The Dreamer

The biochemist and popular author, Isaac Asimov, once said The most exciting phrase to hear in science, the one that heralds new discoveries, is not 'Eureka! ' but 'That's funny….’

This sentence crystallizes my ideal of how science should be done. Another way to put it is: follow the data, especially if it isn’t what you expected. The papers that I value the most in my career are not the ones that are the most cited, or the ones that I am “known for” or defined my career. Rather, they are the ones that started when something unexpected happened and there still remain unanswered questions.

Often, when I was at a loss for a new idea or project, I would simply try to repeat someone else’s published data that I found interesting for some reason or other. Inevitably, I would see something unexpected that would open new avenues. I would strongly advocate that everyone spend a few weeks trying to replicate some published data. I guarantee that you will be left with a handful of new questions.

I joined the Emory faculty in 1976 and began by studying the electrophysiological responses of the pacemaker region of the heart to the autonomic nervous system transmitters. These studies were going well, but one day when I read that ATP was often packaged in synaptic vesicles along with autonomic neurotransmitters, I dumped some ATP (and adenosine as a “control”) onto my sinoatrial node preparation and was shocked to find that both ATP and adenosine had nearly the same effect as acetylcholine. Although the paper that I wrote seems pedestrian by today’s standards (1), I still love it, not only because it was one of my first publications as an independent investigator, but also because I felt that I had seen something no one else had ever seen and because it so clearly shows my naivete. I rigorously tested all the questions that came into my mind, no matter how naïve, about how adenosine might be producing this effect. I used this paper to get my second R01 grant.

I did not pursue the adenosine project very avidly because one of my graduate students had stumbled onto another interesting observation. Autonomic transmitters regulate the force of cardiac contraction by altering the amplitude of the voltage-gated calcium current. It had been proposed that this change in calcium current was mediated by changes in the phosphorylation state of the calcium channel. To test this, we were perfusing hearts with radioactive 32-phosphate and looking for proteins that incorporated 32P in response to norepinephrine. We were hoping to find the calcium channel protein and correlate its phosphorylation state with the force of cardiac contraction. We were very excited to find a very prominent protein band at 160kD that fit the bill almost perfectly. But, as we looked more closely, it seemed there was too much of this protein for it to be the calcium channel and it wasn’t soluble in the detergent Triton-X100, so it was probably not a membrane protein. It turned out that it was not the calcium channel but a myofibrillar protein called myosin-binding protein-C. Quandary: should I move on and continue my search for the calcium channel or figure out what this myofibrillar protein was doing? Actually, I didn’t think about it too hard. The data were crystal-clear and it posed too many interesting questions to ignore. Finally, we discovered that the phosphorylation of this protein was not related to the force of contraction as I first guessed, but to the rate of relaxation of the heart, allowing it to fill with blood more fully as heart rate increased in response to norepinephrine (2). Today it is known that mutations in myosin binding protein-C are a major cause of inherited hypertrophic cardiomyopathy.

However, I still had an interest in how calcium channels worked. So, I started expressing them in Xenopus oocytes, but my voltage clamp recordings were contaminated by a current I did not understand. I wanted to block this current so that I could study the calcium current, but eventually I ended up spending the rest of my career studying this current and the proteins that mediated it. This current was a calcium-activated chloride current carried by anoctamin-1 or TMEM16A and was not only present in Xenopus oocytes, where it is responsible for the fast block to polyspermy, but also many secretory epithelial cells in mammals. I like my first papers on this calcium-activated chloride current because they reveal how utterly confounded I was. I couldn’t figure out whether I was dealing with one type of channel with very odd features or two different channels (3). The main puzzle was that inward currents had different sensitivity to Ca2+ and different gating scheme than outward currents. This was hard to explain if this current was mediated by a single protein. The experiments we designed to sort this out and their interpretations were unnecessarily convoluted. As such, they reveal how I think scientific problems should be approached, even if the thought processes are muddled. It is interesting to me that according to the Journal of General Physiology website, the average number of views of this paper between 1999 and 2003 was less than 20 per month, but in 2018 (nearly 20 years after it was published!), it received over 100 views per month and still had ~25 views per month in 2023. Somehow this validates my decision to study this channel.

The molecular identity of the Ca2+-activated Cl- channel was a major puzzle for us. At one point, we started project to expression clone it but when postdoc Khaled Machaca moved on to a faculty position, the project stalled. Around that time, another lab published that bestrophins were Ca2+-activated Cl- channels. It soon became pretty clear to us that the properties of this channel did not fit the current we were interested in. It did not have this peculiar voltage-dependent Ca2+ sensitivity that we had seen in the native oocyte current. But, again, I was seduced by the data. In this case, it was not a “that’s funny” moment, but rather a “what’s wrong?” moment. It was clear to me that bestrophins were a new kind of channel and that they – in some cell types – were involved in cell volume regulation. But some of my colleagues did not believe it was a chloride channel and the rest did not think it was involved in cell volume regulation. I like these papers, not because we were right, but because of the rigor with which we tested the questions (4,5).   

Finally in 2008, the gene encoding the Ca2+-activated Cl- channel was identified to be TMEM16A (also known as ANO1). Now, from our work and the work of many others, we finally understand the mechanisms of gating of the Ca2+-activated Cl- current that so confused and intrigued me back in 1996. But, once again, I could not discipline myself. There are 10 genes in the TMEM16 family. I thought it would be a simple PhD thesis for one of my graduate students to characterize one of the ANO1 paralogs, like ANO5 or ANO6. But when she couldn’t record any chloride currents from cells expressing these proteins, we were puzzled – until we learned from work from Shigekazu Nagata’s lab that they might be phospholipid scramblases,  which opened a whole new line of investigation (6).