Additional diversity-oriented exploration applied to our two previously disclosed, highly selective M₅ Positive Allosteric Modulator (PAM) probes (ML129 and ML172) initially met with flat Structure-Activity-Relationship (SAR) and negligible improvements. Speculative incorporation of a novel fragment structure from the unconfirmed, single-point High Throughput Screen (HTS) effort concurrently underway at Scripps Research Institute Molecular Screening Center (SRIMSC) initiated a new series of compounds. SAR around this series provided, for the first time, sub-micromolar functional activity and ultimately led to the selection of ML326 as our 3^rd generation M₅ PAM probe molecule (hM₅ PAM EC₅₀ = 550 nM, hM_1–4 > 30 μM, hM₅ ACh fold-shift = 19). This best-in-class compound will find utility in assays directed toward understanding M₅ receptor signaling both in vitro and, to a more limited extent, in vivo.

Probe Structure & Characteristics

ML326

1. Recommendations for Scientific Use of the Probe

2. Materials and Methods

Cell culture and transfections. Chinese hamster ovary (CHO-K1) cells stably expressing rat (r)M₁ were purchased from the American Type Culture Collection and cultured according to their indicated protocol. CHO cells stably expressing human (h)M₂, hM₃, and hM₅ were generously provided by A. Levey (Emory University, Atlanta, GA); hM₁ and hM₄ cDNAs were purchased from Missouri S&T cDNA Resource; rM₄ cDNA was provided by T. I. Bonner (National Institutes of Health, Bethesda, MD). hM₁, hM₄, and rM₄ cDNAs were used to stably transfect CHO-K1 cells purchased from the American Type Culture Collection using Lipofectamine2000. To make stable hM₂, rM₂, hM₄ and rM₄ cell lines for use in calcium mobilization assays, these cells also were stably transfected with a chimeric G-protein (G_qi5) (provided by B.R. Conklin, University of California, San Francisco) using Lipofectamine 2000. rM₁, hM₃, and hM₅ cells were grown in Ham’s F-12 medium containing 10% heat-inactivated fetal bovine serum (FBS), 20mM HEPES, and 50μg/mL G418 sulfate. hM₂–G_qi5 and hM₄–G_qi5 cells were grown in the same medium also containing 500 μg/mL Hygromycin B. Stable rM₄–G_qi5 cells were grown in DMEM containing 10% heat-inactivated FBS, 20 mM HEPES, 400 μg/mL G418 sulfate, and 500 μg/mL Hygromycin B. All cell culture reagents were purchased from Invitrogen Corp. (Carlsbad, CA) unless otherwise noted.

Calcium Mobilization Assays - Potency determinations. Assays were performed within the Vanderbilt Center for Neuroscience Drug Discovery’s Screening Center. CHO cell lines expressing muscarinic acetylcholine receptors were plated (15,000 cells/20 μL/well) in black-walled, clear-bottomed, TC treated, 384 well plates (Greiner Bio-One, Monroe, North Carolina) in Ham’s F-12, 10% FBS, 20 mM HEPES. The cells were grown overnight at 37 °C in the presence of 5% CO₂. The following day, plated cells had their medium exchanged to Assay Buffer (Hank’s balanced salt solution, 20 mM HEPES and 2.5 mM Probenecid (Sigma-Aldrich, St. Louis, MO)) using an ELX405 microplate washer (BioTek), leaving 20 μL/well, followed by addition of with 20 μL of 4.5 μM Fluo-4, AM (Invitrogen, Carlsbad, CA) prepared as a 2.3 mM stock in DMSO and mixed in a 1:1 ratio with 10% (w/v) pluronic acid F-127 and diluted in Assay Buffer for 45 minutes at 37 °C. The dye was then exchanged to Assay Buffer using an ELX405, leaving 20 μL/well and the plates were incubated at room temperature for 10 min prior to assay. Test compounds were transferred to daughter plates using an Echo acoustic plate reformatter (Labcyte, Sunnyvale, CA) and then diluted into Assay Buffer to generate a 2× stock in 0.6% DMSO (0.3% final). Acetylcholine (ACh) EC₂₀ and EC₈₀ were prepared at a 5× stock solution in assay buffer prior to addition to assay plates. Calcium mobilization was measured at 37 °C using a Functional Drug Screening System 6000 or 7000 (FDSS6000 or FDSS7000, Hamamatsu, Japan) kinetic plate reader according to the following protocol. Cells were preincubated with test compound (or vehicle) for 144 seconds prior to the addition of an EC₂₀ concentration of the agonist, ACh. 86 seconds after this addition, an EC₈₀ concentration of ACh was added. Control wells also received a maximal ACh concentration (1 mM) for eventual response normalization. The signal amplitude was first normalized to baseline and then as a percentage of the maximal response to ACh. Microsoft XLfit (IDBS, Bridgewater, NJ) was utilized for curve fitting and EC₅₀ value determination using a four point logistical equation. Compounds showing dose-dependency were assigned ‘Outcome’ = ‘Active’, EC₅₀=‘Value’, and % ACh max=‘Value’.

Calcium Mobilization Assays - Fold-Shift Assays. Assays were performed within the Vanderbilt Center for Neuroscience Drug Discovery’s Screening Center. CHO cell lines expressing muscarinic acetylcholine receptors were plated (15,000 cells/20 μL/well) in black-walled, clear-bottomed, TC treated, 384 well plates (Greiner Bio-One, Monroe, North Carolina) in Ham’s F-12, 10% FBS, 20 mM HEPES. The cells were grown overnight at 37 °C in the presence of 5% CO₂. The following day, plated cells had their medium exchanged to Assay Buffer (Hank’s balanced salt solution, 20 mM HEPES and 2.5 mM Probenecid (Sigma-Aldrich, St. Louis, MO)) using an ELX405 microplate washer (BioTek), leaving 20 μL/well, followed by addition of with 20 μL of 4.5 μM Fluo-4, AM (Invitrogen, Carlsbad, CA) prepared as a 2.3 mM stock in DMSO and mixed in a 1:1 ratio with 10% (w/v) pluronic acid F-127 and diluted in Assay Buffer for 45 minutes at 37 °C. The dye was then exchanged to Assay Buffer using an ELX405, leaving 20 μL/well and the plates were incubated at room temperature for 10 min prior to assay. Test compounds were prepared in Assay Buffer to generate a 2× stock in 0.6% DMSO (0.3% final). Acetylcholine (ACh) concentration responses were prepared at a 5× stock solution in assay buffer prior to addition to assay plates. Calcium mobilization was measured at 37 °C using a Functional Drug Screening System 6000 or 7000 (FDSS6000 or FDSS7000, Hamamatsu, Japan) kinetic plate reader according to the following protocol. Cells were preincubated with test compound (or vehicle) for 144 seconds prior to the addition of a concentration response of the agonist ACh and the fluorescence was monitored for a total of 5 min. The signal amplitude was first normalized to baseline and then as a percentage of the maximal response to acetylcholine. Microsoft XLfit (IDBS, Bridgewater, NJ) was utilized for curve fitting and EC₅₀ value determination using a four point logistical equation. Fold-Shift values were calculated by dividing the ACh EC₅₀ in the presence of 30 μM compound by the ACh EC₅₀ in the presence of vehicle. Compounds showing greater than a 3-fold shift were assigned ‘Outcome’ = ‘Active’, ‘Fold-Shift’ =‘Value’, and ‘% ACh max’ = ‘Value’.

2.2. Probe Chemical Characterization

Probe compound ML326 (CID 57525860, SID 137294184, VU0467903) was prepared according to the above scheme and provided the following characterization. 1-(2-phenoxyethyl)-5-(trifluoromethoxy)indoline-2,3-dione: TLC R_f = 0.79 (hexane/ethyl acetate 1:1); >98% pure by LCMS (M+H⁺ = 352); HRMS calcd for C₁₇H₁₃NO₄F₃[M+H⁺]; 352.0797 found 352.0795; ¹H NMR (400 MHz, CDCl₃ calibrated to 7.26) δ7.52-7.46 (m, 2H), 7.31-7.25 (m, 3H), 6.98 (t, J = 7.4 Hz, 1H), 6.82 (d, 2H), 4.28 (t, J = 5.0 Hz, 2H), 4.17 (t, J = 5.0 Hz, 2H); ¹³C NMR (400 MHz, CDCl₃ calibrated to 77.16) δ182.25, 158.27, 157.93, 150.01, 145.44, 131.02, 129.78, 121.75, 118.32, 114.39, 112.89, 65.94, 40.62.

Solubility: Solubility for ML326 in PBS (@ pH = 7.4, final DMSO concentration 1%) was determined to be 1.5 μM, which is ~3-fold higher than its hM₅ EC₅₀ and better than the 2^nd generation M₅ probe compound ML172 (< 0.10 μM), but slightly worse than the 1^st generation M₅ probe ML129 (3.1 μM).

GSH Conjugates. Not determined due to limitations associated with the ionization properties of ML326 (vide infra) and the miniaturized scale of the assay procedure.

Stability. Stability was determined for ML326 at 23 °C in PBS (no antioxidants or other protectorants, initial ML326 concentration = 100 μM and final DMSO concentration 5%). After 48 hours, 3% of the initial concentration of ML326 remained. However, this is not concerning considering the solubility of ML326 was determined to be 1.5 μM under very similar conditions (vide supra). The 4% additional DMSO used in this stability determination ultimately accounts for the seemingly improved solubility at the later time points (30 minutes and beyond). The relatively constant concentrations seen for time points beyond 30 minutes are consistent with the good chemical stability seen during the preparation of ML326 and are what would be expected from examination of its molecular structure.

View in own window

	Percent Remaining (%)
Compound	0 Min	15 Min	30 Min	90 min.	24 Hour	48 Hour
ML326, CID 57525860	100	14	3.9	3.7	6.6	2.9

This data can also be displayed as a graph:

Figure 1Stability of ML326

Compounds added to the SMR collection (MLS#s): MLS004576079 (ML326, CID 57525860, 25.0 mg); MLS004576080 (CID 57525831, 7.6 mg); MLS004576081 (CID 57525868, 11.5 mg); MLS004576082 (CID 57525867, 12.8 mg); MLS004576083 (CID 57525838, 6.5 mg); MLS004576084 (CID 57525847, 5.0 mg).

3. Results

There are five muscarinic acetylcholine receptor (mAChR) subtypes (M₁ – M₅) widely expressed in both the central nervous system (CNS) and periphery of mammals. These receptors, whose endogenous agonist is acetylcholine (ACh), play critical roles in regulating a variety of diverse physiological processes. Within the CNS, the M₁, M₄ and M₅ subtypes are believed to be the most important with respect to normal neuronal functioning. Of these three, M₅ is the least studied as a combined result of its lower expression levels¹ (< 2% of total muscarinic receptor population within the brain) and, until recently, a near absence of highly selective M₅ receptor ligands. Much of our current understanding surrounding the function of M₅ has come from M₅ receptor localization, M₅ knock-out mice² and experiments conducted with non-selective muscarinic ligands.³ It is intriguing to note that the M₅ receptor is the only muscarinic receptor observed the substantia nigra pars compacta (based on mRNA detection),^4,5 leading to the prediction that M₅ functions in addiction/reward mechanisms. This hypothesis was supported in man through the clinically observed correlation between a specific M₅ gene mutation and an increase in cigarette consumption (+ 27%), as well as, an increased risk for cannabis dependence (+ 290%).⁶ Additionally, M₅ receptors have been localized on the cerebrovascular system and shown to be critical in the ACh-induced dilation necessary for normal blood perfusion.⁷ In a broader sense, M₅ KO mice show decreased prepulse inhibition,⁸ CNS neuronal abnormalities and cognitive deficits,⁹ thus supporting the potential for M₅ ligands in the treatment of numerous CNS disorders including schizophrenia, Alzheimer’s disease, ischemia and migraine.

To date, only two publications have described the development of M₅ selective ligands. Both manuscripts disclosed positive allosteric modulators (PAMs) culminating in the development of two MLPCN probe molecules: ML129¹⁰ and ML172¹¹ (Figure 2). While these molecules demonstrated unprecedented selectivity towards the M₅ receptor, their potency and physical properties may indicate why these probes are unlikely to find broad utility in animal models exploring the function of M₅. While we have entered into at least one collaborative effort to explore the effects of these M5 PAMs on addictive and reinforced behavior in rats, the physical properties of the current probes will necessitate intracerebroventricular (ICV) infusion for those animal studies. This route of administration is both technically challenging and cost-prohibitive, limiting its use by many laboratories. Identification of a more potent M₅ probe that can be administered peripherally would likely result in a broader level of interest throughout the neuropharmacology community.

Figure 2

Prior Art: M₅ PAM probe molecules ML129 and ML172.

Using the M₅ PAM probes ML129 and ML172 (Figure 2) as a starting point, our first goal was to improve M₅ PAM potency by exploring both minor structural modifications and more aggressive changes to the isatin core, while maintaining the excellent muscarinic receptor subtype selectivity demonstrated with ML172. Unfortunately, SAR would continue to be very steep in this series as illustrated by the structures in Figure 3.

Figure 3

Modifications to the isatin core of ML129.

The most promising compound, SID 137294174, derived from nitromethane addition to the isatin carbonyl, failed to produce a potency which surpassed ML129. Other hydroxyl analogs prepared in a similar vein were generally inactive, with the sole exception being SID 137294171. Comparing SID 137294171 with SID 137294215 confirms the steep SAR within this series where changing a fluorine to a methoxy resulted in a total loss of M₅ PAM activity. Surprisingly, even the 3-oxetanyl analog of ML129, SID 137294182, was found to be inactive. This modification, although seemingly minor,¹² indicated the importance of a carbonyl or a tertiary hydroxyl at this location. Another approach we explored was through the incorporation of heterocyclic biaryls in place of the methoxy or phenoxy termini found in ML129 and ML127, respectively. This technique had proven successful in our M₅ PAM¹¹ and M₁ PAM13 programs; however this was not to be the case in this attempt as can be seen from Table 1. Among the more efficacious of these analogs prepared, as judged by their respectable ACh_max values, none provided an EC₅₀ value of less than 10 μM.

Table 1

Active heterocyclic biaryl analogs.

Constantly thwarted by unrelentingly steep SAR, the unpleasant possibility of terminating this project without success was becoming a reality. Fortunately, concurrent with this Extend Characterization Chemical Probe Development Pan (EC CPDP) there was an on-going effort being conducted at SRIMSC to re-screen the expanded MLSMR sample collection against M₁, M₄, M₅ and untransfected CHO cells (1X01 MH077606). During the course of triaging the single-point HTS data, a very weak M₅ activator (agonists and PAMs were not distinguishable at this stage) which contained an isatin core and a novel southeastern region was observed (CID 2145491, SID 24816245, Figure 4).

Figure 4

HTS single-point hit.

Drawing on previously developed SAR from our first M₅ PAM probe ML129¹⁰ we suspected that this 7-methyl isatin would not show a high level of muscarinic selectivity against M₁ and/or M₃ due to the absence of the critical 5-trifluoromethoxy substituent. This suspicion was confirmed, at least for the M₁ receptor, during the concentration response curve (CRC) phase of the HTS hit progression at Scripps, as can be seen from Figure 5.

Figure 5

CRC overlay for CID 2145491.

Within this CRC overlay is included the M₁ agonist trace (blue), the M₄ agonist trace (red), the M₅ agonist trace (green) and the untransfected CHO cell line agonist trace (black), showing that CID 2145491 has equipotent activity at M₁ and M₅, while it only possess base line activity at M₄ (the red trace mostly obscuring the black trace).

Even before we obtained the CRCs in Figure 5, the single-point HTS hit CID 2145491 was pursued at-risk by exploring the position of the methyl group on the isatin ring in conjunction with various linkers between the isatin nitrogen and the pendant phenyl ring. The percent activities for the examples appearing in Table 2 were generated as single-point activities (in triplicate) with the test compounds present at 30 μM to quickly survey a range of linkers and phenyl substitutions.

Table 2

Methyl isatin analogs.

When prepared in-house, the initial HTS hit (SID 137294186, CID 2145491) continued to show a very weak activity of just 33% of the ACh response. Moving down this table, while maintaining the 7-methyl, the SAR was uniformly flat with respect to substitution on the phenyl ring. The most notable exception was seen when the alkyl chain was extended an additional methylene (SID 137294183) to provide a PAM showing 86% ACh_max when tested at 30 μM. Realizing the importance of a substituent at the 5-position of the isatin with respect to M₅ PAM activity, we prepared a similar library of 5-methyl isatin analogs and were pleased to see a general improvement in single-point activities when compared to the 7-methyl analogs (Table 2). Once again, the most notable exception was seen for the n = 3 propyloxyphenyl analog; however, in the context of the 5-methyl isatin core, this modification was now extremely detrimental with respect to M₅ PAM activity, providing a wholly inactive compound (SID 137294189). The next logical step was to incorporate the 5-trifluoromethyoxy group present in our earlier probes, ML129 and ML172, to hopefully improve both M₅ potency and selectivity (Table 3).

Table 3

5-CF₃O isatin analogs.

Much to our surprise, the most interesting, new N-alkyl side chain present in SID 137294183 (Table 2), the propyloxyphenyl moiety, did not provide a highly potent compound in the context of the 5-trifluoromethoxy isatin (Table 3, SID 137294187). Instead, SID 137294187 displayed a hM₅ EC₅₀ = 2.0 μM and moderate efficacy of 46% ACh_max. This unexpected result demonstrated the importance of not assuming that SAR from one series would necessarily track within even a closely related series (commonly referred to as steep SAR), and the importance of preparing libraries of compounds to increase the probability of discovering productive structural modifications when optimizing allosteric ligands. Accordingly, we discovered that the ethoxylphenyl side chain possessed improved potency and provided, for the first time, M₅ PAMs with sub-micromolar potency and robust efficacy. The simplest analog, with an unsubstituted phenyl ring (SID 137294184, CID 57525860, aka ML326), provided a hM₅ EC₅₀ = 550 nM and 86% ACh_max. The phenyl group was shown to be critical for activity since replacing it with a methyl group (SID 137294196) decreased M₅ PAM activity by greater than ~20-fold. Substitution on the terminal phenyl group was simultaneously explored and the most promising examples are included in Table 3. While both the 4-methyl (SID 137294188) and 4-trifluormethyl (SID 137294201) analogs displayed nominally superior potency compared to ML326, their selectivity profiles relative to the other muscarinic receptors was not ideal and showed that the 5-trifluoromethoxy isatin was not a universal solution for providing highly selective M₅ PAMs. Selectivity for the M₅ receptor versus the M₁ – M₄ receptors was where ML326 (CID 57525860) displayed its superiority over all the other compounds appearing in Table 3. A full muscarinic selectivity profile for ML326 (a.k.a. VU0467903) was determined against both the human and rat M₁ – M₅ receptors demonstrating a very high preference for the M₅ subtype (Figure 6, a similar profile for the human subtypes was generated - data not shown). This selectivity experiment for ML326 against the rat receptors also provided potency (rM₅ EC₅₀ = 470 nM) and efficacy (ACh_max = 55%) values which were in good agreement with those determined for the human subtype.

Figure 6

CRC overlay of muscarinic rM₁ – rM₅ selectivity profile for ML326 (VU0467903).

Now focusing on ML326, we determined the extent to which this 3^rd generation probe was able to left-shift (fold shift) the ACh CRC at the rM₅ receptor when present at 30 μM. This experiment determined that ML326 (Figure 7) induced a slightly higher fold shift than ML129 (aka VU0238429) when these two values were determined to be 20-fold and 17-fold, respectively. This represented a marked increase over ML172 which only displays a roughly 5-fold shift in the ACh CRC.¹¹ However, ML172 and ML326 form an attractive pair of highly selective M₅ PAMs; one with a lower fold shift and the other with a higher fold shift profile. Having both types of probes is desirable because it has yet to be determined whether a high fold shift M₅ PAM or a low fold shift M₅ PAM is the optimal treatment for the various disease states in which M₅ PAMs may find utility.

Figure 7

ACh fold shift determinations for ML326 (VU0467903) and ML129 (VU0238429), at 30 μM, on the rM₅ receptor.

Very encouraged by the initial in vitro potency and selectivity profile for ML326 we prepared gram quantities and set about characterizing it through our in-house tier 1 DMPK assays (PPB, rat/human microsomal stability with predicted intrinsic clearance, P₄₅₀ inhibition, rat brain homogenate binding, etc.) and single dose PK/CNS exposure in rat. Many of these experiments were initiated in parallel, including the collection of rat plasma/brain samples; however we encountered insurmountable LCMS/MS analytical quantization issues due to poor ionization of ML326 using ESI, APPI, and APCI ionization probes, which prevented the determination of routine tier 1 DMPK parameters and in vivo rat exposure. Alternative methods including chemical derivatization and UV absorbance also failed to provide the requisite sensitivity for detection of ML326 thus preventing the completion of numerous studies. However, it is worth noting that the inability to quantitate in vivo concentrations of ML326, does not necessarily mean the compound is not present at possibly efficacious levels; that possibility remains to be tested in appropriate animal models. The inability to quantitate ML326 was a truly unexpected, late stage detractor from the generally positive profile for ML326. Future efforts could focus on improving the ionization properties for this class of compounds; potentially through the introduction of weakly basic amines at locations which do not negatively impact potency. Regardless, based on ML326’s superior potency and muscarinic subtype selectivity, this 3^rd generation probe will undoubtedly find utility in electrophysiology experiments, in vitro studies of selective M₅ activation and potentially even animal models where dose-response effects could still be observed, despite not being able to quantitate systemic exposure. In summary, the profile presented here supports ML326 as the current best-in-class M₅ PAM, made freely available to the research community through the MLPCN.

3.3. Profiling Assays

To more fully characterize ML326, and to better inform the scientific community about its potential off target activities, this 3^rd generation M₅ PAM was tested using Eurofins’ (formerly Ricerca’s, formerly MDS Pharma’s) Pan Labs Lead Profiling Screen. This battery of radioligand binding assays consists of 68 common GPCRs, ion channels and transporters where the test compound (ML326) was present at 10 μM. Responses were considered significant if > 50% inhibition was observed. However it should be pointed out that these are only single-point values and that functional selectivity may be significantly better than suggested by these “% inhibitions.” Table 5 presents the Pan Labs results for ML326, which showed a significant response in only ten assays. Also included in Table 5 are some of the previously determined Pan Labs results for ML172 at these ten assays, rather than the full 32 targets at which ML172 showed a significant response. In this respect, ML326 represents an improvement over the 2^nd generation M₅ PAM by virtue of showing only ten off-target responses, rather than 32 as displayed by ML172. Furthermore, six out of the ten % inhibition values for ML326 are even reduced relative to ML172.

Table 5

Pan Labs profiling of ML326 and selected results for ML172.

Additionally, a set of calculated physical properties were determined for ML326 and compared to the 1^st- and 2^nd generation M₅ PAM probes (Table 6). For comparison, this table also shows the averaged values for compounds appearing in the MDDR database (MDL Drug Data Report database, 2010) at two stages of clinical development (Phase I and Launched). ML326 compares favorably with the average values for both Phase I and Launched compounds, and displays a significant improvement over ML172 with respect to cLogP, and LogS (Log of calculated solubility).

Table 6

Calculated property comparison between the M₅ probes and MDDR compounds.

4. Discussion

5. References

1.

Yasuda RP, Ciesla W, Flores LR, Wall SJ, Li M, Satkus SA, Weisstein JS, Spagnola BV, Wolfe BB. Mol. Pharmacol. 1993;43:149–157. [PubMed: 8429821]

2.

Wess J, Eglen RM, Gautam D. Nat. Rev. Drug Discov. 2007;6:721–733. [PubMed: 17762886]

3.

Steidl S, Miller AD, Blaha CD. Yeomans. PLoS ONE. 2011;6(11):e27538. [PMC free article: PMC3216953] [PubMed: 22102904] [CrossRef]

4.

Vilaro MT, Palacios JM, Mengod G. Neurosci. Lett. 1990;114:154–159. [PubMed: 2395528]

5.

Weiner DM, Levey AI, Brann MR. Proc. Natl. Acad. Sci. USA. 1990;87:7050–7054. [PMC free article: PMC54680] [PubMed: 2402490]

6.

Anney RJL, Lotfi-Miri M, Olsson CA, Reid SC, Hemphill SA, Patton GC. BMC Genetics. 2007;8(46) [PMC free article: PMC1978498] [PubMed: 17608938] [CrossRef]

7.

Yamada M, Lamping KG, Duttaroy A, Zhang W, Cui Y, Bymaster FP, McKinzie DL, Felder CC, Deng C-X, Faraci FM, Wess J. Proc. Natl. Acad. Sci. USA. 2001;98:14096–14101. [PMC free article: PMC61174] [PubMed: 11707605]

8.

Thomsen M, Wörtwein G, Fink-Jessen A, Woldbye DPD, Wess J, Caine SB. Psychopharmacology. 2007;192:97–110. [PubMed: 17310388]

9.

Araya R, Noguchi T, Yuhki M, Kitamura N, Higuchi M, Saido TC, Seki K, Itohara S, Kawano M, Tanemura K, Takashima A, Yamada K, Kondoh Y, Kanno I, Wess J, Yamada M. Neurobiol. Dis. 2006;24:334–344. [PubMed: 16956767]

10.

Bridges TM, Marlo JE, Niswender CM, Jones CK, Jadhav SB, Gentry PR, Plumley HC, Weaver CD, Conn PJ, Lindsley CW. J. Med. Chem. 2009;52:3445–3448. [PMC free article: PMC3875304] [PubMed: 19438238]

11.

Bridges TM, Kennedy JP, Cho HP, Breininger ML, Gentry PR, Hopkins CR, Conn PJ, Lindsley CW. Bioorg. Med. Chem. Lett. 2010;20:558–562. [PMC free article: PMC3177601] [PubMed: 20004578]

12.

Wuitschik G, Carreira EM, Wagner B, Fscher H, Parrilla I, Schuler F, Rogers-Evans M, Muller K. J. Med. Chem. 2010;53:3227–3246. [PubMed: 20349959]

13.

Bridges TM, Kennedy JP, Noetzel MJ, Breininger ML, Gentry PR, Conn PJ, Lindsley CW. Bioorg. Med. Chem. Lett. 2010;20:1972–1975. [PMC free article: PMC2834874] [PubMed: 20156687]