Category → Making Molecules
It’s been a busy six months for new Hepatitis C (HCV) meds: first, Merck and Vertex have their drugs approved in May, and then Pharmasset leaks PSI-7977 clinical data. Now, Anadys Pharmaceuticals has just announced Phase IIb results for its clinical candidate setrobuvir (ANA-598). The pill lowered virus levels to undetectable limits in 78% of patients after 12 weeks of combination treatment with either ribavirin or pegylated interferon. Anadys notes only one major side effect, a rash occurring in 1/3 of the ‘598-treated patients. The therapy targets patients in tough-to-treat HCV genotype 1 (gen1), unlike PSI-7977, which targets gen2 and gen3.
The data seems to have convinced Roche, which acquired Anadys last Monday in all-cash deal analysts say represented a 260% premium over Anadys’s Friday stock closing price. Roche, no stranger to the HCV battle, hopes to integrate setrobuvir into a potential oral drug cocktail with its current suite of polymerase and protease inhibitors.
Setrobuvir interacts with N5SB polymerase at the allosteric “palm” binding site, located in the center of the baseball-mitt shaped enzyme. The drug’s sulfur-nitrogen heterocycle – a benzothiadiazine – is the key to virus inhibition; Anadys has installed the motif in all their HCV inhibitors, going back to their 2005 patents.
Chemists have known about the virus-targeting properties of this heterocycle for a while, but most derivatives have been culled in pre-clinical testing (see J. Antimicrob. Chemoth. 2004, 54, 14-16 for a brief review). Interestingly, chemists initially prepared benzodiathiazines, such as those in Merck’s chlorothiazide (c. 1957, according to the Merck Index), as diuretics, which found use in diabetic treatment. Over the next 40 years, modified medicines treated conditions ranging from epilepsy and cognitive therapy to hypertension and transcriptase regulation. Tweaked benzodiathiazines first showed anti-HIV and anti-CMV activity in the mid-1990s.
One final advantageous wrinkle in this structure: unlike PSI-7977, setrobuvir is not nucleoside-derived. This feature changes its binding behavior, pharmacokinetics, and even its intellectual property strategies, since many current antiviral therapies mimic the nucleosides found in RNA and DNA chains.
An announcement hinting at the possibility of an all-oral hepatitis C treatment had researchers abuzz last week. Pharmasset, a Princeton, NJ company specializing in antiviral discovery, alluded to upcoming conference data that suggested a combination of ribavirin (a generic antiviral) and Pharmasset’s experimental pill PSI-7977 lowered viral counts to near-undetectable levels in a ten-patient trial (kudos to Adam Feuerstein of The Street for initial reports. . . here at The Haystack, editor Lisa Jarvis has also tracked HCV drug development for some time now).
Hepatitis C virus (HCV) is a chronic liver virus with an estimated 180 million infected worldwide. Two relatively new extermination options are available: Merck’s Victrelis (boceprevir) and Vertex’s Incivek (telaprevir), approved by the FDA ten days apart last year. Unfortunately, though both drugs are administered orally, each requires co-administration of injected interferon, which can cause severe fatigue and flu-like symptoms. Both oral drugs inhibit the same enzyme: the NS3 protease, which drags down a patient’s immunity and helps the virus to produce new copies of its proteins.
In contrast, the ribavirin and PSI-7977 combination involves no injections, making it easier for patients to follow the appropriate medication schedule, and lessening side effects. The PSI compound also clips a different target: NS5B polymerase, an RNA enzyme that helps viral genetic replication. In addition, the PSI-7977 is “pan-genotypic,” meaning it inhibits several genetically different strains of HCV.
A 2010 article (J. Med. Chem. 2010, 53, 7202) details the full story of PSI-7977’s synthesis. Notice anything interesting? It’s really a nucleotide strapped on to a P-chiral prodrug, a “protected” substance the body later converts to the active drug species. This P-chiral motif is seen more often in asymmetric phosphine ligands (compounds that stick to metal catalysts during reactions to modify catalyst activity) than in drug development – often chemists install drug chirality at carbon or sulfur instead. The initial drug lead was actually a mixture of both phosphorus enantiomers (“Sp” and “Rp”), until process chemists realized they could selectively crystallize out the more potent “Sp” product.
In the meantime, Pharmasset scientists haven’t stopped pushing their HCV portfolio forward: a recent paper (J. Org. Chem., 2011, 76, 3782) details a new lead: PSI-352938, a cyclic phosphate prodrug attached to a purine-fluororibose nucleotide warhead. The team credits this new prodrug design with a 10-100-fold increase in potency over the “naked” adenine drug for NS5B RNA polymerase inhibition. PSI-352938 recently completed a multiple ascending dose Phase I trial, in which a daily 200 mg dose brought HCV titres down below the detection limit in 5 of 8 patients.
Looks like Afinitor (everolimus), a drug marketed by Novartis for various cancers, may soon have a new indication. Already approved for a variety of diseases – kidney cancer, pancreatic tumors, and organ rejection prevention – Afinitor shows new promise for breast cancer patients. Clinical data released Monday demonstrate marked improvement for hormone-resistant breast cancer patients when Afinitor, an mTOR inhibitor, is used in combination with the aromatase inhibitor Aromasin (exemestane). Patients receiving both drugs delayed disease progression an average of 7 months, versus 3 months for Aromasin alone.
Standard therapy for breast cancer includes treatment with estrogen receptor antagonists, such as Aromasin and tamoxifen, which bind in the estrogen receptor pocket of cancer cells, slowing proliferation (see the excellent NCI website for more information on breast cancer treatment). Aromasin itself has a very similar structure to estrone (a natural body hormone that binds to estrogen receptors) except that it irreversibly modifies the receptor pocket upon binding, making Aromasin a so-called “covalent” or “suicide” inhibitor (see Lila Guterman’s article from Sept. 5, 2011 issue of C&EN for more on drugs that bind for keeps).
Like Aromasin, Afinitor follows the trend of being structurally related to a natural binder of a key cancer target protein. mTOR (mammalian target of rapamycin), the protein target of Afinitor and related macrolides, was first discovered through binding studies using rapamycin, a polyketide natural product found in a soil bacterium from Easter Island (its Polynesian name is Rapa Nui, hence, rapamycin). Rapamycin also goes by the generic name sirolimus, of which so many analogues have been prepared that all go by the catch-all “limus drugs.” The attachment of a hydroxyethyl (CH2CH2OH) tail to rapamycin produces everolimus, which compared to sirolimus demonstrates better pharmacokinetic properties, including higher bioavailability (greater proportion of drug reaching target sites) and a shorter plasma half-life (meaning the drug doesn’t stick around as long, which can help curb toxicity or other side effects).
Note: Please see Sally Church’s post on Pharma Strategy Blog for more info on mTOR pathway biology and coverage of ECCO 2011 conference information regarding everolimus.
GlaxoSmithKline recently announced a contract with the Biomedical Advanced Research and Development Authority (BARDA), a US government preparedness organization (Note: it’s not often pharma-relevant press releases come from the Public Health Emergency website!). The award guarantees GSK $38.5 million over 2 years towards development of GSK2251052, a molecule co-developed with Anacor Pharma a few years back, as a counter-bioterrorism agent. The full funding amount may later increase to $94 million, pending BARDA’s future option.
The goal here is to develop “GSK ‘052”, as it’s nicknamed among med-chemists, into a new antibiotic against especially vicious and virulent Gram negative bacteria, such as the classic foes plague (Yersinia pestis) or anthrax (Bacillus anthracis).
So what’s so special about this molecule? Usually, med-chemists “color” with the same atomic “crayons”: some carbon, sulfur, nitrogen, oxygen, and hydrogen, with a few halogens or transition metals every now and then (luckily, the golden age of mercury and arsenic therapies has largely passed on!). But seeing boron ensconced in a lead molecule rings alarm bells . . . you don’t usually see boron in pharmaceutical scaffolds!
Look closely at GSK’052 (shown above): that’s a boron heterocycle there! Anacor, a company specializing in boron based lead compounds, first partnered with GSK in 2007 to develop novel benzoxaborole scaffolds. This isn’t the first company to try the boron approach to target proteins; Myogenics (which, after several acquisitions, became Millennium Pharma) first synthesized bortezomib, a boronic acid peptide, in 1995.
Stephen Benkovic (a former Anacor scientific board member) and coworkers at Penn State first discovered Anacor’s early boron lead molecules in 2001, with a screening assay. The molecules bust bacteria by inhibiting leucyl-tRNA synthetase, an enzyme that helps bacterial cells to correctly tag tRNA with the amino acid leucine. Compounds with cyclic boronic acids “stick” to one end of the tRNA, rendering the tRNA unable to cycle through the enzyme’s editing domain. As a result, mislabeled tRNAs pile up, eventually killing the bacterial cell.
Inhibition of synthetase function turns out to be a useful mechanism to conquer all sorts of diseases. Similar benzoxaborozoles to GSK ‘052 show activity against sleeping sickness (see Trypanosoma post by fellow Haystack contributor Aaron Rowe), malaria, and various fungi.
A heartfelt thank-you to Chemjobber and See Arr Oh for helpful discussions!
CENtral Science’s benevolent overlord, Rachel Pepling, has organized a blog carnival around the theme of “your favorite chemical reaction”. For the Haystack’s contribution, I thought it would be appropriate to write about a reaction medicinal chemists might find familiar. So I re-read See Arr Oh’s post about which types of reactions were really the most common in the med-chem toolkit. I decided on amide formation, which sits just about at the top of the list. I’m not sure it’s my favorite chemical reaction; I’ve got a special place in my heart for the Heck reaction (or Mizoroki-Heck reaction), though I’ve already blogged extensively about it. But every amide bond formation I ran in grad school worked. That’s justification enough for me!
Amides are the chemical ties that bind amino acids together to form peptides and proteins. Amides also turn up in a variety of other small molecules that nature makes. So it’s not surprising that amides are frequently found in drugs. Take a look at University of Arizona chemist Jón T. Njarðarson’s poster of top brand name drugs and marvel at the amide-y goodness.
Amide bond formation isn’t accomplished by a single, archetypical chemical reaction– far from it. I thought I’d provide a brief overview of some classic chemistry in this area and then move into a selection of modern-day additions to the amide-construction toolkit. Continue reading →
Trust your gut . . . scientifically speaking. From belly-button bacteria to classification of signature microflora (all the various microbes that populate the intestinal tract), it feels like recent popular “culture” grows best in a petri dish. Many scientists now classify humans as superorganisms, meaning our survival depends on a host of “good” internal bacteria that digest fiber, make vitamins, and help the immune system. But what happens when these good bacteria suddenly get wiped out by a non-selective antibiotic? This sets the stage for a Clostridium difficile intestinal conquest.
Simple contact transmits this bacterium between patients in hospitals, causing antibiotic-assisted diarrhea, bloating, and potential colitis. When a patient is treated with a broad-spectrum antibiotic, C. difficile survive by forming spores with tough outer coats, only to thrive again when there are few other bugs in the gut with which to compete.
Two new players have recently entered the fight against the difficult C. difficile: first, Optimer Pharmaceuticals’ new narrow-spectrum antibiotic for C. diff. treatment, Dificid (fidaxomicin), approved in May 2011. This antibiotic macrolide belongs to the tiacumicin class of natural products, members of which have been known since Abbott first isolated compounds from fermentation broths in 1987. Dificid specifically inhibits Clostridium RNA polymerase enzymes; without these enzymes, gene transcription halts, and the cells die.
Clearing the infection is great, but wouldn’t it be nice to ease the intestinal pain while the drug takes hold?
Researchers at UTMB-Galveston might have found a good target for drugs that could do just that. In the August advanced online publications at Nature Medicine, Tor C. Savidge at UTMB-Galveston reports on human metabolites that can inhibit C. difficile toxins TcdA and TcdB, the major agents behind painful antibiotic-assisted diarrhea. S-nitroso-glutathione, a nitroso (NO)-conjugated version of glutathione found in stool samples of infected patients, can “pass off” its NO group to the sulfur of a specific cystine amino acid residue in the toxins, shutting down their activity. The authors point out that instead of active site binding, the normal mode of action for most enzyme inhibitors, this NO seems to inhibit the toxins via an allosteric site, meaning they bind somewhere else on the toxin but still impair its function. Potency for in vitro inhibition is still in the high micromolar range (43-57 µm), but the study may point the way to the development of more selective NO-transfer drugs.
Last week, the FDA approved Zelboraf (vemurafenib), co-marketed by Roche and Daiichi Sankyo, for the treatment of melanoma characterized by genetic mutation BRAF V600E, which occurs in a subset of the overall patient population. Treatment of late-stage melanoma patients with Zelboraf increases their survival around five months longer than traditional chemotherapy. Cancer-stricken families believe this extra time justifies the $9400 / month price tag for the treatment, considering the dearth of treatments currently available for these near-terminal patients (for a more detailed look into the people who brought vemurafenib to market, read Amy Harmon’s New York Times article series from 2010).
Vemurafenib went from concept to approval in just six years, lightning-fast for pharma, which usually takes decades to bring a drug to market. So, what’s the secret behind its success?
Vemurafenib, developed initially by San Francisco pharma company Plexxikon (acquired in 2011 by Daiichi Sankyo) shows all the hallmarks of rational drug design. Initial screening of a 20,000-member compound library against the ATP-binding site of 3 kinases (Pim-1, CSK, and p38) yielded a 7-azaindole lead structure. This approach, known as fragment-based lead discovery (FBLD) – the concept that a drug can be built up from a tiny piece as opposed to a high-potency binder – may represent a first for the industry, as pointed out by Dan Erlanson of blog Practical Fragments. Further synthetic modification of this azaindole fragment, supported by computer binding studies, showed that a hydrophobic (nonpolar) pocket on the enzyme surface could best be filled by a difluoro-phenylsulfonamide group. Biochemical assays confirmed that a ketone linker (in place of the 3-aminophenyl group shown above) between the azaindole and the sulfonamide increased potency. Additionally, a 5-chloro residue on the azaindole eventually became a 4-chlorophenyl group; it’s unclear how this relatively non-polar group helps improve binding, since early active-site models suggest it faces out towards the watery cell cytoplasm.
How is Zelboraf halting melanoma growth? It all comes down to kinase inhibition, a topic covered with both a story and a Haystack post here at C&EN last year. B-RAF, a common gene overexpressed in melanoma cells, produces a protein kinase that is selectively inhibited by Zelboraf. Once shut off, this pathway reinstates a “lost” negative feedback loop for the BRAF V600E tumor cells, resulting in a cascade failure of growth factors further down the line. Cell growth arrest or apoptosis (cell death) follows, but only for the targeted melanoma cells, with no effect on non-cancerous cells.
In an interesting twist, a review published in July shows that inhibitors of Raf kinases (the family of kinases that includes the product of the B-RAF gene) can be developed for either the “activated” or “resting” forms of the enzyme. These two forms of the same enzymatic target show remarkably different clinical applications: Zelboraf targets the “activated” Raf kinase. Bayer’s Nexavar (sorafenib), a “resting”-form Raf inhibitor, was approved in 2005 for treatment of kidney and liver cancer, but shows little activity against BRAF V600E melanoma.
Update (4:30PM, 8/25/11) – Deleted “in silico” from screening description. Assays were run in vitro using AlphaScreen beads (PerkinElmer).
In this week’s issue of C&EN, I’ve written about the search for new anesthetic drugs, as well as the accompanying quest for a better understanding of how anesthetics work. Anesthesia is safer than it’s ever been because highly trained physicians and nurses can manage its complications. The drive to improve anesthetics is nowhere near as strong as it is for other drug classes such as oncology drugs, as Imperial College biophysicist Nick Franks told me. But that doesn’t mean the drugs in use are perfect.
Take propofol, or 2,6-diisopropylphenol, which is marketed as Diprivan by AstraZeneca. It’s arguably the most commonly used injectable anesthetic for surgeries in developed nations. It even has a nickname around the operating room, “milk of amnesia”, because of its effects on memory, and because of the milky appearance the sparingly water soluble compound takes on in the oil-water emulsion needed to deliver it to the bloodstream.
But propofol has side effects. Several firms have made adjustments to propofol or its formulation in order to address the limitations, and they’re finding out whether those chemical tweaks translate into benefits for patients. Continue reading →
Ironwood Pharmaceuticals and Forest Laboratories last week announced submission of an NDA for linaclotide, a peptide macrocycle for treatment of irritable bowel syndrome (IBS). This is the first new drug application for Ironwood, a 13-year old Cambridge, MA company, and it could validate other companies’ strategies for large-ring drugs (covered recently by Carmen Drahl in C&EN). There’s an enormous potential market for this drug; by Ironwood’s count, a combined 45 million people in the US suffer from IBS and related chronic constipation (CC), yet few drugs are approved for these conditions. So, how does linaclotide help IBS sufferers, um . . . go?
This 14-amino acid peptide ring, taken orally, arrives at the intestinal lumen, where, according to Ironwood patent literature, it docks with a receptor enzyme called guanylate cyclase C (GC-C). The extracellular domain (part that sticks out of the cell membrane), upon binding, initiates the intracellular domain (inside the cell) to begin production of guanosine-3’, 5’-cyclic monophosphate (cGMP), a signaling molecule that induces changes in the intestinal wall. In short, cGMP prompts the intestinal surface to release chloride and bicarbonate ions into the intestinal tract, which decreases sodium uptake and increases fluid secretion (Note: interestingly, this is similar to the body’s response upon E.coli infection; a bacterial toxin called ST-peptide causes traveller’s diarrhea). In Ironwood’s own words, these physiological changes “accelerate intestinal transit,” which helps to move solid waste and decrease overall pain by acting on local nerve responses.
Update (3:20PM, 8/17/11) – Changed “nearly 45 million people in the US alone suffer from IBS, yet few drugs are approved for this condition” to “combined 45 million people in the US suffer from IBS and related chronic constipation (CC), yet few drugs are approved for these conditions.”
As reported by Nature News and Forbes’ The Medicine Show on July 20, dapagliflozin, a BMS-developed diabetes drug marketed with partner AstraZeneca, was given a “thumbs-down” by an FDA review panel on July 19. After the 9-6 final vote, panel members commented favorably on the drug’s new mechanism, but evidently felt that the safety profile could not be overlooked: the FDA committee meeting statement mentions increased risk of breast and bladder cancer, increased genital infections, and perhaps most seriously, potential for drug-induced liver injury (DILI).
Dapagliflozin has been one of the rising stars of the new class of Sodium-Glucose cotransporter 2 (SGLT2) inhibitors for diabetes treatment, whose development roster includes Johnson & Johnson, Astellas, Boehringer Ingelheim, Roche, GSK, and Lexicon (Note: see Nat. Rev. Drug Disc. 2010, 551 for a full recap). The excitement behind these drugs comes from a relatively new idea for diabetes treatment: inhibition of the SGLT2 enzyme stops the kidney from reabsorbing sugar, leading to excretion of the excess glucose in the urine, which in turn lowers blood sugar. Dapagliflozin, like most SGLT2 inhibitors, is a glucose molecule with a large aromatic group attached to the carbon atom in the spot chemists call the anomeric position. Such so-called C-glycosides are thought to have improved staying power in the bloodstream relative to O-glycosides (where the linkage point is at an oxygen atom, a more common scenario in sugars), since they are less susceptible to enzymatic breakdown.