Wednesday, May 13, 2015

Linking genes to cell function

In the 1990's, as scientists were conducting the human genome project, there was a belief that knowing the genome would explain all the mysteries of the body.  Once we could decode our genome, we could customize medicine for our unique makeup.

This turned out to be a little optimistic.

Genes contain the codes for proteins, and that's all they do.  Just knowing the code for a protein doesn't tell you much about that protein, though.  Sometimes we are lucky, and we find that a disease is linked to a protein whose function we already knew.  Most of the time it doesn't work like that.

Most of the time, a geneticist will study a group of people who have the same traits and look for common elements in their genes.  The story of alpha synuclein, the protein that many scientists now believe is closely tied to the pathogenesis (i.e., cause) of Parkinson's, offers a nice illustration of how this happens.

As I wrote earlier, alpha synuclein was discovered in 1994.  In 1995, researchers figured out what gene coded for alpha synclein and then, in 1996, a collaboration of NIH researchers with colleagues from the Robert Wood Johnson medical school and Seconda Universita degli Studi di Napoli showed that, in a large Italian family, they could link the family members who developed Parkinson's with a specific genetic location.  Soon after, again under the lead of Mihael Polymerapoulos at NIH, the team showed that this was the gene that coded for alpha synuclein.

Even when it was discovered, Polymerapoulos and his colleagues knew that the mutation they had discovered could not explain very much of the Parkinson's disease that neurologists were treating in their clinics - few people would have that specific mutation.  However, it didn't have to.  Knowing that the gene for alpha synuclein was important meant that it was likely that the protein alpha synuclein was important - but how?  In short order the next breakthrough came from the lab of Cambridge's Michel Goedert and his team: Lewy bodies, known to be present in the brains of people with Parkinson's since 1912, were composed of alpha synuclein.

Since then, further research has shed more light on the puzzle.  A group of leading neurologists figured out that Parkinson's seemed to be transmitted into transplanted cells.  This fueled the development of the prion hypothesis, that Parkinson's disease might be a prion disease.  Northwestern's Dimitri Krainc, when he was at Harvard, tracing the clues from another gene associated with Parkinson's, explained a part of this mechanism.  Others have provided more important clues.  

The importance of alpha synuclein to current research is well documented.  What's important about this story is that the genetic research didn't explain a genetic cause for the 15% of the disease that seems to be hereditary.  Instead, a mystery that started with one Italian family with a rare genetic defect started us on the path to understand all of Parkinson's disease.

The idea that knowing your genes will explain your health hasn't panned out.  Finding a genetic defect that causes Parkinson's isn't important just so that we can make genetic tests.  Understanding genetic defects that are linked to disease can identify mechanisms that are important in those diseases. We originally thought that genetic tests would tell us how to personalize medicine for you.  Instead we are finding that by understanding what genes are linked to aspects of each individuals health, we can identify ways to help everyone.

Sunday, May 3, 2015

On the challenge of "cure"

We talk about "curing" Parkinson's, but what does "cure" mean?

If you talk about curing smallpox, I know what you mean.  You mean that a person might have the smallpox virus in his or her body, and you want me to make a drug that removes it.  I can visualize that: we need a chemical that will bind to the smallpox virus and activate the body's own systems for getting rid of stuff.  If you have cancer and talk about curing cancer, I know what that means: it means that you have some cells that carry mutated DNA that makes them cancerous, and you want me to get rid of those cells.

When we talk about a cure for Parkinson's, I don't think we all mean the same thing.

When you have Parkinson's, neurons in your brain get sick.  They stop working the way that they should.  The connections they make with other cells atrophy and disappear.  Eventually, the neurons die.  To be diagnosed with Parkinson's disease, you very likely have many neurons dead and gone.  (We all lose neurons as we age, just with Parkinson's some specific ones die faster.)  Other neurons get sick and atrophy, losing the connections that are central to brain function.

If a cure for Parkinson's means restoring the brain (motor? cognitive?) function to a level of functioning that is typical of people of the same age, sex, education, and activity level, that isn't a cure like a cure for smallpox or a cure for cancer.  It isn't just removing the other -- for example, a virus or defective cells.  It is removing an other -- misfolded forms of a protein -- and also replacing lost function.

Thus, mechanistically there is no way that one therapy would achieve both results -- stopping the pathology and replacing lost function.

We are pretty close to stopping the pathology.  There are trials right now (May 2015; EDITED 10/19/2015 -- the trial lined under the word "trial" has had very positive results published) hoping to stop or slow Parkinson's pathology.  Each of these is based on a good understanding of basic biology and/or population studies.  Each of these studies promises to at least shed new light on Parkinson's disease, even if they don't directly lead to a therapy.  It is worth noting that even a failed therapy provided the new insight to put us on track to today's most promising therapeutic targets.

In terms of replacing lost function, dopamine replacement therapy is the gold standard and represented a revolution in Parkinson's treatment that won its developers the Nobel Prize.  We can do pretty well at replacing dopamine.  We do much worse at restoring the function lost as other systems are attacked by Parkinson's.  Cell-based therapies, despite the hype and enthusiasm, don't work in controlled trials.  Outliers seem to benefit, but not better than outliers in studies of medical therapy.

The future of Parkinson's is most likely this: we will develop therapies that stop the progression, with some recovery of atrophied neurons (leading to some recovery-related dyskinesia and possibly hallucinations).  Lost neurons will not recover.  Once we can stop the pathology of Parkinson's disease, we will treat people at their earliest signs and we will be done with Parkinson's.

As far as restoring lost function, we do really well with medication today. As far as cellular solutions, if we could replace the lost cells in Parkinson's disease, we could replace the lost cells with natural aging.  It would be a fountain of (brain) youth.  We're probably not there yet.  Your brain ages for an evolutionary reason, and I don't think we're quite there in reversing this yet.

The change in our mental models of Parkinson's from being a disease of dopamine to being a disease of alpha synuclein have changed the meaning of cure.  I can imagine stopping the pathology of alpha synuclein, thus curing Parkinson's.  However, having cured the disease of alpha synuclein, people will still have symptoms of their alpha synuclein disease that linger on, for example, their lost dopamine cells.  A cure for the underlying pathology of Parkinson's won't eliminate the burden of the disease -- people will probably still need their sinemet -- but it will change our view of the future for people with the disease.

What is Parkinson's?

The Parkinson's disease community's rallying cry is "cure."  What is a cure for Parkinson's?  What do we mean by cure?  The first question we need to ask to answer this question is, What is Parkinson's?

To the best of our knowledge, most Parkinson's disease is what we call a synucleinopathy.  This means that it is a disease associated with a protein called alpha synuclein, or, sometimes in scientific shorthand, asyn.  Why alpha synuclein? The answer is a bit of a technical aside (you can safely skip the rest of this paragraph), you might find that people sometimes talk about proteins being "alpha helical" or "beta sheet" folded proteins.  This has nothing to do with the "alpha" in "alpha synuclein."  Alpha synuclein was one of two forms of synuclein identified in 1994.  One was called "alpha" and the other, "beta" -- thus, the word "alpha" doesn't really mean anything.  Alpha synuclein can form an insoluble structure called an amyloid, which sounds similar to the Alzheimers protein amyloid beta, but the word amyloid refers to a structure, not a protein, and the beta in amyloid beta and the alpha in alpha synuclein are also not related.  This is just one of those confusing things that scientists are taught and laypeople have to try to figure out, like why the galaxy in the constellation Perseus is called NGC 1277 while the galaxy in Andromeda is called M31.  It's just history, not science.

Back to Parkinson's.

In 1817, James Parkinson wrote the article that first described the disease, called An Essay on the Shaking Palsy.  This essay -- and our understanding of the disease from the earliest days until the 1970's, focused on the major clinical symptoms of Parkinson's that emerge from how the disease impacts the dopamine system, notably the dopamine-producing neurons of the part of the brain called the substantia nigra, which basically means, "black stuff," named in the days when anatomists just cut up corpses and named what they saw with little (and often wrong) insight.

In the late 1960's and 1970's, clinical researchers figured out that they could treat Parkinson's with dopamine replacement therapy.  Parkinson's disease was thus a disease of dopamine and therefore, we could focus on the dopamine system in treatments of Parkinson's.  End of story, right?

Wrong.

Focusing on dopamine for Parkinson's is like saying that harbor pollution is a disease of clams.  Replacing the shellfish in a harbor doesn't address the problem of pollution, and replacing dopamine (or the cells that produce dopamine) doesn't cure Parkinson's.  Parkinson's is a disease of cellular pollution (a topic for a later post), and to stop/fix/cure Parkinson's, we need to address that cellular pollution.

The better we address the dopamine system, the more clearly we can see how the synucleinopathy impacts other systems.  Bill Dauer of the University of Michigan recently received a Udall Center grant from NIH to study how Parkinson's affects the cholinergic system.  The cholinergic system is downstream -- affected after -- from the dopamine system, and we have no good treatments for its dysfunction.

Treating the dopamine system is critical to helping people with Parkinson's to deal with their symptoms.  However, focusing on the dopamine system, the substantia nigra, or other motor features of Parkinson's is a distraction from efforts to "cure" Parkinson's.

Why?

As a synucleinopathy, Parkinson's is disease of diffusion: misfolded alpha synuclein slowly but, so far, inexorably diffuses throughout the brain reaching far beyond the basal ganglia (the part of the brain associated with motor symptoms and containing the substantia nigra and other structures we talk about in Parkinson's).

The first breakthrough in Parkinson's treatment came from understanding how different parts of the brain were associated with how Parkinson's presented in the clinic and resulted in treatment by levodopa.  (Breakthrough 1.1 was when doctors realized that carbidopa made levodopa more tolerable: levodopa made people nauseous and carbidopa prevented that.)  The second breakthrough came from understanding how different parts of the brain interacted with each other and resulted in DBS.  Drugs like dopamine agonists and MAO-B inhibitors came from understanding how cells interacted.  The next breakthrough will come from understanding how molecules in cells function and this will herald a new opportunity to change the course of Parkinson's. (Linking genes to cell function will be another topic: now I've written that.)