Basic Biology: The Complex Core of Drug Discovery
The hardest task in major league sports is hitting a baseball. Those who can successfully do this even three out of every ten times over the course of a career are likely to find themselves enshrined in Cooperstown at the Hall of Fame. Unfortunately, scientists who strive to discover new medicines can only fantasize about a 30 percent success rate for their efforts. Drug discovery is a remarkably complicated process, and the success rate for advancing any potential medicine through the first three stages of clinical trials alone is less than 10 percent. Even then, few potential drugs ever get to the clinical trial stage due to toxicity, poor absorption or distribution into tissues, or a host of other problems.
Why is drug discovery and the development of new medicines so difficult?
There are some 23,000 genes within the human genome that carry the instructions for making a wide spectrum of proteins, many of which interact with each other in an interwoven mesh of complicated networks. Trying to understand how all of these molecules work together to develop a human being is a future dream, but understanding the biological role of even a single protein recalls Churchill’s famous quote about “a riddle wrapped in a mystery inside an enigma”. Let me share a simplified summary of the biology of a single gene I used to work on named fms (rhymes with rims) to illustrate the arduous challenge that medical researchers face.
Back in 1971, a young veterinarian in Philadelphia was trying to figure out why a cat under her care had developed a sarcoma (a type of cancer). This biological question initiated more than 40 years of research studies, yielding novel and important insights in not just cancer but in at least five other fields of biology. Susan McDonough, the veterinarian, was able to isolate a virus from the cat’s tumor that, when injected into another cat, resulted in the formation of a new tumor. Eventually the virus, now known as a feline sarcoma virus, was found to contain a gene, designated v-fms, that conferred on the virus the ability to cause tumors. What was it about this gene that enabled it to cause malignant cancers to form? Would there be the potential here for the development of an anti-cancer treatment for cats, and maybe humans as well?
Years later, my late colleague Joe Woolford showed that the v-fms gene in the virus was actually a modified version of a fms gene that is normally present in all cat cells (and many other vertebrate species as well, including humans). Somehow the virus had “captured” and integrated part of this gene, a process that turned the virus from a relatively benign state to one that could transform a normal cell into a cancerous one. The protein encoded by the normal fms gene is a member of a family of tyrosine kinase receptors, which are expressed on the surface of different types of cells. These proteins normally function by activating various intracellular pathways in response to binding a specific signaling molecule, which at the time was unknown for fms. The identification of the molecule that bound to and activated the fms receptor was finally determined in 1985. It turned out to be a previously identified blood cell growth factor known as CSF-1. This protein had previously been shown to stimulate the growth and development of blood cells known as monocytes and macrophages, which are important components of the immune system. The scientific finding raised an important new question: Would there be potential here to develop a drug that stimulated the production of these cells as a way to fight infections?
The story didn’t end here. As it turns out, if you remove the gene that encodes either the fms or CSF-1 proteins from developing mice using molecular biology techniques (the popular “knock-out” approach), the animals that develop have not only a deficiency in certain blood cells, they also suffer from a rare bone disorder known as osteopetrosis. This manifests itself as bones that are abnormally thick and heavy, which also makes them relatively brittle. Results of the knock-out experiment suggested that CSF-1 binding to fms played a role in normal bone development. Osteopetrosis is not just seen in these bioengineered mice; it afflicts humans as well. Again, this result raised a provocative question: Would there be potential here for the development of one or more drugs that could either block the development of osteopetrosis, or conversely, stimulate bone production in people with osteoporosis, the much more common bone-breakdown disorder ?
The fms knock-out mice also showed clear evidence of reproductive problems as well. Mammary glands in female mice failed to develop normally, and estrus cycling times were more infrequent and irregular than normal. Male mice showed decreased testosterone levels, lowered sperm counts, and they mated less frequently than normal mice. Would there be the potential here to develop a drug that affected contraception or fertility in either sex?
Now we had a clear understanding of a protein, CSF-1, that had biological effects on several distinct cell types that were mediated by its interactions with the fms receptor. Well, the story got more complicated when a discovery was made by colleagues of mine at Immunex in 1998. Melanie Spriggs and coworkers were studying a gene contained in the Epstein-Barr virus that had been given the wonderfully descriptive name of BARF1. Their experiments demonstrated that the protein made by this gene also bound to CSF-1, even though it was not related to fms. This observation suggested that there was an evolutionary advantage for this virus to carry a gene whose protein product could bind CSF-1 made by human cells. Would there be some potential here for the development of a treatment for Epstein-Barr virus, which causes infectious mononucleosis and can also lead to the development of two different types of cancer?
At last, the full function of the fms gene was known. It bound CSF-1 (as did BARF1) and was involved with the formation of blood cells and helped to regulate bone density and tissues in the reproductive system. That was the story until 2008, when yet another research group of people at FivePrime Therapeutics, Schering-Plough and DigitAB showed that a second growth factor, IL-34, was also capable of binding to and activating the normal fms gene. This raised the question as to why there would be two different proteins that seemed to function through the same receptor. Interestingly, CSF-1 and IL-34 do not appear to have identical biological activities. Perhaps CSF-1 and IL-34 are produced by different cell types at different times in development to regulate distinct processes? Would there be potential here for the development of IL-34 as a treatment for blood, bone, or reproductive disorders, or for treating Epstein-Barr virus or other infections?
Well, that was the story until several months ago. Our understanding of the biology changed again when an international team of researchers was trying to determine the cause of a particularly rare type of dementia (that affects people in their 40’s and 50’s) with the unwieldy name of hereditary diffuse leukeoencephalopathy with spheroids (HDLS). The molecular biology tool kit was once again brought out, and the results were entirely unexpected. It turns out that this disease results from genetic mutations in fms that eliminate its enzymatic activity, leading to the loss of a particular type of brain cell and dementia. Researchers wondered: Would there be potential here for the development of either a new drug, or the transplantation of a particular type of cell as a treatment for HDLS or other brain disorders?
Looking back, an inquisitive veterinarian taking care of a sick cat launched a lengthy journey of discovery whose end has likely still not been reached. This investigative process has stretched out over four decades, involved countless research groups, produced thousands of science papers, and led to biological insights in oncology, hematology, bone formation, reproductive biology, virology, and brain development. The biology of this single gene turned out to be much more complicated than many would have predicted after it was initially discovered. This level of complexity is not rare; the closely related kit receptor regulates the levels of a different set of blood cells, affects pigmentation and fertility, and also plays a role in tumor formation.
At any one of a number of different time points fms might have been viewed as a viable drug target for a variety of distinct disorders. A key question is whether the entire biological “story” really needs to be known in order to successfully develop a drug. The answer is clearly no, since you can never be sure that you have a complete understanding of the biology of any potential drug target or molecule. Having said that, any drug that seeks to either inhibit or stimulate this receptor will likely affect a number of different tissues. This would certainly complicate its potential development path. I’ve long been a strong proponent of using research collaborations to acquire additional data that will facilitate making the most informed decision possible regarding the potential development of a drug. More data is always helpful, but there will always be a tension between accumulating additional information and pulling the trigger for a drug development program. At some point, drugmakers have to accept the biological unknowns they confront, and commit themselves to the significant time, expense, and risk that comes with mid-and-late stage drug development.
For those of you who have wondered why coming up with new medicines is so difficult, the answer should now be apparent. Biology is amazingly complex, and just when you think you have it all figured out, something new pops up to muddy the waters or point you in a new direction. Physicist Yaneer Bar-Yam summarized it nicely when he said “To understand the behavior of a complex system we must understand not only the behavior of the parts but how they act together to form the behavior of the whole.”