Albert Calmette, Bronchoalveolar lavage (BAL), Calmette Guerin, Delayed Type Hypersensitivity (DTH), Granuloma, Infectious Droplet Nuclei, Interferon-gamma (IFN-g), John Stanford, Lalita Ramakrishnan, Mantoux test, Original Antigenic Sin, The Lubeck disaster, Tumor Necrosis Factor-alpha (TNF-a)
The difficulty is not so much with tuberculosis (TB), the disease, as with obfuscating dogma that has prevented Mucosal Immunity taking center stage in TB research and vaccine development. Further, over the past century, even as biomedical research and development expanded exponentially, TB‘s domain shrank to primarily encompass the poor who also tend to be the dispossessed and malnourished. Careerist Science and then amply filled the ensuing lacuna in scientific prestige attendant to TB.
1. Post-1970s scientific fashion has emphasized reliance on poorly predictive animal models, such as the mouse, which do not recapitulate human TB disease (1).
2. We use unsuitable vaccine routes of administration such as intradermal, which do not recapitulate natural TB infection which is airborne. Are we using the currently available vaccine(Bacille Calmette Guerin) appropriately? ‘s history is an important guide for novel TB vaccine development. Different countries administer different sub-strains, and even use different routes, potentially resulting in different immunogenicity and efficacy. An examination of suggests that if we misread its history and abandon it wholesale, we might be throwing the baby out with the bath water.
3. We have forgotten the object lesson of the Original Mycobacterial Sin (2), i.e. immune response imprinting by Non-Tuberculous Mycobacteria (NTM). An evocative phrase, it’s a riff on the older phrase, Original Antigenic Sin (3, 4). Today, we would conceptualize similar ideas under the rubric,. More narrowly, with reference to T and B cell responses, we would call the phenomenon Different parts of the world harbor different NTM in their environment. Some are prevalent in soil, others in water (5), some even form biofilms in man-made structures such as showerheads and plumbing. The degree and route of our exposure to NTM depends on our geography. How does NTM influence anti-TB immunity?
4. We have under-explored the biology of TB latency. What is the biology of long-lived TB infection in the absence of overt disease (latency)? A third of the human population is presumed infected with TB. Is this accurate? If so, what is different about latent TB infection? How do human genetic polymorphisms and NTM exposure differences influence the ensuing immune responses associated with conversion of latent infection to active disease?
5. Why have we still not predictably characterized protective human immune responses against TB? Why is TB biomarker development languishing? We rely on poorly predictive tests such as the outdated skin-associated(Tuberculin Sensitivity Test; TST) and the irrelevant systemic immune response readouts. Assuredly, the long shadow of , the discoverer of TB, and recent Immunology dogma about the so-called Th1 response ( ) impede our efforts to identify and develop qualified biomarkers for predicting TB infection, TB disease, and (lung) TB control.
6. TB is typically a disease of poverty and attendant malnutrition, ergo the afflicted population has little political traction to dictate better science.
7. Expanded and more stringent (Western) regulatory environment mandates testing a whole range of vaccine candidates (subunit, live attenuated, non-replicating vector-based). Each requires different efficacy assessment and biomarker approaches.
Even Longer Answer (Proceed with caution. Not for the faint-hearted. You will need time and patience from hereon).
If you’re up for it, let’s take a roller coaster ride through TB land.
TB by the numbers
It’s common to see estimates suggesting a third of the human population is infected with TB. What’s the real risk of TB? About 10 to 20% of individuals do not become infected even after sustained high-level exposure, 5% develop clinical TB within 2 years of exposure, and the remaining do not develop overt disease but are considered to have latent TB infection (LTBI; Latency). Among the latter, 5 to 10% develop clinical TB in their lifetime, typically due to some kind of immunodeficiency (e.g. HIV infection, anti-TNF treatment). Thus, in 2012, there were approximately 8.6 million new TB cases and 1.3 million deaths (6).
However, the numbers hide a much more tragic story and as the cliche goes, a picture speaks louder than words.
Today, TB is largely a Third World disease propagated by poverty, its attendant malnutrition, and the unregulated practice of medicine in many poor countries (7, 8, 9).
A journey through the TB genome is to witness a counter-intuitive exercise in minimalism. Here’s an organism, a rarity within its genus, being an obligate pathogen among mainly saprophytic species (34), and yet it harbors fewer, rather than more, genes (35).
At least 6 different TB lineages appear to have co-evolved and adapted to human populations (36, 37, 38, 39). Circulating TB strains could be different in their virulence and induction of immunity (40, 41). Natural human mutations reveal that mutations that impair IFN-g immunity may be important for opportunistic mycobacterial disease control (42). However, natural human mutations have so far yielded no clear insight into human susceptibility for pulmonary TB (6). Typically, persistent bacteria tend to rely on a ploy called Antigenic Variation to evade the host’s immune responses (43). As the term suggests, Antigenic Variation refers to gene sequence modifications, specifically of sequences that are the target of host immunity. This does not appear to happen in TB. In fact, CD4 T cell epitopes, sequences that are the target of a host’s CD4 T cell response, among 21 TB strains showed the opposite, being highly conserved (44). Unfortunately, Immunology easily lends itself to military terminology. In that light, pathogen-host relationships are couched as evolutionary arms races. TB genome analysis runs somewhat counter to this, similar to the case with another intracellular bacterium, Salmonella eneterica serovar Typhii (45). How to explain this? Within the dogma of an arms race, options become severely truncated at this point, and authors postulate that more conserved sequences drive host-beneficial immunity while diverse ones drive pathogen-beneficial immunity (44, 46). What if the selection were elsewhere, not on CD4 T cell epitopes? Maybe on B cell epitopes? Maybe on gamma-delta T cell epitopes? Maybe the scenario is not an arms race so much as an entente between TB and humans? Maybe even a detente?
Latency refers to the capacity for the TB bacterium to remain viable within the host without overt signs and symptoms of active disease. Usually latency is inferred by the . Newer tests such as do not offer substantial improvement (47), and neither test accurately distinguishes latent infection from active disease, nor reactivation of latent infection from reinfection (48). Recently, a blood-derived genetic signature claims to distinguish latent infection from active disease (49). However, this signature overlaps with that of another disease, Sarcoidosis (50), disease of unknown etiology thought to be related to TB.
Vitamins and TB
Vitamin D deficiency increases susceptibility to TB (62). Vitamin C recently showed direct anti-TB bactericidal activity (63). Such studies emphasize the link between poverty, its attendant malnutrition and TB, and suggest simple dietary improvements such as vitamin sufficiency among poverty-stricken populations could greatly ameliorate TB resistance.
A Novel Model upends Century-old TB dogma
Over the past decade or so, among others, Lalita Ramakrishnan at the University of Washington in Seattle has developed a Zebrafish model of mycobacterial infection (Infection with the fish pathogen, Mycobacterium marinum). This model has thus far upended one of the central dogmas of TB. Since the days of , we have believed that the Granuloma, or walling off, response is a host protective response designed to imprison whatever induces the Granuloma, be it indigestible material or bacteria like TB. However, this zebrafish model (74, 75, 76) instead suggests that Mycobacterium can actively sculpt the host’s Granuloma response, ensuring it be populated with precisely those host cells most amenable for bacterial invasion (a particular subset of host macrophages), proliferating within them and using them to spread even as they deliberately evade those endowed with the capacity to kill them.
What is protective anti-TB immunity?
Can it be assessed in the blood? Should it? Why, if the action is taking place in the lung? The following table highlights perils attendant to these differences (51).
Mouse knockout studies place T cells center stage in TB control (52, 53, 54). Why does TB cause mortality? It does not secrete toxins. We deem an exuberant, poorly regulated immune response, i.e. immunopathology, the culprit. This umbrella term encompasses a variety of tissue destructive processes by the host’s own immunity that cause irreversible damage, especially to delicate lung parenchyma in pulmonary TB, eventually leading to death. In humans, circulating blood-derived CD4 T cells from active TB patients but not from asymptomatic latently infected, appear to have a propensity for exuberant anti-TB TNF-a (55). In the past, IFN-g response to the same antigens could not make the same distinction. However, this study has three caveats. One, they examined circulating blood mononuclear cells, two, they stimulated with peptide pools. The latter bypasses complex biological processes (Antigen Processing and Presentation) with the potential for artifacts (56). Three, anti-TB CD4 T cells are sparse, not abundant, in circulation (51). This necessitates overnight stimulation. Such an assay is not a snapshot of in situ immunity. On the other hand, contacts of active TB patients have robust TB-specific responses in their bronchoalveolar lavage (BAL) but not in their peripheral blood (57). Rapid recall responses are also seen in BAL (58). Maybe we should also examine local lung immunity and not just peripheral blood? Why then are such studies rare and not the norm (51)? Is it because examining peripheral blood is much easier even if the ensuing information is less relevant?
Is There A Role for B cells and Antibodies in TB?
TB is an obligate intracellular organism. Dogma dictates B cells and antibodies are irrelevant against it. Yet circulating antibody titers correlate with disease prognosis for other intracellular bugs such as Chlamydia, Leishmania (70) as well as with BCG. BCG also elicits long-lived memory B cells (71) and oral BCG induces an antibody switch from IgG to IgA (72). B cells and antibodies remain controversial topics in TB (73) yet how to explain this phenomenology? Could an effective antibody response prevent TB spreading from one infected cell to another? Could anti-TB IgA prevent TB spreading from one infected individual to another?
How Do Co-Infections Influence TB outcome?
Here too, some data flout dogma. Dogma states that an effective anti-TB response is Th1, i.e. a response dominated by mediators like IFN-g and TNF-a. The same dogma states that certain organisms like the helminthes drive a Th2 response, dominated by mediators like IL-4 and IL-13, with Th1 and Th2 working against each other. Thus, if BCG infected the lung at the same time that the lung was acutely infected with a helminth, the BCG infection should do worse. Yet the opposite happened in a mouse model study (67). Contrary to dogma, T cell responses are much more complex and inherently more plastic than previously envisaged. A single measure such as the capacity of circulating CD4 T cells to make anti-TB IFN-g does not correlate with protection (68). However, neither does expanding from single to multifunctional circulating CD4 T cell capacity (69). Again, maybe we need to comprehensively examine anti-TB immunity in the airways and lung secretions?
is the live, attenuated vaccine approved for TB. The development of this vaccine reads like a adventure novel with the attendant highs and lows. The developers were and . In Lille, France, starting in 1908 with a virulent bacterium (Mycobacterium bovis, not Mycobacterium tuberculosis) isolated from the udder of a tuberculous cow, and donated to them by , they diligently, methodically, and heroically sub-cultured this bacterium every 3 weeks in their special glycerin-potato medium, adding ox bile to the mix when they noticed their bacterial cultures tended to clump, which fortuitously led to a reduction in virulence. Why heroically? Heroic because they continued their sub-cultures unabated through the German occupation of Lille, even through increasing scarcity of potato and ox bile! By 1919, after about 230 subcultures over 11 years, they had in their hands a tubercle bacillus that did not cause TB in guinea pigs, rabbits, cattle, or horses (10). Starting in 1921 and continuing to date, it’s estimated that more than 5 billion people got the BCG vaccine.
The Lubeck disaster
In 1930, in Lubeck, Germany, 250 infants were given oral BCG. 73 developed TB and died within a year (10). How did this happen? Unfortunately, in the Lubeck laboratory, BCG and virulent TB were accidentally cultured in the same incubator with the tragic consequence that the children got vaccinated with TB-contaminated BCG (10, 11).
Lubeck was however only a temporary setback. Resurgence of TB during the 2nd World War restored public confidence in BCG by inducing its massive scale use. Starting in 1974, intradermal BCG vaccination at birth has resulted in an estimated 3 billion cumulative vaccinations worldwide and approximately 100 million vaccinations per year (12, 13). However, BCG efficacy in vaccine trials ranges from 80% in the UK to 0% in South India (59). Why is BCG vaccine efficacy so variable?
Could BCG sub-strains explain differences in vaccine efficacy?
and supplied their BCG to various laboratories around the world. Unfortunately, freeze-drying technology only emerged in the 1960s so early cultures were maintained by repeated sub-cultures, leading to differences over time. Substantial differences have been documented between the different BCG strains (13). Today, we mainly use five BCG strains, Pasteur 1173 P2, Danish 1331, Glaxo 107, Tokyo 172-1 and Russian BCG-1 (11, 13), while Brazil uses Moreau RD (14). BCG sub-strains can be grouped by date:
Group 1: 1921-1925; BCG Moscow, BCG Moreau, BCG Tokyo
Group 2: 1920-1931; BCG Sweden, BCG Birkhaug
Group 3: 1931; BCG Glaxo, BCG Copenhagen (also BCG Danish)
Group 4: 1934; BCG Tice, BCG Connaught
Data suggest that early BCG is more immunogenic (14, 15). A criticism of BCG refers to documented differences and deletions (11, 31, 32) in BCG compared to TB, deletions which house immunodominant TB proteins such as ESAT-6 and CFP-10, purported important in driving anti-TB immunity. However, tweaking BCG to express such TB antigens doesn’t seem to improve upon BCG either (33). On the other hand, in another Sabatini-esque twist, BCG Tice and Connaught are used for bladder cancer treatment in Europe and US (16). We don’t know how it does it but it works better than existing chemotherapy (17, 18). Yet another intriguing mystery in TB land.
Maybe we should consider administering BCG to mimic the natural route of TB infection?
Initially, chose to give BCG orally because he believed infection occurred through that route. Brazil continued oral BCG vaccination until the mid-1970s using BCG Moreau (11). From the beginning, oral BCG vaccination was under pressure from the need to generate an “allergic” skin response as evidence of vaccine “take”, which doesn’t happen with oral BCG. Why the need for this skin response? Here looms the long shadow of , who spent the latter part of his scientific career in the quest for a magic bullet cure for TB. His idea for a cure was a purified extract of TB, , injected into the skin. This “treatment” itself ended in disaster but its legacy survives as the , used till date as a “biomarker” for mycobacterial exposure, for TB infection and for BCG vaccine “take”. This skin response is actually a certain type of immune response called the Delayed Type Hypersensitivity (DTH; ). Today, we know that a BCG-induced DTH does not correlate with protection against TB (19) because the cells and antigens involved in this response are different from those suspected to be involved in protection, and yet, we don’t have a more accurate test for vaccine “take”. Again, maybe we should examine immune responses to TB, to BCG, to novel TB vaccine candidates at more relevant sites, say in the lung airways and in lung secretions, perhaps?
TB is an airborne pathogen. A remarkable series of experiments (20) in the 1950s proved that TB spreads through Infectious Droplet Nuclei (coarse respiratory droplets expelled by an infected person in an enclosed space and breathed in by another). They housed colonies of healthy guinea pigs in a separate room on a separate floor in a hospital ward housing highly infectious and very sick TB patients. The room housing these uninfected guinea pigs received its air supply directly from the hospital ward housing the sick TB patients. Most of these guinea pigs developed TB but not if the air from the sick patients’ room was pre-treated with ultraviolet light that killed the TB bacilli. What is the point of this anecdote? It emphasizes we get TB from breathing in infected air. Yet, BCG vaccination does not mimic the route of TB infection. Why? Would BCG work better as an aerosol or orally? I am not alone in posing this question (21). Since BCG is a live, attenuated bacterium, TB aficionados would warn that a potential danger of oral or any other mucosal route targeting the upper airways is risk for cervical adenitis. However, this danger is well allayed by the fact that no cervical adenitis cases were ever reported from Brazil where BCG Moreau was given orally to newborns from 1945 to 1977 (11, 19). In fact, aerosol BCG protects monkeys from airborne TB (22) and aerosol BCG can even be safely given to humans (23). Even though these latter groundbreaking observations date back 40 years or more, we have apparently made little subsequent effort to capitalize on them.
Mucosal vaccination has several advantages for TB:
1. Easy to administer.
2. More practical since it does not need trained personnel, and there’s no risk of contaminated needles.
3. Easier for people so it would engender greater compliance.
4. Easier to produce. Injectables need to be highly pure and need to be especially free of microbial products such as endotoxin. On the other hand, mucosal surfaces already abundantly harbor microbes. Thus, mucosal vaccines need not be highly pure. This leads to enormous cost savings in manufacturing and quality control.
5. Mucosal vaccination induces both mucosal and systemic immunity (24). On the other hand, systemic immunization such as intramuscular and subcutaneous, do not trigger robust mucosal immunity (25). Mucosal vaccination also offers many choices such as oral, intranasal, pulmonary, rectal and vaginal (11, 26).
Ancillary advantages of BCG
BCG and Trained Immunity
BCG appears to improve immunity against other microbes such as Staphylococcus aureus, Candida albicans, and unrelated neonatal and infant vaccines such as Hepatitis B vaccine, Polio Vaccine, Tetanus Toxoid (28, 29).
BCG can be given at birth
BCG is one of a handful of vaccines given at birth, others being Hepatitis B and Polio. This is an enormous advantage because in large parts of the world, birth is the single most reliable point of contact of an individual with the public health system. Inability to vaccinate at birth is a huge missed opportunity for many vaccine preventable diseases. Such is obviously not the case with TB. Babies vaccinated at birth with BCG make robust, polyfunctional anti-TB CD4 T cell responses (30). So much for the idea that babies have an immature immune system!
NTM (Non-Tuberculous Mycobacteria) or EM (Environmental Mycobacteria)
There are at least 55 species of NTM (34). The largest human trial for BCG in Chingleput, South India, showed 0% efficacy against adult pulmonary TB (59). Local prevalence and ensuing early life exposure to NTM were largely implicated in these results (60). Since then, prior exposure to NTM has been extensively explored (61), as an immunomodulator, either blocking (preventing) or masking (doing the job itself) BCG. NTM are differentially distributed around the world, being more prevalent closer to the equator (60). However, even after decades of research on this topic, we have more questions than answers about the role of NTM in anti-TB immunity.
1. Which NTM species are important for protective anti-TB immunity?
2. Exposure how? Oral, nasal, other?
3. Starting when? At birth?
4. Do different NTM species and/or routes imprint different types of anti-TB immunity? Are some imprints beneficial, others harmful? How does nutritional status influence this imprinting process?
Here is another intriguing tale from TB land. In the 1970s, John Stanford, a British microbiologist, traveled to Uganda, where BCG proved quite effective. There, he isolated from the “hippo” mud on the shores of Lake Kyoga, an NTM he named Mycobacterium vaccae (). Testing suggested it could cure autoimmune disease, no less! In 1994, with his student, Graham Rook, he established a biotech company, Stanford Rook, to capitalize on this fortuitous finding. However, the product failed initial test as a TB drug ( ) and share prices plummeted. This product or others like it may yet succeed in re-calibrating immunity gone haywire. In the meantime, their early promise as BCG replacements has faded.
How Does Microbiota influence TB and vice-versa?
Studies are just beginning to map this terrain (64, 65, 66).