First discovered in 1882 byRobert Koch,M. tuberculosis has an unusual, waxy coating on its cell surface primarily due to the presence ofmycolic acid. This coating makes the cells impervious toGram staining, and as a result,M. tuberculosis can appear weakly Gram-positive.[3]Acid-fast stains such asZiehl–Neelsen, orfluorescent stains such asauramine are used instead to identifyM. tuberculosis with a microscope. The physiology ofM. tuberculosis is highlyaerobic and requires high levels of oxygen. Primarily a pathogen of the mammalianrespiratory system, it infects the lungs. The most frequently used diagnostic methods for tuberculosis are thetuberculin skin test,acid-fast stain,culture, andpolymerase chain reaction.[2][4]
Growth ofMycobacterium tuberculosis on Löwenstein-Jensen (A) and Ogawa medium (B), after six weeks at 37°C.
Other bacteria are commonly identified with a microscope by staining them withGram stain. However, the mycolic acid in the cell wall ofM. tuberculosis does not absorb the stain. Instead, acid-fast stains such asZiehl–Neelsen stain, or fluorescent stains such asauramine are used.[4] Cells are curved rod-shaped and are often seen wrapped together, due to the presence of fatty acids in the cell wall that stick together.[11] This appearance is referred to as cording, like strands of cord that make up a rope.[8]M. tuberculosis is characterized in tissue by caseatinggranulomas containingLanghans giant cells, which have a "horseshoe" pattern of nuclei.[citation needed]
M. tuberculosis can be grown in the laboratory. Compared to other commonly studied bacteria,M. tuberculosis has a remarkably slow growth rate, doubling roughly once per day. Commonly usedmedia include liquids such asMiddlebrook 7H9 or 7H12, egg-based solid media such asLowenstein-Jensen, and solid agar-based such asMiddlebrook 7H11 or7H10.[8][12] Visible colonies require several weeks to grow on agar plates.Mycobacteria growth indicator tubes can contain a gel that emits fluorescent light if mycobacteria are grown. It is distinguished from other mycobacteria by its production ofcatalase andniacin.[13] Other tests to confirm its identity includegene probes andMALDI-TOF.[14][15]
Analysis ofMycobacterium tuberculosis viascanning electron microscope shows the bacteria are2.71±1.05 μm in length with an average diameter of0.345±0.029 μm.[16] Theouter membrane and plasma membrane surface areas were measured to be3.04±1.33 µm2 and2.67±1.19 µm2, respectively. The cell, outer membrane, periplasm, plasma membrane, and cytoplasm volumes were0.293±0.113 fl (= μm3),0.006±0.003 fl,0.060±0.021 fl,0.019±0.008 fl, and0.210±0.091 fl, respectively. The average totalribosome number was1672±568 with ribosome density about716.5±171.4/(0.1 fl).[16]
Humans are the only known reservoirs ofM. tuberculosis. A misconception is thatM. tuberculosis can be spread by shaking hands, making contact with toilet seats, sharing food or drink, or sharing toothbrushes. However, major spread is throughair droplets originating from a person who has the disease either coughing, sneezing, speaking, or singing.[18]
When in the lungs,M. tuberculosis isphagocytosed byalveolar macrophages, but they are unable to kill and digest the bacterium. Its cell wall is made ofcord factor glycolipids that inhibit the fusion of thephagosome with thelysosome, which contains a host of antibacterial factors.[19][10]
Specifically,M. tuberculosis blocks the bridging molecule, early endosomal autoantigen 1 (EEA1); however, this blockade does not prevent fusion of vesicles filled with nutrients. In addition, production of the diterpeneisotuberculosinol prevents maturation of the phagosome.[20] The bacteria also evades macrophage-killing by neutralizing reactive nitrogen intermediates.[21] More recently,M. tuberculosis has been shown to secrete and cover itself in 1-tuberculosinyladenosine (1-TbAd), a specialnucleoside that acts as anantacid, allowing it to neutralize pH and induce swelling in lysosomes.[22][23]
InM. tuberculosis infections,PPM1A levels were found to be upregulated, and this, in turn, would impact the normal apoptotic response of macrophages to clear pathogens, as PPM1A is involved in the intrinsic and extrinsic apoptotic pathways. Hence, when PPM1A levels were increased, the expression of it inhibits the two apoptotic pathways.[24] With kinome analysis, theJNK/AP-1 signalling pathway was found to be a downstream effector that PPM1A has a part to play in, and the apoptotic pathway in macrophages are controlled in this manner.[24] As a result of having apoptosis being suppressed, it providesM. tuberculosis with a safe replicative niche, and so the bacteria are able to maintain a latent state for a prolonged time.[25]
Granulomas, organized aggregates of immune cells, are a hallmark feature of tuberculosis infection. Granulomas play dual roles during infection: they regulate the immune response and minimize tissue damage, but also can aid in the expansion of infection.[26][27][28][29][30]
The ability to constructM. tuberculosis mutants and test individual gene products for specific functions has significantly advanced the understanding of itspathogenesis andvirulence factors. Many secreted and exported proteins are known to be important in pathogenesis.[31] For example, one such virulence factor iscord factor (trehalose dimycolate), which serves to increase survival within its host. Resistant strains ofM. tuberculosis have developed resistance to more than one TB drug, due to mutations in their genes. In addition, pre-existing first-line TB drugs such as rifampicin and streptomycin have decreased efficiency in clearingintracellularM. tuberculosis due to their inability to effectively penetrate the macrophage niche.[32]
JNK plays a key role in the control of apoptotic pathways—intrinsic and extrinsic. In addition, it is also found to be a substrate of PPM1A activity,[33] hence the phosphorylation of JNK would cause apoptosis to occur.[34] Since PPM1A levels are elevated duringM. tuberculosis infections, by inhibiting the PPM1A signalling pathways, it could potentially be a therapeutic method to killM. tuberculosis-infected macrophages by restoring its normal apoptotic function in defence of pathogens.[24] By targeting the PPM1A-JNK signalling axis pathway, then, it could eliminateM. tuberculosis-infected macrophages.[24]
The ability to restore macrophage apoptosis toM. tuberculosis-infected ones could improve the current tuberculosis chemotherapy treatment, as TB drugs can gain better access to the bacteria in the niche.[35] thus decreasing the treatment times forM. tuberculosis infections.
Symptoms ofM. tuberculosis include coughing that lasts for more than three weeks,hemoptysis, chest pain when breathing or coughing, weight loss, fatigue, fever, night sweats, chills, and loss of appetite.M. tuberculosis also has the potential of spreading to other parts of the body. This can cause blood in urine if the kidneys are affected, and back pain if the spine is affected.[36]
Typing of strains is useful in the investigation of tuberculosis outbreaks, because it gives the investigator evidence for or against transmission from person to person. Consider the situation where person A has tuberculosis and believes he acquired it from person B. If the bacteria isolated from each person belong to different types, then transmission from B to A is definitively disproven; however, if the bacteria are the same strain, then this supports (but does not definitively prove) the hypothesis that B infected A.[citation needed]
Until the early 2000s,M. tuberculosis strains were typed bypulsed field gel electrophoresis.[37] This has now been superseded byvariable numbers of tandem repeats (VNTR), which is technically easier to perform and allows better discrimination between strains. This method makes use of the presence of repeatedDNA sequences within theM. tuberculosis genome.[citation needed]
Three generations of VNTR typing forM. tuberculosis are noted. The first scheme, called exact tandem repeat, used only five loci,[38] but the resolution afforded by these five loci was not as good as PFGE. The second scheme, called mycobacterial interspersed repetitive unit, had discrimination as good as PFGE.[39][40] The third generation (mycobacterial interspersed repetitive unit – 2) added a further nine loci to bring the total to 24. This provides a degree of resolution greater than PFGE and is currently the standard for typingM. tuberculosis.[41] However, with regard to archaeological remains, additional evidence may be required because of possible contamination from related soil bacteria.[42]
Antibiotic resistance inM. tuberculosis typically occurs due to either the accumulation of mutations in the genes targeted by the antibiotic or a change in titration of the drug.[43]M. tuberculosis is considered to be multidrug-resistant (MDR TB) if it has developed drug resistance to both rifampicin and isoniazid, which are the most important antibiotics used in treatment. Additionally, extensively drug-resistantM. tuberculosis (XDR TB) is characterized by resistance to both isoniazid and rifampin, plus anyfluoroquinolone and at least one of three injectable second-line drugs (i.e.,amikacin,kanamycin, orcapreomycin).[44]
M. tuberculosis (stained red) in tissue (blue)CordingM. tuberculosis (H37Rv strain) culture on the luminescent microscopy
The genome of theH37Rv strain was published in 1998.[5][45] Its size is 4 million base pairs, with 3,959 genes; 40% of these genes have had their function characterized, with possible function postulated for another 44%. Within the genome are also sixpseudogenes.[citation needed]
Fatty acid metabolism. The genome contains 250 genes involved infatty acid metabolism, with 39 of these involved in thepolyketide metabolism generating the waxy coat. Such large numbers of conserved genes show the evolutionary importance of the waxy coat to pathogen survival. Furthermore, experimental studies have since validated the importance of a lipid metabolism for M. tuberculosis, consisting entirely of host-derived lipids such as fats and cholesterol. Bacteria isolated from the lungs of infected mice were shown to preferentially use fatty acids over carbohydrate substrates.[46]M. tuberculosis can also grow on the lipidcholesterol as a sole source of carbon, and genes involved in the cholesterol use pathway(s) have been validated as important during various stages of the infection lifecycle ofM. tuberculosis, especially during the chronic phase of infection when other nutrients are likely not available.[47]
PE/PPE gene families. About 10% of the coding capacity is taken up by thePE/PPE gene families that encode acidic, glycine-rich proteins. These proteins have a conserved N-terminal motif, deletion of which impairs growth in macrophages and granulomas.[48]
Antibiotic resistance genes. In 2013, a study on the genome of several sensitive, ultraresistant, and multiresistantM. tuberculosis strains was made to study antibiotic resistance mechanisms. Results reveal new relationships and drug resistance genes not previously associated and suggest some genes and intergenic regions associated with drug resistance may be involved in the resistance to more than one drug. Noteworthy is the role of the intergenic regions in the development of this resistance, and most of the genes proposed in this study to be responsible for drug resistance have an essential role in the development ofM. tuberculosis.[51]
Epigenome.Single-molecule real-time sequencing and subsequent bioinformatic analysis has identified threeDNA methyltransferases inM. tuberculosis,MycobacterialAdenineMethyltransferases A (MamA),[52] B (MamB),[53] and C (MamC).[54]All three areadenine methyltransferases, and each are functional in some clinical strains ofM. tuberculosisand not in others.[55][54]Unlike DNA methyltransferases in most bacteria, which invariably methylate theadenines at their targeted sequence,[56] some strains ofM. tuberculosis carry mutations in MamA that cause partial methylation of targeted adenine bases.[54] This occurs as intracellular stochastic methylation, where a some targeted adenine bases on a given DNA molecule are methylated while others remain unmethylated.[54][57] MamA mutations causing intercellular mosaic methylation are most common in the globally successful Beijing sublineage ofM. tuberculosis.[54] Due to the influence of methylation on gene expression at some locations in the genome,[52] it has been hypothesized that IMM may give rise to phenotypic diversity, and partially responsible for the global success of Beijing sublineage.[54]
TheMycobacterium tuberculosis complex (MTBC) evolved in Africa and most probably in theHorn of Africa.[58][59] In addition toM. tuberculosis, the MTBC has a number of members infecting various animal species, includingM. africanum,M. bovis (Dassie's bacillus),M. caprae,M. microti,M. mungi, M. orygis, andM. pinnipedii. This group may also include theM. canettii clade. These animal strains of MTBC do not strictly deserve species status, as they are all closely related and embedded in theM. tuberculosis phylogeny, but for historic reasons, they currently hold species status.[citation needed]
TheM. canettii clade – which includesM. prototuberculosis – is a group of smooth-colonyMycobacterium species. Unlike the established members of theM. tuberculosis group, they undergo recombination with other species. The majority of the known strains of this group have been isolated from the Horn of Africa. The ancestor ofM. tuberculosis appears to beM. canettii, first described in 1969.[60]
The established members of theM. tuberculosis complex are all clonal in their spread. The main human-infecting species have been classified into seven lineages. Translating these lineages into the terminology used for spoligotyping, a very crude genotyping methodology, lineage 1 contains theEast African-Indian (EAI), the Manila family of strains and some Manu (Indian) strains; lineage 2 is theBeijing group; lineage 3 includes theCentral Asian (CAS) strains; lineage 4 includes theGhana andHaarlem (H/T),Latin America-Mediterranean (LAM) and X strains; types 5 and 6 correspond toM. africanum and are observed predominantly and at high frequencies inWest Africa. A seventh type has been isolated from the Horn of Africa.[58] The other species of this complex belong to a number of spoligotypes and do not normally infect humans.[citation needed]
Lineages 2, 3 and 4 all share a unique deletion event (tbD1) and thus form a monophyletic group.[61] Types 5 and 6 are closely related to the animal strains of MTBC, which do not normally infect humans. Lineage 3 has been divided into two clades: CAS-Kili (found inTanzania) and CAS-Delhi (found in India andSaudi Arabia).[citation needed]
Lineage 4 is also known as the Euro-American lineage. Subtypes within this type include Latin American Mediterranean, Uganda I, Uganda II, Haarlem, X, and Congo.[62]
A much cited study reported thatM. tuberculosis has co-evolved with human populations, and that themost recent common ancestor of theM. tuberculosis complex evolved between 40,000 and 70,000 years ago.[63][61] However, a later study that included genome sequences fromM. tuberculosis complex members extracted from three 1,000-year-old Peruvian mummies, came to quite different conclusions. If themost recent common ancestor of theM. tuberculosis complex were 40,000 to 70,000 years old, this would necessitate an evolutionary rate much lower than any estimates produced by genomic analyses of heterochronous samples, suggesting a far more recent common ancestor of theM. tuberculosis complex as little as 6000 years ago.[64][65]
An analysis of over 3000 strains ofM. bovis from 35 countries suggested an Africa origin for this species.[66]
There are currently two narratives existing in parallel regarding the age of MTBC and how it has spread and co-evolved with humans through time. One study compared theM. tuberculosis phylogeny to a human mitochondrial genome phylogeny and interpreted these as being highly similar. Based on this, the study suggested thatM. tuberculosis, like humans, evolved in Africa and subsequently spread with anatomically modern humans out of Africa across the world. By calibrating the mutation rate of M. tuberculosis to match this narrative, the study suggested that MTBC evolved 40,000–70,000 years ago.[59] Applying this time scale, the study found that theM. tuberculosiseffective population size expanded during theNeolithic Demographic Transition (around 10,000 years ago) and suggested thatM. tuberculosis was able to adapt to changing human populations and that the historical success of this pathogen was driven at least in part by dramatic increases in human host population density. It has also been demonstrated that after emigrating from one continent to another, a human host's region of origin is predictive of which TB lineage they carry,[67][68] which could reflect either a stable association between host populations and specificM. tuberculosis lineages and/or social interactions that are shaped by shared cultural and geographic histories.
Regarding the congruence between human andM. tuberculosis phylogenies, a study relying onM. tuberculosis and humanY chromosome DNA sequences to formally assess the correlation between them, concluded that they are not congruent.[69] Also, a more recent study which included genome sequences fromM. tuberculosis complex members extracted from three 1,000-year-old Peruvian mummies, estimated that themost recent common ancestor of theM. tuberculosis complex lived only 4,000 – 6,000 years ago.[70] TheM. tuberculosis evolutionary rate estimated by the Bos et al. study[70] is also supported by a study on Lineage 4 relying on genomicaDNA sequences from Hungarian mummies more than 200 years old.[71] In total, the evidence thus favors this more recent estimate of the age of the MTBC most recent common ancestor, and thus that the global evolution and dispersal ofM. tuberculosis has occurred over the last 4,000–6,000 years.[citation needed]
Among the seven recognized lineages ofM. tuberculosis, only two are truly global in their distribution: Lineages 2 and 4. Among these, Lineage 4 is the most well dispersed, and almost totally dominates in the Americas. Lineage 4 was shown to have evolved in or in the vicinity of Europe, and to have spread globally with Europeans starting around the 13th century.[72] This study also found that Lineage 4 tuberculosis spread to the Americas shortly after the European discovery of the continent in 1492, and suggests that this represented the first introduction of human TB on the continent (although animal strains have been found in human remains predating Columbus.[70] Similarly, Lineage 4 was found to have spread from Europe to Africa during theAge of Discovery, starting in the early 15th century.[72]
It has been suggested that ancestral mycobacteria may have infected early hominids in East Africa as early as three million years ago.[73]
DNA fragments fromM. tuberculosis and tuberculosis disease indications were present in human bodies dating from 7000 BC found atAtlit-Yam in theLevant.[74]
M. tuberculosis is a clonal organism and does not exchange DNA viahorizontal gene transfer. Despite an additionally slow evolution rate, the emergence and spread of antibiotic resistance inM. tuberculosis poses an increasing threat to global public health.[75] In 2019, the WHO reported the estimated incidence of antibiotic resistant TB to be 3.4% in new cases, and 18% in previously treated cases.[76] Geographical discrepancies exist in the incidence rates of drug-resistant TB. Countries facing the highest rates of antibiotic resistant TB include China, India, Russia, and South Africa.[76] Recent trends reveal an increase in drug-resistant cases in a number of regions, with Papua New Guinea, Singapore, and Australia undergoing significant increases.[77]
Multidrug-resistant tuberculosis (MDR-TB) is characterised by resistance to at least the two front-line drugsisoniazid andrifampin.[78][76] MDR is associated with a relatively poor treatment success rate of 52%. Isoniazid and rifampin resistance are tightly linked, with 78% of the reported rifampin-resistant TB cases in 2019 being resistant to isoniazid as well.[76] Rifampin-resistance is primarily due to resistance-conferring mutations in the rifampin-resistance determining region (RRDR) within therpoB gene.[79] The most frequently observed mutations of the codons in RRDR are 531, 526 and 516. However, alternative more elusive resistance-conferring mutations have been detected. Isoniazid function occurs through the inhibition of mycolic acid synthesis through theNADH-dependent enoyl-acyl carrier protein (ACP)-reductase.[80] This is encoded by theinhA gene. As a result, isoniazid resistance is primarily due to mutations within inhA and the katG gene or its promoter region - acatalase-peroxidase which is required to activate isoniazid.[80] As MDR inM. tuberculosis becomes increasingly common, the emergence of pre-extensively drug resistant (pre-XDR) and extensively drug resistant (XDR-) TB threatens to exacerbate the public health crisis. XDR-TB is characterised by resistance to both rifampin and Isoniazid, as well second-line fluoroquinolones and at least one additional front-line drug.[76] Thus, the development of alternative therapeutic measures is of utmost priority.[citation needed]
An intrinsic contributor to the antibiotic resistant nature ofM. tuberculosis is its unique cell wall. Saturated withlong-chain fatty acids or mycolic acids, the mycobacterial cell presents a robust, relatively insoluble barrier.[81] This has led to its synthesis being the target of many antibiotics - such as Isoniazid. However, resistance has emerged to the majority of them. A novel, promising therapeutic target is mycobacterial membrane protein large 3 (MmpL3).[82] The mycobacterial membrane protein large (MmpL) proteins are transmembrane proteins which play a key role in the synthesis of the cell wall and the transport of the associated lipids. Of these, MmpL3 is essential; knock-out of which has been shown to be bactericidal.[82] Due to its essential nature, MmpL3 inhibitors show promise as alternative therapeutic measures in the age of antibiotic resistance. Inhibition of MmpL3 function showed an inability to transport trehalose monomycolate - an essential cell wall lipid - across the plasma membrane.[82] The recently reported structure of MmpL3 revealed resistance-conferring mutations to associate primarily with the transmembrane domain.[83] Although resistance to pre-clinical MmpL3 inhibitors has been detected, analysis of the widespread mutational landscape revealed a low level of environmental resistance.[83] This suggests that MmpL3 inhibitors currently undergoing clinical trials would face little resistance if made available. Additionally, the ability of many MmpL3 inhibitors to work synergistically with other antitubercular drugs presents a ray of hope in combatting the TB crisis.[citation needed] Genome modeling ofM. tuberculosis has also highlighted potential synthetic lethal interactions that could inform new therapeutic strategies. Specifically, knock-out of the gene Rv0489 has been shown to render gene Rv2156c essential. Given that Rv2156c is involved inpeptidoglycan synthesis, thissynthetic lethality could be exploited to sensitizeM. tuberculosis toβ-lactam antibiotics, which typically target cell wall synthesis. This opens up the possibility of repurposing existing β-lactam drugs as part of combination therapies to more effectively treat resistant tuberculosis strains.
Another cause forM. tuberculosis drug resistance areefflux pumps that can be found in many bacteria. Efflux pumps are used by the bacteria for some endogenous functions, for example lipid transport from the inner- to the outer membrane, but also as a way of reducing the drug concentration inside the cell. The EfpA efflux pump that can be found inM. tuberculosis is a part of themajor facilitator superfamily (MFS). InM. tuberculosis the EfpA transports endogenous lipids and possibly drugs via a staircase-flips model from the inner to the outer membrane. This mechanism is a possible drug target, as inhibition of it leads to the death of the bacteria, seemingly mostly by blocking its endogenous functions.[84]
The nature of the host-pathogen interaction between humans andM. tuberculosis is considered to have a genetic component. A group of rare disorders calledMendelian susceptibility to mycobacterial diseases was observed in a subset of individuals with a genetic defect that results in increased susceptibility to mycobacterial infection.[85]
Early case and twin studies have indicated that genetic components are important in host susceptibility toM. tuberculosis. Recentgenome-wide association studies (GWAS) have identified three genetic risk loci, including at positions 11p13 and 18q11.[86][87] As is common in GWAS, the variants discovered have moderate effect sizes.[citation needed]
As anintracellular pathogen,M. tuberculosis is exposed to a variety of DNA-damaging assaults, primarily from host-generated antimicrobial toxic radicals. Exposure to reactive oxygen species and/or reactive nitrogen species causes different types of DNA damage including oxidation, depurination, methylation, and deamination that can give rise to single- and double-strand breaks (DSBs).
DnaE2 polymerase is upregulated inM. tuberculosis by several DNA-damaging agents, as well as during infection of mice.[88] Loss of this DNA polymerase reduces the virulence ofM. tuberculosis in mice.[88] DnaE2 is an error-prone DNA repair polymerase that appears to contribute toM. tuberculosis survival during infection.
The two major pathways employed in repair of DSBs arehomologous recombinational repair (HR) andnonhomologous end joining (NHEJ). Macrophage-internalizedM. tuberculosis is able to persist if either of these pathways is defective, but is attenuated when both pathways are defective.[89] This indicates that intracellular exposure ofM. tuberculosis to reactive oxygen and/or reactive nitrogen species results in the formation of DSBs that are repaired by HR or NHEJ.[89] However deficiency of DSB repair does not appear to impairM. tuberculosis virulence in animal models.[90]
M. tuberculosis has existed throughout history, but the name has changed frequently over time. In 1720, though, the history of tuberculosis started to take shape into what is known of it today; as the physicianBenjamin Marten described in hisA Theory of Consumption, tuberculosis may be caused by small living creatures transmitted through the air to other patients.[93]
More than 100 million people around the world have died from being infected with mycobacterium tuberculosis due to unavailability of the vaccine in some parts of the world.
TheBCG vaccine (bacille Calmette-Guerin), which was derived fromM. bovis, while effective against childhood and severe forms of tuberculosis, has limited success in preventing the most common form of the disease today, adult pulmonary tuberculosis.[94] Because of this, it is primarily used in high tuberculosis incidence regions, and is not a recommended vaccine in the United States due to the low risk of infection. To receive this vaccine in the United States, an individual is required to go through a consultation process with an expert inM. tuberculosis and is only given to those who meet the specific criteria.[95]
Research indicates there may be a correlation between BCG vaccination and better immune response toCOVID-19.[98]
The DNA vaccine can be used alone or in combination with BCG. DNA vaccines have enough potential to be used with TB treatment and reduce the treatment time in future.[99]
^abRyan KJ, Ray CG (2004). "Mycobacteria".Sherris Medical Microbiology: an Introduction to Infectious Diseases (4th ed.). New York: McGraw-Hill. p. 439.ISBN978-0-83-858529-0.
^Fu LM, Fu-Liu CS (1 January 2002). "Is Mycobacterium tuberculosis a closer relative to Gram-positive or Gram-negative bacterial pathogens?".Tuberculosis.82 (2–3):85–90.doi:10.1054/tube.2002.0328.PMID12356459.
^abcFitzgerald DW, Sterline TR, Haas DW (2015). "251 – Mycobacterium tuberculosis". In Bennett JE, Dolin R, Blaser MJ (eds.).Mandell, Douglas, and Bennett's principles and practice of infectious diseases. Elsevier Saunders. p. 2787.ISBN978-1-4557-4801-3.OCLC903327877.
^Murray PR, Rosenthal KS, Pfaller MA (2005).Medical Microbiology. Elsevier Mosby.
^abHunter RL, Olsen MR, Jagannath C, Actor JK (2006). "Multiple roles of cord factor in the pathogenesis of primary, secondary, and cavitary tuberculosis, including a revised description of the pathology of secondary disease".Annals of Clinical and Laboratory Science.36 (4):371–386.PMID17127724.
^Flynn JL, Chan J (August 2003). "Immune evasion by Mycobacterium tuberculosis: living with the enemy".Current Opinion in Immunology.15 (4):450–55.doi:10.1016/S0952-7915(03)00075-X.PMID12900278.
^Ramakrishnan L (April 2012). "Revisiting the role of the granuloma in tuberculosis".Nature Reviews. Immunology.12 (5):352–366.doi:10.1038/nri3211.PMID22517424.S2CID1139969.
^The ability to restore macrophage apoptosis toM. tuberculosis infected ones could improve the current tuberculosis chemotherapy treatment, as TB drugs can gain better access to the bacteria in the niche (M),
^Zhang H, Li D, Zhao L, Fleming J, Lin N, Wang T, et al. (October 2013). "Genome sequencing of 161 Mycobacterium tuberculosis isolates from China identifies genes and intergenic regions associated with drug resistance".Nature Genetics.45 (10):1255–60.doi:10.1038/ng.2735.PMID23995137.S2CID14396673.
^Chalut C (September 2016). "MmpL transporter-mediated export of cell-wall associated lipids and siderophores in mycobacteria".Tuberculosis.100:32–45.doi:10.1016/j.tube.2016.06.004.PMID27553408.