School of Pharmacy biotech symposium features case histories of future pharmaceuticals

The 2nd Annual Bay Area Biotechnology Symposium, presented by the UCSF School of Pharmacy’s Industry Outreach Program in coordination with the UCSF Postdoctoral Scholars Association at Mission Bay in late May 2011, fully lived up to its billing: “Pharmaceuticals of the Future: Case Histories and Challenges.”

The symposium featured slide presentations on major drug research programs by a half-dozen research executives from pharmaceutical companies actively engaged in the pharmacy school outreach program.

“The program goal is to increase the interactions between the faculty here and scientists in industry,” said School of Pharmacy Associate Dean of External Scientific Affairs Daniel Santi, MD, PhD, in his opening remarks.

The symposium’s speakers spotlighted new therapeutics aimed at some of today’s major medical challenges, from Alzheimer’s Disease and cancer, to hemophilia and obesity.

The speakers described years of painstaking iterative lab work, modifying molecules to improve potency, pharmacokinetics, and safety. They highlighted the complex challenges posed when taking potential drugs from lead discovery and optimization, through in vitro laboratory work and multiple pre-clinical animal species, and finally into early-stage testing in humans.

Alzheimer’s: holding the line on amyloid

Indeed, the day’s first presenter, Alexander Kamb, PhD, vice president of research in neuroscience at Amgen in South San Francisco and a former UCSF postdoctoral fellow, made a strong case for perseverance at rational drug discovery—and the most strategically effective use of research dollars—to meet the large and burgeoning (100 million patients in 40 years) unmet medical need represented by Alzheimer’s Disease.

As Kamb explained, Amgen’s focus has been on reducing amyloid plaques, which are insoluble extra-neuronal aggregates of A-beta peptides characteristic and hypothetically causative of the brain-wasting illness, by reducing A-beta production by as much as possible.

The logic of this approach, he asserted, is genetic as well as histological, since mutations that either cause early onset Alzheimer’s or greatly increase one’s odds of developing it (e.g., APOE4, Down’s Syndrome) affect secretase processing enzymes, yielding increased amounts of A-beta.

There are two such enzymes—beta- and gamma-secretase. Kamb argued that inhibiting the former, which was cloned by Amgen scientists back in 1998, was the most rational route to take because it can be readily expressed and crystallized for structural study. More notably, knockout rodent models, which inhibit the target enzyme by eliminating it completely, do not reveal potential toxic side effects.

But beta-secretase has proved to be “a very tough target,” said Kamb. Its active site is highly charged. Molecules designed to bind to and inhibit its activity must overcome the efflux mechanisms of the blood-brain barrier. And given the cognitively compromised patients, a drug must be stable enough for once-a-day dosing and very safe.

Moving from cell-based assays and chemical optimization to avoid off-target enzymes, through rats and into monkeys, Kamb reported that Amgen had been able to get a drug into the cerebrospinal fluid that bathes the brain and reduces A-beta production. Next step: Phase I human safety and pharmacokinetic safety testing—and perhaps, ultimately, a real test of the so-called amyloid hypothesis of Alzheimer’s causation.

“You need to be hopeful, because to be otherwise is not helpful,” Kamb concluded. And responding to audience questions, he noted: “We’re controlling everything we can, but humans are seven percent different at the nucleotide level. We’re not monkeys. So there are lots of unpredictable things and, of course that’s what Phase I clinical studies are for.”

Regulation of appetite and body weight by TrkB agonists

Indeed, Kamb was only one of several speakers to highlight the challenges posed and insights gained from pre-clinical animal model testing.

In a talk that drew its share of laughter for the sometimes surprising detours of preclinical investigation, John Lin, MD, PhD, executive director, experimental medicine at Rinat, Pfizer Inc., in South San Francisco, described efforts to develop an anti-obesity drug.

At first, the strategy seemed relatively straightforward. Based on studies of genetic mutations in both mice and human patients, it was found that lost function of TrkB receptors in the hypothalamic part of the brain—specifically in key neuronal clusters that respond to hormones like insulin and leptin to regulate hunger and satiety—led to overeating and obesity.

TrkB receptors are tyrosine kinase enzymes located in cell membranes that bind to growth factors called neurotrophins, which enhance neuron function.

Applying this rationale, Rinat researchers created a TrkB activator, a recombinant form of the potent neurotrophin 4 (i.e., NT4). “We thought if we supplied some TrkB agonist it would bypass some of the problems of leptin resistance associated with the obese patients,” said Lin.

The strategy worked flawlessly in mice. Injections of NT4 led to appetite suppression and reduced body weight. A much smaller dose of the drug injected directly into the rodents’ hypothalamuses had the same effect. A similar reduction in food intake was seen in rhesus monkeys when the drug was given directly into the brain.

But when the NT4 was given systemically to a group of baboons at the Southwest National Primate Research Center in San Antonio, the effect was, instead, exactly the opposite. Already obese, the baboons maxed out their daily biscuit allowances and managed to gain an additional 15 percent in body weight. When the drug administration stopped, so did the overeating.

Trying to make sense of this contradictory finding, which was repeated in cynomolgus monkeys, Lin and his colleagues tried a different, longer-lasting sort of TrkB agonist—a monoclonal antibody—that had the same effect. They also looked at whether dosing was too high, leading to a paradoxical down-regulation of the TrkB pathway—but lower dose frequency resulted in the same increases in body weight.

Lin concluded his talk with the experimentally-derived hypothesis that “there are two pharmacologically distinct compartments in the hypothalamus of non-human primates”: a more “peripherally accessible” compartment in which neurotrophins increase appetite and a more central TrkB system where they decrease it.

In such a scenario, the opposite pharmacological effects may reflect the limits of diffusion of the tested NT4 and agonist antibodies, which would reach a wider anatomical region of the hypothalamus in a rodent than in larger animals like baboons. And while this hypothesis may await more testing toward an anti-obesity treatment, Lin noted this “peripheral” TrkB system might yet be targeted to treat illnesses such as anorexia and wasting syndromes.

Discovery and development of PI3K inhibitors

Michael Varney, PhD, senior vice president for research and small molecule discovery at Genentech in South San Francisco, traced some of that company’s efforts to intervene in “cancer’s heartland,” inhibiting key enzymes in cell proliferation pathways involved in multiple forms of the disease.

Specifically, in the case of the compounds Varney described, Genentech researchers sought to inhibit the PI 3-kinase (i.e., PI3K)-AKT-mTOR pathway that becomes overactive, both due to overactive PI3K mutants and to mutations / deletions of the PTEN gene that normally acts as a brake on the pathway.

The small molecule drugs developed by Genentech aim to “hit [both] PI 3-kinase and mTor as a way of shutting down this entire arm of the arm of the [cell] growth signal,” said Varney.

The research executive then traced a “tour de force” of drug optimization, tweaking a PI3K-alpha inhibiting compound so that it would be a more potent inhibitor of various PI3K isoforms, as well as mTOR, while simultaneously solving a Rubik’s Cube of pharmacokinetics including lowering in vivo clearance, while increasing free fraction and solubility and maintaining maximal oral bioavailability (i.e. stability).

This under-the-molecular-hood manipulation, aided by an analysis of key structural and binding features of both a PI3K active site and the lead compound, was followed by iterative alterations to the latter.

For example, the addition of an aminopyrimidine at a key location on the original inhibitor added potency for both PI3K and mTOR while also effectively rendering the molecule more highly oxidized and thus increasing its stability relative to common metabolizing enzymes (i.e., CYP450s), leaving a higher percentage of the drug available to be active in vivo.

But pre-clinical work, moving from rats to dogs, showed a need for further optimization to reduce so-called unbound clearance (i.e., liver clearance of drug not bound to plasma proteins). Chemical changes to improve stability then raised solubility issues that, in turn, had to be addressed.

The optimized compound, dubbed GDC-980, not only has better pharmacokinetic properties than the starting molecule, but also increased potency via its multiple enzyme inhibition. This made it 150 times more dose-potent in in vitro testing against a prostate cancer cell line.

The experimental drug is now undergoing Phase I trials in humans where pharmacokinetics appear consistent with animal models. It also appears effective in a pharmacodynamics model; inhibition of PI3K activity as reflected in reduced downstream phosphorylation of AKT.

Different dosing regimens and combinations are being tried in up to a half-dozen different tumor types, said Varney. Phase II trials are anticipated to start this year.

Engineering of small molecule drugs to regulate biodistribution, reduce C_max and regulate CNS entry

Two other symposium presenters, both with research and development presences at UCSF’s Mission Bay campus, described attaching polyethylene glycol polymers (i.e., PEGs)—either to established therapeutic proteins or to small molecule drugs as a way to significantly enhance treatments’ bioavailability and even their biodistribution within the body.

Stephen Doberstein, PhD, senior vice president and chief scientific officer for Nektar Therapeutics, asserted that his company mitigates the “target biology and compound risk” that comes from trying to discover wholly new drugs by re-engineering existing therapies to make them safer, more efficacious, and even applicable to entirely new indications.

Indeed, Nektar manipulates small molecule drugs by focusing on the architecture of the polymers it joins to them and even the chemical linkers used for those conjugations, which are eventually broken down by esterase or non-enzymatic hydrolysis. As Doberstein put it: “We want to start with well-understood drugs and put them in the right place, at the right time, at the right dose in the body.”

Take the powerful cytotoxic oncology drug, Irinotecan, which is used to fight colo-rectal cancer but which has by severe side effects including diarrhea and neutropenia, which reduces vital infection-fighting white blood cells.

As a result of those side effects, the prodrug whose active form (SN38) has a half-life of about 50 hours can only be infused about once every three weeks: The drug benefit is thus diluted by the dosage limitations. Also, with no measurable SN38 for a week or two between doses, tumor cells have the opportunity to divide, mutate, and become more resistant.

To tackle this problem, a new polymer-drug conjugate dubbed NKTR-102 was designed with Irinotecan attached at the ends of four-armed PEGs. The polymer’s relatively large size and shape (i.e., long, flexible, able to adopt varied structural states, thus a large hydrodynamic radius), plus the ester-linkages to the prodrug, all promote a longer circulation for the effective drug, increasing half-life from 50 hours to 50 days. In addition, the peak concentration of the drug (i.e., C-max) was reduced, which Nektar researchers hypothesized should reduce side effects.

In addition, while NKTR-102, because of its size and shape is contained by normal vasculature—and thus circulates longer—it was hypothesized that it would also tend to accumulate and persist in leaky areas such as tumor sites. Infrared dye studies in mice support that “enhanced permeability and retention effect” as well as tumor suppression.

Indeed, NKTR-102, with its reduced side effects, altered distribution of cytotoxin—both over time and in the body—showed promise in a Phase II study of 70 patients with late-stage metastatic breast cancer. About 41 percent of the subjects had reduced tumor burden or stable disease after six months, with 30 percent of the subjects showing a significant reduction in tumor size and a handful of “complete responses” which meant, if not a cure, complete reduction of target lesions.

“This points us to a very high level of activity in patients who are really quite sick, “ said Doberstein, who expects a single agent Phase III trial of the drug to begin around the end of this year.

“One could try to apply this technology to almost every small molecule cytotoxic,” he added. “There’s a fairly large number of projects in preclinical development right now that follow this same general concept of [using] prodrugs to increase therapeutic index.”

Indeed, Nektar is not limiting its PEG biodistribution-shifting efforts to oncology medications. Another potential use has emerged from a serendipitous observation: Stable short-chain PEGs attached to small molecules can reduce or eliminate their ability to cross the blood-brain barrier.

Thus NKTR-118 was developed. This is a drug that seeks to address a major side effect limiting the use of opioids to treat chronic pain—the tendency of opioids to induce debilitating constipation by reducing gut motility.

NKTR-118 adds a short-chain polymer to naloxone, which is a generic opioid antagonist typically used in emergency room injections to reverse heroin overdoses. It creates a pill form of the drug that will not enter the brain or affect opioids’ analgesia there, but will instead act peripherally to counteract and treat their negative effect on bowel function.

Doberstein presented Phase II data showing NKTR-118 as both bioavailable and effective even as central nervous system pain relief is maintained. The drug is now in a large worldwide phase III trial in partnership with AstraZeneca.

In addition, “rate of entry across the blood-brain barrier appears to be tunable,” said Doberstein. “The architecture and size of the polymer conjugate that you attach to the small molecule can actually change the rate.”

This has significant applications, he said, because opioids’ swift rate of entry into the brain is associated euphoria and addiction on the one hand, and sedation on the other; whereas actual pain relief is more due to a steady-state concentration in the brain.

The company has engineered NKTR-181, which is just starting Phase I clinical trials, to test an abuse/addiction/sedation-avoiding opioid analgesic. He concluded by noting the potential for applying the slowed rate-of-brain-entry architecture to reducing sedation effects from antihistamines as well as wooziness from tricyclic anti-depressants used for neuropathic pain applications.

Rational design of an active, long-acting Factor VIII for Hemophilia A treatment

Another prime application PEGylation was described by John Murphy, PhD, director of molecular biology and protein expression at Bayer Healthcare and another former UCSF postdoctoral fellow. Bayer researchers at Mission Bay and in Berkeley are applying the technology to increase the circulating half-life of Factor VIII, which is the blood clotting factor deficient in the most common form of hemophilia.

Where it can be afforded, a recombinant or plasma-derived form of the Factor VIII protein is given prophylactically. This is especially relevant for patients in whom the factor otherwise falls below one percent of normal and who are thus at risk for severe bleeds from injury as well as joint bleeds, which result in crippling arthropathies, oozing soft tissue bleeds, and a much higher risk of intracranial hemorrhage.

But because recombinant Factor VIII has a 13-hour half-life, patients must be infused three times a week. And since hemophilia is a life-long disease, that means subjecting young children to frequent intravenous injections and implanted ports that increase infection risk and other complications.

To get adequate protection for a full week with the current supplemental protein would require more than 70 times normal circulating levels, an expensive and potentially thrombotic proposition. Alternatively, Murphy explained: “Doubling the half-life has about the same effect as increasing the dose 70-fold.”

Factor VIII is a large 2,332-amino acid protein that undergoes extensive post-translational modification: It is activated by, and then contributes to, the amplification of thrombin in the blood coagulation pathway. Its relatively rapid clearance—even as it mostly circulates with its chaperone Von Willebrand factor—is due to active clearance by hepatocytes and Kupffer cells mediated in part by LDL-related LRPs (i.e., lipoprotein receptor-related proteins).

Thus Bayer researchers seeking to extend Factor VIII’s half-life had to thread a therapeutic needle of “disrupting the clearance while still maintaining the activity,” Murphy explained. This was complicated by “overlap of regions involving clearance and function.”

But the company’s protein chemists noted that there were no free cysteines on the molecule’s surface—they were either in disulfide bonds or buried within it. So they engineered the amino acid onto Factor VIII surfaces in order to do site-specific conjugation of polymers that would block clearance while maintaining activity.

Five of these engineered Factor VIII variants were selected, with polymers attached near known binding sites for clearance receptors. They were then further screened for both activity and for PEGylation efficiency (i.e., relative commercial viability).

Eventually, effective activity against bleeding was confirmed in vivo in both mice and dogs. The circulating half-life extension, which included looking at different sizes and numbers of polymers, was also tested versus wild-type Factor VIII. It was increased two-fold in various species.

Ultimately, studies in knockout mice compared the activity of PEGylated Factor VIII dosed 48 hours pre-injury versus the unmodified protein given one day before. Dosed in this manner, the compounds offered the same protection against bleeding, indicating a two-fold prolongation of protection.

The new extended-life Factor VIII has gone into clinical trials this year.