CHEMISTS FROM pharmaceutical companies came together at the American Chemical Society's national meeting in Salt Lake City last month to unveil a new crop of potential medicines and share the insights they harvested along the drug discovery pathway.
Six speakers participated in the Division of Medicinal Chemistry's "First-Time Disclosure of Clinical Candidates" symposium, which was supported by Sanofi-Aventis and organized by Albert J. Robichaud, senior director of chemical sciences at Wyeth Research, in Princeton, N.J. Each presenter revealed to the public the structure of a small molecule that is currently progressing through clinical trials. The talks described potential treatments for Alzheimer's disease, migraine, type 2 diabetes, endometriosis, attention deficit hyperactivity disorder, and sarcopenia.
Each presentation was a story about transforming a hit compound into a drug, and relatively few chemists working in drug discovery end up being a visible part of those stories, even after decades of work, Robichaud told C&EN. "Each of the presenters has been fortunate enough, through the efforts of hundreds of their colleagues, to develop a chemical entity that has a good chance of helping people," he said.
John E. Macor, executive director of neuroscience discovery chemistry at Bristol-Myers Squibb, in Wallingford, Conn., described BMS-708163, a potential treatment for Alzheimer's disease.
Alzheimer's is a fatal disease that destroys brain cells, ravaging memory, thinking, and behavior. It's thought to affect up to 5.3 million Americans, according to the Alzheimer's Association. There is no cure, but physicians can manage the disease with acetylcholinesterase inhibitors or drugs that block N-methyl-D-aspartic acid receptors. "All of those drugs fight the symptoms but not the underlying causes of the disease," Macor told C&EN. By getting at the cause of Alzheimer's, BMS hopes it can develop a drug that stops the disease's progression.
One of the hallmarks of Alzheimer's is brain lesions composed mostly of a peptide called amyloid-beta (Aβ). It's thought that elevated levels of Aβ are neurotoxic and cause Alzheimer's, Macor said. That idea, known as the β-amyloid hypothesis, hasn't been proven, he added. Aβ comes from a larger precursor protein, which is cleaved by two enzymes, β- and γ-secretase. Blocking either of those enzymes is an attractive approach to treating Alzheimer's, Macor said.
BMS-708163 inhibits γ-secretase, which is an enzyme complex that processes many proteins in the body. That promiscuity was a key concern for the BMS team, especially since one of γ-secretase's other substrates is Notch, an important cell-fate regulator. To avoid side effects in tissues where Notch controls cell fate, such as the digestive tract, finding γ-secretase inhibitors that block Aβ cleavage while permitting Notch cleavage was critical.
BMS chemists began their task inspired by some structural features of BMS-299897, a γ-secretase inhibitor that was an earlier clinical candidate for treating Alzheimer's. There were several challenges along the way, Macor said. "In addition to getting Notch selectivity, we had all the problems you usually have with a neuroscience program, including getting the compounds across the blood-brain barrier," he said. Eventually, medicinal chemistry efforts yielded BMS-708163, a potent γ-secretase inhibitor that's highly selective for cleavage of Aβ over that of Notch.
In animal tests, BMS-708163 significantly reduced Aβ levels in the brain, and in Phase I clinical trials, healthy volunteers taking the drug had a significant reduction in Aβ levels, as measured in their cerebrospinal fluid. BMS-708163 is administered orally and is now in Phase II clinical trials, which will assess the drug's safety and tolerability in patients. "We think that BMS-708163 is a molecule that has a very good chance of testing the β-amyloid hypothesis," Macor said.
CHRONICLING EFFORTS to treat a different neurological disorder, Ian M. Bell, senior investigator in the department of medicinal chemistry at Merck Research Laboratories, in West Point, Pa., described MK-3207, a potential treatment for migraine. About 13% of the U.S. population suffers from migraines, which cause moderate to severe pain, often accompanied by nausea and oversensitivity to light and sound.
Tryptamine-based agents that activate specific 5-hydroxytryptamine receptors, also known as serotonin receptors, are the established therapies for migraine attacks, Bell said. But the drugs constrict blood vessels and are not recommended for patients with cardiovascular disease, he added.
Bell and his coworkers aimed to find a drug that blocks the ability of calcitonin gene-related peptide (CGRP) to activate its receptor. CGRP is a 37-amino-acid neuropeptide that has been implicated in migraine. Agents that block activation of the CGRP receptor might block pain transmission and nerve-related inflammation, as well as normalize dilated blood vessels, thus relieving migraine pain and associated symptoms without causing adverse cardiovascular effects.
The most potent blockers from screens and early medicinal chemistry efforts tended to be poorly soluble in aqueous solutions, which limited their ability to be taken orally. By focusing on incorporating polar functionality while retaining high affinity for the CGRP receptor, the team eventually arrived at MK-3207, Bell said.
The compound was highly potent in preclinical experiments, and initial clinical studies in healthy volunteers began in 2007. MK-3207 is administered orally and is now in Phase IIb clinical trials for the acute treatment of migraine. On April 21, Merck announced that it would delay seeking Food & Drug Administration approval for another CGRP receptor antagonist, telcagepant. Some study patients taking the drug daily to prevent migraines developed high levels of liver transaminase enzymes, a possible marker of acute liver damage.
THE THIRD neurological disorder discussed at the symposium, attention deficit hyperactivity disorder (ADHD), is often diagnosed in children. That led to an extra-high bar for safety in Pfizer's drug-development efforts, said Travis T. Wager, a senior principal scientist at Pfizer's Groton, Conn., site. He described the company's drug candidate for ADHD, PF-3654746.
ADHD affects 3–5% of children in the U.S., according to the National Institute of Mental Health. Kids with ADHD tend to be hyperactive and have trouble concentrating. If left untreated, ADHD can have long-term effects into adulthood. The drugs most prescribed for ADHD are stimulants, which act in a matter of hours to help patients focus but have abuse potential. A nonstimulant drug is also available, but it can take days to weeks to work and has been associated with rare cases of liver toxicity. Pfizer set out to develop a drug that has the fast-acting traits of stimulants and the best safety profile possible, Wager told C&EN.
Pfizer's agent selectively blocks the histamine H3 receptor. Histamine receptors exist in several distinct types, and the H3 receptor is believed to play a role in neurotransmission. Blocking H3 receptors in animals boosts release of histamine and other neurotransmitters, thereby improving cognitive function. Blocking H3 receptors in humans is therefore an appealing strategy for treating ADHD, Wager said.
With a high-throughput screen for H3 receptor binding, Wager and his colleagues identified biaryl diamines as lead compounds. Biaryl compounds can intercalate into DNA and lead to gene loss or mutated gene function. To avoid that possibility, the team replaced one of the aryl rings with a cyclobutyl ring. "Due to their three-dimensional shape, a cyclobutyl-containing molecule can't intercalate as easily, but the H3 receptor still recognizes it as an antagonist," Wager said. In vitro, those cyclobutyl agents led to phospholipidosis, a storage disorder in which excess phospholipids accumulate within the lysosomes of cells. Further adjustments, including making their candidates less basic and less lipophilic, helped the Pfizer team eliminate that problem, Wager said.
In animal tests, PF-3654746 increased neurotransmitter levels in the brain. On the basis of favorable safety outcomes in rats and dogs, the team chose to move it into clinical trials, Wager said. In Phase I trials on healthy adult volunteers, orally administered PF-3654746 was safe and well-tolerated, and it reached H3 receptors in the brain effectively even at low doses of less than 3 mg. It is currently in Phase II clinical trials for the treatment of ADHD.
Across the ocean, another Pfizer team developed PF-2413873, which is a potential new treatment for endometriosis. The disease occurs when endometrial tissue, which forms the inside lining of the uterus, grows in places it is not supposed to, usually on the surfaces of organs in the pelvic and abdominal areas. The disease can lead to pelvic pain and infertility, and it affects more than 5.5 million women in North America alone, according to the National Institutes of Health.
Treatment options for endometriosis include surgery to remove the excess tissue. Estrogen promotes the growth of endometrial tissue, so reducing estrogen levels with pharmaceuticals is another effective option. Injectable drugs are available that completely shut down female hormones such as estrogen, inducing a menopauselike state, said Karl R. Gibson, an associate research fellow at Pfizer's Sandwich, England, site. However, the injectable drugs can lead to a loss in bone mineral density and therefore are restricted to six to 12 months of use, after which time endometriosis could recur, he said.
Clinical trials have shown that an orally administered progesterone receptor blocker can also stop the growth of endometrial tissue caused by estrogen, thereby reducing symptoms of endometriosis, Gibson said. However, those agents are steroids and can have other targets besides the progesterone receptor, which could lead to side effects, Gibson added.
Gibson and his teammates aimed to block the progesterone receptor with a nonsteroidal compound. A high-throughput screen for progesterone receptor binding resulted in a lead compound, but the compound contained an amide group that easily hydrolyzed in animals, resulting in a short half-life. "We found we could replace the amide group with a sulfone, which doesn't hydrolyze but retains potency," Gibson told C&EN in a phone conversation. The team then fine-tuned the lipophilicity of their compound series to arrive at PF-2413873.
In animal tests, PF-2413873 blocked the estrogen-driven growth of endometrial tissue. So far, tests in healthy male volunteers have shown that orally administered PF-2413873 has an excellent half-life. Even though this drug is intended for female patients, drug candidates are typically safety tested in males first as a general practice, Gibson explained. It is currently in Phase I clinical trials for treatment of endometriosis. "By virtue of its distinct mechanism of action, we believe that PF-2413873 is unlikely to affect bone mineral density," Gibson told C&EN.
ANOTHER nonsteroidal compound holds promise, this time for another therapeutic target: diabetes. A team at Amgen, in collaboration with Biovitrum AB, a Sweden-based pharmaceutical company, has developed AMG 221 as a potential therapy for type 2 diabetes. Diabetes affects 23.6 million people in the U.S., or 8% of the population, with type 2 being the most common subtype, according to the American Diabetes Association. At the meeting, Christopher H. Fotsch, a research director in medicinal chemistry at Amgen, in Thousand Oaks, Calif., disclosed AMG 221's structure.
The Amgen team's diabetes target is an enzyme called 11β-hydroxysteroid dehydrogenase type 1 (11β-HSD1). This enzyme reduces cortisone to cortisol. Both molecules are steroid hormones, but unlike cortisone, cortisol can act on the glucocorticoid receptor (GR), which is found in almost every cell in the body and regulates genes, controlling development, metabolism, and immune response. Cortisol binding to the GR stimulates glucose biosynthesis in the liver and fat-cell development.
Tests in mice have indicated that 11β-HSD1's cortisol-making activity plays a role in type 2 diabetes, as has clinical evidence from humans. For instance, patients with Cushing's syndrome, who have excess cortisol, also can have high blood sugar and abdominal obesity. Furthermore, glucose biosynthesis in the liver is abnormally high in type 2 diabetics. These findings suggest that selectively blocking 11β-HSD1, and therefore the production of cortisol, might benefit type 2 diabetics, Fotsch said.
Biovitrum's high-throughput screen for 11β-HSD1 inhibitors turned up an aryl thiazolone hit with moderate activity. Medicinal chemistry efforts led to potent analogs that blocked 11β-HSD1 in mice. The team obtained an X-ray crystal structure of one of the analogs with 11β-HSD1 (Bioorg. Med. Chem. Lett. 2007, 17, 6056).
Unfortunately, some aryl thiazolone prospects had a tendency to stick to plasma protein from human serum, which could reduce potency and absorption in humans. To remedy those problems, the team changed the aryl ring to a bulkier lipophilic norbornyl ring, and they alkylated the thiazolone at carbon 5 to prevent epimerization of an important chiral center, Fotsch said. These changes gave AMG 221, Amgen's clinical candidate.
In mice that were pushed into obesity by diet, orally administered AMG 221 blocked 11β-HSD1 activity in liver and fatty tissue and reduced glucose levels. In obese human volunteers, AMG 221 inhibits 11β-HSD1 activity in fatty tissue after a single oral dose, Fotsch said. AMG 221 is in Phase I clinical trials for type 2 diabetes, according to Amgen's website.
BACK AT MERCK, a team is tackling a different steroid-related challenge, one that could lead to a drug for sarcopenia. This ailment, which affects the elderly, causes progressive skeletal muscle loss, leading to impaired mobility and loss of independence. Robert S. Meissner, director of medicinal chemistry at Merck's West Point, Pa., site, disclosed the structure of MK-0773, which is being developed to build muscle mass to limit the effects of sarcopenia.
Up to 25% of people over the age of 60 are estimated to have sarcopenia, which means the disease could affect as many as 200 million people worldwide. However, sarcopenia is not well-established or recognized as a disease, and there are no approved pharmaceutical treatments. Androgens such as testosterone promote muscle growth and are a possibility, he said, but they have serious side effects, such as excessive body hair growth in women and prostate growth in men, limiting their usefulness as drugs.
Meissner and coworkers sought an agent that would avoid the detrimental effects of androgens while maintaining the beneficial growing and strengthening effects on muscle and bone. The team began the search with an azasteroid androgen ligand closely related in structure to finasteride, an FDA-approved drug developed by Merck for treating prostate enlargement and male-pattern baldness.
The chemistry team incorporated basic nitrogen-containing heterocycles to arrive at a set of molecules that offered high potency, water solubility, and oral availability. A subset of those prospects had the desired selectivity for the androgen receptor, which works in concert with other proteins to regulate a network of genes involved in metabolism and development. Further modifications to enhance druglike properties led to MK-0773, which is a partial agonist, or activator, of the androgen receptor.
In animals, MK-0773 promoted robust bone and muscle growth with minimal effects on the skin and uterus in females and the prostate in males. At the molecular level, the team showed that it regulates genes to a different degree compared with testosterone.
Clinical trials to date indicate that orally administered MK-0773 is safe and tolerated well. Furthermore, in postmenopausal women, it promotes muscle growth without causing the skin effects observed with testosterone treatment. MK-0773 is currently in a Phase II clinical trial for sarcopenia in women.
Asked to comment on the significance of the annual first disclosures session, Pfizer's Wager noted that it's a great place for learning how drug companies solve problems. "I like to take what I've learned at the symposium and say, 'How can I develop a strategy so that I don't have to face the problems the presenters faced down the road?' " he said.
Session organizer Robichaud agreed that the symposium is a top-notch learning experience. "You learn things in that session that they don't put in books. It can take 10, 20, or 30 years of work in an area to gain that kind of knowledge."