For the patient who has swallowed months of tuberculosis pills only to relapse weeks after the last dose, the enemy was never fully killed. A small population of bacteria had gone quiet, riding out the antibiotics in a dormant, drug-tolerant state, then reawakening once the treatment stopped. Those survivors, known as persisters, are the reason tuberculosis therapy stretches across half a year and still fails a meaningful share of the people who complete it. A team at Johns Hopkins now says it has built a vaccine aimed squarely at that hidden reservoir, and that it clears infection faster and blocks relapse in animals.
According to ScienceDaily and Johns Hopkins Medicine, scientists at Johns Hopkins Medicine and the Johns Hopkins Bloomberg School of Public Health developed a therapeutic intranasal DNA vaccine that fuses two genes to steer the immune system toward the drug-tolerant persisters that survive conventional antibiotics. The work, framed by its authors as a preclinical proof of concept, drew fresh coverage on July 4, 2026, and was published in the Journal of Clinical Investigation.
Persisters that outlast the pills
Tuberculosis remains the deadliest infectious disease on the planet. The World Health Organization estimated that roughly 1.23 million people died of the disease in 2024 and that about 10.7 million fell ill, figures that have kept the pathogen at the top of the mortality tables even as newer threats capture headlines. A central reason the disease is so stubborn is biological. Mycobacterium tuberculosis can slip into a survival mode when it faces antibiotics, low oxygen or scarce nutrients, and in that state it tolerates drugs that would otherwise destroy it.
The Johns Hopkins team traced that behavior to a specific genetic switch. According to ScienceDaily, the bacterium carries a gene called relMtb, which produces a protein that helps the microbe endure hostile conditions and enter its persistent, drug-tolerant form. Those persisters are not resistant in the classical, mutation-driven sense; they simply wait. When treatment ends, they can resume growth and seed a relapse, which is one of the mechanisms that forces standard regimens to run for months rather than days.
Fusing two genes into one instruction
The vaccine's design rests on a deliberate pairing. The researchers linked relMtb to a second gene, Mip3a, so that the immune system would be pointed directly at the persister population. According to Johns Hopkins Medicine, the relMtb component supplies the target, the very protein the dormant bacteria rely on to survive, while the Mip3a component produces a signal that recruits immature dendritic cells.
Those dendritic cells are the couriers of the immune system. They collect fragments of the tuberculosis bacterium and present them to T cells, the coordinators that mount a focused attack. By fusing the two genes, the vaccine effectively hands the immune system both a wanted poster and a summons, drawing the presenting cells to the site and equipping them to flag the specific bacteria that antibiotics leave behind. Delivered through the nose, the construct is intended to raise defenses at the airway, the route by which tuberculosis enters and establishes itself.
Delivery through the airway
Intranasal delivery is not incidental to the strategy. Tuberculosis is an airborne, lung-centered disease, and mucosal immunity at the respiratory surface can meet the pathogen where it lands. According to the Johns Hopkins account relayed by ScienceDaily, the nasal vaccine generated immune activity both locally in the airways and more broadly through the body, a combination the researchers presented as evidence that the approach engages the tissues that matter most for tuberculosis control.
Results in mice and monkeys
This report is open to every reader. Subscribers unlock the full Speedway Scene archive and keep independent, rigorous journalism on the forces that move markets and power on its feet. Get the Briefing
The strongest data came from mice. According to ScienceDaily, when the intranasal DNA fusion vaccine was administered alongside standard tuberculosis drug therapy, infected animals cleared the bacteria faster, showed reduced lung inflammation and, critically, avoided relapse after treatment ended. The researchers reported greater recruitment of dendritic cells, better-organized lung tissue and durable T-cell responses, the immunological signatures they had engineered the vaccine to produce.
The team extended the work into a second species. In rhesus macaques, the vaccine elicited tuberculosis-specific immune responses in the bloodstream and airways that persisted for at least six months, according to ScienceDaily. The macaque experiments measured immune response rather than protection against active infection, so they establish durability and reach rather than efficacy on their own. Even so, seeing consistent signals across rodents and primates is the kind of cross-species agreement that developers look for before contemplating human trials.
Boosting a regimen built for resistant strains
Perhaps the most consequential finding for global tuberculosis care concerns drug resistance. According to ScienceDaily and Johns Hopkins Medicine, the vaccine improved the performance of the drug combination of bedaquiline, pretomanid and linezolid, a regimen central to modern treatment of drug-resistant tuberculosis. If a therapeutic vaccine can make that combination work better or faster, the implication is that it could shorten therapy or improve cure rates precisely in the cases that are hardest to treat.
That matters because drug-resistant tuberculosis is a persistent gap in the global response. The World Health Organization has reported that only about two in five people with drug-resistant disease accessed treatment in 2024, and multidrug-resistant strains remain a health-security concern. A tool that attacks persisters directly, rather than relying solely on antibiotics that persisters can wait out, would address a mechanism that current chemotherapy struggles to reach.
Distance still to travel
The Johns Hopkins researchers were careful about the stage of the work. According to ScienceDaily, they emphasized that additional preclinical study is required before the vaccine can move into human clinical trials. Nothing in the reported data establishes safety or efficacy in people, and the macaque results, while encouraging, did not test protection against live infection. This is a laboratory advance with a plausible clinical logic, not a product nearing approval.
Stakes for patients and health systems
Set against the scale of tuberculosis, the appeal of the approach is straightforward. The people who bear the disease are concentrated in low- and middle-income countries, where lengthy regimens are difficult to complete and relapse compounds the burden on families and clinics. A therapy that shortens treatment or reduces relapse would ease that pressure at both the individual and system levels, freeing programs to reach more patients with fewer resources.
The persister problem also explains why incremental gains in tuberculosis care have been so hard won. Antibiotics that clear the actively growing bacteria leave the dormant fraction largely untouched, which is why regimens are long and why stopping early invites recurrence. A vaccine engineered to recruit the immune system against that dormant fraction attacks the disease along a different axis, and pairing it with existing drugs, as the Johns Hopkins team did, reflects a combination strategy rather than a replacement for chemotherapy.
For now, the finding sits where many promising tuberculosis candidates begin, with clean animal data, a coherent biological rationale and a long verification path ahead. According to the coverage that circulated in early July 2026, the researchers have positioned relMtb and Mip3a as the working core of a therapeutic vaccine that could eventually complement the antibiotics tuberculosis has learned to outlast. Whether that promise holds in humans is the question the next phase of study, and human verification of these early claims, will have to answer.