for 1 h, washed extensively and then treated with atropine for 23 h. Subsequently, cells were incubated in monolayer with increasing concentrations of [3H]NMS (0.01–6.5 nmol/l) for 1 h at 37°C. Non-specific binding was determined using 10 μmol/l atropine. Reactions were terminated as described above. Protein determinations were performed according to the method of Bradford . Bound radioactivity (dpm) was quantitated by liquid scintillation spectrometry.
In the second protocol, siRNA transfections were performed as noted above. The transfection protocol was unsuccessful in cells seeded in 24-well plates due to significant cell death. We currently do not have an explanation for this phenomenon. Therefore, transfections were performed in 10-cm dishes. Subsequently, cells were incubated for 1 h at 37°C in the absence or presence of xanomeline (10 μmol/l) followed by washing and immediate use or further incubation for 23 h in control medium. Additional cells were treated with xanomeline for 24 h prior to washing. Cells were then removed from the 10-cm dishes by trypsinization and centrifuged (300 g, 2 min) and resuspended in Hepes buffer (3 times). Subsequently, cells were incubated with increasing concentrations of [3H]NMS (0.04–8.5 nmol/l) in suspension for 1 h at 37°C using 105 cells/assay tube in a final volume of 1 ml. The reaction was terminated by filtration on Whatman GF/C filters (Whatman Schleicher & Schuell, Keene, N.H., USA) using a Brandel cell harvester (Brandel, Inc., Gaithersburg, Md., USA). Filters were washed 3 times with 4-ml aliquots of ice-cold saline and dried before radioactivity (dpm) was measured. Non-specific binding was determined using 10 μmol/l atropine and protein content was determined using the method of Bradford .
Two different protocols were used to assess the functional response of the M3 receptor following exposure to various agonists. In all cases, cells in monolayer were loaded 24 h prior to use with myo-[3H]inositol (1 μCi/ml) at 37°C. In the first protocol, cells were pretreated for 1 h with carbachol (10 μmol/l) or pilocarpine (10 μmol/l), washed 3 times with Hepes buffer and used immediately or after 23 h incubation in control media. Additional groups of cells were pretreated with carbachol (10 μmol/l) or pilocarpine (10 μmol/l) for 24 h prior to washing 3 times with Hepes buffer. In the functional assay, cells were incubated in buffer with 10 mmol/l LiCl at 37°C for 1 h in the presence of carbachol (10 nmol/l to 10 mmol/l), pilocarpine (10 nmol/l to 1 mmol/l) or xanomeline (1 nmol/l to 100 μmol/l). The reaction was stopped with 0.3 mol/l HClO4 and neutralized with 0.15 mol/l K2CO3. The samples were centrifuged (1,500 g, 15 min) and total inositol phosphates were separated by ion-exchange chromatography (AG1-X8 resin). In all cases, [14C]inositol-1-phosphate was used as an internal recovery standard. Radioactivity (dpm) was quantitated by liquid scintillation spectrometry.
The second protocol was the same as stated above with the following exceptions. First, only the prolonged xanomeline treatment groups (10 μmol/l; 1 h/wash/23 h wait) were used. Secondly, a receptor saturating concentration of atropine (10 μmol/l) was present concurrently with xanomeline during the initial pretreatment or during the 23-hour waiting period after washing away free xanomeline.
All analyses were conducted using Prism 4.0 (GraphPad Software, Inc., San Diego, Calif., USA). Inhibition of [3H]NMS binding isotherms was analyzed via non-linear regression to derive estimates of IC50 (midpoint location or potency parameter). Data were refitted according to both one- and two-site mass-action binding models, and the better model was determined by an extra sum-of-squares test. The quasi-irreversible nature of xanomeline binding did not permit transforming IC50 values to inhibition constants because such transformation assumes reversible competitive interaction. Saturation binding isotherms were analyzed via non-linear regression to derive individual estimates of Bmax (total cell-surface receptor density) and Kd (radioligand-receptor equilibrium dissociation constant). Even though radioligand depletion was minimal, all calculations were based on the concentration of free radioligand (added dpm minus bound dpm). PI hydrolysis isotherms were analyzed via non-linear regression. Data shown are the means ± SEM. Comparisons between mean values were performed by unpaired t tests, paired t tests or one-way analysis of variance (ANOVA), as appropriate. A probability (p) value <0.05 was taken to indicate statistical significance.
Previous research demonstrated that brief treatment of cells expressing the M3 receptor with xanomeline resulted in a monophasic inhibition curve of [3H]NMS binding whereas long-term treatment produced a biphasic binding inhibition curve . Experiments were designed to confirm that the observed effects of xanomeline are due to its unique interactions with the M3 receptor rather than a general property of agonists. Cells were exposed to increasing concentrations of the full agonist carbachol or the partial agonist pilocarpine  as described in Methods, and the specific binding of 0.2 nmol/l [3H]NMS was determined. In untreated cells, carbachol and pilocarpine inhibited [3H]NMS binding in a concentration-dependent manner when present concurrently with the radioligand, with complete inhibition of binding at high agonist concentrations (98 ± 0.6% carbachol; 99 ± 5.4% pilocarpine). The binding isotherms of both agonists were best described by a single-site binding site and both exhibited a lower potency (pIC50 = 3.6 ± 0.05 carbachol; 4.7 ± 0.11 pilocarpine) than xanomeline (pIC50 = 7.1 ± 0.1) (fig. (fig.1a1a and b, respectively ). When carbachol or pilocarpine were present for 1 h followed by washing and immediate use in the binding assay or incubation in control media for 23 h prior to the binding assay, no decrease in [3H]NMS binding was observed (fig. (fig.1a,1a, b). This is in contrast to results to the wash-resistant effects of xanomeline that are potentiated by prolonged waiting after washing off free xanomeline . However, when cells were incubated with carbachol or pilocarpine for 24 h followed by washing, there was a concentration-dependent inhibition of [3H]NMS binding that was best fit by a single-site binding isotherm with potencies of pCI50 = 7.8 ± 0.19 for carbachol and 6.7 ± 0.05 for pilocarpine, with a maximal inhibition of [3H]NMS binding of 48 ± 1.3 and 40 ± 5.4%, respectively (fig. (fig.1a1a and b, respectively). Once again, this is in contrast to the effects of xanomeline under similar conditions whereby inhibition of [3H]NMS binding results in a biphasic curve and reaches a maximum of nearly 100% . Previous research demonstrated a decrease in protein content following xanomeline pretreatments at the M1 receptor . Thus, experiments were undertaken to determine if there were similar alterations in protein content following the various carbachol or pilocarpine pretreatments described above at the M3 receptor. Measurement of protein content using the method of Bradford  did not detect any differences between the aforementioned treatment and control groups (data not shown).
Previous research has shown that prolonged exposure of cells expressing muscarinic receptors to carbachol leads to a decrease in cell-surface receptor expression with no change in radioligand affinity [16,17,18,21,22,23]. To confirm these results in our experimental paradigm, we determined the effects of long-term exposure of M3 receptors to this agonist. CHO cells expressing the M3 receptor were subjected to the various pretreatments with 10 μmol/l carbachol as detailed above, washed, then exposed to increasing concentrations of [3H]NMS. When cells were treated with carbachol for 1 h followed by washing and incubation in control media for 23 h, there was no change in the maximal binding (Bmax = 56,600 ± 4,200 dpm/well) or radioligand affinity (Kd = 0.28 ± 0.01) compared to control (Bmax = 65,000 ± 2,900 dpm/well; Kd = 0.34 ± 0.01) (fig. (fig.2).2). In contrast, pretreatment with xanomeline under the same conditions resulted in a decrease in cell-surface receptor density . In agreement with previous literature [16,17,18, 23], pretreatment with carbachol for 24 h resulted in a significant (p < 0.05) decrease in maximal binding (Bmax = 26,200 ± 2,700 dpm/well). However, there was an unexpected significant (p < 0.05) increase in radioligand affinity (Kd = 0.23 ± 0.03) compared to control (fig. (fig.2).2). This is in stark contrast to the marked decrease in radioligand affinity observed following pretreatment with xanomeline for 24 h .
Pretreatment of cells with xanomeline either acutely or for long term resulted in antagonism of the functional response to subsequent stimulation with agonists . Thus, experiments were conducted to ascertain if the observed delayed effects of wash-resistant xanomeline binding on receptor function were unique to xanomeline or are due to general effects elicited by long-term exposure to all muscarinic agonists. A 10-μmol/l concentration of either carbachol or pilocarpine was selected, since at this concentration both agonists produce maximal receptor activation and inhibition of [3H]NMS binding following prolonged exposure.
When cells were pretreated with carbachol or pilocarpine for 1 h followed by washing and incubation in control media for 23 h, no change in efficacy or potency of carbachol in simulating phosphoinositide hydrolysis was observed (fig. (fig.3a,3a, b; table table1).1). In contrast, pretreatment with carbachol for 24 h followed by washing resulted in a significant (p < 0.05) decrease in the maximal response and potency of a second addition of carbachol (fig. (fig.3a;3a; table table1).1). A similar long-term treatment with pilocarpine did not alter either concentration-response parameters of carbachol (fig. (fig.3b;3b; table table1).1). In addition, pretreatment with pilocarpine for 1 h followed by washing and waiting did not modify the concentration-response relationship of pilocarpine. However, pilocarpine treatment for 24 h prior to washing followed by a second addition of pilocarpine demonstrated significantly (p < 0.05) decreased potency as well as marked decrease in efficacy (fig. (fig.3c),3c), similar to what was observed with carbachol pretreatment followed by carbachol stimulation (fig. (fig.3a3a).
Further experiments were performed to determine the effects of 24-hour carbachol pretreatment on the ability of xanomeline to stimulate a response (fig. (fig.3d).3d). Pretreatment of cells for 24 h with carbachol followed by washing and stimulation with increasing concentrations of xanomeline resulted in minimal production of inositol phosphates observed only at the highest concentration of xanomeline used (100 μmol/l) (fig. (fig.3d3d).
Prolonged agonist exposure can lead to receptor internalization, downregulation or uncoupling of the receptor from its G-protein [24, 25]. The current set of experiments was conducted to differentiate the long-term effects of xanomeline on internalization and downregulation of the M3 receptor. QNB is a lipophilic muscarinic receptor antagonist that is able to bind to both cell-surface and internalized (but intact) receptors [26, 27]. In contrast, NMS (a permanently-charged quaternary amine) is only able to bind to receptors expressed on the cell surface . We compared effects of the various xanomeline treatment paradigms on the specific binding of both radioligands. Receptor-saturating concentrations of [3H]QNB and [3H]NMS were used to confine effects of xanomeline to alterations in the number of receptors without interference from changes in radioligand affinity.
Cells were exposed to increasing concentrations of xanomeline for the various treatment conditions and the binding of either 1.4 nmol/l [3H]QNB or 2.8 nmol/l [3H]NMS was determined. These concentrations were chosen because they result in similar receptor occupancy by either ligand (approximately 93% of total receptors). In untreated cells, inclusion of xanomeline in the binding assay resulted in complete concentration-dependent inhibition of binding of both radioligands with similar potency (fig. (fig.4a,4a, b; table table2).2). Cells pretreated with xanomeline for 1 h followed by washing and immediate use in the binding assay showed wash-resistant xanomeline binding that exhibited a lower potency against either radioligand compared to when xanomeline was included in the binding assay (p < 0.05) (fig. (fig.4a,4a, b; table table2).2). Additionally, inhibition of [3H]NMS binding was incomplete following this treatment. When cells were pretreated with xanomeline for 1 h followed by washing and waiting 23 h, there was a significant (p < 0.05) increase in the apparent potency of xanomeline as compared to its acute wash-resistant effects. Surprisingly, this increase was greater for NMS compared to QNB binding. Noteworthy, inhibition of binding under these treatment conditions was incomplete for both radioligands resulting in a maximal inhibition of approximately 35–45% (fig. (fig.4a,4a, b; table table22).
An additional group of cells was incubated with increasing concentrations of xanomeline for 24 h followed by washing and immediate use in radioligand binding assays. This condition resulted in biphasic xanomeline inhibition curves of either radioligand, with the high-potency site resulting in approximately 42–43% of the binding sites for both radioligands. The apparent high- potency site under these conditions demonstrated a markedly higher potency than that observed in the case of cells that were briefly treated with xanomeline, washed and used immediately or further incubated in control media for a day (table (table2).2). Once again the inhibition of binding was incomplete for both radioligands under these conditions (table (table22).
Experiments were designed to determine the role of xanomeline interaction with the receptor's orthosteric domain in the development of its long-term effects. A receptor-saturating concentration of atropine (10 μmol/l), a muscarinic receptor antagonist, was added either concurrently with 10 μmol/l xanomeline pretreatment or during the waiting period following xanomeline washout. [3H]NMS saturation binding isotherms were constructed to examine the effects of blocking the orthosteric site with atropine on xanomeline-induced changes in both radioligand affinity and maximal binding. Since atropine has been shown to induce its own effects [23, 29], comparisons were made between cells treated with both atropine and xanomeline and those treated with atropine alone. The simultaneous presence of atropine during incubation of cells with xanomeline for 1 h followed by washing and waiting 23 h in control medium did not influence the significant (p < 0.05) decrease in [3H]NMS maximal binding induced by xanomeline pretreatment alone (fig. (fig.5a,5a, b; table table3).3). However, when atropine was present during the 23-hour incubation following xanomeline pretreatment and washing, the long-term changes in radioligand maximal binding were no longer evident (fig. (fig.5c;5c; table table3).3). In contrast, either protocol of atropine treatment abolished the long-term effects of xanomeline on radioligand affinity. All treatment groups had similar protein content.
Experiments were conducted to further examine the role of receptor activation on the long-term effects of xanomeline treatments on [3H]NMS saturation binding parameters using another experimental approach. It is known that the M3 receptor exerts its functional response via activation of the Gq/11 class of G-proteins . Thus, we utilized siRNA technology to block expression of these G-proteins. Cells were transfected with 50 μmol/l siRNA targeting Gq and G11 G-proteins in an effort to block receptor function. Western blot analysis confirmed that transfection with siRNA targeting Gq/11 resulted in a nearly 100% decrease in the detectable expression of Gq/11 class of G-proteins (fig. (fig.6a).6a). In addition, loss of receptor function following transfection was assessed by determining agonist-mediated production of inositol phosphates using two concentrations of carbachol (1 and 100 μmol/l) or xanomeline (100 nmol/l and 10 μmol/l). The response to carbachol was decreased by 62 and 27%, respectively, and the response to xanomeline was reduced by 78 and 82%, respectively (fig. (fig.6b).6b). The ability of agonists to stimulate a functional response subsequent to siRNA treatment suggests that although the levels of Gq/11 G-proteins were no longer detectable following siRNA treatment, some Gq/11 G-proteins were still present. A [3H]NMS saturation binding paradigm was applied to determine the effects of blocking receptor function on xanomeline-induced changes in both radioligand affinity and maximal binding. However, while attached cells were pretreated with xanomeline, binding experiments were conducted in suspension due to technical limitations (see Methods). In control non-siRNA-transfected cells, a trend toward a decrease in radioligand affinity was observed following 1-hour immediate-use xanomeline pretreatment as well as pretreatment with xanomeline for 24 h prior to washing. Additionally, a trend toward a decrease in maximal binding resulted with 1-hour xanomeline pretreatment followed by washing and waiting 23 h or treatment for 24 h with xanomeline followed by washing and no waiting. These effects on radioligand affinity and maximal binding were similar, although less pronounced, than those obtained from [3H]NMS saturation binding performed in monolayer  (fig. (fig.6c;6c; table table4).4). This is similar to previous work by Thompson and Fisher  in which there was a greater loss of receptors in cells that were pretreated with carbachol in monolayer compared to cells pretreated in suspension prior to the radioligand binding assay. Cells transfected with siRNA displayed a trend toward a decrease in radioligand affinity for all xanomeline pretreatment groups except the 24-hour treatment group that had a significant (p < 0.05) decrease in radioligand affinity compared to control. There was no change in maximal binding for any of the siRNA-treated groups compared to control (fig. (fig.6d;6d; table table4).4). Protein content remained the same across the various pretreatment groups (data not shown).
Similar to the previously described [3H]NMS binding experiments, functional experiments were conducted to determine the effects of blocking the primary binding site with a receptor-saturating concentration of atropine (10 μmol/l), either concurrently with 10 μmol/l xanomeline pretreatment for 1 h or during the 23-hour waiting period following xanomeline pretreatment and washing on the ability of carbachol to stimulate the production of inositol phosphates. In contrast to results obtained from binding experiments, simultaneous presence of atropine with xanomeline for 1 h followed by washing and waiting for 23 h in control medium abolished the effects of xanomeline on carbachol potency (fig. (fig.7a,7a, b; table table3).3). The presence of atropine during the 23-hour incubation following xanomeline treatment and washout also eliminated xanomeline-induced changes in carbachol potency, although a decrease in efficacy was observed compared to control (fig. (fig.7c;7c; table table33).
Shortening the hydrophobic chain of xanomeline has previously been shown to result in diminishing its wash-resistant binding and functional properties at the M1 receptor . Thus, the propyl analog of xanomeline (three carbons shorter than xanomeline) does not display wash-resistant binding or stimulate a functional response at the M1 receptor. At the M3 receptor, this analog exhibited wash-resistant binding, albeit with very low potency compared to xanomeline (data not shown). While this analog produced 45% of the PI response to xanomeline when present in the assay of receptor function, it did not display any wash-resistant receptor activation (data not shown). We therefore utilized the propyl analog to discern the role of persistent receptor activation, versus wash-resistant binding per se in the observed long-term effects of xanomeline at the M3 receptor. When cells were exposed to the propyl analog at 100 μmol/l for 1 h followed by washing and overnight incubation in control media, there was only a minute decrease in cell-surface receptor density (Bmax 51,500 ± 3,100 dpm/well) in conjunction with a slight increase in radioligand affinity (Kd 0.74 ± 0.01 nmol/l) compared to control (Bmax 73,400 ± 4,900 dpm/well; Kd 0.74 ± 0.01) (fig. (fig.8).8). In contrast, 24-hour exposure of cells to the propyl analog prior to washing resulted in an anomalous significant (p < 0.05) increase in cell-surface receptor density (133,100 ± 10,500 dpm/well) along with a significant (p < 0.05) increase in the Kd of [3H]NMS to 5.73 ± 0.74 nmol/l compared to control (fig. (fig.88).
Previous literature has shown that wash-resistant inhibition of [3H]NMS binding by xanomeline occurs with markedly lower potency at 4°C than at 37°C . However, functional assays demonstrated that no functional response was elicited by xanomeline at 4°C (data not shown). Therefore, experiments were conducted at 4°C to further demonstrate the need for a functional receptor to observe long-term effects of xanomeline. Increasing concentrations of [3H]NMS were used to determine cell-surface receptor density and radioligand affinity. Exposure of cells to xanomeline for 1 h at 37°C followed by washing and 23-hour incubation at 4°C in the absence of free xanomeline did not result in changes in maximal cell-surface receptor density (33,400 ± 3,600 dpm/well) or radioligand affinity (Kd 0.34 ± 0.03 nmol/l) compared to the 4°C control (Bmax 35,000 ± 2,800; Kd 0.24 ± 0.004 nmol/l) (fig. (fig.9).9). However, when cells were exposed to xanomeline at 4°C for 24 h prior to washing there was a significant (p < 0.05) concentration-dependent increase in [3H]NMS Kd (1.08 ± 0.06 and 6.25 ± 0.44 nmol/l, respectively) in the absence of a change in cell-surface receptor density (35,300 ± 4,100 and 32,500 ± 2,000, respectively) compared to the 4°C control (fig. (fig.99).
In the literature, xanomeline has been proposed to be a functionally selective M1/M4 muscarinic agonist [4, 32]. In agreement with previous findings, we have demonstrated that xanomeline acts as a partial agonist at the M3 receptor [4, 13]. Inadvertent activation of the M3 receptor by xanomeline may account for some of the side effects that were seen in clinical trials such as gastrointestinal tract problems and hypersalivation [4, 5]. In addition, we have demonstrated that long-term changes in M3 receptor binding and function are evident following acute pretreatment with xanomeline . The current study was undertaken to discern the mechanisms underlying long-term regulation of the M3 receptor by xanomeline.
Pretreatment with pilocarpine or carbachol followed by washing did not result in acute or delayed effects on inhibition of [3H]NMS binding, suggesting that the wash-resistant effects of xanomeline are ligand-specific rather than a general property of brief agonist binding to the receptor (fig. (fig.1).1). Exposure of cells to either carbachol or pilocarpine for 24 h prior to washing resulted in a monophasic concentration-dependent inhibition of [3H]NMS binding that exhibited markedly higher agonist potencies than those obtained when the agonists were added directly to untreated cells in the binding assay medium. In contrast, treatment with xanomeline for 24 h prior to washing resulted in a biphasic binding curve . While the profile of xanomeline treatment is very different from that of other agonists, the high-potency component of inhibition of [3H]NMS binding following xanomeline treatment resembles that of the monophasic attenuation of binding caused by 24-hour pretreatment with other agonists. This similarity suggests that the former might be the result of a common receptor response to long-term activation. In support of this notion, incubation of cells with carbachol or pilocarpine for 24 h prior to washing resulted in incomplete inhibition of [3H]NMS binding that was similar in magnitude to the proportion of high-potency binding induced by long-term incubation of xanomeline-pretreated and washed cells . This might be due to an inherent limited susceptibility of the receptor to agonist-induced regulation. In contrast, 24-hour pretreatment with higher concentrations of xanomeline resulted in complete inhibition of [3H]NMS binding , further suggesting that xanomeline regulates the receptor in a distinct manner compared to other conventional agonists.
The specific effects of wash-resistant xanomeline binding were also observed in [3H]NMS saturation binding experiments. Exposure of cells to carbachol for 1 h followed by washing and incubation in control media for 23 h resulted in radioligand binding parameters that were not different from control (fig. (fig.2).2). However, pretreatment with xanomeline under the same conditions led to a decrease in cell-surface receptor density . This suggests that long-term receptor regulation by xanomeline is the result of its persistent binding to the receptor rather than the initial interaction with the receptor prior to washing. Similarly to previously published results [16,17,18, 23], exposure of cells to carbachol for 24 h prior to washing resulted in a decrease in receptor density. However, an increase in radioligand affinity was also observed, which is in contrast to previous findings. At present we do not have a reasonable explanation for this anomalous observation. Pretreatment with xanomeline for 24 h prior to washing also resulted in a decrease in cell-surface receptor expression, but this effect was accompanied by a significant decrease in radioligand affinity . These observations may be explained by interference of wash-resistant xanomeline with [3H]NMS binding in an allosteric manner.
The observed delayed effects of acute xanomeline wash-resistant binding were associated with attenuation of the receptor response to activation by carbachol. Control experiments with the conventional agonists carbachol and pilocarpine confirmed that the functional changes observed following xanomeline pretreatments were due to unique ligand-specific characteristics rather than being a general property of agonist binding to the receptor. When cells were pretreated with pilocarpine for 1 h followed by washing and waiting 23 h or for 24 h prior to washing, no changes in carbachol-induced receptor stimulation were observed. Similarly, pretreatment with carbachol for 1 h followed by washing and waiting 23 h did not alter the ability of carbachol to elicit a response. However, when cells were treated for 24 h with carbachol prior to washing, there was a depression in the maximal response in addition to a reduction in potency of the carbachol-mediated response (fig. (fig.3).3). Literature supports that the observed decrease in the functional response may be due to internalization of cell-surface receptors in addition to uncoupling of the receptor from G-proteins [16, 18]. Pretreatment of cells with pilocarpine produced changes in its ability to subsequently activate the receptor similar to those described above for carbachol. Contradictory reports exist regarding the effects of prolonged exposure to pilocarpine on receptor expression [33, 34]. However, in our paradigm 24-hour exposure of cells to pilocarpine prior to washing resulted in a decrease in receptor expression (data not shown). Thus, the change in the functional response to pilocarpine can be attributed to a decrease in cell-surface receptor density following continuous agonist exposure, because a partial agonist needs near or full receptor occupancy to stimulate a maximal response .
Previous research has documented that prolonged exposure to an agonist can lead to receptor internalization, downregulation or uncoupling of the G-protein . Experiments were therefore conducted to determine whether the observed long-term effects of xanomeline are due to receptor internalization or downregulation. These possibilities were differentiated by comparing the effects of xanomeline treatment on the specific binding of [3H]QNB and [3H]NMS (fig. (fig.4).4). QNB is a lipophilic antagonist that is able to bind to both cell-surface and internalized (but intact) receptors [26, 27], whereas NMS (a permanently-charged quaternary amine) is only able to bind to receptors located on the cell surface . The absence of significant differences between the binding profiles of [3H]QNB and [3H]NMS as a result of 24-hour treatment with xanomeline prior to washing supports the notion that the effects observed in both displacement and saturation binding experiments are mainly due to receptor downregulation rather than internalization. Interestingly, treatment with xanomeline for 1 h followed by washing and waiting 23 h indicated an apparent higher potency of xanomeline in decreasing [3H]QNB than [3H]NMS binding. An opposite profile would have been expected if a major proportion of the receptors were being internalized rather than downregulated. One possible explanation of these data is that xanomeline wash-resistant binding to the M3 receptor may impart differential allosteric modulation of NMS and QNB binding. The magnitude of cooperativity between the orthosteric and an allosteric site on a receptor is dependent on the specific receptor conformations effected by binding of various ligands . Therefore, the effects of a given allosteric modulator are ligand specific.
While xanomeline wash-resistant binding takes place at an allosteric site on the muscarinic receptor, the persistent functional response likely involves activation of the receptor primary binding domain by xanomeline's active head group [9, 12, 37]. While the formation of xanomeline wash-resistant binding takes place in the presence of atropine, the resulting wash-resistant receptor activation is blocked by atropine . Thus, further experiments were conducted to determine the role of receptor activation in the long-term changes observed following acute exposure of the receptor to xanomeline (fig. (fig.5).5). Blocking the primary binding site with atropine during brief treatment with xanomeline did not prevent long-term reductions in receptor availability induced by persistently-bound xanomeline, suggesting that activation of the receptor during the initial treatment period is not necessary in order for xanomeline to induce long-term changes. These data also indicate that access to the orthosteric site is not necessary for wash-resistant binding of xanomeline to take place. This is in agreement with previous studies at the M1 receptor showing that this interaction takes place at a secondary receptor binding site [12, 14, 38]. However, blocking the primary binding site with atropine during the long waiting period following washing off free xanomeline blunted the long-term effects of xanomeline on [3H]NMS binding. Together, these results indicate that receptor activation is necessary to produce the long-term consequences of xanomeline wash-resistant binding.
To further explore the importance of receptor activation in the long-term effects of xanomeline, receptor function was inhibited using siRNA targeted against the Gq and G11 G-proteins (fig. (fig.6).6). Functional assays and Western blots confirmed that receptor function and detectable Gq/11 G-protein expression were significantly reduced. When siRNA-transfected cells were subjected to the various xanomeline treatments, no differences in receptor density on the cell-surface were observed. This is in contrast to non-transfected cells in which prolonged xanomeline pretreatments resulted in decreased cell-surface receptor density. These results further support the importance of receptor function in the development of xanomeline's long-term effects.
Functional experiments were conducted to determine the effects of suppressing receptor activation on the delayed actions of xanomeline (fig. (fig.7).7). The concomitant presence of atropine with xanomeline during the initial pretreatment period, followed by washing and prolonged incubation in ligand-free medium, prevented the change in carbachol potency in stimulating inositol phosphate production observed following xanomeline treatment alone. This is in contrast to the results obtained in [3H]NMS saturation binding experiments in which the long-term effects of wash-resistant xanomeline binding were still evident under these conditions. In comparison, the presence of atropine during the 23-hour incubation following acute xanomeline treatment resulted in a decrease in carbachol efficacy. Once again, this is in contrast to the observed antagonism by atropine of xanomeline-induced changes in [3H]NMS saturation binding. These discrepancies may be due to unique interactions of atropine with the receptor. While commonly described as an antagonist, atropine also exhibits properties of an inverse agonist [39, 40]. For example, previous research found that continuous treatment with atropine can lead to an increase in receptor expression [23, 29]. Thus, the effects of atropine itself may be concealing the functional effects of long-term xanomeline exposure.
Previous experiments at the M1 receptor have shown that changing the structure of xanomeline, specifically shortening the carbon side chain, results in alterations in its persistent binding and receptor activation . We confirmed that the propyl analog of xanomeline, which has three fewer carbons in the side chain, exhibits low-potency wash-resistant binding at the M3 receptor in the virtual absence of concomitant receptor activation. Using a [3H]NMS saturation binding paradigm (fig. (fig.8),8), pretreatment with the propyl analog for 1 h followed by washing and prolonged incubation resulted in a very small decrease in cell-surface receptor expression compared to that elicited by xanomeline, without a change in radioligand affinity. In contrast, 24-hour treatment with the propyl analog prior to washing resulted in an unexpected significant (p < 0.05) increase in cell-surface receptor density in conjunction with a large reduction in radioligand affinity. This is in stark contrast to the effects induced by xanomeline treatment for 24 h prior to washing, whereby a marked decrease in cell-surface receptor expression was observed. Taken together, these results suggest that wash-resistant binding in the absence of persistent receptor activation is not the major contributor to the long-term effects of xanomeline binding.
Additional [3H]NMS saturation binding experiments were conducted at 4°C, which has been shown to prevent receptor activation by agonists [15, 41]. Treatment of cells with xanomeline for 1 h at 37°C followed by washing and 23-hour incubation at 4°C did not result in changes in binding parameters compared to untreated control cells incubated at 4°C. However, when cells were treated with 10 or 100 μmol/l xanomeline for 24 h at 4°C prior to washing, there was a concentration-dependent decrease in radioligand affinity (fig. (fig.9).9). Taken together, these results suggest that xanomeline avidly bound at the receptor allosteric site requires receptor activation for the induction of its long-term effects on receptor expression. Interestingly, while the effects of xanomeline on maximal cell-surface receptor density seem to be dependent on receptor activation at the orthosteric site, xanomeline's effects on radioligand affinity are not. This differential effect on Bmax and Kd is consistent with other protocols that inhibit receptor function. It has to be noted, however, that changing temperature is known to elicit general changes in the conformation of many cellular proteins in addition to altering receptor conformation.
Even though xanomeline is a partial agonist at the M3 receptor , it exerts long-term changes in receptor regulation similar to those observed at the M1 receptor [13, 14], albeit to a lesser extent. This may suggest a shared mechanism of agonist-induced regulation for the M3 and M1 receptors. This is not surprising since functional responses at both the M3 and M1 receptors are elicited via the same class of G-proteins .
In summary, we have demonstrated that the long-term consequences of xanomeline binding at the M3 receptor are due to unique properties of the mode of interaction of xanomeline with the receptor. These effects of xanomeline are most likely the result of receptor downregulation, allosteric modulation of the receptor or a combination of both. Additionally, long-term receptor modulation by xanomeline is dependent on an intact functional receptor response.
This work was supported by National Institutes of Health grant RO1-NS25743 and grant No. T32DE007288 from the National Institute of Dental & Craniofacial Research. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institute of Dental & Craniofacial Research or the National Institutes of Health.