By Cynthia Robbins-Roth

BioWorld Today Columnist

How do you develop drugs for diseases that you can't see until it's too late? Where the cause is buried in parts of the body that defy most experimental methods? Where biomarkers don't exist?

How do you begin to unravel complex cascades of interweaving molecular networks and come up with treatments that can provide sufficient clinical benefit at a reasonable cost?

The biotech industry began as a way to convert natural human proteins into therapies - turning rapidly growing cells into little factories churning out cytokines, growth factors and monoclonal antibodies.

But to date, only human growth hormone (HGH) and insulin have performed completely as expected. The anticancer properties of cytokines and monoclonal antibodies are still mostly untapped, diabetic foot ulcers still resist growth factor treatments and neurodegenerative diseases have not succumbed to the various neuronal factors.

Why don't these proteins do what they are supposed to do? How can we improve the odds for new treatments, especially those aimed at complex biology?

Neuropharmacologists believe they are hampered by excessive focus on the same old targets (mostly serotonin- and dopamine-related), compound libraries with strong bias to those targets, and downright silly animal models that are not predictive of human response.

The focus on limited targets also means adverse side effects can come out of left field. Recent reports suggested that the huge weight gain and increased risk of diabetes seen with the newer mood stabilizers may come from their indirect effect on AMP kinase, which monitors energy use and appetite regulation. Who knew?

A big breakthrough came in 1998 when Rusty Gage (Salk Institute) showed that adult brains are able to make new neurons. The breakthrough for patients with depression came in 2003 when Rene Hen and colleagues at Columbia University showed animal evidence that neurogenesis - generation of new neurons - was important in how drugs relieved depression. That supported the evidence that depressive episodes cause physical brain damage.

BrainCells Inc. (BCI), of San Diego, was founded in 2003 by Gage and Harry Hixson, and merged with NeuroGenix (founded by the Columbia team) to use those discoveries to find better drugs for psychiatric illnesses. Because the clinical symptoms of psychiatric illnesses are behavioral, finding meaningful objective measurements was a priority. A $17.7 million Series A round helped move things along.

The team set to work optimizing a novel set of biological screens based on neurogenesis. Neurogenesis is a multistep process: stem cell division; differentiation into neurons, astrocytes and glial cells; migration to the correct spot in the brain, making the right connections with existing structures; and surviving long enough to be integrated into the brain.

With any complex process, researchers tend to compartmentalize the different steps to allow controlled study. But the more you simplify the system to control variability, the further you get from the natural in vivo setting - making it harder to know what the results mean.

Carrolee Barlow, BCI chief scientific officer, said it's the sum total of those events that she wants to capture - not just proliferation of new cells. Studying neurogenesis in a cell-based assay instead of neurotransmitter modulation at a molecular level lets you see important biological activities - even if you don't know the molecular mechanism behind them.

The hard part was moving the assays into a human neural stem cell setting, using non-transformed primary cell cultures grown in a milieu that mimics what they would see inside a human brain. That means no fetal calf serum!

The team confirmed that neurogenesis occurred in vivo when the in vitro screens were positive. Running marketed antidepressants through the screens showed a common pattern of proliferation and differentiation into neurons, plus suppression of certain cell types. Interestingly, many neurological drugs showed a U-shaped dose curve, with higher doses adversely affecting neuronal growth.

BCI CEO Jim Schoeneck said that other companies' depression drugs that failed in Phase III showed no neurogenesis in BCIs' assays. He warned that while neurogenesis is necessary, it may not be sufficient. Positive results in the company's assays probably increase the odds of clinical success, but certainly don't guarantee it.

Leapfrogging NCE discovery

Schoeneck is well aware of today's requirement for getting products to the clinic in a shorter time frame, with cost constraints built into the company strategy. Repurposing is a key focus for BCI, leapfrogging over new molecule discovery for now.

The BCI screens have given the team hints of potential molecular targets beyond the classic dopamine/serotonin/norepinephrine bias. Barlow said her group used the Sigma catalog to find chemicals known to act on those targets, then went hunting for existing drug candidates that were structurally related to the positive neurogenic hits.

That approach led to a 2006 deal with Mitsubishi for a shelved Phase II non-psychiatric CNS drug that was safe but not effective. BCI plans to put the drug into new Phase II studies in a mood disorder this year.

On a parallel track, the company collects hundreds of drugs on the market or in late-stage development that are safe enough to use chronically. The compounds, regardless of current indication, get run through the BCI screens.

Organon thought enough of BCI's screens to cut a collaboration deal last year that will test shelved Organon compounds with good safety data for potential neurogenesis properties. The collaborators will work together to develop any resulting candidates.

Combination Therapy

Many psychiatric patients do not respond well to current treatments. Many believe that combination treatments are required - but how to find them?

BCI's screens can look for drug combinations with a true synergistic effect in a large matrix of complementary mechanisms - something not possible using in vivo screening. Some unexpected combos have shown a 2-log improvement in potency when used together. Animal models are used to confirm in vivo effect, and classic pharmacology follows to tweak the structures for optimal properties.

Those combo assays also can reveal unexpected drug-drug interactions that can kill a promising drug candidate. For example, certain folks stop responding to their psychiatric meds when taking allergic rhinitis drugs. Wouldn't it be nice to know about those interactions ahead of time?

BCI has identified a combination therapy that should move into the clinic later this year.

Nothing up my sleeve!

And all of this happens without really understanding everything that's happening in vivo. Schoeneck is keeping the team focused on using the screens primarily on compounds ready for clinical development. "The science may lead to new areas. But we can't be the ones leading for now."

Barlow isn't too worried. BCI's collaborators and colleagues are very involved in trying to understand molecular mechanisms behind those serious psychiatric diseases, and finding biomarkers to subdivide the patient populations. The company's screens and the information that emerges will support those efforts. Some day in the not-too-distant future, sales revenue will allow company scientists to dive back into solving this mystery.

Robbins-Roth, PhD, founding partner of BioVenture Consultants, can be reached at Her opinions do not necessarily reflect those of BioWorld Today.

BioWorld Today  February 26, 2007