Bioprospecting for antituberculosis natural products – A review

2.1 Tuberculosis

Tuberculosis is an ancient disease of humans caused by the Mycobacterium tuberculosis complex [16]. The disease is known to be associated with immunosuppression, malnutrition, poverty, and overcrowding and results in the death of more than 1 million people annually around the world [17,18]. The genus Mycobacterium consists of several species, some of which are adapted to humans and various animals as shown in Figure 1. Those that cause human tuberculosis (M. tuberculosis sensu stricto and M. africanum) are obligate human pathogens. Although M. bovis and M. caprae have been found in several animals and humans, there is rarely human-to-human transmission [16].

Figure 1

Species of Mycobacterium and their various adapted hosts. Some might infect multiple hosts but M. africanum and M. tuberculosis sensu stricto are known to infect only humans.

M. tuberculosis is a very fastidious bacterium with a typical generation time of 12–24 h when growth conditions are optimal [19]. The bacterium is about 0.8–4 µm in size, acid-fast, non-motile, and aerobic [17]. A sizeable number of individuals will develop latent tuberculosis, which may be well contained in some persons and not progress to TB immediately. However, 5–15% of this group of individuals with latent tuberculosis will progress to active TB [20,21].

Tuberculosis is spread through the inhalation of infectious droplet nuclei that contain viable M. tuberculosis [22]. Churchyard et al. [23] gave a simple pathway by which tuberculosis is transmitted. This pathway entails a source case of tuberculosis, generation of infectious particles, survival of particles in the air, inhalation by susceptible individuals, and ultimately potential to develop tuberculosis.

Clinical manifestation of TB is presented in various forms. TB could manifest with mild and non-specific symptoms as observed in primary tuberculosis, which is most often self-resolving. In pulmonary tuberculosis, patients present with night sweats, fever, and weight loss, and about 95% of individuals have non-remitting cough that may be accompanied by chest pain in patients with dyspnoea or subpleural involvement [22]. Besides pulmonary tuberculosis, extrapulmonary TB (EPTB), which is tuberculosis outside the lungs, also occurs. It is characterized by possible weight loss and fever and could manifest as lymphatic TB that involves any regional lymph node, meningeal TB that presents with headache and altered mental status or neurologic defects in advance, and TB of the joints and bones with swelling and persistent localized pain [24].

The paucibacillary nature of EPTB largely contributes to a setback in diagnosis due to lack of highly sensitive and specific testing kits, the emergence of multi- and extensively drug-resistant strains, limited vaccine efficacy, and HIV have exacerbated cases of tuberculosis globally and tend to stall eradication efforts [25].

2.2 Antituberculosis chemotherapies

Although the treatment strategy and guidelines of drug-sensitive TB (DS-TB) have not obviously evolved over the past 35 years, which requires a minimum of six months of treatment with a combination of four basic treatment regimens comprising isoniazid (INH) (1), rifampin (RIF) (2), pyrazinamide (PZA) (3), streptomycin (4) as well as ethambutol (EMB) (5), which are first-line antituberculosis drugs as shown in Table 1. While second-line drugs consist of fluoroquinolones (ofloxacin (6), moxifloxacin (7), levofloxacin (8), and ciprofloxacin (9)), injectable (capreomycin (10), kanamycin (11), and amikacin (12)) and p-aminosalicyclic acid (13), ethionamide (14) prothionamide (15) and cycloserine (16) terizidone (17) [26]. Bedaquiline (18) a diarylquinoline, linezolid (19) an oxazolidinone, delamanid (20) and pretomanid (21), which belong to the nitroimidazole class of antibiotics, and clofazimine (22) are newer and repurposed drugs used for tuberculosis treatment some of which are components of shorter-course regimen [27,28]. The treatment of multidrug resistance (MDR) and extensively drug-resistant (XDR) tuberculosis has, however, changed over time in repurposed and new drugs being introduced into the treatment regimens with treatment duration of up to 24 months [29,30]. The treatment of tuberculosis is in phases that consist of the initial phase that kills most of the actively growing bacteria with eradication of clinical symptoms and rapid sputum conversion. This is followed by the continuation phase of treatment that takes care of slow-growing strains or persistent tubercle bacilli [26]. All structures of antituberculosis chemotherapies are shown in Figure 2.

Figure 2

Structure of antituberculosis chemotherapies.

4.1 Plant-derived antimycobacterial natural products

In many regions of the world, plant-derived anti-TB recipes are well known in folklore medicaments. Plants are an important source of natural products with diverse bioactivities and have the enormous therapeutic potential [14]. Numerous structurally unique bioactive compounds have been isolated from plant extracts [49]. This has driven a lot of drug-discovery research on plants that have previously been known to be used traditionally for the treatment of infections. Infusions of leaves from Lippia javanica (Verbenaceae), an aromatic plant, are used in Mozambique as a decoction to treat several ailments like flu symptoms, cough, stomach aches, and many others [48].

Triterpene euscaphic acid was obtained from this herb and found to exhibit antimycobacterial bioactivity against M. tuberculosis with a minimum inhibitory concentration recorded as 50 μg/mL [50]. Some compounds such as diospyrin (23), shinanolone (24), and 7-methyljuglone (25) were isolated from the roots of the plant Euclea natalensis (Ebenaceae) and displayed significant activity against drug-sensitive and -resistant strains of M. tuberculosis [51]. This plant has been known historically to treat infections associated with chest conditions, toothache amongst other uses, and lends credence to its ethnomedicinal use [52]. A. afrawas was also reported to exhibit several biological activities such as antidepressant, antioxidant, cardiovascular, and antimycobacterial, in vitro [53]. This plant is commonly available in Eastern and Southern Africa and has been used traditionally to treat respiratory and oro-pharygeal illnesses such as colds, coughs, asthmatic conditions, to mention a few [54]. Ursolic acid (26) and hydroquinone (27), which were isolated from Artemisia capillaris Thunb. (Asteraceae), a plant that was traditionally known for the treatment of malaria, was assessed for its antimycobacterial activity against MDR-TB strains and was found to have MIC values ranging from 0.0125 to 0.025 mg/mL [55]. Another plant, Curcuma longa L. (Zingiberaceae), which has been traditionally associated with the treatment of whooping cough, was assessed for its antimycobacterial potentials. Isoxazole (27) analogues of curcuminoids were isolated from this plant and were found to have bioactivity against MDR-TB at MIC values ranging from 0.0019 to 0.00312 mg/mL [56].

Knowltonia vesicatoria (Ranunculaceae), a plant commonly available in South Africa, has been traditionally known for the treatment of tuberculosis. Antimycobacterial drug discovery efforts led to the isolation of 2 compounds, 5-(hydroxymethyl)dihydrofuran-2(3H)-one (28) and 5-(hydroxymethyl)furan-2(5H)-one (29), with a minimum inhibition concentration of 50.0 mg/mL. Crude extracts of this plant were also used together with isoniazid (INH) (1) to determine whether there would be improved antimycobacterial activity due to probable synergistic effect, thus validating its traditional use in the treatment of tuberculosis [52,57].

4.2 Microbes as sources of antimycobacterial natural products

Microorganisms have been recognized as robust sources of chemotherapeutics. However, their potentials as a repertoire of useful antibiotics have not been fully exploited as only a very limited number of microbial diversities is known globally [58]. The discovery of the compound streptomycin several decades ago (4) by Selman Waksman was a major breakthrough in drug discovery as the first effective antibiotic against tuberculosis [59]. Selman Waksman’s commitment to the isolation and screening of soil bacteria in the search for bioactive small molecules, which could be potential antibiotics, was validated by the discovery of streptomycin [60].

Fungi have been implicated to exhibit antimycobacterial potentials. The fungus Mortierella alpine FKI-4905 isolated from Japanese soil was found to have antimycobacterial activity against M. tuberculosis and M. smegmatis with MIC values of 12.5 and 0.78 µg/mL [61]. Another natural anthraquinone compound, 4-deoxybostrycin (30), was obtained from the mangrove endophytic fungus Nigrospora sp., which was isolated from the South China Sea. This compound showed good anti-TB activity in vitro. Further studies on the mechanisms of action on M. tuberculosis H37Rv showed that 4-deoxybostrycin (30) could affect the expression of its genes that are involved in coenzymes, lipids, nucleotides, and energy to mention a few [62]. However, some other researchers [63] explored the antimycobacterial potentials of freshwater actinobacteria isolated from Lake Michigan through screening of a library of compounds derived from aquatic bacteria. A fraction from one of the bacteria (Micromonospora sp.) exhibited significant inhibitory activity in vitro in comparison to clinically used anti-TB agents. From this strain, two new secondary metabolites, diazaquinomycins H (DAQH) (31) and J (DAQJ) (32), were isolated and characterized with MIC values of 0.04 and 0.07 μg/mL, respectively, against M. tuberculosis. Manikkam et al. [64] also isolated actinomycetes from diverse Indian soil ecosystems with a significant number exhibiting antimycobacterial activity.

5.1 Uncultured microbes and novel anti-TB natural products

The number of microorganisms that have been isolated using traditional culturing techniques is less than 1% of microorganisms, making the extensive majority of microorganisms as well as their biochemical metabolic pathways unavailable [71,72,73]. The metagenomics approach remains promising in drug discovery from uncultured microorganisms. These microorganisms have proven an important source of novel antibiotics through analysis of their DNA isolated from environmental samples [65]. The advent of metagenomic procedures as a culture-independent technique has therefore made various tools available to evaluate the level of the uncultured microbial diversity in full, consequently, enabling access to the biochemical metabolic pathways in these uncultured microbes [74,75]. Culture-independent molecular approach has however confirmed the presence of indigenous actinobacteria through different studies [76,77,78]. These approaches focus on nucleic acids directly extracted from the samples [79] and typically involve the amplification of DNA from RNA extracted from environmental samples by PCR followed by further analysis of the diversity of the amplified molecules (community fingerprinting) [9]. These amplicons could otherwise be cloned, sequenced, and the rare/novel actinobacteria available in the sample identified and enumerated [80,81]. Monciardini et al. [76] developed selective primer sets for PCR amplification of the16S rDNA obtained from Actinomycetales families Thermomonosporaceae, Streptosporangiaceae, Streptomycetaceae, and Micromonosporaceae as well as from Dactylosporangium sp. Each designed primer, evaluated on the basis of genomic DNA from reference strains, indicated high specificity and sensitivity. However, the use of the primers to process environmental samples indicated the discovery of these groups of actinomycetes in high frequency and also showed sequences that could be ascribed to novel actinobacterial groups [76]. Other molecular approaches such as terminal restriction fragment length polymorphism (T-RFLP) analysis, evaluates the terminal restriction fragments from a PCR-amplified marker based on the size polymorphism [82] and the method links at the minimum, three technologies such as comparative genomics/RFLP, PCR, and electrophoresis. Electrophoresis techniques such as the denaturing gradient gel electrophoresis (DGGE) as well as TGGE (thermal-GGE) are also procedures by which short segments of DNA of different sequences but the same length can be resolved electrophoretically [82]. In this case, separation of DNA fragments leverages on the decreased electrophoretic movement of a double-stranded DNA, which is partially denatured either by a denaturing reagent or increased temperature and resolved on polyacrylamide gels. The PCR-DGGE/TGGE method was extensively utilized to determine the abundance and diversity of actinomycetes in environmental samples [73,77,82,83].

Metaproteomic and metagenomic technologies enable newer techniques towards gene, genome, protein as well as the discovery of metabolic pathways [72]. The number of yet-to-be-cultured marine microbiota that synthesizes secondary metabolites has greatly reduced accessibility to a vast genetic diversity linked to untapped chemical resources required in therapeutic and other industrial applications [9,84]. This includes complex marine (e.g. sponge, dinoflagellates, tunicates) also terrestrial (e.g. human gut, biofilm, plant-microbe) microbial consortia in which case the presence of huge populations of varying microbiota as well as related genomes that constitute natural product gene clusters yet are unexplored [72,73]. This novel chemical and metabolic diversity resource will help increase fundamental knowledge and make a contribution to current drug discovery efforts against several diseases.

6.1 Agar diffusion method

The agar diffusion test of natural products is typically done on the agar medium surface by placing the filter paper disks previously immersed in extracts on the plate inoculated with the test organism and incubated at 37°C. This enables the test compounds to diffuse into the agar consequently inhibiting the growth of the inoculum. After incubation, the mean diameters of inhibition zones of the test compounds against the inoculum are recorded. The model strain, M. marinum, has been used for antimycobacterial natural products screening due to its ubiquitous nature. Diffusion assay is an ideal technique of choice for mixtures of compounds such as essential oils or polar molecules [85]. This method ensures rapid screening of crude extracts using M. marinum as a model strain [86,87].

6.2 Agar microdilution assay

This method developed by Golus et al. [88] involves the screening of natural products with different concentrations against pathogenic test strains in a 96-well microplate containing the agar medium. The assay enables the easy, fast, and cost-efficient sample preparation including the provision of a stable and uniform dispersion of test samples without differentiation of the coloured plant extracts into oil–water phases, which is a frequent occurrence in methods involving liquid media. The agar microdilution method also enables the easy, reliable, and cost-efficient MIC determination of such agents.

6.3 LJ proportion method

This proportion method uses the Löwenstein–Jensen medium (LJ), which is an egg-based medium supplemented with malachite green dye. Incorporation of the antimycobacterial drugs in the LJ medium is done at varying concentrations before the medium is solidified by a process known as inspissation. The solid medium is later inoculated with 100 µL of the inoculum of up to 1.0 MacFarland Standard and incubated at 37°C for 42 weeks. Natural products are generally not heat-stable, hence this method could serve as a disadvantage if the crude extracts are incorporated into the medium prior to inspissation, which requires heat of up to 80°C. Alternatively, the natural product could be added to the surface of the solid medium after inspissation and cooling. Although the advantage of using this method is its cost efficiency, particularly in resource-limited laboratories that are unable to afford more expensive culture media, the disadvantage of using this method is the long period of incubation to assess the resistance or susceptibility of the strain to the compounds through growth detection [86].

6.4 Micro-broth dilution assay

Bioactivity screening of natural products with the use of a 96-well microplate offers the advantage of minimal sample requirements, reduced cost, and high-throughput processes, including the potential for automation [89,90,91,92]. This method involves the growth of mycobacteria in growth media such as Middlebrook 7H9 broth supplemented with growth supplements including oleic acid, albumin, dextrose, and catalase. The growth of several mycobacterial strains could be determined quantitatively in a liquid medium by turbidity but a major drawback in this test is the possibility of mycobacteria clumping together. The metabolic compounds or extracts derived from natural products may also confer some turbidity on the culture medium thereby serving as a challenge in the determination of results. The use of an indicator dye that works with the principle of oxidation–reduction such as Alamar blue, resazurin, tetrazolium, malachite green, and crystal violet, makes the micro-broth dilution a technique of choice for speed and sensitivity [93,94].

6.5 Radiometric BACTEC 460 assay

This assay determines the inhibition or growth of M. tuberculosis through the measurement of ¹⁴CO₂ generated by M. tuberculosis cultivated in broth containing ¹⁴C-labeled palmitic acid. This liquid growth medium is Middlebrook 7H12 medium, and mycobacterial growth is measured using the BACTEC 460 instrument for HTS assay and the results can be observed in less than 2 weeks [95]. The percentage inhibition of ¹⁴CO₂ production relative to drug-free controls could be extrapolated, or large numbers of extract concentrations could be screened and their MIC values determined [92,96,97]. Results are recorded usually at 24 h intervals, thus, enabling accurate evaluation of growth kinetics or inhibition of the test mycobacterial. The main disadvantage of this assay is the cost such as in disposal of the isotope in certain countries as well as the substantial amount of growth medium required, which consequently utilizes a large quantity of samples for the assay [86].

6.6 BACTEC MGIT 960 assay

This assay requires the use of modified Middlebrook 7H9 broth containing a fluorescent sensor ruthenium chloride pentahydrate as a growth medium in a non-radiometric automated system. The method is an excellent, high-throughput procedure, which is advantageous in terms of provision of very fast and reliable results in less than 2 weeks. [55]. However, a major disadvantage to this method is the high costs as well as large volumes of the medium required to run the assay, which consequently requires a large amount of the test sample (metabolic extract) to be screened.

6.7 Reporter gene assays

These assays are generally centred on the use of reporter genes associated with the red fluorescent protein, luciferase, and green fluorescent protein. The proteins are typically derived from different species of insects (fireflies), crustaceans, beetles, and bacteria (Vibro harveyi) and allow the rapid determination of bacterial viability, which is extrapolated by expressing a luminescent or a fluorescent protein [92,98]. The assay can be applied in a multiwell format, which brings about improved high-throughput detection. A major advantage in the use of the reporter gene assays is its cost-effectiveness as it is inexpensive for routine use in the HTS of antitubercular compounds with minimal commercial applications [99,100]. The major disadvantage of using this method is that its use for commercial applications is usually restricted by patent challenges; hence, limiting the number of mycobacteriology laboratories adopting this method for susceptibility screening of natural products against M. tuberculosis [92,98].

6.8 Wayne’s hypoxia model

For decades, latent tubercle bacilli of M. tuberculosis have continued to challenge TB therapy efforts, which has become a cause for concern thereby resulting in prolonged treatment of active TB. The currently available anti-TB drugs are ineffective against the dormant forms of M. tuberculosis and the fact that there was no screening bioassay for chemical molecules that have activity against dormant tubercle bacilli has been an impediment to the development of novel chemotherapeutics against latent TB [92]. A model was therefore developed by Wayne and Sramek [101] as the first in vitro model for Mycobacterium tuberculosis for a specific time frame (usually 24 days), thereby inducing hypoxic conditions thus leading to the sensitivity of hypoxia-induced latent bacilli for metronidazole with the introduction of a single stress factor.

6.9 The rapid anaerobic model

Some other modified versions of the Wayne model that are aimed to yield more consistent and accurate results have been developed such as the rapid anaerobic model [102,103]. The major difference between this model and the Wayne model is the size of stir bars, which are larger, as well as the much faster stirring rate in others to increase the homogeneous population of bacilli, and oxygen distribution within the culture. The available oxygen is depleted rapidly ensuring anaerobiosis within a limited period in comparison with the Wayne model [104].

6.10 Betts starvation model

A nutrient starvation model, which involves the simulation of M. tuberculosis in physiologically dormant states and the introduction of one stress factor, was developed by Betts et al. [105]. The procedures require the transfer of bacilli into a nutrient-deficient medium followed by incubation at 37°C for 6 weeks [105]. In this model, the rate of respiration is used to inhibit the growth of the test strain. Ultimately, this model provides evidence for deceleration of the energy metabolism, transcription apparatus, lipid biosynthesis, and cell division including induction of the stringent response and many other genes that may lead to long-term survival within the host.

6.11 Low-oxygen recovery assay (LORA)

The search for novel antimicrobials through screening for bioactivity is usually carried out using microbes that are actively replicating. It has however been established that antimicrobial tolerance in various bacterial infections is a result of a physiological state of non-replicating persistent (NRP) dormant forms of M. tuberculosis. For the treatment of tuberculosis, the possibility of shortening the typical 6-month regimen is highly dependent on the NRP subpopulation. LORA is an assay developed for the bioactivity of compounds against non-replicating M. tuberculosis utilizing an M. tuberculosis pFCAluxAB strain. This luminescence-based assay uses high-throughput measures to screen test bioactive compounds against NRP forms of M. tuberculosis [106,107]. In this assay, the MICs of antimicrobial agents are evaluated using the microplate method [106]. This is a multiple stress model that works by applying a combination of stress factors including high CO₂ (10%), low oxygen (5%), low acidic pH (5.0), and nutrients. The advantage of adopting this model is its efficiency in generating bacilli that meet all criteria of dormancy, and this multiple stress method is easily applicable to HTS of natural products [108,14].

6.12 A low pH and nutrient stress assay

The need for easy and more rapid assessment of novel anti-tubercular agents against non-replicating bacilli facilitated the development of a method for the detection of the bactericidal activity of new bioactive natural compounds within 7 days by Early et al. [109]. This method utilized the incubation of M. tuberculosis strains at sub-optimal pH, bringing about induction of a non-replicating state. Antimicrobial susceptibility testing of bacteria against compounds is determined by the evaluation of luminescence. The advantages of this simple assay are that it does not require an outgrowth period, and results are easily extrapolated with minimal manipulations, which could be adapted for bactericidal activity screening of natural products against non-replicating M. tuberculosis.

6.13 Anti-TB ex vivo assays

This procedure involves the use of mouse/human macrophages (usually THP-1 cells, J774 cells), which evaluates the intracellular M. tuberculosis killing or sterilizing activity of test compounds [14]. In most cases, macrophages are infected with M. tuberculosis with a multiplicity of infection (MOI) of 10, and intracellular anti-TB activity is determined by exposure to test compounds for 3–7 days. The antimycobacterial effects of the test compounds are typically monitored by determining the bacterial load within macrophages by lysing them with hypotonic buffer (1.5 mM MgCl₂, 10 mM HEPES, and 10 mM KCl). The samples are spread on Dubos agar plates at varying time intervals and colonies are enumerated after incubation for 21 days. The bacterial viability is determined by the enumeration of colony-forming units [110,111,112].