(C) PLOS One
This story was originally published by PLOS One and is unaltered.
 . . . . . . . . . .



Tuberculosis caused by Mycobacterium africanum: Knowns and unknowns [1]

['Marta L. Silva', '- Instituto De Investigação E Inovação Em Saúde', 'University Of Porto', 'Porto', 'Ibmc - Instituto De Biologia Molecular E Celular', 'Doctoral Program In Molecular', 'Cell Biology', 'Icbas - Instituto De Ciências Biomédicas Abel Salazar', 'Baltazar Cá', 'Inasa - Instituto Nacional De Saúde Pública Da Guiné-Bissau']

Date: 2022-08 

Abstract Tuberculosis (TB), one of the deadliest threats to human health, is mainly caused by 2 highly related and human-adapted bacteria broadly known as Mycobacterium tuberculosis and Mycobacterium africanum. Whereas M. tuberculosis is widely spread, M. africanum is restricted to West Africa, where it remains a significant cause of tuberculosis. Although several differences have been identified between these 2 pathogens, M. africanum remains a lot less studied than M. tuberculosis. Here, we discuss the genetic, phenotypic, and clinical similarities and differences between strains of M. tuberculosis and M. africanum. We also discuss our current knowledge on the immune response to M. africanum and how it possibly articulates with distinct disease progression and with the geographical restriction attributed to this pathogen. Understanding the functional impact of the diversity existing in TB-causing bacteria, as well as incorporating this diversity in TB research, will contribute to the development of better, more specific approaches to tackle TB.

Citation: Silva ML, Cá B, Osório NS, Rodrigues PNS, Maceiras AR, Saraiva M (2022) Tuberculosis caused by Mycobacterium africanum: Knowns and unknowns. PLoS Pathog 18(5): e1010490. https://doi.org/10.1371/journal.ppat.1010490 Editor: Antje Blumenthal, University of Queensland, AUSTRALIA Published: May 26, 2022 Copyright: © 2022 Silva et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Funding: This work was financed by the FCT-Aga Khan Development Network grant #333197025 to MS. MLS is funded by FCT PhD scholarship 2020.05061.BD; MS is funded by FCT Estímulo Individual ao Emprego Científico CEECIND/00241/2017. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing interests: The authors have declared that no competing interests exist.

TB caused by M. africanum versus M. tuberculosis Several studies have been performed with the aim of unveiling associations between clinical and epidemiologic data and the infecting bacteria, i.e., M. tuberculosis or M. africanum. It is important to note that these studies have been performed in different countries, at different times, and using different methodologies. Therefore, multicentric studies are still in need, and it is somehow not surprising that some discrepant results are seen across different reports, as further discussed below. No marked differences were found in the chest X-ray presentation of TB caused by M. africanum or M. tuberculosis [60]. Furthermore, both pathogens were shown to respond similarly to the standard 4 first-line drugs in TB treatment, although patients diseased with M. tuberculosis L4 strains responded faster to TB treatment than those with M. africanum L6 strains [61]. The slow clinical recovery of M. africanum-infected patients as compared to M. tuberculosis-infected ones may result from a higher content of persister-like M. africanum bacilli in sputum at diagnosis [62]. Despite the similar clinical presentation of TB, several studies associated infections with M. africanum strains with more vulnerable hosts. Studies conducted in The Gambia found M. africanum infections to be more common in HIV-coinfected patients, as well as in older individuals and individuals presenting severe malnutrition [60]. The association between M. africanum infection and elder patients was also reported in Ghana [63] and with lower body mass index individuals in Mali [64]. However, a clear association between M. africanum and HIV coinfection is still controversial. A recent study in Ghana showed no significant differences between the prevalence of M. tuberculosis or M. africanum infections in individuals with diabetes, another important comorbidity in TB [65]. Furthermore, it is possible that M. africanum infections correlate with slower progression to active TB. A study from The Gambia showed that despite similar rates of transmission, individuals exposed to M. tuberculosis strains were more likely to progress to active TB disease than those infected with M. africanum ones [66]. This was supported by another study in Mali associating a longer time between symptom onset and TB diagnosis in M. africanum infections [64]. In line with these findings, infection of mouse models with M. africanum strains showed a slower progression of the disease [51,67,68] with mild lung pathology even in mice lacking IFN-γ, which are highly susceptible to M. tuberculosis infection [51]. Of note, whether transmission rates are equivalent between M. tuberculosis and M. africanum is not fully set, as a study in Ghana associated M. africanum strains with reduced recent transmission rates [69]. Importantly, whereas in The Gambia, the prevalent M. africanum lineage is L6 [66], in Ghana, it is L5 [69], and so differences in transmission rates may reflect the specific characteristics of L5 or L6 strains. Thus, all in all, M. africanum strains present several differences when compared to M. tuberculosis ones (Table 1) and is generally viewed as a less virulent pathogen than M. tuberculosis. As mentioned before, it is as yet not possible to establish comparisons between the recently identified L9 strains and those of L5/L6 or M. tuberculosis.

Host immune responses to M. africanum Innate immune responses Infection of human monocyte-derived macrophages with distinct strains of the MTBC showed variation of the induced cytokine responses including between the 2 isolates of M. africanum tested, with 1 inducing strong cytokine release and another inducing a weak response [70]. Interestingly, in the same study, both M. africanum isolates seemed to grow less inside resting macrophages than their M. tuberculosis counterparts [70]. Another study, focusing on the pathogen transcriptional adaptation upon macrophage infection, reported distinct MTBC lineage signatures, including the failure of M. africanum strains to induce the phthiocerol dimycocerosate (PDIM) locus, a complex cell wall lipid unique to mycobacteria associated with its virulence [71]. More recently, an isolate of M. africanum L6 was shown to induce considerably less IFN-β by infected bone marrow–derived macrophages than M. tuberculosis strains from L2 or L4 [72]. Although the M. africanum isolate also triggered cGAS and STING, infections of macrophages by this pathogen induced less mitochondrial stress, thus decreased production of mitochondrial reactive oxygen species that contributed to less type I IFN being produced [72]. The in vivo effect of IFN-αβ signalling during infection by M. africanum strains was subsequently studied in mouse models. In agreement with the detrimental role of type I IFN in TB [73], lack of type I IFN receptor signaling led to reduced lung bacterial burdens and less severe histopathological findings upon M. africanum infection [67]. These results highlight that even the lowest levels of IFN-αβ induced during chronic M. africanum infection are potentially pathogenic [67]. Collectively, these studies and others [51] demonstrate that M. africanum strains infect macrophages, inducing a cytokine response, while adapting to the host cell. The molecular mechanisms underlying these responses, such as the recognition of M. africanum strains by pattern recognition receptors, remain however elusive. M. africanum strains were shown to bind recombinant human mannose-binding lectin (MBL), a plasma opsonin, more efficiently than M. tuberculosis strains and a protective association between TB and the human MBL2 G57E variant, associated with lower MBL levels, was described, only in TB caused by M. africanum [74]. It is possible that the stronger binding of M. africanum strains to MBL may favour the bacteria uptake by macrophages, promoting the establishment of infection in vivo, and thus the protective MBL deficiency may have been selected in the human population in regions endemic for M. africanum. Another study has identified increased levels of TLR9 expression in unstimulated blood of patients infected with M. africanum isolates as compared to other MTBC strains infections [75], which may suggest a role for TLR9 in innate immune responses to M. africanum strains. Of note, the levels of IL-12p70 and IL12A were also significantly higher in M. africanum-infected patients, while those of IL-15, IL8, and MIP-1α were higher in M. tuberculosis-infected patients [75]. A broader study comparing peripheral blood gene expression profiles between M. africanum- and M. tuberculosis-infected patients showed no differences at diagnosis, although there were distinct signatures associated with each infection posttreatment, predominantly associated with immune responses and metabolic diseases [76]. Adaptive immune responses Comparison of T cell responses from M. tuberculosis- or M. africanum-infected TB patients before chemotherapy and following overnight stimulation of whole blood with ESAT-6/CFP-10 or with purified protein derivative (PPD) showed higher single-TNF-α-producing CD4 and CD8 T cells and lower single-IL-2-producing T cells in the case of M. africanum infections [77]. Additionally, a persistently high proportion of activated T cells was reported in M. africanum-infected individuals posttreatment [77]. However, the frequencies of PPD-specific polyfunctional CD4 T cells did not differ between the 2 infections [77], both before and after treatment, suggesting an overall uniform immune response triggered by either pathogen. This is in line with the abovementioned studies on peripheral blood transcriptomic and metabolic profiles obtained at diagnosis [76]. Interestingly, stimulation of whole blood with ESAT-6/CFP-10 stimulation after treatment induced significantly higher production of pro-inflammatory markers, such as IFN-γ, in the case of M. tuberculosis-infected TB patients [75,76]. In the mouse model of infection, a modest immune response has been reported upon infection with a M. africanum isolate, also associated with restricted lung pathology [51]. Taken all this together, it is possible that a lower immune response takes place upon M. africanum infection, which although precluding the clearance of the pathogen, may protect the host from tissue immune pathology. This hypothesis is compatible with a slower progressing infection and may be explained by pathogen-associated factors. Pathogens belonging to the MTBC are known to have remarkably hyperconserved T cell epitopes, suggesting that ensuring T cell responses is more important to these agents than evading them [78]. Interestingly, a study showed that the L6 strains of M. africanum were significantly more genetically diverse than the L5 ones, including in predicted T cell epitopes [79]. Additionally, even though the majority of the T cell epitopes were conserved between the 2 lineages, a higher ratio of nonsynonymous to synonymous single nucleotide variation was detected in the epitopes from L6 strains relatively to L5 ones [39]. Thus, it is tempting to speculate that the evolutive pressure to hyperconserve T cell epitopes may be weaker in the case of L6 strains, leading to lower T cell responses and favouring the persistence of the pathogen in its host population. Further studies are required to address these hypotheses and investigate the contribution of T cell responses to TB caused by M. africanum strains.

Geographic restriction of M. africanum: A case of immune adaptation? A specific adaptation of M. africanum to the host population, particularly to the host immune response, is a conceivable hypothesis to explain the geographic restriction of this pathogen. Previous studies provide compelling evidence towards this hypothesis in the case of M. africanum L5 strains. In a study in Ghana, M. africanum was significantly more common in TB patients belonging to the Ewe ethnic group an association that was mainly driven by L5 strains [80]. Possible interactions between M. africanum infection and human genetic diversity were also described in other studies. A polymorphism in the exonic allele (g.760A) of the ALOX5 gene (which encodes for 5-lipoxygenase, an important regulator of the immune response in TB [81,82]) was associated with higher risk of TB in Ghana, an association that was stronger in infections caused by M. africanum L6 strains [83]. Furthermore, another study identified a highly frequent variant of the human immunity–related GTPase M (IGRM), a regulator of the autophagic process, in the Ghanaian population and associated it with protection against M. tuberculosis L4 strains, but not against M. africanum isolates [84]. Thus, higher frequencies of genetic variants conferring increased susceptibility to M. africanum strains in West African individuals may at least partially explain the geographical restriction of this pathogen to this region. Still, more studies linking human and pathogen genetic diversities are needed to validate this hypothesis. In this line, specific HLA genetic associations may be of potential interest to explain the geographic distribution of M. africanum versus M. tuberculosis infections. A study conducted in Mali identified various class I HLA-A and HLA-B alleles associated with active TB disease caused by either pathogen. However, several class II HLA-DR variants were found to be associated with M. tuberculosis but not M. africanum strains, with only the variant DRB1*03:01 being associated with both groups [85]. It is tempting to speculate that specific associations between HLA variants and M. africanum strains may reflect variations affecting T cell epitopes in M. africanum, which as described above are not as hyperconserved as in M. tuberculosis. More recently, the hypothesis that differences in the intestinal microbiota of patients infected with M. africanum isolates could contribute to the high susceptibility of West African individuals to infections with this pathogen has been proposed [86]. Patients infected with M. africanum strains presented less microbiome diversity than individuals infected with M. tuberculosis isolates or healthy controls and were enriched in bacteria from the Enterobacteriaceae phylum Proteobacteria as compared to healthy controls [86]. Since a positive correlation between the abundance of Enterobacteriaceae and an inflammatory gene expression profile was reported, differences in the intestinal microbiome may contribute as host-associated factors predisposing to infections by M. africanum. Other 2 hypotheses may explain the geographic restriction of M. africanum infections, which are less related to the host immune response. It is possible that M. africanum is an attenuated member of the human-adapted TB-causing bacteria, being therefore outcompeted by M. tuberculosis. This hypothesis is supported by the reduction of TB cases caused by M. africanum strains over time for some countries [21–23], although it is not observed in other countries [17,23] as discussed above. Finally, M. africanum L6 strains share a common ancestor with animal (nonhuman) adapted strains (Fig 1). Animal-adapted lineages are composed of Mycobacteria that infect different species of animals as preferential hosts, including nonhuman primates and other mammals. The ancestry of M. africanum L6 allows raising the hypothesis that, despite having become a human pathogen, strains of this lineage may still be adapted to an animal that could function as a reservoir in West Africa. Although M. africanum strains have been isolated from several animal species, including pigs and cows [87], an animal reservoir has never been identified. Of note, the animal reservoir hypothesis is less likely to prove valid in the case of L5 strains since this lineage is phylogenetically less related to animal adapted members of MTBC [41] (Fig 1).

Conclusions Up to 50% of the TB cases in West Africa have been attributed to M. africanum strains. A striking feature characterizing these TB-causing bacteria is its geographical restriction, which contrasts with the widespread distribution of M. tuberculosis strains and remains largely unexplained. Adaptation of M. africanum strains to the West African population, perhaps mediated through differential modulation of the immune response, is a likely hypothesis. Importantly, infections of humans and experimental models with strains of M. africanum are generally more attenuated than those with M. tuberculosis strains. This offers an opportunity to learn from M. africanum and its interactions with the host, with the aim of better controlling M. tuberculosis. There are several outstanding questions that would advance our knowledge in this field towards better strategies to tackle TB: What is the actual origin of M. africanum? Are there nonhuman reservoirs relevant in supporting human transmission? This would provide important information on the evolution of specific members of the MTBC, as well as potentially guide measures to mitigate M. africanum infections. Are M. tuberculosis and M. africanum distinct entities? This remains a matter of debate, as although recent studies suggest that both pathogens belong to the same species, phenotypic differences between M. tuberculosis and M. africanum strains are well documented. Additional studies are required to fill this knowledge gap further informing similarities and particularities of different MTBC members of relevance for TB management. Are there coinfections caused by M. africanum and M. tuberculosis strains? The fact that both pathogens are endemic in the same geographic area would suggest a scenario where coinfections are possible. Clarifying this question would be interesting to understand which pathogen would impact more disease presentation or if different clinical/immune response characteristics would prevail. What are the differences between TB and the immune response during infection with L5, L6, and L9 strains? How does this correlate with M. tuberculosis infections? Does M. africanum modulate innate or T cell responses in specific populations? Is this associated with increased susceptibility of the population or decreased virulence of the pathogen? Disclosing the immune signatures of M. africanum infections and correlating those with the disease manifestation will provide valuable knowledge to develop potential immune interventions in TB, including vaccines. Are there differences in latency establishment, duration, or reactivation between M. tuberculosis and M. africanum strains? Elucidating this question is hampered by our inability to categorize the pathogen species in latent cases. However, by revealing immune signatures of M. africanum infections, it may be possible to then look at latent and progressing cohorts in an attempt to establish latent signatures specific of the different MTBC members. This will further our knowledge on the natural history of TB, again potentially offering novel targets to control TB. Answering these questions to understand the differences between M. africanum and M. tuberculosis strains will provide valuable knowledge towards identifying the cellular and molecular determinants allowing the widespread of the M. tuberculosis strain lineages, which are a global threat. This knowledge will also advance our understanding on the biology of M. africanum and its interactions with the human host, which is highly relevant considering the TB burden in West Africa. Furthermore, consistently stratifying for the type of infecting bacteria in human-based studies will contribute to a better interpretation of novel TB intervention tools, including diagnosis and vaccines. All this will in turn inform the development of better, more specific approaches to tackle TB.

[END]
---
[1] Url: https://journals.plos.org/plospathogens/article?id=10.1371/journal.ppat.1010490

Published and (C) by PLOS One
Content appears here under this condition or license: Creative Commons - Attribution BY 4.0.

via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/