Upending the dogma of TB latency: Will it offer new clues for vaccine research?

There are many reasons why HIV is such a perplexing virus to combat. One of them is that there are incredibly few people — only a few, if that, of the nearly 86 million people to acquire HIV to date — who have naturally cleared the virus after becoming infected. Once an HIV infection occurs, it almost always persists in the body for a lifetime.

For decades, the same was thought to be true for tuberculosis (TB), the disease caused by infection with Mycobacterium tuberculosis. In the absence of lengthy treatment with multiple antibiotics, it was thought that TB often persisted in the body indefinitely. Since the 19th century, when scientists discovered evidence of the tuberculosis bacterium in the lungs of individuals who had died of causes other than TB, it was thought that there were two forms of TB, acute and latent. Those with acute TB experienced symptoms of the disease and maybe even died from it. By the 21st century, the definition of latency was cemented — it encompassed asymptomatic individuals who had a positive skin test indicative of immunoreactivity to Mycobacterium tuberculosis. And, depending on their location and circumstances, some of these latently infected individuals were likely treated to prevent them from developing active TB, which remains one of the world’s most common causes of death due to an infectious disease (rivalled only by SARS-CoV-2).

Until recently, estimates suggested that 2 billion people were latently infected with Mycobacterium tuberculosis. This put about a quarter of the world’s population at risk of “reactivating” their latent infection to develop TB at any time, even decades after their initial infection.

I recently spoke with Ramakrishnan about these findings, its implications for future vaccine research, and her ongoing research on the basic biology of TB in zebrafish, a transparent host that allows for great insights. An edited version of our conversation appears below.

What are the implications of the finding that most people clear TB?

In our epidemiological work, we’ve shown that most people undoubtedly clear TB and they clear it reasonably fast, so, the key question is: What is different about the 5-10% of people who don’t clear it? It is important to devise longitudinal contact studies to figure that out. If one follows serum immunological and transcriptomic profiles of TB contacts who develop immunoreactivity, one can then link them to getting disease versus not. Since most disease develops within a year or two, that amount of follow-up would be sufficient to discern differences. The people who don’t clear TB don’t all have a single mutation that makes them unable to clear the infection. Rather, there are likely multiple paths to genetic susceptibility that the TB bacteria can exploit using myriad virulence deterrents.

TB, like any infectious disease, represents a game between the pathogen and the host. If you just consider it from an evolutionary perspective, then you have a situation where this bacterium has survived for millennia so it has evolved multiple ways to evade the host’s immune defenses. Understanding this is key to figuring out why some people are unable to clear TB on their own. Studying the earliest stages of TB infection in zebrafish has been a real boon because the epidemiological studies suggest that the outcome of the host-pathogen battle in TB begins to be determined early after infection.

Why zebrafish? What is the advantage of using this model of human TB infection and pathogenesis?

Our zebrafish work has illuminated multiple steps of TB that would not be accessible in your traditional animal models because we can follow infection live and watch the interactions in a see-through host. And we can do so from the very first host-pathogen encounters, for instance when the infecting bacteria first encounter host macrophages This model system also provides us with the ability to genetically manipulate both the bacterium and the host simultaneously, which has proved powerful.

What are some of the key findings you’ve made on the early infection process in zebrafish?

We have had insights into literally each step of infection, including how the bacteria survive the initial interaction with host macrophages by using two surface lipids in a coordinate fashion to counter the immune system. These lipids allow the bacteria to stave off microbicidal macrophages and instead recruit more permissive macrophages that the bacteria can infect and use to get ferried into the deeper tissues.

Bacteria can then alter the macrophages’ endocytic trafficking of the cell to gain a favorable niche. They also use several determinants to survive in the macrophages. For instance, we’ve found that the mycobacteria can activate efflux pumps in macrophages, which help them survive and also make them tolerant to antimicrobial agents, making it a double whammy for the host.

Another intriguing aspect is the formation of the granuloma — the hallmark structure that forms in TB. The granuloma consists of infected macrophages that undergo a specialized differentiation, recruiting many other types of immune cells to form this very complex structure. Granulomas can be sites of enhanced host immunity and eradicate bacteria in many cases. But in a substantial minority of cases, the bacteria can exploit the granuloma for their expansion. Paradoxically, they do this by accelerating the kinetics of granuloma formation, changing a host-protective structure into a harmful one. This is also something we have been able to discover by monitoring and manipulating granuloma kinetics in transparent larvae. We’ve found that the TB bacteria kill the infected macrophage eventually and meanwhile recruit new macrophages to the granuloma that engulf the dying cells. Thus, the TB bacteria from each dying macrophage will get distributed into two or three new macrophages, which provides them with new growth niches. The bacteria are using the granuloma to spread exponentially, and, as a result, this process makes the disease more pathogenic, more transmissible, and more morbid.

Have you shown that similar processes are taking place in human TB infection?

Yes, in many cases, we have been able to link these findings to human disease. I am an infectious disease clinician and so I need to see an understanding of human disease coming from our little fish eggs.

In the case of one of the genetic mutants we’ve identified, we found that it is linked with a hyperinflammatory state because of an overproduction of the cytokine TNF [tumor necrosis factor]. TNF is a critical host protective cytokine in TB. But what we discovered in the fish is that an excess production of TNF is as bad as having too little, and possibly even worse. We have been able to detail the mechanism behind this. We’ve then studied this finding in human TB cohorts and shown that humans who overproduce TNF, probably through a slew of different genetic variations, are hypersusceptible to death from tuberculous meningitis, a particularly deadly form of TB. Corticosteroids are used in conjunction with antibiotics to attempt to tamp down this immune response and improve survival in TB meningitis. In keeping with our findings in the fish where we showed the deleterious inflammatory pathway caused by TNF, we found that in TB meningitis cohorts, only those who overproduced TNF benefitted from steroids. We’ve done studies in Vietnam and Indonesia where we’ve shown this to be the case and now there is a formal trial going on where steroids will be administered only to those with the specific genotype that makes them overproduce TNF. This is a just a bird’s eye view, or rather a fisheye view, of how some of our discoveries are being applied to human TB disease.

What was the response to your research showing latent TB may be much less common than previously thought?

I think people were very concerned that by revealing that many fewer than 2 billion people are infected with TB, we were minimizing the problem of TB. But that is certainly not the case. A disease that continues to kill over a million and a half people per year, despite the existence of antibiotics for more than 60 years, is nothing to be scoffed at. We are actually upgrading TB in terms of its virulence because if you assume 2 billion people are currently infected, then the number of people who die is very small. But if you now consider that your denominator is really only 10 million people, then you suddenly have a much higher fatality rate, a shockingly high fatality rate given we have effective treatment.

We also used to think that latency affected a quarter of the world — we taught this to our medical students and even wrote this in opening paragraphs of our papers. Then bit by bit we started to realize that this simply was not the case. Which brings me to another point, which is that none of this work is really ours, we’re just pulling together beautiful, rigorous work that has been done by other people over the decades. The dogma has just persisted despite existing longstanding evidence to the contrary.

I think the fact that the skin test can remain positive in people who cleared the infection was what led to much of the confusion over latency. But if you think about it like a proper immunologist, you recognize that you can retain an immune response to an infection that you have cleared. Categorizing people with a positive immunological response to the TB bacterium having latent TB is a big problem because in fact most of these people will have cleared their infection. We don’t think that the terms latent TB and active TB are useful; they are confusing and we have suggested simplified terminology in our paper that is also accurate.

What are some implications of this finding for TB vaccine research?

If I were running a vaccine trial, I’d be very happy to see the analyses we’ve provided because what they’re telling you is that instead of having to wait years and years to see if the vaccine worked or not, you should be able to discern efficacy within a year. That’s a huge benefit and would be a huge cost savings if you’re designing a trial.

However, if you are going to test a vaccine to prevent someone with latent infection from progressing to active TB, which is what is happening in the Phase III vaccine trial of the M72 vaccine candidate being conducted by Wellcome and the Gates Foundation, you have to be able to separate the people who have a positive skin test and are still infected, from those who have a positive skin test but have completely cleared the infection. To just lump all the skin test positives together and run a trial without recognizing that most of the people, or a significant portion of the people, with a positive skin test are actually cured is really just ignoring the epidemiological data and will muddy the interpretation of the results. And if anything, evidence suggests that people who had a positive skin test and cleared the infection are more resistant to TB. So, if you treat all people with a positive skin test as infected and really 90% of them have already cleared the infection, how are you going to discern what you’re actually seeing in response to this vaccine? It is not clear to me why the trial is seeking to examine vaccine efficacy in preventing latency reactivation rather than preventing infection or primary disease.

“If you think that a quarter of the world population — around 2 billion people — is infected with M. tuberculosis, it is a daunting challenge for eliminating TB. The realization that the true number, although still substantial, is much lower makes the problem more solvable.”

~ Lalita Ramakrishnan, Disease Models & Mechanisms

---

Read more:

• In this paper, Ramakrishnan and colleagues look specifically at the idea of latent TB infection in individuals living with HIV.
• The latest report on TB from the World Health Organization, which has altered its language on latency based on Ramakrishnan’s and colleague’s findings.
• The 7th Global Forum on TB Vaccines is taking place October 8-10, 2024, in Rio de Janeiro, Brazil.