Tuberculosis and rational vaccine design

Commenting on a recent study from her institution (McMaster University, Ontario, Canada), published in the journal Mucosal Immunology, Dr. Stephanie Swift, my blog colleague at Scilogs and a postdoctoral scientist working in host-pathogen interaction, wrote the other day about how vaccines can be used to train the innate immune system to recognize and repel dreaded pathogens, such as Mycobacterium tuberculosis (“M.tb”), the bacterium responsible for tuberculosis (‘TB’). [I shall highlight a few related points here, but do go read her post: it is informative and interesting.]

Immunologically speaking, the body’s defence mechanisms are dual layered. The first line of defence, called “innate” immune response, is a non-specific, general system; it is comprised of barrier mechanisms that hinder the entry of micro-organisms, as well as an immune cell-based (‘cellular’) and a non-cellular, protein-based (‘humoral’) compartment. The second line of defence – a more robust, specific and sustained response – is provided by another system, called “adaptive” immunity, which also employs many cellular and humoral mechanisms. As the names suggest, innate immunity is ‘always on’; when microbes try to invade, innate immune mechanisms stave off the first attack, and help recruit the components of adaptive immunity. Adaptive immunity serves as immunological memory that primes and programs the defence mechanisms, so that in the event of a second attack, the immune system responds in a focused manner, targeting the specific invading pathogen.

Tuberculosis remains a significant threat to public health globally; co-infection with HIV has driven its resurgence despite efforts to reduce its impact. An estimated 1.4 million people succumb to TB each year (WHO, 2011), with the largest burden observed in Africa and Southeast Asia. Although anti-tubercular drug discovery has revived itself lately with renewed vigor, focusing on adaptive immunity has been the mainstay of immune-based TB therapies – as Dr. Swift pointed out in her essay, too.

And is there a good reason for this focus? Yes, it appears, there is, but we need to remember some fundamental facts about this pervasive pathogen. M.tb, which likely evolved from soil-dwelling ancestors to become a human pathogen around 10000 years ago, is endowed with unique abilities geared towards survival and persistence:

• M.tb is considered aerobic-to-facultative-anaerobe, which means it normally grows in presence of oxygen, but if oxygen is deficient, it can shift to other mechanisms of respiration that doesn’t require oxygen.
• When oxygen and nutrients are plentiful, and temperature is 37˚C (as, say, in the lungs), M.tb makes more copies of itself in 18-24h (3-4 weeks on artificial medium in vitro). For a pathogen, this is an extremely slow rate of division.
• When conditions are not propitious, M.tb has the unique ability to enter a dormant, non-replicative state with low respiratory rate or metabolic activities. When in this zen-like state, it is not affected by either host immune mechanisms or anti-tubercular drugs (which otherwise kill the growing bug). It also modifies its metabolism, via genetic changes, to adapt to the nutrient limitations.
• M.tb possesses an unusual cell wall, thick and impermeable, composed of complex sugars (polysaccharides), amino acid-containing sugars (peptidoglycans), sugar-containing fatty acids (glycolipids) and long-chain fatty acids (e.g. mycolic acid). It is these latter lipid-components that are considered to protect the bacteria against weak disinfectants and desiccation.

M.tb revealed with a special stain specific for Mycobacteria, acid-fast Ziehl-Neelsen stain; Magnified 1000X. Image courtesy: CDC/Dr. George P. Kubica (1979).

Macrophages, members of first-line defence, are immune cells that engulf microbes via a process called phagocytosis, and kill them internally by:

• Creating an acidic (low pH) environment inside the bubble-like enclosure, called a phagosome;
• Producing hyper-reactive (and corrosive) chemical derivatives of oxygen and nitrogen, which causes damage to microbial DNA, lipids, proteins and other structural components;
• Flooding the microbes with enzymes which, under the acidic condition, can break down (‘hydrolyze’) lipids and proteins in the outer layers of the microbes;
• Releasing peptides (small proteins) with potent antimicrobial properties (such as cathelicidin, hepcidin, etc.), which pokes holes into the outer layers of the microbe, and
• Undergoing a regulated suicide process known as apoptosis, in case the microbe is able to escape its enclosure and wade into the cytoplasm, the fluid-filled intracellular space, of the macrophage itself.

Immune messenger proteins – called cytokines, notably Interferon-γ – secreted by certain immune cells influence these defensive process by acting on and activating the same and other immune cells. Macrophages and another immune cell, known as dendritic cells, can pick up pieces of the destroyed microbes and their products, and display it to yet another immune cellular component, the T-lymphocytes, at which point the immune response traverses to the adaptive side; T-lymphocytes instruct B-lymphocytes to produce specific antibodies against the microbe, and create memory B-lymphocytes and memory T-lymphocytes in preparation of subsequent attacks; other immune cells are also appropriately instructed to recognize the pathogenic components for unleashing their destructive power in a focused manner on the microbe in the event of the next attack.

But… M.tb’s unique abilities include not only escaping elimination, but also surviving inside macrophages. It is thought that their unusual cell-wall composition allows them to invade resting macrophages silently. Once inside the resting or active macrophages, M.tb can shut down the internal processes by which macrophages kill and digest microbes. In addition, virulent M.tb can prevent the macrophage from undergoing apoptosis by blocking the self-destruct signal and making quick repairs to the already-destroyed cellular structures. (Note: If you are familiar with Stargate SG-1, this is an easy parallel with Replicators!)

In addition, inside activated macrophages, M.tb can tolerate the low pH environment by reducing the macrophage protein pumps responsible for acidification, up-scaling the production of its own urease enzyme, which produces ammonia to neutralize the excess acid, and programming a set of its own genes called aprABC, whose functions allow it to adapt to the acidic micro-environment inside the host cells. Therefore, M.tb essentially reprograms the host cells after entry to prevent its own destruction and ensure its persistence.

And that is not all. Once inside macrophages, M.tb has the ability to use a host-derived carbon source (most likely, cholesterol and glycerol) for its metabolic needs;  it modifies both its lipid metabolism and its toxic waste disposal mechanism to suit its environment. To prevent damage from corrosive ionic derivatives of oxygen and nitrogen, M.tb enhances the production of several key enzymes (such as, superoxide dismutases, peroxidases, and reductases) that can neutralize these ions, as well as repair the damages caused. There is some evidence that macrophages may use excess zinc ions inside the phagosome to kill M.tb, but recently it was discovered that the bacteria can safely pump out the zinc ions. It even counteracts the effect of the antimicrobial peptides by neutralizing their negative charge via a positively-charged lipid molecule on its membrane.

But wait, there is more! Ordinarily, macrophages with engulfed microbes migrate to tissue sites, where other cell types, such as monocytes, lymphocytes, and neutrophils are signaled and recruited to create a confined environment, called a granuloma, where a delicate balance is established between the host immune cells and the pathogen. Formation of well-organized granulomas, comprising different immune cells, is a characteristic feature of several microbes, mainly some respiratory fungal pathogens (such as Cryptococcus, Histoplasma etc.), and – of course – M.tb. In TB, a primary lesion or point of damage becomes a solid granuloma involving macrophages, monocytes, dendritic cells, T- and B-lymphocytes; solid granulomata in lungs may represent foci in which M.tb remain in a latent, quiescent condition – but not dead. Under certain circumstances, especially if the host immunity falters for some reason, they can be reactivated, and progress to active disease. The granuloma in this case becomes necrotic or caseous (filled with dead tissue and dissolved material), which allows for the bacteria to escape its immune-prison chamber, and spread elsewhere via blood, and to other persons via breathed/sneezed out air from the lungs.

Image: Necrotizing granulomas (G) localized around an airway (A). (B) is a blood vessel. ©Dr. Yale Rosen; you can view more of the pulmonary pathology photos by fabulous Dr. Yale Rosen on Flickr.

It is clear that while the host innate immune system, led by the macrophages, can boast of a remarkable arsenal of microbicidal mechanisms in general, when it comes to M.tb, their efficacy is not guaranteed; during co-evolution with mammalian hosts for thousands of years, M.tb has been conditioned to defeat the hostile intracellular environment and persist in the host. Therefore, it is hardly surprising that most of the current vaccine candidates are not aimed at preventing or eliminating primary TB infection, but focus on priming the players of adaptive immunity to stave off the emergence of active disease by targeting the metabolically active, replicating pathogen.

Vaccine antigens are antigens which, when used to immunize, prime the immune system to recognize the same or similar structures in the invading pathogens and mount an immune response. A booster is one or more subsequent doses of the same or similar vaccine antigens, designed to further hone the ability of the immune system to focus on the target. The desired effect of this strategy is the establishment of a state of immunological memory, which allows the immune system to respond more rapidly and effectively to previously-encountered pathogens. The memory responses – which are often the principal antimicrobial response in an immunized host – are dependent on antigen-specific B- and T-lymphocyte populations made up of genetically identical (‘clonal’) cells, and are known to also differ qualitatively from primary responses, in terms of both antibody production and T-cell memory.

For the longest time, we have had the BCG vaccine, derived from a cousin of M.tb, the Mycobacterium bovis strain Bacillus Calmette-Guérin. Following WHO recommendations, BCG is routinely administered to neonates or infants in parts of the world endemic for TB; BCG is at least partially effective in preventing development of serious TB manifestations in children, such as TB-meningitis and disseminated (“miliary”) TB. However, it appears that for some reason, BCG vaccination is unable to sustain immunological memory, the sine qua non for adaptive immunity; this leads to highly variable protection from pulmonary TB as far as adults are concerned. To counter this, various investigators have started considering combining BCG vaccine with another vaccine antigen in a prime-boost strategy designed to maximize the benefits of both. Recombinant BCG vaccines, with improved immunogenicity profile, are in clinical trials now.

In addition, several new candidate vaccine antigens are being studied. Incidentally, a team of researchers from my institution (Johns Hopkins University, Baltimore) also published this month an interesting study in PLoS One, in which they identified a potential vaccine antigen that can protect against TB meningitis. This antigen is an Mtb protein, known as PknD, whose function involves helping the bacteria cross the barrier – known as the blood-brain barrier – that ordinarily keeps microbes from entering the brain. They had earlier discovered that an antibody that specifically recognizes this protein is able to neutralize its action. Therein lies the rationale of this vaccine program, which can potentially produce such an antibody in an immunized person, as well as produce T- and B-lymphocyte memory of this protein. This would likely protect specifically against TB meningitis, a high mortality outcome of disseminated TB.

However, as I indicated before, during the period of latency/dormancy, M.tb modifies its genetic program; as a consequence, the character of the antigens usually associated with the bacterium changes, and separate antigens expressed by the dormant pathogen predominate. Therefore, many investigators now consider that a sustained and effective control of dormant M.tb would require the vaccine-induced immune response to recognize target antigens on the latent bacterium. Mouse studies are ongoing, with encouraging results, using a latency-associated M.tb antigen, a protein called Rv2660c, which is selectively expressed during nutrient starvation.  

The prime-boost immunization strategies, focused towards developing immunological memory, have shown promise in battling M.tb, but the mechanism of their actions, particularly how they impact innate and adaptive immunity is yet unclear.

In the McMaster study that Dr. Swift commented on, the researchers used two engineered (recombinant) viral vaccines; the gene for an immunodominant early secreted M.tb protein Ag85A, was inserted into either of two viral vectors, derived from the genomes of adenovirus (Ad) and vesicular stomatitis virus (VSV). Once these vectors are taken up into host cells, this protein would be continuously produced within the host cells, and serve as the vaccine. Using the prime-boost strategy in a mouse model, they found both elicited similar M.tb-specific T-lymphocyte responses (that is, adaptive immunity) in the lung and spleen.

However, the picture changed drastically when they challenged the immunized mice with the bacterium, and studied the signatures (a.k.a. markers) of protection. The Ad-based boosting regimen was able to greatly clear M.tb from the lungs and spleen of the mice, and elicit high levels of a macrophage-stimulating cytokine called IL-12 in the lungs. In contrast, the VSV-based boosting showed no enhancement of protection against TB, and even elicited high levels of another cytokine called IL-10, which profoundly shuts down the anti-M.tb activity of macrophages. These cytokines were released from phagocytic cells that represent innate immunity. Therefore, clearly, Ad-based boost strategy enhanced the activation of innate immune system in the lung, which presumably resulted in the greater protection to TB.

The researchers figured out that the main difference between the two vector vaccines was in how they interacted with – and primed – the innate immune cells at the site of immunization. IL-12 is known to enhance the production of interferon-γ, as well as of the reactive Nitric Oxides, from these immune cells. In short, the immune cells exposed to the Ad-based boost became more efficient killers of M.tb.

In contrast, the researchers found that exposure to VSV-based boost led to increased release of another cytokine, called interferon-β, which has grossly different actions compared to interferon-γ. Increased interferon-β severely blunted the beneficial IL-12 responses in the M.tb-infected phagocytic innate immune cells; in addition, exposure to interferon-β made the immune cells inefficient in controlling the numbers of engulfed (phagocytosed) M.tb.

Whereas T-lymphocyte mediated adaptive immunity is essential and beneficial for many microbial pathogens, effective elimination of pathogens that hide in plain sight within immune cells itself, such as M.tb and HIV, requires the engagement of innate immunity which encounters them first, independent of the T-cells. To succeed against M.tb, a vaccine would need to engage both wings of the defence. As Dr. Swift pointed out, each of the viruses from which the respective vectors were derived:

“… enters lung cells in a certain way, attaching at different sites, activating different intracellular pathways, expressing different viral products and consequently engaging different parts of the immune system.”

Viral vectors offer an attractive choice for engaging adaptive immunity against different pathogens. This study shows how they can be used to strengthen innate immunity, too. These are important considerations for a rational vaccine design.

Two major studies discussed:

• Jeyanathan M, Damjanovic D, Shaler CR, Lai R, Wortzman M, Yin C, Zganiacz A, Lichty BD, & Xing Z (2013). Differentially imprinted innate immunity by mucosal boost vaccination determines antituberculosis immune protective outcomes, independent of T-cell immunity. Mucosal immunology, 6 (3), 612-25 PMID: 23131783
• Skerry, C., Pokkali, S., Pinn, M., Be, N., Harper, J., Karakousis, P., & Jain, S. (2013). Vaccination with Recombinant Mycobacterium tuberculosis PknD Attenuates Bacterial Dissemination to the Brain in Guinea Pigs PLoS ONE, 8 (6) DOI: 10.1371/journal.pone.0066310