Spotlight on mycobacterial lipid exploitation using nanotechnology for diagnosis, vaccines, and treatments

Abstract

Tuberculosis (TB), historically the most significant cause of human morbidity and mortality, has returned as the top infectious disease worldwide, under circumstances worsened by the COVID-19 pandemic's devastating effects on public health. Although Mycobacterium tuberculosis, the causal agent, has been known of for more than a century, the development of tools to control it has been largely neglected. With the advancement of nanotechnology, the possibility of engineering tools at the nanoscale creates unique opportunities to exploit any molecular type. However, little attention has been paid to one of the major attributes of the pathogen, represented by the atypical coat and its abundant lipids. In this review, an overview of the lipids encountered in M. tuberculosis and interest in exploiting them for the development of TB control tools are presented. Then, the amalgamation of nanotechnology with mycobacterial lipids from both reported and future works are discussed.

Graphical abstract

Mycobacterium tuberculosis, the microbial cause of tuberculosis, synthesizes atypical and abundant lipids with potential applications in diagnostics, vaccines and therapies. While the hydrophobic nature of most lipids had importantly limited their exploitation, especially at the large-scale, nanotechnology approaches, including innovative supports or biomimetic assemblies, are currently emerging to fill that gap.

Introduction

With the decline of COVID-19 cases and deaths, tuberculosis (TB), the deadliest disease in human history, is once again the major infectious killer worldwide. According to the last report of the World Health Organization (WHO) on TB, M. tuberculosis, the causal agent of TB, caused 10 million new cases and 1.6 million deaths in 2021, just below those attributed to COVID-19 during its deadliest year.¹ Unfortunately, with the focus on halting coronavirus infection, control measures against TB have been seriously neglected, with devastating effects. Deaths caused by TB increased for the first time in a decade, while the number of diagnoses, reports and treatments fell to unexpected levels.² One of the major obstacles to the control of TB is the complexity of the life cycle of M. tuberculosis: in most people, M. tuberculosis infection is efficiently confined inside a lung granuloma, resulting in a latent stage that may last for decades or until death; however, under immunosuppressive conditions, the infective node turns into a highly contagious respiratory disease.³ Although M. tuberculosis was identified more than a century ago, tools to prevent, diagnose and treat the disease at its multiple stages are scarce. On the other hand, M. tuberculosis is only the main species involved in human TB, as closely related species, namely M. africanum, M. canetti, M. bovis, and M. caprae, are being identified more and more in human TB cases in various countries.4., 5. It is noteworthy that the different lineages of these species, which are part of the M. tuberculosis complex (MTC),⁴ display phenotypic differences that should be considered in the design of control measures.

The current vaccine consists of a live attenuated mycobacterial strain, the so-called Bacillus Calmette-Guérin (BCG), which induces protection against the most severe forms of TB, especially in children, but it is highly ineffective against adult TB in the highly infectious, spreading stage.⁶ New vaccine candidates are currently being explored, with proposals ranging from recombinant BCG or promising attenuated mycobacterial strains,7., 8. to nucleic acids carried by novel nanostructured platforms.⁹ It is expected that new vaccines may be available during the next decade, for which nanotechnology would represent an important component. Regarding diagnostics, more efforts are needed to address the detection of both latent and active TB promptly in the most affected populations. While the development of affordable methods for detecting active cases is one of the top targets in public health, since it would significantly impact the spread of the disease,¹⁰ diagnosis of the latent infective form is also needed to focus on personalized medicine and epidemiological measures.¹¹ As pointed out above, millions of cases of active TB go undiagnosed each year, leading to unacceptable preventable deaths. Some survey reports indicate that only 20 % of active TB cases are currently diagnosed.¹² At present, the most common methodology used to detect active TB is sputum smear microscopy, a traditional smear test that requires skilled personnel and the manipulation of the pathogen. With no appropriate methodologies for diagnosis, the most affected settings worldwide frequently misinterpret TB symptoms (such as fever, chest pain, weight loss or night sweats) with those of other common diseases. One of the top priorities in TB research is therefore the development of friendly and low-cost diagnostic platforms, a challenge to which nanotechnology seems the natural alternative. On another side, a poor understanding of the interactionships between M. tuberculosis and host cells is now limiting the engineering of tools to fight against TB. The biology of the bacillus is extremely complex: both activators and suppressors of immunity are expressed by the pathogen in a still poorly comprehended spatiotemporal way. In this regard, the potential of nanotechnology to engineer biomimetic tools at the appropriate sizes, textures, and forms would be of the highest interest as well.

Fat: the molecular key feature in the tubercle bacillus

TB disease is probably as old as humans are. Tuberculosis-like pathologies have been found in ancient mastodons, and M. tuberculosis DNA has been identified in a Pleistocene bison.13., 14. In hominids, there is data supporting that early M. tuberculosis progenitors might have been infecting individuals as long as three million years ago.¹⁵ The question arises, therefore, about the basis for such infective success, with the tubercle bacilli apparently surpassing mankind's evolution. From a molecular point of view, M. tuberculosis is a highly complex microorganism that involves multiple pathogenesis factors, with proteins, lipids and glycans of the highest specificity.16., 17., 18. Although all molecular types in M. tuberculosis have been found to be involved in pathogenesis, we can affirm that the most outstanding feature of the bacillus is, with no doubt, the unusual waxy composition of the cell coat. Indeed, the extremely fatty coat of M. tuberculosis importantly delayed the consistent staining (and visualization) of the bacilli for Robert Koch to finally report his discovery in 1882.¹⁹ Again, it is because of the highly impermeable cell wall that most common antibiotics fail to kill the bacilli, leading to late use of chemotherapy against TB, as compared to other bacterial diseases. In 1998, the first annotated genomic sequence had revealed that M. tuberculosis contains examples of every known lipid and polyketide biosynthetic system, including all those found in bacteria, mammals and plants.¹⁶ The amount of fatty material in M. tuberculosis is huge as well: as a basis for comparison, the variety of enzymes involved in fatty acid metabolism in the tubercle bacillus is 5-fold greater than that in E. coli, giving rise to lipid content (ca. 40 % of the cell dry mass²⁰) paralleled only by oleaginous microalgae when subjected to stress.²¹ Many studies have addressed the ultrastructure of M. tuberculosis cell-walls through various microscopy techniques.18., 20. To date, we know that mycobacteria are coated by a structure with a thick polysaccharide network and at least two lipid bilayers. Importantly, lipids with no covalent linkage to this structure are known to be exported outside of the bacterial cell during infection.²² In Fig. 1 , a model based on those reports is depicted.

Fig. 1.

A model of the cell-wall of M. tuberculosis. The multiple layers of mycobacterial cell walls include: the plasma membrane (PM), a polysaccharide network composed of peptidoglycan (PG) and arabinogalactan (AG), which are highly immunogenic structures. Very long fatty acyls, the mycolic acyls (MA) are covalently linked to AG. MA feature an extremely hydrophobic inner leaflet where a set of mycobacteria-specific, amphipathic lipids are assembled non-covalently, forming an outer leaflet (OL). This atypical outer membrane is also known as mycomembrane (MM). Image created with Biorender.

From a structural point of view, mycobacterial lipids show some of the most atypical features, including skeletons up to 80C in length, polymethyl branches, or fatty esters whose rupture must be achieved through unusually heavy lytic procedures.²⁰ That is why the presence of mycobacterial lipids has been used as a hallmark, along with DNA, in the identification of TB in ancient remains.²³ Not surprisingly, the encounter between such original chemicals and human cells allows a set of interactions such as their recognition by cell surface receptors, the rise in antibodies during active infections, or the disruption of cell membranes upon the liberation of the so-called “free” lipids.²⁴ While the mechanisms linked to biological effects of mycobacterial lipids on host cells are poorly understood, growing evidence highlights their importance in pathogenicity (see Box 1 ). For instance, the delipidation of mycobacterial cells has been associated with both immune activation and suppression²⁵; from gene manipulation experiments, we know that the impairment of biosynthetic lipid pathways may lead to a tremendous decline in virulence of mycobacterial strains26., 27.; regarding individual lipids, some of them have been recognized for activating innate immune mechanisms (through the involvement of pattern recognition receptors), and others are specifically recognized by T cells via the CD1 antigen presenting molecule.28., 29. Altogether, these cumulative data have shaped mycobacterial lipids as central actors in TB disease.

Box 1.

Top-ten facts about mycobacterial fats.

•M. tuberculosis genome encodes for 5-fold more lipid metabolism enzymes than Escherichia coli, and contains lipid metabolic pathways encountered in bacteria, mammals and plants.16 •M. tuberculosis produces the largest fatty acids known in nature (with lengths up to 90C atoms).18., 20. •Mutations in lipid synthases strongly decrease the virulence of M. tuberculosis strains.26 •Differences of cell-surface lipid profiles in strains from distinct genomic backgrounds significantly influence M. tuberculosis pathogenicity.40., 57. •During infection, M. tuberculosis lipids with no covalent attachment to cell-wall polysaccharides are exported to human cell membranes of infected and bystander cells.22 •During infection, the so-called free mycobacterial lipids interact with pattern-recognition receptors, allowing them to activate the innate immune response.29 •Through still poorly understood mechanisms,58 mycobacterial lipids induce the production of specific antibodies, useful as biomarkers in TB diagnosis.30., 32., 33., 59., 60. •A set of amphipathic lipids from M. tuberculosis are recognized as specific T-cell antigens, through the involvement of essentially nonpolymorphic, non-canonical CD1 class of antigen presenting molecules.28 •Mycobacterial fatty coats are the preferential target for anti-tuberculous drugs.61 •Most bioactive lipids of M. tuberculosis are not addressable in aqueous milieux for biological interactions to occur, but only through nanotechnology approaches.

From this knowledge, the applicability of lipids in tools against TB appears evident, especially for the development of diagnostics, vaccines or immune stimulators. Mycobacterial lipids have been described as among the most valuable reagents for TB immunodiagnosis.³⁰ They have also been proposed as valuable biomarkers of infection.³¹ Opportunities to obtain mycobacterial lipids either from alternative bacterial sources32., 33., 34. or through synthetic approaches35., 36., 37. have been described, and chemical probes for tagging mycobacterial lipids are available as well.³⁸ However, working with lipids has not been an easy task, especially at bench-to-bedside levels or large-scale applications. For many years, the biological studies of lipids have involved cumbersome procedures. In some studies, lipids were put in contact with cells in the form of films at the bottom of cell culture dishes.39., 40. Other strategies had included the use of either detergent-containing media,⁴¹ or toxic solvents such as dimethylsulfoxide and ethanol.⁴² Now, with the advancement of nanotechnology, a myriad of possibilities has arisen for the design of lipid-displaying materials. In this review, pioneering works using nanotechnology to explore the biomedical applications of mycobacterial lipids will be overviewed. General advances in nanotechnology to fight against TB will be briefly presented, and the opportunities to exploit lipids will be discussed.

Nanotechnology, nanomedicine, and TB

Nanomedicine is the subcategory and application of nanotechnology to medicine, a field that represents one of the major interests in materials science. One of the main factors allowing these two disciplines to interact easily is the nanostructured nature of biology: every biomolecule (protein, lipid, sugar) or biological structure in the cell is at the nanometric scale.⁴³ Therefore, the potential to obtain synthetic structures at commensurable scales to those found in cells, and to engineer materials at the molecular level, constitutes a unique opportunity. Nanotechnology allows for the development of materials sized from 1 to 100 nm in size, with made-to-measure features: structures with very large surface-to-volume ratios, nanoporous carriers, or particles showing a highly controlled biodegradation; their physicochemical properties can be controlled by tailoring sizes, shapes, chemical composition, structure, morphology, and textural properties. For instance, nanoparticles used in nanomedicine can be readily fabricated from either soft (organic and polymeric) or hard (inorganic) materials, and the interactions between such materials and cells can be controlled through functional chemistries (Fig. 2 ). Examples are the use of specific antibodies for targeting organs or tissues and the use of coatings that provide stealthiness vis-à-vis the reticuloendothelial system, both having clear applications in drug delivery.⁴⁴ Another application of nanotechnology enables the design of materials with electrochemical conductivity or with optical activity useful in the development of biosensors. Related to this, the combined search for DDS and optical activity within a unique nanomaterial is a promising strategy being explored for diagnostics, including that intended for simultaneous in-vivo sensing and therapy, an approach known as theragnostics.⁴⁵ Altogether, these possibilities have pushed nanomedicine as an emerging field in TB research, with expectations for improvements in drug-delivery, diagnosis, treatment, and monitoring of infection.46., 47., 48., 49.

Fig. 2.

Nanotechnology makes it possible to obtain structures with controlled physicochemical properties, including size, shape, charge, textural properties, or specific bioactivities. Figure reproduced with permission by.⁶²

The past two decades have witnessed an increasing interest in the development of nanotech-based tools against TB, especially for drug delivery, biosensing and vaccine applications. Nanotech-devices for TB control can be categorized as follows:

• (1)Vaccine or vaccine subunits composed of powerful immune protective protein antigens,
• (2)Drug carriers intended to improve the pharmacological features of anti-TB treatments, as compared to traditional systems.
• (3)Biosensors based on proteins, peptides or nucleic acids known to bind human biomarkers, including molecular targets from M. tuberculosis or anti-M. tuberculosis antibodies.

Nanotechnology-based vaccines against TB

BCG, the only vaccine approved against TB was developed more than a century ago. However, the immune response to mycobacteria is not always associated with protection. No better options than BCG have been found, and a search for novel strategies is urgently needed. The use of nanotechnology in anti-TB vaccine development seems promising, as non-replicative and safe, though immunogenic formulations may be obtained by engineering materials composed of selected antigens, adjuvants and nanocarriers. Moreover, the possibilities to target specific cells, tissues and trafficking pathways through nanotechnology have been extended to non-protein biomolecules, such as sugars.⁵⁰ Vaccine developments have recently included only a handful of proteins: a 30–31 kDa protein (also known as Ag85A), the PPE44 antigen, a heat shock protein (HspX), and two secreted, short-length proteins known as Culture Filtrate Protein (CFP)-10 and ESAT-6.51., 52., 53. Interestingly, the unexpected success of nanotech-based vaccines developed against COVID-19 might represent a tour-de-force for the development of novel formulations against other important diseases, including TB. At present, a variety of vaccine strategies are under exploration, including carriers based on viral capsids or virus-like particles (VLP), biomimetic nanoassemblies, and DNA or mRNA vaccines.51., 52., 53., 54., 55. Most of these non-living vaccine alternatives include adjuvant components, such as oligonucleotides, antimicrobial peptides, monophosphoryl lipid A (MPLA) or the mycobacterial incomplete Freund adjuvant (IFA).⁵⁶

Drug carriers for anti-TB treatments

With nanotechnology, the free form of a drug can be formulated into nano/microstructures leading to more efficient treatments than conventional pharmaceutical forms. Many types of nanostructures have been explored for antituberculous DDS, including liposomes, solid lipid nanoparticles, nanostructured lipid carriers, nano and microemulsions, as well as various kinds of nanoparticles for encapsulating first- or second-line antituberculosis drugs.46., 47., 48., 49. Importantly, all these nanostructures may be modified on their surfaces in order to avoid side-effects or increase efficacy. Targeting agents such as DNA probes, proteins, peptides, saccharides, or lipids, as well as molecular mimics (aptamers or peptide mimotopes) have been used to improve pharmacokinetics. These decorations allow nanotech-based DDS to increase the uptake of drugs in the bloodstream, to overcome microbial resistance and to address drug treatments through different administration routes.63., 64. Nanostructured lipid carriers were the firstly DDS approved and have been widely explored in treatments against TB. At present, other platforms are being extensively studied, including polymeric and protein nanoparticles, micelles and lipid complexes.⁶⁵ Some systems have been approved for clinical use in the treatment of other diseases, thus leading to their interest in the field of TB; this is the case of polylactic-co-glycolic acid (PLGA) particles, which represent one of the most promising options for drug delivery and have even started undergoing clinical trials.⁶⁶ Another interesting construct was proposed by Trousil et al., where rifampicin, a first line anti-TB drug, was encapsulated in polymer nanoparticles that also contained a Förster resonance energy transfer (FRET) system to measure drug release in real time.⁶⁷ In conclusion, the availability of efficient DDS for TB treatment would be a breakthrough, given that the durations and schedules of conventional treatments (3–4 drugs must be given for a 6–9 month-period) frequently lead to poor patient compliance and adherence to treatments.⁶⁸

Biosensors for TB diagnosis

A very active area for the application of nanotechnology against TB is in the development of biosensors. Virtually all the protein antigens involved in vaccine development have also been explored as reagents for diagnostic purposes, along with some valuable diagnostic antigens, namely the Ag85B, the MPB83 antigen, and a lipoglycan molecule, the lipoarabinomannan.³⁰ Two major strategies for active TB diagnosis have been proposed (Fig. 3 ): (i) the detection of secreted mycobacterial components either in biological fluids (sputum, urine, saliva) or in exhaled breath particles, and (ii) the search for human biomarkers associated with the disease, which mainly include tuberculosis-associated antibodies.

Fig. 3.

Biosensors to diagnose active TB cases are based on the identification of either molecular fingerprints secreted early on by the pathogen (proteins, lipids), or biomarkers secreted by host cells during an active infection, such as inflammation-associated mediators or antibodies. Figure created with Biorender.

The detection of molecules secreted by M. tuberculosis is promising, since an essential set of compounds is known to be secreted at the early stages of the disease. This is the case of the early secreted antigenic target of 6 kDa (ESAT-6) and the 10 kDa culture filtrate protein (CFP10), which are highly specific antigens of M. tuberculosis with genes in loci that have been deleted from mycobacteria other than M. tuberculosis. ¹¹ Not surprisingly, various efforts are being undertaken in the development of CFP10-ESAT6-detecting biosensors. Importantly, the same antigens are currently used to look for the presence of blood memory T-cells in the diagnosis of latent TB (a method usually performed through conventional cell assays).¹¹ As mentioned in the following section, alternative molecular signatures secreted by M. tuberculosis at early stages of infections are the cell-surface free lipids, though only one of them, the lipoarabinomannan, has been thoroughly explored and used in commercially available tests.⁶⁹ Most sensors that detect mycobacterial molecules are based on the detection of antigen-antibody reactions, i.e. they use specific antibodies for capturing the target molecule. Nanotechnology has allowed the development of very interesting methods to evidence antigen-antibody reactions. Applied to TB diagnosis, transduction strategies include naked-eye detection through nanogold-based lateral immunochromatography,⁶⁹ optical immunosensors fabricated on liquid crystals or plasmonic surfaces,70., 71., 72. and a set of electrochemical immunosensors, which have been based on amperometric, impedimetric, potentiometric and conductometric measurements, sometimes fabricated on innovative substrates.73., 74., 75., 76., 77. For their part, DNA-based tests are expensive and less likely to be suitable in point-of-care settings; regardless, the use of whole genome sequencing has led to the identification of regions associated with resistance to multiple drugs in M. tuberculosis clinical isolates, including single nucleotide polymorphisms (SNPs), insertions and deletions.⁷⁸ Recent studies have proposed a targeted sequencing approach for the identification of such mutations, which has been successfully achieved in sputum samples.79., 80. Therefore, the development of nanotechnology-based sequencing methods to address this approach constitutes a worthy challenge.

On the other hand, the detection of biomarkers secreted by host cells is an attractive strategy, especially for point-of-care settings where biosafety facilities are absent and bacterial-containing samples would pose a risk. Because the spreading and most common presentation of active TB is a chronic pulmonary disorder associated to inflammation, various proinflammatory markers have been explored as a signature of the disease. According to original and systematic analyses, good diagnostic performances are achieved only with multiple markers,81., 82., 83. which have been detected through proteomic technologies.⁸¹ The development of rapid tests based on these markers seems very attractive; however, because other inflammatory diseases give rise to increased levels of inflammatory molecules as well, the specificity of the tests can be compromised. Incidentally, two recent studies have shown that the diversity of gut microbiota is clearly altered during pulmonary TB,⁸⁴ leading to discriminatory metabolic profiles in faecal samples between healthy and infected individuals.84., 85. Although more research is needed to identify the metabolites with the highest prognostic value, the diagnostic/monitoring potential is relevant, especially for smear- and culture-negative patients.

Among the many molecular interactions known in biology, antigen recognition by antibodies represents one of the most specific. During active TB, the constant release of M. tuberculosis antigens gives rise to antibody responses of little value for fighting the disease, though they present a great opportunity for diagnostics. Antibody responses against the bacilli are known to be a hallmark of TB disease, and there have been antibody signatures specifically associated with active TB infection.30., 32., 33., 59., 60., 81. As for some other diseases, gold-based lateral flow immunochromatography has been a common point-of-care platform in the search for serum antibodies. The naked-eye test is based on the use of colloidal (colored) gold for tagging the total antibodies of a human sample, and the immobilization of control and test antibodies onto functional lines of lateral flow chromatography (Fig. 3). The technique is rapid and simple, and has been used in the development of commercial serological tests for the diagnosis of TB.⁸⁶ Unfortunately, the early spread of such tests without international guidelines led to terrible misdiagnosis problems. The tests were unable to consistently achieve enough sensitivity or specificity, and a policy statement published by the WHO encouraged further research to identify new alternatives for serological tests before they can be recommended.⁸⁶ At present, antibody signatures against some selected antigens appear among the best biomarkers to achieve the sensitivity and specificity combination accepted by the WHO for a point-of-care diagnostic test.⁸¹ However, important issues remain to be addressed. Some of the best protein antigens for inducing antibodies are those known to evoke strong protective responses. This is the case, for instance, with the CFP10 and the ESAT-6 (as pointed out above). Thus, the search of antibodies against potential components of future vaccines restrains the development of new diagnostics based on these markers. The heterogeneity of antibody responses to proteins in human populations represents another problem: it is known that the high polymorphic nature of antigen-presenting molecules (the MHC molecules) leads to very different anti-protein antibody responses,87., 88., 89. hampering the selection of signatures for new tests. Finally, although the use of nanotechnology could significantly improve the sensitivity of serological tests, until now only a few attempts have addressed the development of novel platforms in the search for tuberculosis-associated antibodies.

Monitoring of TB treatment outcome

Most of the technologies mentioned above have also been explored to monitor the outcomes of TB drug treatments,⁹⁰ with one of the top interests being the prevention of contact exposure to highly infectious aerosols from active TB. At present, nanotechnology strategies are emerging to allow the implementation of friendly techniques for TB monitoring. One of the most promising approaches is the use of nanodevices to concentrate biomarkers without requiring special equipment. A great example is the use of copper & polymer nanocages that selectively harvest biomarkers before the application of conventional immunoassays.⁹¹ Additional studies are still needed to specify a molecular fingerprint of TB treatment success,⁹⁰ but lipids will probably be part of the basis of such a predictive signature.

Current advances in nanomedicine to exploit mycobacterial lipids

As described in the last section, the possibilities offered by nanotechnology for fighting TB are expanding rapidly. However, most of the explored approaches were based on molecular hydrosoluble components of the bacilli, while the outstanding variety and originality of lipids has been largely neglected. Mycobacterial lipids have qualities that allow them to be useful in different products, as they constitute important drug targets, highly specific markers of disease and well-described modulators of innate immunity, or highly specific antigens. Moreover, the exploitation of lipids as modulators of the immune response may lead to developments against additional infectious or chronic diseases, well beyond the field of TB research. During the past decades, the use of polar compounds easily managed in aqueous solutions has dominated in biomedical research. Now, the advances in nanotechnology allow for controlled design of non-hydrosoluble materials. In the following sections, we overview the strategies used to exploit M. tuberculosis lipids in distinct applications. We present different sections about pioneering works to exploit lipids through nanotechnology, according to the type of application intended. Then we discuss trending nanotechnology approaches that should open the gates to combatting TB in the future.

Lipids for vaccines and immunotherapies – modulators of the immune response

The capacity for mycobacteria to activate the immune response has been recognized since the discovery of the bacillus. Preparations from killed bacteria and cell-wall mycobacteria are commonly used as adjuvants in research and animal immunizations,⁵⁶ i.e. as boosters of vaccine immunogenicity. The chemical structure of mycobacterial cell-wall components is particularly original, sometimes with poor likelihood of enzymatic degradation by human cells. They have, for instance, arabinosyl and galactosyl residues of 5- instead of 6-member rings (as is common in mammalian cells), lipopeptides with D-aminoacids, unique sugar moieties, and very atypical fatty acyl tails.17., 93. Including these molecular features, various amphipathic, free lipids have been identified as ligands of innate immune receptors linked to proinflammatory signaling pathways, such as the Toll-like receptor (TLR)-2, Dectin-2, the DC-SIGN family of receptors and the Mannose Receptor (MR). M. tuberculosis lipids identified as ligands of these mannose-recognizing receptors are the lipoarabinomannan (LAM), the lipomannan (LM) and the phosphatidyl-inositolmannosides (PIMs), all located in the M. tuberculosis cell-wall and known to be trafficked into phagocytic and toward bystander cell-membranes.22., 94. Other lipids, namely a dimycoloyl trehalose (DMT) and some simpler synthetic analogs, are known to bind the C-type lectin receptor Mincle, another proinflammatory receptor of innate immunity.95., 96. On the contrary, some other amphipathic lipids from M. tuberculosis (or structural variants of those known as immune activators) are known to downmodulate immune response mechanisms such as T-cell proliferation, the release of proinflammatory cytokines, the expression of cellular receptors, or the phagosomal maturation pathway.39., 94., 97., 98., 99. These molecules are therefore thought to act as virulence factors during M. tuberculosis infections, and they offer applications for therapeutics/vaccine interventions, either to increase or downmodulate the immune response. In contrast to heavily proinflammatory bioactive lipids, such as the lipopolysaccharides from E. coli, ¹⁰⁰ mycobacterial proinflammatory lipids are part of therapeutic preparations currently in use, including the BCG vaccine, which is widely used to vaccinate newborns and as immunotherapy against bladder cancer.¹⁰¹ While the macromolecular lipoglycans, LAM and LM, are easily manipulated in water, special methodologies are needed to use the more hydrophobic molecules in biological systems.

At present, some studies have been done to exploit mycobacterial lipid immunomodulators through nanotechnology. Using purified lipids, our group proposed a biomimetic vaccine carrier composed of a PIM-decorated silica-based mesoporous structure.¹⁰² In this work, PIMs were extracted from M. tuberculosis H37Rv type strain and purified to formulate large unilamellar liposomes along with phosphatidylcholine. Then, colloidal PIM-displaying silica was prepared by the vesicle fusion method (Fig. 5). The resulting particles were found to interact readily with antigen-presenting cells (THP1-derived macrophages). Both TLR-2 involvement and a rapid particle internalization process were promoted with this colloidal preparation, showing its potential as a subunit vaccine carrier.

Fig. 5.

M. tuberculosis phosphatidylinositol mannosides (PIMs), activators of innate immunity, were displayed in lipid bilayers to decorate mesoporous SBA-15 silica, to obtain biocompatible particles with adjuvant activity for vaccine purposes. Figure reproduced with permission from.¹⁰²

Lipids and vaccines – antigens that specifically activate the cellular immune response

Contrary to what was thought for many decades, the variety of antigens recognized by T-cells to specifically activate the immune response importantly includes non-protein microbial molecules, namely amphipathic lipids.¹⁰³ The recognition of these nonprotein antigens is mediated by CD1, a noncanonical antigen-presenting molecule, and is involved in the response against a set of microbial pathogens. In M. tuberculosis, the lipids identified as CD1-restricted T-cell antigens are mycolic acids (MA), glucose- and glycerol-monomycolates (GMM, GroMM), LAM, PIMs, diacyltrehalose (DAT) and mannosyl-phosphomycoketide (MPM),36., 37., 95., 103. with fine specificities for precise structures of the lipid tails.36., 104., 105., 106. We know now that lipid-specific T-cells are activated and proliferate upon M. tuberculosis infection, or after BCG vaccination, suggesting the usefulness of mycobacterial lipids as vaccine subunits, along with the possibility of monitoring latent TB through the detection of memory T cell responses to particular lipids.107., 108., 109. In view of this, various nanotechnology approaches are being explored to develop preparations for properly displayed lipids. Shang et al. prepared micellar nanocarriers comprised of polyethyleneglycol-polypropylene sulfide copolymers to deliver the highly hydrophobic lipid antigens MAs.¹¹⁰ A potent T cell response to the lipid was found using this approach, although a mixture with hydrophilic antigens could not be achieved. The design of mycobacterial lipid-displaying liposomes has been one of the most explored and promising strategies, especially since the description of synthetic approaches to obtain mycobacterial lipid analogs. Liposomes composed of various antigenic mycobacterial lipids, also known as mycosomes, have been widely explored. Mycosomes have included sulfatides, PIMs and MA synthetic analogs, sometimes in combination with protein antigens.111., 112., 113., 114., 115.

All the studies using mycosomes have led to strong immune responses due to the lipid components, either in vitro¹¹³ or in vivo.111., 116., 117. Liposomes decorated with dendritic cell-targeting molecules have been reported as well.¹¹⁸ Moreover, the use of mycobacterial lipids has been shown to induce a stronger immune response than that achieved with currently used adjuvants, such as monophosphoryl lipid A (MPLA) or alumina.¹¹⁶ With another approach, Das et al. prepared M. tuberculosis lipid-loaded chitosan nanoparticles, where isolated lipids were physisorbed to the polysaccharide matrix, to obtain highly immunogenic constructs.¹¹⁹ Finally, a combination of mycobacterial lipid-containing liposomes with inorganic nanostructured carriers has been proposed by our group. The resulting hybrid material has interesting advantages, namely increased stability of the lipid-displaying assemblies, as compared to liposomes, and the ease of cargo encapsulation through a simple adsorption process.¹⁰²

TB diagnosis – lipids as markers of disease

As mentioned above, M. tuberculosis lipids have unique structural features which may lead to the unequivocal detection of the bacilli in human samples (Fig. 4 ). Lipids are known to be secreted by infected cells,²² and recent reports have highlighted the interest in detecting MA, tuberculostearic acid (TSA) and a phosphatidylinositol as biomarkers of M. tuberculosis infection.120., 121., 122., 123., 124. The identification of these lipids can be performed rapidly in blood, sputum or exhaled breath aerosols; however, their analyses require sophisticated equipment, such as mass spectrometers, as well as highly trained personnel. Current technologies are not suitable in the search for hydrophobic molecules, which are commonly released within lipid micro- and nanovesicles. That is why only LAM, a hydrosoluble lipoglycan, has been used as a biomarker of TB disease in rapid tests. During an infection, LAM is released in urine, thus serving as an efficient marker for urinary LAM-based tests.¹²⁵ Commercial point-of-care tests, based on lateral flow assays are available from Abbott Diagnostics (Lake Bluff, USA), the AlereLAM test, and from Fujifilm (Tokyo, Japan), the “FujiLAM”. These tests use one or two drops of urine on a strip, and polyclonal or monoclonal antibodies tagged with nanogold to allow naked-eye detection (Fig. 6). Until now, the search for LAM in urine samples has been the only method recommended by the WHO for TB diagnosis, specifically in HIV-positive patients. Meta-analyses involving the urine AlereLAM test have shown specificities up to 98 %, but poor sensitivities (<50 %).¹²⁵ Therefore the development of improved LAM-based tests is underway, with the “FujiLAM” as the first, advanced, next-generation LAM test. “FujiLAM” is based on the combination of two high-affinity antibodies, along with a silver-based amplification step, which significantly increases test sensitivity (Fig. 6).¹²⁶ One of the main approaches for improving the LAM-based test in urine is identifying determinants of LAM immunogenicity. Because sugar heterogeneity is common throughout M. tuberculosis genotypes, the search of key glycan epitopes and high-affinity antibodies is currently an objective.¹²⁷

Fig. 4.

The lateral flow immunochromatography test for the detection of tuberculosis-associated antibodies. In this test, antibody-containing samples are placed on a nitrocellulose strip, where the antibodies are labelled with gold nanoparticles. After labelling, the colored molecules follow a capillarity-induced flow through the pad, where functional lines will immobilize biomarker and/or control antibodies, allowing a result visible to the naked-eye. Figure reproduced with permission by.⁹²

Fig. 6.

(a) Pre-January 2014 LAM strip test manufacturer´s reference card illustrating visual intensity grades 0–5; (b) January 2014 new LAM strip test manufacturer’s reference card illustrating visual intensity grades 0–4. In this reference card, the first positive band corresponds to the grade-2 intensity band in the old, pre-January 2014 reference card. Permission granted by Abbot Corporation to publish this figure.

Using a distinct approach, Hiatt & Cliffel developed a quartz crystal microbalance (QCM) immunosensor, where anti-LAM rabbit antibodies were immobilized to capture the lipoglycan antigen, and mass changes were detected through the oscillation of the piezoelectric material. Preliminary binding studies showed that using this method, LAM can be detected in <20 min.⁷¹ For their part, Crawford et al. developed a sandwich immunoassay by immobilizing monoclonal antibodies against mannose-capped LAM on a glass-supported 200 nm-thick gold layer. Then, the capture lipoglycan was detected through surface-enhanced Raman spectroscopy (SERS)-active nanoparticles via a second antibody (Fig. 7). The method showed promising results in a small collection of sera from TB-positive and healthy individuals.¹²⁸

Fig. 7.

Detection of LAM through a SERS-double sandwich immunoassay. LAM is captured on an antibody-functionalized gold surface (A). Then, SERS-active gold nanoparticles (B, C) are bound to LAM through a second anti-LAM antibody to promote the generation of LAM-specific Raman shifts (D). Figure reproduced with permission by.¹²⁸

Recently, a novel interferometry platform was developed to detect LAM in undiluted urine. The specific capture of LAM is carried out by using a monoclonal antibody and a Mach Zehnder interferometer chip, integrated within a microfluidic chamber, is used to detect the macromolecular binding. By optimizing several parameters, this system reached a limit of detection in the range of picograms (475 pg/mL), encouraging its further exploration as a point-of-care system.¹²⁹ Finally, a strategy for enhancing the sensitivity of LAM detection was proposed through a copper complex dye within a hydrogel nanocage, able to sequester the lipoglycan with a very high affinity, allowing its further characterization through an immunoassay. The technology was able to detect LAM in urinary samples from HIV negative individuals with active TB.¹³⁰

The interest in LAM as a useful biomarker of active TB has promoted the development of interesting techniques for harvesting the glycan biomarker to obtain highly concentrated samples: using a copper complex dye within polymeric nanocages, Paris et al. developed a bait-chemistry technique to harvest LAM from urine samples, leading to highly sensitive detection, which has been applied to other mycobacterial and host disease biomarkers.¹³⁰ For their part, Zheng et al. identified the presence of LAM in the surface of extracellular vesicles (EVs) released from bacteria during active TB, then used conventional EV-recovery techniques to obtain concentrated samples where conventional assays were performed with great sensitivity results.¹³¹

In addition to the developments described for detecting LAM, some other studies have addressed optical detection methods for MA, the long-chain specific carboxylic acids of mycobacteria, with highly atypical chemical structures.

Interestingly, MA has been shown to give specific Raman signals when using surface enhanced Raman spectroscopy methods. Mühlig et al.¹³² demonstrated that silver nanoparticle-mediated SERS analysis of whole bacteria gives rise to specific Raman signals of MA bearing different functional groups. Furthermore, the authors proposed a closed system for easily disrupting mycobacterial cells, followed by an optical analysis within a lab-on-a-chip device (Fig. 8). Using isolated bacteria to obtain lipid samples, the identification of MA using silver nanoparticle-mediated SERS has been confirmed by Perumal and coworkers.¹³³ Nonetheless, these SERS-related techniques have not been validated in clinical specimens yet, where the number of bacterial cells and complexity of the samples may be restrictive.

Fig. 8.

A promising closed system for the identification of mycobacterial MAs has been developed through the combination of a module to disrupt bacterial cells with a lap-on-a-chip (LOC)-SERS device. Figure reproduced with permission by.¹³²

TB diagnosis – the search for TB-associated antibody responses against lipids

The identification of antibodies in blood or serum from individuals with active TB stands out as one of the most suitable diagnostic techniques for point-of-care settings. Antibody detection may be adapted to a variety of nanotechnologies, through different transduction biosensing methods, such as electrochemistry, a variety of optical techniques (visible and Raman spectroscopies, interferometry) and microbalances. Methods can be easily adapted to measure antibody classes found in distinct samples, such as serum or saliva. However, diagnostic approaches using antibodies as biomarkers have failed so far to fit the requirements for sensitivity and specificity needed to replace the traditional microscopy observation of stained smears.⁸¹ In our opinion, two main issues are responsible for such demotivating results: (a) the antigens used as immunoreagents for antibody tests are either nonspecific or insufficient to provide sensitivity by themselves, and (b) there is still a scarcity of technologies for exploiting the variety of non-protein mycobacterial antigens to which antibody responses are known to appear. Most serological studies have pointed to the convenience of using multiple antigen cocktails, instead of individual reagents. Anti-protein antibodies are extremely variable between individuals, in part due to the polymorphism of the antigen-presenting molecules involved in their production.⁸⁷ In contrast to these facts, antibody response to non-protein lipid antigens relies on processes that are non-significantly polymorphic across human populations,²⁸ and antibody responses during TB are stronger against lipids than proteins.¹³⁴ In spite of this, the use of nanotechnology to exploit the large variety of lipid antigens from the tubercle bacillus has been poor. Although glycolipids have been identified among the best reagents to capture antibodies associated to active TB,30., 32., 33., 59., 60. there is a lack of methods aimed at exploiting the non hydrosoluble reagents. Regarding this, the following section briefly describes a set of emerging nanotechnologies through which lipids can be assessed for the development of TB control tools.

Emerging nanotechnologies – an opportunity to exploit M. tuberculosis lipids

The classical definition of lipid was “Hydrophobic compounds, which are soluble in organic solvents as chloroform, ethers or alcohols”. At present, although lipids are defined according to their biosynthetic and functional issues,¹³⁵ most of them are indeed hydrophobic. While this feature represents the basis of cell individuality and compartmentalization (with lipid membranes as our structural divisions), dealing with lipids in the laboratory has not been easy, as common methodologies for cell biology and biochemistry studies occur in aqueous environments. With advances in nanotechnology, the possibility of obtaining synthetic materials or self-assembled arrays to produce well controlled nanostructures is greatly enhancing the potential to exploit lipids. Pioneering works with mycobacterial lipid-containing liposomes have resulted in powerful immunogenic structures, also known as mycosomes.112., 113., 114., 115., 116., 117., 119. A set of advances in biophysics and nanotechnology are opening up novel possibilities for the study of lipid-protein interactions. One example is the use of bicelle systems i.e., lipid bilayer-containing disc-like particles produced in the presence of detergents.¹³⁶ Bicelle structures are useful for nuclear magnetic resonance (NMR) studies and were useful in studying the location of proteins in the bilayer and the effect of different lipids on enzyme binding in Escherichia coli membranes.¹³⁷ A recent work by Patrick et al. demonstrated the great potential of novel bicelle structures in the study of protein-lipid interactions in mycobacterial mimetic membranes.¹³⁸ Supported lipid bilayers (SLBs) constitute another promising approach to the display of lipids in either fundamental studies or technological applications, such as diagnostics or targeted therapies.139., 140. Recent advances have led to friendly and scalable procedures for obtaining SLBs with different lipid compositions and a wide variety of solid surfaces, including porous, conducting or optical active materials.140., 141., 142. Using this approach, our group developed PIM-containing mycosomes on the surface of SBA15-type mesoporous silica, a biocompatible material, to generate a stable lipid-displaying construct with high loading capabilities, leading to specific interactions with cells from the innate immune system.¹⁰² In recent years, the use of SLBs had been extended to biogenic extracellular vesicles (EVs) i.e., nanosized lipid vesicles that, secreted by cells, better represent the complexity of natural membranes.¹⁴³ This approach has facilitated the development of natural vesicle-supported lipid bilayers, where the study of lipid interfaces seems promising, and interesting technological applications of mycobacterial EVs¹⁴⁴ may be readily exploited.

Finally, the exploration of specific materials for the simple immobilization of mycobacterial lipids (through physisorption or chemisorption) is worthy of mention. Cano-Velázquez et al. proposed a method to monitor anti-mycobacterial antibodies with drop-casted antigenic lipids on polydimethylsulfoxide (PDMS), a hydrophobic transparent polymer commonly used in miniaturizable devices, along with interferometric measurements.¹⁴⁵ For their part, Das et al. demonstrated the successful immobilization of M. tuberculosis lipids on chitosan nanoparticles, allowing the production of a powerful immunogenic material for vaccine applications.¹¹⁹

Conclusions

Current knowledge about the biology of M. tuberculosis has pointed to lipids as the key molecular features of the microbe. Outstandingly abundant and structurally unique, lipid-containing compounds of M. tuberculosis are involved in cell recognition, activation or suppression at various steps of the immune response and antigenicity. Biogenic or synthetic lipid analogs have a great potential in the development of tools against TB. Nonetheless, if the search for tools against TB has been neglected for decades (with a >100-year-old vaccine, an increasing number of multidrug resistant strains, and no friendly diagnostic tests for the most affected populations), the study of lipids is in an even worse situation. Working with hydrophobic, bioactive molecules has been associated with cumbersome steps and techniques which are not easily implemented at the large scale. In contrast, nanotechnology provides a variety of strategies allowing for easy management of lipids: from the production of colloidal structures, where lipids can be displayed and managed in aqueous milieux, to the design of hybrid particles where lipid bioactivities may determine the fate of nanoconstructs intended for biological systems. Molecular engineering, as achieved through bottom-up nanotechnology, may therefore represent a solution to finally making use of lipids, the most neglected useful markers of the tubercle bacillus.

Financial support

The authors thank the financial support from CONACyT (Mexico) through Grant CF2019-53395, and from DGAPA (UNAM) through Grants IT200421 (PAPIIT) & PE201622 (PAPIME).

Conflict of interest

The authors declare no conflict of interest.