Welcome to FOCUS In Sound, the podcast series from the FOCUS newsletter published by the Burroughs Wellcome Fund. I’m your host, science writer Ernie Hood.
On this edition of FOCUS In Sound, we’re going to learn about a devastating infectious pathogen – Mycobacterium tuberculosis, a bacterium with incredible staying power and, unfortunately, considerable killing power as well. Although we might think of it as a somewhat old-fashioned disease, tuberculosis is still an enormous global health problem, with 15 million new cases every year and as many as 3 million deaths worldwide. Also, up to one-third of the world’s population, that is, as many as 2 billion people may be latently infected with tuberculosis. Despite the many advances in biomedical research over the past several decades, it is still the case that very little is known about how the tuberculosis bacterium works. Joining us on FOCUS In Sound is a young investigator who is working to change that by conducting basic research on the tuberculosis bacterium. By characterizing the pathogenesis of the disease and enhancing the ability to diagnose it, the ultimate goal of her group’s work is nothing less than the ultimate eradication of this infectious scourge.
Dr. Sarah Fortune is the Melvin J. and Geraldine L. Glimcher Associate Professor of Immunology and Infectious Diseases at the Harvard School of Public Health. She received a BS in biology from Yale University and her MD from Columbia University’s College of Physicians and Surgeons. She completed her residency in Internal Medicine and fellowship in Infectious Diseases at the Brigham and Women’s Hospital and Massachusetts General Hospital. Dr. Fortune is supported by awards from Howard Hughes Medical Institute, the Doris Duke Charitable Foundation, the National Institute of Allergy and Infectious Diseases and a New Innovator Award from the National Institutes of Health. In 2012, she was the recipient of a 5-year, $500,000 grant as one of the Burroughs Wellcome Fund’s Investigators in the Pathogenesis of Infectious Disease.
Sarah Fortune, welcome to FOCUS In Sound…
Thank you so much. It’s a pleasure to be here.
Why has there been so little progress at this point in combating tuberculosis?
That is a good question, and in some ways, surprising. We have made enormous advances in many fields. Tuberculosis has been under-studied, in part because it affects the rest of the world, not so much the United States. In your introduction, you mentioned something like a third of the world is estimated to be latently infected with TB. Most of those people are in Africa, India and China, so tuberculosis hasn’t been a funding priority until relatively recently in the United States. And then, tuberculosis has been really hard to work with. When we work with tuberculosis, we work in what’s called a Biosafety Level 3 facility, which is not quite like the containment facilities in the movies Outbreak or something, but is pretty close. Everybody’s wearing a space suit and works in very scripted, stereotyped fashion to be safe, and progress is slow. So the combination of those – a lack of funding and the constraints of working in a Biosafety Level 3 facility – have really made progress slow.
Why do we not have a better handle on tuberculosis? We don’t have a better handle on tuberculosis because we lack all of the critical tools that we take for granted in other areas of medicine to control the disease. So, we can’t diagnose people well, once we can diagnose people, people have to take antibiotics for something like 6-12 months to be cured of their infection, and there is no effective vaccine. So the whole armamentarium of tools that we would use to fight an infection, we don’t have for tuberculosis.
Why does TB diagnosis remain poor? I understand that in many parts of the world it currently takes about six weeks to get an initial diagnosis, and another six weeks or so to determine whether it’s a drug-sensitive or drug-resistant strain.
You are absolutely right. In many parts of the world, like Boston, Massachusetts, where I live, the tuberculosis bacterium grows very slowly, and traditionally in microbiological diagnostics, we diagnose an infection by growing the bacterium up and then figuring out what it is, and then exposing it to the antibiotics that you would take and asking if that bacterium lives or dies. Because the tuberculosis bacterium grows so slowly, that whole process takes something like three months. And that’s not a resource allocation issue; that is literally how long that process takes. So it would take the same amount of time in Mass. General Hospital as it would in Johannesburg, and to solve that problem, therefore, we need new tools. We need really completely revolutionary new ways of diagnosing the infection that don’t rely on standard culture-based diagnostics
I see. Well, where do we currently stand in our ability to treat tuberculosis? Have there been advances, or is it still very challenging to treat?
TB is still challenging to treat, although there has been investment over the last decade in new drug programs, and that is beginning to show fruit. For the first time in actually I think almost three decades there are new antibiotic compounds coming up, and they are specifically for TB. So it’s possible we’re going to have new drugs as of two or three years from now, and that is great. However, all of the new drugs that are emerging are drugs that will still take a six month- or nine month- or a year-long course to cure somebody. So what you’d really love is to treat TB the same way you’d treat something like your ear infection. You go, you get a seven-day course of antibiotics, and you’re done. And for that we’re going to need really different kinds of drugs, and we’re not even close to that.
I know that drug resistance is also very much of an issue. What is it about the tuberculosis bacterium that makes it so incredibly resilient, and so often resistant to drug therapies?
That’s a great question, and the short answer is, we don’t have any idea. And just to kind of give you a sense of how incredibly hardy this bacterium is, we can expose the bacterium to antibiotics that you would take to which the organism should be sensitive, and you will only kill a portion of the population all of the time, even in a culture dish where everything is perfect. We really don’t understand, in that population of bacteria, what makes one bacterium susceptible to antibiotics and one bacterium tolerant of those antibiotics. And really that’s one of the major focuses of our research, is to try to understand essentially bacterial individuality – what makes one bacterial cell different from another, and why you can kill off some easily and what makes those kind of long-lived cells special. We hope by doing that we’d be able to develop better interventions to really get rid of those long-lived cells quickly.
With that in mind, let’s move on to your laboratory’s microbiological research on tuberculosis. I understand your efforts are actually concentrated in three specific areas, so I’d like you to give us a brief overview of each one, starting with the genetic and epigenetic variation.
I kind of introduced the idea that we think individual bacterial cells are special and distinct from one another, and just like you and I are distinct from one another, or really the cells in our body. Our hair cells are different than our skin cells. We kind of take that for granted. But I think when you look at a population of bacterial cells it’s easy to imagine they are all the same, essentially. So we’ve been trying to understand the mechanisms by which bacterial cells differentiate, and a lot of what we have done is build on paradigms that have been developed in people studying often eukaryotic cells like the cells in our body, and trying to understand why our hair cells are different than our skin cells. That might be at a genetic level – their DNA is different. That might be at an epigenetic level, so there is something that generates long-lived differences between cells such as histone changes in the proteins that bind the DNA. Or there could be higher frequency mechanisms, and we have broadly looked at all of those. So we started kind of at the foundation, by looking at genetic diversification, and used a whole range of approaches; a lot of whole genome sequencing of isolates from different conditions, to try to comprehensively map their rates and locations of genetic diversification. The upshot of that is, the rates are higher than we’d expect. They kind of occur at different stages of disease than we might have expected, but they’re not sufficient to explain the diversity of cells that we see if we just look at cells under a microscope, where there we see really they’re very phenotypically different. And so we’ve been looking at other mechanisms by which those cells might differentiate, and that has brought us to epigenetic diversification and high-frequency diversification.
In the course of that line of research, what have you been able to find out about the diversification that affects drug resistance, for example?
The cells differentiate at many different rates using many different mechanisms. So what’s been totally startling is that when we started, the presumption I think in the field was that a population of mycobacteria were very homogeneous and really not very plastic. So they didn’t really actually have a lot of capacity for diversification. And we have uncovered multiple mechanisms by which that diversity arises. One of the simplest is that it turns out that in a mycobacterial cell, which kind of you can think of as like a cucumber, is not symmetric. So one end of the cell and the other end of the cell are not the same, so the stem end of the cucumber and the butt end of the cucumber are not the same, and so when that cell divides, and it would kind of break in the middle, it creates daughter cells that are different. They inherit different cellular contents, and they function in different ways. We’ve been talking about antibiotic susceptibility, and one of the very simple ways in which they’re different is they’re differentially susceptible to different antibiotics. That kind of diversity arises incredibly quickly, and so if you start with one cell, by the time you have a little flask of cells, you have daughter cells that are really, really different from one another in terms of their susceptibility to drugs, just from this very simple mechanism. And then there are layered on top of that slower mechanisms of diversification that are a little bit more stable, that allow the cell to have a longer-lived identity, but that’s still not changing its DNA. And then at the kind of slowest level, the cell is still mutating, and mutating in ways and at rates that we wouldn’t have expected.
So this is actually kind of a unique challenge for this type of research – I take it it’s a very tricky bacterium.
I suspect every bacterium is tricky in its own ways, but because TB has established such a successful relationship with us, as human hosts – TB only lives in people, it is exquisitely adapted to survive in people, and all of those mechanisms of diversification really play into that relationship. So its uniqueness is that all of that diversity really is part of its ability to cause prolonged infection. I think that for many people, in the United States especially, it may not be clear what a problem drug resistance is. It’s just important to recognize that in Mumbai over the past year strains of tuberculosis have emerged that are resistant to every antibiotic we have. It’s a real and immediate threat in parts of the world, and something that we have to, as new drugs come on line, as these new compounds are really coming to the market, we have to understand in parallel the process of drug resistance such that we can protect the few drugs that we have. Otherwise we’re going to have an untreatable infection spread from person to person, just kind of willy-nilly.
What types of methodologies do you and your colleagues employ in working to define these bacterial determinants of variability that you’ve been discussing?
My lab has used a whole range of technologies, ranging from microscopy to whole genome sequencing. We’re really enchanted, honestly, with the new high-throughput technologies, because one of the funny things about tuberculosis is that it’s not really closely related to a lot of bacteria that have been well-studied. You may have heard of E. coli or Bacillus subtilis, and those are what we call model bacteria, and they are super well-studied, and we know a lot about them. Mycobacteria are as closely related to E. coli as we are to plants. And a lot of the paradigms about how a cell works in E. coli is really just on hold in mycobacteria. So we have a huge, huge, huge knowledge gap that we have to in-fill quickly. And honestly, we just can’t be sitting around for the next twenty years figuring out how that cell works. And what these new technologies and new high-throughput technologies at least hold the promise of, is really rapidly filling in the knowledge gap in a kind of comprehensive fashion without doing the one-by-one-by-one experiments that have built this beautiful picture of the workings of a model organism. So we’ve really been trying to make use of them and figure out how to try to put together all of that complex data in a way that makes sense to us and that allows us to ask questions about how the cell works.
One thing that’s coming through loud and clear in what you’re telling us is a real sense of urgency with this disease.
Yes. TB is really an epidemic in some parts of the world. If you go to Durban, South Africa, at any given point in time 1 in 100 people have TB, active tuberculosis. Here in Boston, Massachusetts, it’s about 1 in 100,000 people. So the rates in some parts of the world are extraordinary, and that’s active disease. If you go to Durban, South Africa, I think you could assume that almost everybody, a huge portion of the population, is latently infected, and there’s just this epidemic of active disease. And this is a region of the world where highly drug-resistant strains are emerging. So it is an enormous and immediate problem. For people who don’t remember this, TB is an aerosol infection. That means, I cough, the bacteria hang out in the air, and somebody else breathes them in, and that’s how it’s transmitted. It’s transmitted from person to person in a very promiscuous way, and that gives really the capacity for very large-scale infection.
I understand that one result of the new technologies you’ve been able to deploy, as you described, as in so many other fields, is a mountain of data to sift through. Tell us about that problem and the unique approach you’ve undertaken to help solve it…
Any of these new technologies, be they the whole-genome sequencing technologies or the live cell imaging technologies, generate huge amounts of data, and we have to be able to resolve that into some sort of biologic picture that makes sense. And so it poses all sorts of challenges for us – computational challenges, and then very simple challenges like, we take movies of bacteria and we try to understand why one bacterial cell is different from another by watching them in movies. If you think about how you would do that, in a normal movie you would just kind of with your eye track those and say, okay, I can see this one has turned green and this one has turned red – something simple like that. But if you’re going to do that for hundreds of thousands of images, you need an automated way to do that, and it’s really hard, really hard to develop an automated way that you can then customize for your next question. So one of the ways that we’ve been exploring is crowdsourcing that image identification in order to enable people’s eyes to do what your eyes do well, which is to track things and find differences, and what it is really hard to teach a computer how to do, even in this day and age.
Is the crowdsourcing beginning to bear fruit and actually give you the information you need?
It has been very exciting, and for those people who have been thinking about crowdsourcing, actually crowdsourcing is being used a lot in commercial image analysis. For example, companies will have stacks of text that the scanner can’t quite read, and so people will go in and say, is this a w or an n or something. Or they’ll have faces, and they’ll want to know, are these people happy or sad? And they’ll have crowdsourcing teams that will do that. Those are big commercial companies, Fortune 500-level companies that have a lot of image processing needs that are interfacing with the crowdsourcing world, and what we’re trying to figure out now is how the average academic lab, which is working at a much smaller scale, quite frankly with much more beautiful and interesting images; how we can interface in an effective manner with what is essentially a commercial crowdsourcing world. It’s been a little bit of a challenge, but it’s exciting.
The initial efforts have totally borne fruit and given us a first level of analysis, and we have then essentially wanted to push that to the next level and be able to develop a suite of customizable tools that we could ask the crowdsourcing worker at home to do customizable tasks, and that’s what we’re trying to do now.
How might your research efforts – not only the crowdsourcing but everything you’re up to in your lab – lead to new diagnostic capabilities or even improvements in treatment?
We care a lot about new diagnostics, and actually if I had a goal for the lab, it would be that we would really, at least as part of a collaborative effort, change the face of TB diagnostics. TB diagnostics sound not so glamorous, but as I alluded to earlier, it’s critically important that people are diagnosed early, before they have a chance to spread this infection to lots of people; that the diagnostic be easy and cheap and fast, and essentially that just doesn’t exist now. So what we’re trying to leverage is in our ability to work with individual bacteria and handle them and track them easily, and use that technical capacity, in collaboration with engineers, and sort of translate that technical capacity into a simple diagnostic that would essentially be able to see small numbers of bacteria and recognize them as tuberculosis quickly.
And when would you expect that line of research to bear translational fruit?
The process of deciding when you’re ready to say you can do something like start a company is not entirely clear to me, so we’re kind of at that point. Are we ready? We’re deciding whether we’re ready or not. And I think we’re not quite sure if we’re ready or not. We’ve made a lot of progress, to the point where we can ask the question, are we ready to go? Are we ready to form a company and build a device, and I don’t have an answer for it yet. I would love to say yes, but it may be we have a little more work to do.
I’m sure that’s going to keep you busy in the laboratory and keep you coming to work every day.
Well, actually science is wonderful in terms of keeping you coming to work every day, because every day is new, and every day you get to decide what you’re going to do and try something new. It’s fantastic.
How did you happen to find yourself in this particular line of research?
I am a doctor by training, and I went to medical school at Columbia, which is in New York City, in the early 1990s. It seems a whole long time ago, but in the early 1990s, that was when HIV was epidemic in New York, and we didn’t have antiretroviral drugs. We had lots and lots of new HIV infections, and exactly at that time we had outbreaks of multidrug-resistant tuberculosis, where the United States had thought that we had really solved the TB problem. So in our hospitals and the armory across the street and the prisons in New York, HIV-infected people and then doctors and nurses and staff were getting drug-resistant tuberculosis and dying. It was incredibly formative for me, and at the time, from sort of a naïve perspective, I saw a lot of attention being paid to HIV, and I kind of thought, we’ll have that problem well under control by the time I finish all of my medical school and clinical training and post-doctoral fellow training, and what is not clear to me is whether we’re going to have the tuberculosis problem solved, so that’s what I’m going to work on. And that’s where I am.
Last but not least, Sarah, where is your research headed from here? What are your next steps?
I’ve talked a lot about bacterial individuality and how that may determine whether one bacterium dies in the face of drugs or survives in the face of drugs. The other big question in terms of bacterial individuality is how that changes the course of disease. One of the really interesting things about TB is how variable the clinical course of disease is. If you have a hundred infected people, about ten of those people will get active disease immediately, and the rest will be what we call latently infected. So when we say a third of the world is latently infected, we mean they’re infected but they’re not sick. Of those ninety people who are latently infected, only another ten, let’s say, will get active disease during the course of their lifetime. That begs the question of what distinguishes the people who get sick from the people who don’t. And how could we make those people who get sick more like the people who don’t? The field has really thought of that question in terms of human differences – so I’m different than you, and that might make me susceptible or you susceptible. But we have been wondering whether or not it’s actually in part determined at the level of the bacterium. When people get infected, they’re getting infected by one or two bacteria, and so whether the one bacterium that I get is different than the one bacterium you get sets up a fundamentally different course of disease that plays out in these very different ways at the clinical level. If I had to articulate the next big goal, it would be to understand how bacterial individuality shapes disease course.
Sarah, you and your group are doing some fascinating and important work, and we wish you the best of luck for continued success. Thanks so much for joining us today on FOCUS In Sound…
Thank you so much, it’s been a pleasure.
We hope you’ve enjoyed this edition of the FOCUS In Sound podcast. Until next time, this is Ernie Hood. Thanks for listening!