Key roles in modulating cognition and behavior made the M1 muscarinic acetylcholine receptor (mAChR) a prime target for Alzheimer disease drug developers dating back nearly two decades. However, untold millions of big pharma dollars have yet to produce an M1 drug that works well and avoids adverse events from non-specific activation of other mAChR subtypes. To trim the side effect profile, renewed efforts have taken a different tack. Instead of focusing on the acetylcholine binding site, which is highly conserved across all mACh receptors, researchers have found ways to activate individual mAChR subtypes by targeting unique allosteric sites away from the substrate-binding action. Reporting today in the Journal of Neuroscience, researchers led by P. Jeffrey Conn of Vanderbilt University in Nashville, Tennessee, have applied this strategy to develop a highly selective M1 activator. In addition to its lack of agonist activity against any of the other mAChR subtypes, the compound had anti-amyloidogenic effects in a rat neuronal cell line and alleviated symptoms in rat models of psychosis. These data raise the possibility that selective M1 activators could someday reach the clinic as a treatment for schizophrenia and AD symptoms.
Among those compounds that met their demise in failed mAChR drug development programs, the one that made it furthest was an Eli Lilly drug called xanomeline. In a Phase 3 trial of 343 people with mild to moderate AD, this M1/M4-preferring mAChR agonist improved performance on two widely used cognitive scales (ADAS-Cog and CIBIC-plus) and reduced a number of behavioral disturbances, including hallucinations, delusions, and vocal outbursts (Bodick et al., 1997). On the downside, the high doses that brought these clinical benefits came with side effects (predominantly gastrointestinal) that caused more than half the participants to stop using the drug. Overall, though, Conn and others saw in the Phase 3 data a ray of hope that mAChR agonists might eventually be useful for relieving both cognitive and behavioral symptoms of AD and other disorders. “That paper is really what grabbed our attention and made us focus very heavily on this even though most other companies, including Lilly, had decided to drop it,” Conn told ARF.
The allosteric strategy worked well when Conn and colleagues were fishing out modulators of group II metabotropic glutamate receptors (Hemstapat et al., 2007), so he figured he should give it a try for muscarinic acetylcholine receptors, too. Since the dead Lilly drug (xanomeline) had agonist activity at both M1 and M4 receptors, Conn focused on these two subtypes in his efforts to develop more selective compounds. For M4, allosteric targeting has succeeded in his hands (see ARF related news story and Brady et al., 2008) and for Eli Lilly, which recently reported a similar M4 potentiator (Chan et al., 2008). The current paper extends this line of success to M1 receptors.
The newly characterized M1 agonist is TBPB (or, more long-windedly, 1-(1’-2-methylbenzyl)-1,4’-bipiperidin-4-yl)-1H-benzo[d]imidazol-2(3H)-one). Discovered five years ago while Conn was at Merck & Company Inc. in West Point, Pennsylvania, the compound was first presented at the 2006 annual meeting of the American College of Neuropsychopharmacology (Kinney, 2006) and characterized further after Conn moved to Vanderbilt.
First author Carrie Jones and colleagues established TBPB’s pharmacological properties by measuring agonist-induced intracellular calcium increases in cell lines expressing wild-type rat M1 or various rat or human M1 mutants. The researchers reasoned that TBPB functions as an allosteric agonist based on several lines of evidence: 1) it was able to activate an M1 mutant that is insensitive to acetylcholine or orthosteric agonists, and 2) its effects were blocked in a non-competitive manner by a competitive orthosteric antagonist (atropine). To address selectivity, they tested TBPB in cell lines expressing each of the five mAChR subtypes. For comparison, they threw in AF267B, reportedly an M1 agonist that recently revitalized the field with its ability to reduce Aβ and tau pathologies in the triple transgenic (3xTg) AD mouse model (see ARF related news story). In Conn’s new study, AF267B was not selective for M1—it also activated M3 and M5 receptors. On the other hand, TBPB was highly selective for M1.
However, despite its excellent selectivity for the M1 mAChR, the possibility that TBPB may also bind allosteric sites shared by other G protein-coupled receptors (GPCR) cannot be dismissed, wrote Abraham Fisher of the Israel Institute for Biological Research in Ness-Ziona, in an e-mail to ARF. AF267B and related M1 agonists were originally identified in Fisher’s lab. (See full comment below and recent reviews Fisher, 2008 and Fisher, 2008).
In a rat neuronal cell line (PC12) overexpressing human amyloid precursor protein (APP) and M1, TBPB shifted APP processing toward the non-amyloidogenic pathway. With help from coauthor Allan Levey of Emory University in Atlanta, Georgia, this showed up as a 58 percent increase in production of the α-secretase cleavage product and a 61 percent drop in Aβ40 in TBPB-treated versus vehicle-treated cells.
TBPB also appeared to have antipsychotic-like effects—demonstrated in rats at doses that did not elicit peripheral adverse effects commonly seen with other mAChR agonists. Moving into the cognitive realm, Conn said his group is now testing TBPB and newer selective M1 agonists in AD mouse models. He stressed that though AD is in large part a disease of cognition, the behavioral benefits of TBPB and related compounds should not be downplayed. In caring for his father, who recently died of AD, Conn said he and other family members felt that “the psychosis was in many ways a bigger challenge than the cognition.”
Levey calls TBPB “a novel compound that represents a new generation of highly specific drugs.” In an e-mail to ARF, he writes, “The study is important for the AD field because this drug will allow the role of the M1 muscarinic receptor in AD to be more clearly defined, including its potential for AD therapeutics.” (See full comment below.)
Acadia Pharmaceuticals Inc. has recently published animal data on its own selective M1 drug (Vanover et al., 2008), as has Merck, which presented preclinical work on another such compound (benyzl quinolone carboxylic acid, or BQCA) at this year’s Keystone conference on AD in Keystone, Colorado (see ARF related news story). Among TBPB and these two compounds, all target allosteric sites, but only TBPB and the Acadia drug are agonists—that is, they activate M1 in the absence of acetylcholine. BQCA is an allosteric potentiator, which means it does not directly activate M1 but potentiates the effect of acetylcholine. Because cholinergic neurons die in early AD, one concern with potentiators is that they might not be as effective as agonists in the context of reduced endogenous acetylcholine activity, Conn said.
Fisher’s compound, AF267B, is being tested by TorreyPines Therapeutics in a Phase 2 study (under the name NGX267) of dry mouth associated with Sjorgren’s syndrome to get data on the drug’s safety, tolerability, and M1 activity. The small San Diego biotech had been developing the compound as a possible AD treatment. However, the company has been running low on cash and last month reorganized its efforts to focus on developing a migraine drug. Acting CEO Evelyn Graham wrote in an e-mail to ARF that TorreyPines is “seeking a development partner for NGX267” and has “no current plans to initiate AD studies.” The end of TorreyPines’s AD genetics collaboration with Eisai Co. preceded its shutdown of discovery research (see press release). Graham also noted that “the shutdown of our discovery operations was a separate strategic decision from how we are addressing our development plan for NGX267.”—Esther Landhuis