Mycobacterium tuberculosis complex and human coadaptation: a two-way street complicating host susceptibility to TB

Abstract

For centuries, the Mycobacterium tuberculosis complex (MTBC) has infected numerous populations, both human and non-human, causing symptomatic tuberculosis (TB) in some hosts. Research investigating the MTBC and how it has evolved with its host over time is sparse and has not resulted in many significant findings. There are even fewer studies investigating adaptation of the human host susceptibility to TB and these have largely focused on genome-wide association and candidate gene association studies. However, results emanating from these association studies are rarely replicated and appear to be population specific. It is, therefore, necessary to relook at the approach taken to investigate the relationship between the MTBC and the human host. Understanding that the evolution of the pathogen is coupled to the evolution of the host might be the missing link needed to effectively investigate their relationship. We hypothesize that this knowledge will bolster future efforts in combating the disease.

Introduction

To this day, tuberculosis (TB) plagues human populations, causing 1.5 million deaths worldwide per year (1). In South Africa, TB is the leading cause of death due to a single infectious agent (1,2). TB is further compounded by comorbidities such as HIV, diabetes and other non-communicable diseases. Understanding the various factors influencing TB progression and outcome is vital in efficiently combating the disease.

One of these factors is the coadaptation of the host and pathogen. Coadaptation in this context refers to the adaption of the host and pathogen simultaneously and sequentially, in a way that one attempts to overcome the other. This evolutionary process is a hallmark of most organisms that interact and governs the resulting symbiotic relationship.

Here we review research focusing on the adaptation of the host and the mycobacteria that bring about TB. We discuss the history of the disease and how the causative pathogen has evolved, in addition to how the human host has developed resistance (or in contrast, heightened susceptibility) to this pathogen over time. This review, therefore, aims to unravel a complex tangle of knowledge on a two-way street that is TB adaptation.

The History of the MTBC and TB in Africa

The main causative agents of TB in humans are phylogenetically similar bacilli called Mycobacterium tuberculosis (M.tb), Mycobacterium africanum (M.afr) and Mycobacterium bovis (M.bovis). All of these bacilli are part of the larger Mycobacterium tuberculosis complex (MTBC). Various lines of evidence have been used to study the origin of the MTBC and TB as a disease, including genomics, archaeology and paleopathology.

The progenitor species from which the MTBC evolved (Mycobacterium prototuberculosis) has been estimated by comparative genomics to be as old as 3 million years (3). It is hypothesized that these tubercle bacilli caused TB in early East African hominids and has been coevolving with their human host ever since (3). It, therefore, appears as though TB has been afflicting human populations for hundreds of thousands of years.

Based on synonymous substitution rates, genetic studies at the start of this millennium suggested that members of the MTBC differentiated approximately 35 000 years ago (4). With the emergence of whole-genome sequencing technologies, this has been pushed back to between 40 000 and 70 000 years ago using coalescent analyses (5). This coincides with the dates of the Out-of-Africa migration and fits into the hypothesis of an African origin for M.tb, M.afr and TB overall (6,7).

Archaeological remains have also been informative in the hunt for the origin of the MTBC. This evidence is scattered across the world and is in the form of human skeletal remains, animal remains, written documents and artworks. Pott’s lesions are a skeletal anomaly characteristic of TB. The earliest evidence of these lesions can be seen in remains from Israel (7250–6160 bce) (8). In Africa, the earliest evidence of Pott’s lesions has been found on Egyptian mummies dating back to 3400 bce (9,10). Spoligotyping of samples from these remains suggested the presence of M.bovis and M.tb (9,10). M.afr was identified in samples from the same area which was dated to 2050–500 bce (11,12). Although biological evidence of TB was found in ancient Egypt, no mention of TB or its characteristic lesions were found in Egyptian papyri. The oldest written documents that do mention TB were found in India (dated to 3300 years ago) and China (dated to 2300 years ago) (13,14).

The establishment of various migratory routes in to and within Africa ushered in foreign pathogens such as the introduction of M.bovis by pastoralists migrating from East to South Africa (15), and potentially more virulent strains of M.tb previously not found in the area (16). This can be clearly seen upon European colonization in the 17th century whereby the hyper-susceptible indigenous populations suffered an immense loss of life in stark contrast to the Europeans that appeared to largely be resistant to TB (17,18). There are some caveats to this hypothesis that to this day remain unexplored, namely, the contribution of other infectious diseases during that time, the presence (if any) and strain of M.tb prior to colonization and the lack of detailed information regarding the lineage/strain introduced by the Europeans.

Presently, M.tb lineages still show some association with specific geographical regions (19). M.tb lineages 2 (including the Beijing strain which is widespread throughout East Asia), lineage 3 (found largely in East Africa), lineage 4 (termed the ‘Euro-American’ lineage, including the Harlem strain) and lineage 7 (found largely in Ethiopia) are found in most parts of Africa along with another MTC member M.afr (part of lineage 5 and 6) (19). Interestingly, M.afr accounts for the majority of TB cases in West Africa and is largely absent from other geographical locations (19). Africa houses the largest diversity of M.tb lineages, reminiscent of the largest host genetic diversity which underlines the potential for host-pathogen coadaptation.

As indicated by the distribution of lineages worldwide as well as the ability for a host to be highly susceptible to a specific lineage, there appears to be a direct association between the pathogen and the human host. This relationship supports the concept of coadaptation between the pathogen and host; understanding this relationship as well as other independent factors is crucial in combating this disease on the African continent.

Human Host Susceptibility to TB

There are various factors that play a role in a host’s susceptibility to TB. Variation in the host genome plays a definitive role, and the heritability of TB susceptibility has been shown to be substantial (between 40 and 70%) (20–22), although recent studies suggest that these initial estimates lack generalizability across populations (21,23–27). Genetic regions potentially associated with the phenotype have been identified using candidate gene and genome-wide association studies (28–34). These regions house genes linked to both the innate and adaptive immunity suggesting that they play a collective role in the host defense against TB infection and progression from latent TB to symptomatic; the outcome, severity and localization of which largely depends on the interaction between the host, pathogen and their respective environments.

Not all individuals progress to disease after exposure to and infection with M.tb. At present, two tests—each with their own limitations—are used to infer infection (35–37), namely the interferon-gamma release assay (IGRA) and tuberculin skin test (TST). Some individuals test persistently negative when using these tests (so-called ‘resisters’) despite antibody evidence of M.tb exposure (38,39), but show no evidence of clinical disease and there appears to be a genetic component involved (40). Genetic loci linked to resistance to TST conversion include loci on chromosomes 11p14, 2q21-2q24, 5p13-5q22 and 1q32 (27,41–44). Apart from the region on chromosome 11, these regions appear to be population specific. Exposure to M.tb can also result in what is termed latent infection i.e. infection has occurred, but the patient is not exhibiting clinical signs and symptoms of TB, although having a positive TST or IGRA. A subset of these latently infected individuals will progress to active, symptomatic disease—usually within two years after infection (45). Whether or not a host will progress to active disease or remain latently infected appears to be governed by numerous loci with intermediate effect sizes (28–33,46). Genetic regions associated with increased or decreased susceptibility have been found in thousands of genes across a wide variety of ethnic populations. As with variants associated with the resister phenotype, most of the variants associated with TB susceptibility have largely remained unreplicated. This is, however, not the case for variants in chromosome 20q13 and chromosome 11p13; both of these regions are associated with an increased susceptibility to TB, in populations that share a recent genetic history (47–50). These results collectively suggest that the TB susceptibility phenotype is governed by population-specific genomic variation. This can be directly seen in studies investigating genomic variation in African populations and how this leads to a stronger inflammatory immune response (51–53). Multiple loci in the human leukocyte antigen (HLA) region have been associated with TB susceptibility (and even pathogen strain specificity i.e. M.afr versus M.tb infection) across populations, yet also exhibit distinctive population structure reminiscent of local selection and/or genetic drift (49,54–58). This is discussed in more detail in sections that follow.

Table 1

Lineage-dependent genetic associations with human TB

Location of study population	M.tb lineage	Study design	Phenotype	Sample number	Gene	Variant	Odds ratio [95% CI]	Reference
Vietnam	Beijing	Candidate gene	PTB, TBM	424 cases	TLR2	T597C	1.57 [1.15–2.15]	(88)
Uganda	L4-Ugandan	GWAS	Disease severity	113 cases, 121 controls	SLC11A1	rs17235409	2.63 [1.25–4.00]	(86)
Indonesia	Beijing	Candidate gene	PTB	336 cases	SLC11A1	rs17235409	2.15 [1.25–3.70]	(89)
						rs17235416	2.4 [1.19–4.83]
South Africa	Beijing, LAM, LCC, Quebec	Candidate gene	PTB	300 cases	HLA-A, -B and -C	A*01	1.58 [1.04–2.40]	(90)
						A*03	1.65 [1.08–2.54]
						B*07	0.49 [0.25–0.89]
						B*08	2.32 [1.02–5.13]
						B*27	0.35 [0.16–0.68]
						B*44	2.07 [1.22–3.52]
						B*58	2.69 [1.27–5.75]
						C1	0.60 [0.40–0.89]
						C2	2.03 [1.35–3.08]
Vietnam	Beijing	Candidate gene	PTB	370 cases	MARCO	rs2278589	1.7 [1.20–2.30]	(91)
						rs6751745	1.5 [1.10–2,10]
Thailand	Non-Beijing lineage	GWAS	Old age onset of PTB	686 cases	CD53	rs1418425	1.62 [1.35–1.93]	(92)
Ghana	M.afr West Africa-2	Candidate gene	PTB	1916 cases	ALOX5	g.760G/A	1.7 [1.20–2.50]	(93)
Ghana	BeijingCAS superclade	GWAS	PTB	3311 cases	PDZRN4	rs41472447	2.56 [1.48–4.41]	(85)
	Ghana superclade						2.94 [1.71–5.08]
	T_U superclade						0.59 [0.44–0.79]

Studies investigating the functional consequences of genomic variation on host susceptibility to TB largely focus on transcriptomics, microRNAs and epigenetics (59). Transcriptional profiles of resisters compared to TST positive individuals identified pathways controlled by histone deacetylase as a possible mechanism for resistance to TST conversion and active, symptomatic disease (60). The conversion of latent to active TB has been investigated using RNA-sequencing of latently infected controls and active cases. Numerous signatures governing disease conversion were identified, many of which are involved in the regulation of interferon-gamma (61–65). Until recently, the role of epigenetics in host susceptibility to TB was largely unknown (66). Associations between histone modifications and mTOR dependent regulation of glucose and glutamine metabolism has been identified in monocytes and macrophages trained with bacilli Calmette–Guerin (67–70). Upon M.tb infection, DNA methylation regulates the reprogramming of innate immune cells as well as the overall regulation of transcription pathways (71,72). MicroRNA’s play diverse roles in TB pathogenesis and are able to differentiate between latent and active TB as well as other microbial infections (73–80) and appear to modify the TB susceptibility phenotype (81–84).

These associations may reflect distinct evolutionary processes that occurred over time indicative of the unique environment the host found themselves in. However, as with most infectious diseases, there is another side to the war; the pathogen itself.

Evidence for M.tb Evolution and Interaction With the Human Host

As previously discussed, the M.tb lineages, and thus bacterial genetic diversity, exhibit population structure correlating with geographical boundaries and thus host population diversity (19). This is evident from recent studies where interactions between the two genomes were associated with TB susceptibility and TB severity (Table 1). More specifically, these interactions were found to be between specific M.tb lineages and variants in the host’s genome, particularly in the HLA region, IL12B, SLC11A1, PDZRN4, STARD4 and RARA genes (85,86). Although we clearly should not discount the role of the evolution of the host in this scenario, it would also suggest that there are associated molecular changes within the bacteria that correlate with lineage prevalence and distribution.

MTBC members are genetically similar (~99%), the 1% differentiation that enables classification of mycobacterial lineages can explain the observed distribution and prevalence (87). The most likely genomic area that would selectively evolve in a pathogen would be genes encoding antigens. These adaptations aim to evade host immunity and are a hallmark of most pathogens, particularly bacterial. However, in the case of M.tb, the human T-cell epitope region is hyper conserved, with 95% of the total epitopes remained unchanged (94–96). While this may be surprising, researchers argue that this might be of benefit to the bacteria in that they are recognized by the host and the ensuing immune response results in tissue destruction and cavity formation, which increases the potential for transmission (97). This, however, is not mutually exclusive from evolution and possible genomic variation in other parts of the M.tb genome.

Indeed, although there is minimal genetic diversity within T-cell epitopes, there is genetic diversity between M.tb lineages in other parts of the genome reminiscent of local adaptation (16,87,98). The modern strains of M.tb appear to be more virulent and successful compared with ancient strains and this can largely be attributed to host population expansion, migration and resulting transmission (99–102). Since TB virulence is correlated with the transmission, the theory suggests that increased virulence will be favored, in addition to a shorter latency period (99–102). The characteristic latency period of TB allows the pathogen to infect a greater number of hosts without decimating the population and adversely affecting its chances for transmission. As the number of hosts increases and densifies, and transmission is more likely, M.tb can adapt by becoming more virulence.

Another phenotypic adaptation by M.tb is the ability of the bacilli to inhibit endosome maturation and phagosome-lysosome fusion (103). This adaptation enables the bacilli to survive and even multiply within macrophages leading to a larger bacterial load. Furthermore, since M.tb lacks the usual virulence factors (toxins), it has adapted its immune-escaping mechanisms via the modulation of lipid metabolism and various transport proteins, including those that inhibit the antimicrobial effectors of macrophages (104). Macrophages have, therefore, been used to investigate the differential host immune response to MTBC strains and indications are that macrophage infection with more modern strains results in decreased cytokine and chemokine production as compared to more ancient strains (105).

Piecing together M.tb adaptation clearly shows the tendency of modern strains to become highly virulent through various phenotypic strategies as well as the impact human demography has on this evolutionary process.

Conclusion

Recent evidence suggests that the human host and the MTBC have been associated with each other for longer than previously thought, perhaps through the majority of modern human history. As this hypothesis gains traction, it is becoming increasingly plausible that coadaptation between the host and M.tb occurred, with some mechanisms being identified. It is highly likely that the lack of recognition, understanding and inclusion of all factors associated with TB as well as how these factors adapt over time has led to the lack of reproducibility across the fields of host genetics and TB genomics. We, therefore, recommend that researchers should move away from studies that look at the host or the pathogen in isolation and rather incorporate the relationship between the two, or at least study both the genomes involved, as was done by McHenry et al. in the context of disease severity and by Muller et al. in the context of disease progression (85,86). This may in the future lead to novel results that enable TB prevention, increase in treatment efficacy, increase in vaccine efficacy and more importantly, personalized treatment for patients.

Acknowledgements

We thank all the study participants that took part in the studies referenced in this review.

Conflict of Interest Statement: None declared.

Funding

This research was funded (partially or fully) by the South African government through the South African Medical Research Council and the National Research Foundation. C.U. was supported by a fellowship from the Claude Leon Foundation.