Adding insult to injury: Exacerbating TB risk with smoking

Abstract

Inhaled environmental pollutants, most prominently from cigarettes, confer an increased risk of tuberculosis. A recent study published in Cell by Berg, Levitte, et al., (2016) using the zebrafish model of Mycobacterium marinum infection provides new insights into the role of macrophage lysosomal engorgement in compromising host defense against mycobacteria.

Nearly 70 years after the development of effective chemotherapy, tuberculosis remains among the ten leading causes of death in the world, and the leading cause of death due to a single infectious pathogen. In 1952, Selman Waksman was awarded the Nobel Prize in Medicine or Physiology for his discovery of the first effective anti-tuberculosis antibiotic, streptomycin; in his acceptance speech, Waksman predicted that the elimination of tuberculosis was within sight, perhaps only a few years away. However, in that same year, Rene and Jean Dubos wrote a scientific and cultural history of tuberculosis, entitled The White Plague, in which they noted that, because of the importance of what we now call the ‘social determinants of disease,’ measures like antibiotics and vaccination would never, on their own, be enough to eliminate tuberculosis. Despite optimism about new drugs, time has thus far proved them prescient.

Now, new findings published in Cell from Lalita Ramakrishnan’s lab provide important mechanistic insights into the contribution of one such social determinant of TB susceptibility: smoking. The manuscript provides a mechanism by which a seemingly disparate group of conditions, which all share the property of macrophage vacuolar engorgement, impair macrophage-dependent host defense (Berg, Levitte, et al., 2016). In a screen for host mutations that alter zebrafish susceptibility to M. marinum infection, the authors identified a loss of function mutation in snapc1b, which encodes a component of the basal transcriptional apparatus. Phenotypic characterization of snapc1b mutant fish revealed several indications of macrophage dysfunction, including defective macrophage migration and a vacuolated morphology with engorgement from cellular debris. Transcriptional profiling of snapc1b-deficient fish revealed underexpression of the lysosomal hydrolases cathepsin G and L. Re-expression of cathepsin L in the snapc1b mutant background reversed the vacuolated morphology, migration defects, and mycobacterial susceptibility, indicating that these phenotypes are attributable to under-expression of lysosomal hydrolase(s). Surprisingly, engorged macrophages were not intrinsically deficient in restricting mycobacterial growth, suggesting that the susceptibility phenotype did not reflect a bacterial-killing defect. Further, the authors showed that the mycobacterial-susceptibility and macrophage-migration phenotypes are recapitulated in genetic models of other lysosomal storage diseases, and even when macrophages are loaded with inert particles, indicating that macrophage dysfunction is a common accompaniment of lysosomal engorgement of diverse etiologies. Finally, the authors showed that vacuolated alveolar macrophages from smoker’s lungs have similar defects in migration compared to non-vacuolated macrophages from the same smoker.

These findings indicate a potentially widely distributed mechanism by which macrophages become dysfunctional and thereby permissive for mycobacterial infection. The most immediate potential translational impact is in providing a mechanistic underpinning of how social and environmental factors may impact TB susceptibility. One hundred years ago, the social determinants of disease included things like malnutrition and overcrowded living conditions in homes with little sunlight and ventilation. In many parts of the world today, these factors still contribute to the global tuberculosis pandemic. However, additional social factors have been increasingly recognized as playing substantial roles in the stubborn persistence of tuberculosis. One of the most important of these is tobacco use. In recent years, the epidemiologic link between cigarette smoking and tuberculosis has become firmly established in a large number of well-characterized cohorts. Smoking increases the chance that persons exposed to tuberculosis will become infected, that infected persons will develop active disease, and that persons with active disease will have poorer outcomes when treated with otherwise effective antibiotic therapy.

In a recent analysis of data from the National Health and Nutrition Examination Survey (NHANES) in which tobacco use and exposure was confirmed by urinary cotinine measurements, a history of active tobacco use in adults was associated with a more than doubled risk of having latent tuberculosis infection (LTBI), even when controlling for age, gender, socioeconomic status, race, birthplace (U.S. vs. foreign-born), household size, and a history of having ever lived with someone with TB (Lindsay et al., 2014). Among foreign-born persons, passive cigarette smoke exposure was also associated with a significant risk of LTBI; among U.S.-born persons, there was a marked trend in this direction as well. In the PREVENT-TB trial of the Tuberculosis Trials Consortium, in which 8000 patients in North America were randomized to one of two effective regimens for the treatment of latent infection, 22 patients nonetheless ultimately developed active tuberculosis disease (Sterling et al., 2011). Analysis of the risk factors that predicted development of active disease demonstrated that current tobacco use was the strongest single risk factor, increasing risk by nearly five-fold.

Treatment outcomes also appear to be worse for smokers. In a well-characterized cohort of patients in Uganda, Maciel and colleagues found that smokers had a three-fold increased risk of remaining culture positive after two months of antibiotic treatment (Maciel et al., 2013). Similar findings were reported by Leung and colleagues in Hong Kong when they reported on the outcomes of over 16,000 consecutive patients. As in Uganda, smoking predicted a much higher rate of persistent culture positivity despite treatment, and the Hong Kong investigators were additionally able to demonstrate a 60% increase in risk for relapse after treatment completion among current smokers (Leung et al., 2015).

On a global level, the co-occurrence of smoking and tuberculosis is widespread. Just three countries in the world, China, India and Indonesia, account for over 40% of all cases of active tuberculosis each year. Smoking prevalence among men in these countries is 45%, 23% and 57% respectively (Eriksen, et al., 2015). On a global level, Lonnroth estimates that tobacco accounts for 21% of all cases of tuberculosis among adults (Lönnroth et al., 2010).

In the context of these strong epidemiologic observations, the results by Berg, Levitte, et al., (2016) beg several major questions. Berg et al conceptualize the defect in macrophage migration that results from engorgement as conferring susceptibility to initial infection due to an inability of engorged alveolar macrophages to migrate to the infected alveolus. This fits well with the epidemiologic data from NHANES cited above in which smoking conferred susceptibility to initial infection. However, as noted, smoking also confers risk of reactivation from latency, as well as a poor response to antimycobacterial therapy in active disease, two effects that are at present less clearly linked to the macrophage migration defect identified in the new study. It is possible that smoking has additional effects on antimycobacterial immunity that explain the effects on reactivation and/or treatment response not attributable to macrophage engorgement; or it is possible that the macrophage phenotype reported here has unanticipated roles in restricting reactivation from latency and treatment response. Distinguishing these possibilities will require further study but has the potential to illuminate the role for macrophages and granuloma structure in control of latency and treatment response.

Finally, it is tempting to speculate that additional environmental causes of macrophage engorgement, such as particulate air pollution, could confer similar risk to TB. Early epidemiologic evidence linking both indoor and outdoor air pollution to tuberculosis risk was mixed, perhaps because of the difficulty of determining individual exposures, though several recent studies indicate the possibility of a link, and the mechanism described by Berg, Levitte et al. (2016) would seem to be relevant to this as well (Lai et al., 2016; Lin et al., 2014; Smith et al., 2016). This study may stimulate further efforts to examine whether particulate air pollution, independent of smoking, confers risk of TB infection, a finding that would have significant public health implications.

Thus, the paper by Berg, Levitte, and colleagues adds an important mechanistic explanation for the link between tobacco use and tuberculosis previously revealed in epidemiologic studies. This link is undeniably real and underscores the urgency, importance and impact of limiting tobacco (and likely, other particulate matter) exposure as a means of improving tuberculosis control and global health.