Identification of Mycobacterium tuberculosis adherence-mediating components: a review of key methods to confirm adhesin function.

Anti-adhesion therapy represents a potentially promising avenue for the treatment and prevention of tuberculosis in a post-antibiotic era. Adhesins are surface-exposed microbial structures or molecules that enable pathogenic organisms to adhere to host surfaces, a fundamental step towards host infection. Although several Mycobacterium tuberculosis adhesins have been identified, it is predicted that numerous additional adherence-mediating components contribute to the virulence and success of this pathogen. Significant further research to discern and characterize novel M. tuberculosis adhesins is, therefore, required to gain a holistic account of M. tuberculosis adhesion to the host. This would enable the identification of potential drug and vaccine targets for attenuating M. tuberculosis adherence and infectivity. Several methods have been successfully applied to the study and identification of M. tuberculosis adhesins. In this manuscript, we review these methods, which include adherence assays that utilize wild-type and gene knockout mutant strains, epitope masking and competitive inhibition analyses, extracellular matrix protein binding assays, microsphere adhesion assays, M. tuberculosis auto-aggregation assays, and in silico analyses.

Introduction

Between 1990 and 2013, the global tuberculosis (TB) prevalence and mortality rates decreased by 41% and 45%, respectively (1). Despite the apparent progress in TB control, 9 million people developed TB and 1.5 million people succumbed to the disease in 2013, reiterating that this disease is still a severe global public health challenge (1). The World Health Organization’s post-2015 global TB strategy aims to reduce the 2015 worldwide TB incidence and mortality rates by 90% and 95%, respectively, by the year 2035 (1). The achievement of these goals is dependent on the identification of novel biomarkers and targets for which new innovations for TB diagnosis, treatment, and prevention can be elucidated. These goals may be realized by intensifying scientific research into the interaction of the TB etiologic agent, Mycobacterium tuberculosis, with its human host.

A vital precursor to host colonization and the deploying of bacterial virulence factors, such as toxins and effector proteins, is the adhesion of the pathogen to host cells. The initial binding to host cells occurs via weak and non-specific forces that are brought about by the physicochemical properties of the host surface. Thereafter, specific and transient receptor interactions enable a stronger anchorage of the pathogen to the host surface. Lastly, specific and high-affinity attachment is achieved by the interaction of bacterial adhesins with host receptors (2).

Adhesins are monomeric or filamentous cell-surface molecules or structures that facilitate bacterial attachment to host surfaces. This process is a pivotal precursor to bacterial invasion and colonization of the host, by limiting the clearance of the pathogen through shear stress and by activating key virulence factors that are associated with infection (3). Furthermore, bacterial adhesion to host cells results in a manipulation of host cell signalling, which promotes the spread of the pathogen and allows for the microbe to evade the host immune response (3). Adhesins also bind to receptors of the extracellular matrix (ECM), thereby activating signal transduction cascades in the host cell. These adhesins, termed microbial surface components recognizing adhesive matrix molecules (MSCRAMM), use receptors (such as integrins) to adhere to ECM molecules, including bone sialoprotein, collagens, elastin, fibronectin, glycosaminoglycans, laminin, thrombospodin, and vitronectin (4).

The ability of environmental bacteria to adhere to natural biopolymers (e.g., cellulose fibres) can be beneficial to these organisms, such as by providing access to nutrient-rich sources. In addition, the adhesion of bacteria to synthetic biomaterials (e.g., medical devices), such as those biomaterials that have been exposed to body fluids and contain adhered serum proteins, results in severe clinical consequences (4). Adhesin-mediated adhesion to host surfaces also enables biofilm formation (5), a multicellular lifestyle of bacteria that is associated with bacterial tolerance to antimicrobial agents, host defences, and other environmental challenges.

Due to their fundamental role in the host-pathogen interaction, bacteria have evolved a large variety of adherence-mediating components and strategies to enable their adhesion to the host. Pili are polymeric hair-like surface structures that contain an adhesin at its distal tip. They are the most researched class of bacterial adhesins, particularly in Gram-negative bacteria. Gram-positive organisms were also subsequently found to elaborate pili, and their function in adhesion and pathogenesis is being deciphered (6). A large arsenal of non-pilus adhesins have also been identified in bacteria. These adhesins are capable of interacting with host molecules, including transmembrane proteins and ECM components (7).

It is well-established that during infection, M. tuberculosis interacts with and enters host cells, including macrophages, epithelial cells, and dendritic cells. This process is crucial to the organism’s survival, replication, and dissemination (8). However, the adhesins mediating these fundamental processes has been largely unexplored and underappreciated (9). Although several M. tuberculosis proteins have been documented to function in the attachment to host components (Table 1), computational tools have predicted an abundance of additional uncharacterized M. tuberculosis adhesins (Table 2).

Table 1

A list of the currently known Mycobacterium tuberculosis adhesins

Adhesin	Gene (s)	Mediates adhesion to	Reference (s)
19-kDa antigen	Rv3763	Monocytes and macrophages	(24, 25)
Alanine- and proline-rich antigen (Apa)	Rv1860	Pulmonary surfactant protein A and macrophages	(24, 35)
Antigen 85 complex	Rv0129c, Rv1886c, Rv3803c, and Rv3804c	Fibronectin and macrophages	(17, 24)
Cpn60.2 molecular chaperone	Rv0440	Macrophages	(16, 36)
Curli pili	Rv3312A	Laminin, M. tuberculosis, macrophages, and epithelial cells	(13, 14, 19, 28)
DnaK molecular chaperone	Rv0350	Macrophages	(36)
Early secreted antigen ESAT-6	Rv3875	Laminin	(37)
Glutamine synthetase A1	Rv2220	Fibronectin	(23)
Glyceraldehyde-3-phosphate dehydrogenase	Rv1436	Possibly fibronectin (as occurs in group A streptococci)	(38)
Heparin-binding hemagglutinin adhesin	Rv0475	M. tuberculosis and epithelial cells	(12, 27)
Laminin-binding protein	Rv2986c	Laminin	(20)
L,D-transpeptidase	Rv0309	Fibronectin and laminin	(18)
Malate synthase	Rv1837c	Fibronectin, laminin, and epithelial cells	(21)
Membrane protein	Rv2599	Collagen, fibronectin, and laminin	(18)
Mycobacterium cell entry-1 protein	Rv0169	Epithelial cells	(26)
N-acetylmuramoyl-L-alanine amidase	Rv3717	Fibronectin and laminin	(18)
PE-PGRS proteins	Rv1759c	Fibronectin	(22)
	Rv1818c	M. tuberculosis and macrophages	(11)
Protein kinase D	Rv0931c	Brain endothelia and laminin	(15)
PstS-1 (38-kDa antigen)	Rv0934	Macrophages	(24)
Type IV pili	Rv3654c-Rv3660c	Possibly macrophages and epithelial cells	(39)

Table 2

A list of the putative Mycobacterium tuberculosis adhesins identified by computational analyses [a]

Annotation	Putative adhesin-encoding genes
ESAT-6-like proteins	Rv1037c and Rv3890c
Hypothetical proteins	Rv0538, Rv0584, Rv0679c, Rv0988, Rv1115, Rv1116A, Rv1269c, Rv2190c, Rv2730, Rv3036c, Rv3067, Rv3207c, Rv3337, Rv3491, Rv3705c, Rv3811, and Rv3822
Immunogenic protein Mpt64	Rv1980c
Low-molecular-weight antigen CFP2	Rv2376c
Major secreted immunogenic protein Mpt70	Rv2875
Mce-family protein	Rv0590
PE family proteins	Rv1089 and Rv1788
PE-PGRS family proteins	Rv0578c, Rv0872c, Rv0977, Rv0980c, Rv1450c, Rv1803c, Rv2126c, Rv2487c, Rv2853, Rv3345c, Rv3507, and Rv3514
Periplasmic phosphate-binding lipoprotein	Rv0932c
Possible chitinase	Rv1987
Possible conserved lipoproteins	Rv1881c and Rv3576
PPE family proteins	Rv0304c, Rv0354c, Rv0755c, Rv1548c, Rv2353c, Rv2356c, and Rv3159c
Probable cutinase precursor	Rv1984c
Probable lipoprotein aminopeptidase	Rv0418
Probable penicillin-binding protein	Rv0050
Soluble secreted antigen Mpt53 precursor	Rv2878c
Superoxide dismutase	Rv3846

Putative adhesins identified from Kumar et al (18) and the MycobacRV web server (http://mycobacteriarv.igib.res.in)

A major challenge to the treatment of TB patients is the escalating rates of drug-resistant M. tuberculosis strains. A novel strategy for TB treatment and prevention involves targeting bacterial determinants of infection, rather than those of survival. This would incapacitate the pathogen and ensure a natural clearance of the bacilli, without imposing selective pressure and amplifying the number of drug-resistant forms of the bacterium. In addition, it would prevent the release of harmful products (e.g., endotoxins) from lysed bacterial cells, which occurs with the use of bactericidal agents (2). However, due to a plethora of M. tuberculosis adhesion molecules, anti-adhesion TB therapeutics should target a combination of the major adhesins, possibly as an adjunct to the conventional TB therapeutics, to be effective. For adhesion inhibition to be a viable intervention for TB treatment and prevention, further clarity on the entire repertoire of M. tuberculosis adhesins is needed.

The purpose of this work is to review the major methods available for the identification of M. tuberculosis adhesins. These methods include adherence assays that utilize wild-type and gene knockout mutant strains, epitope masking and competitive inhibition analyses, ECM protein binding assays, microsphere adhesion assays, M. tuberculosis auto-aggregation assays, and in silico analyses.

Adherence assays that utilize wild-type and gene knockout mutant strains

Comparing the adhesion capacity of wild-type and gene knockout mutant strains is a widely used approach to identify genetic determinants of adhesion and infection. In this assay, wild-type and gene knockout mutant strains infect host cells (macrophages or epithelial cells) in microtitre plates. A multiplicity of infection of between 1:1 and 10:1 (bacilli to host cells) and an infection time of 1-4 hrs (usually 1 hr) is generally used. This is followed by washing to remove non-adherent bacilli and a lysis step (usually with 0.1% Triton X-100) to disrupt the host cells. Thereafter, serial dilutions of the cell lysates are plated onto agar plates to enumerate viable adherent bacterial cells, including those that have invaded the host cells. Since intracellular bacilli would have had to adhere prior to invading the host cells, the invaded bacteria also represent the total adhered population. However, one may distinguish between the adherent versus invaded fraction using the aminoglycoside protection assay (10) to quantify the invaded bacilli.

This colony-forming unit (CFU)-based adherence assay is advantageous in that it enables the cell association efficiency to be calculated from the total number of bacilli in the infection inocula. In addition, it is a simple and relatively low-cost method that requires no specialized equipment. This assay has been successfully used to identify the adhesin function of the Pro-Glu (PE) polymorphic GC-rich repetitive sequence (PGRS) protein encoded by Rv1818c (11), the heparin-binding hemagglutinin adhesin (HBHA) (12), and curli pili (MTP) (13, 14). In these studies, the deletion of the encoding genes (Rv1818c, Rv0475, and Rv3312A) resulted in a significant reduction in M. tuberculosis adhesion to macrophages and/or epithelial cells.

This assay has also been used to measure the binding of M. tuberculosis to ECM proteins. Be et al (15) incubated wild-type and ΔpknD mutant M. tuberculosis strains with laminin for 1.5 hr. The wells were washed to remove unbound bacteria. The matrices were then lysed and plated onto agar plates for CFU assessment. A two-fold reduction in the adhesion to laminin was observed for the mutant, indicating that protein kinase D is a laminin-binding adhesin.

Epitope masking and competitive inhibition analyses

The deletion of essential (and even some non-essential) M. tuberculosis genes pose a technical challenge to the generation of gene knockout mutants. This prevents the utility of the previously described adherence assay to assess the possible adhesin function of the products that these genes encode. Epitope masking and competitive inhibition analyses are a useful alternate method in such cases. This strategy makes use of purified/recombinant proteins or antibodies against the purified/recombinant protein to block the binding of M. tuberculosis to the host components.

Hickey et al (16) used M. tuberculosis that was either untreated or treated with anti-Cpn60.2 anti-bodies to infect macrophages. Unattached bacteria were removed by washing and attached bacilli were identified by CFU counting and microscopy. The authors found that the M. tuberculosis strains with masked Cpn60.2 displayed a significant decrease in the adhesion to macrophages compared with the unmasked M. tuberculosis. The authors also used macrophages that were either untreated or treated with recombinant Cpn60.2 protein, prior to the addition of M. tuberculosis. The results indicated that M. tuberculosis displayed a decreased adhesion to Cpn60.2-treated macrophages compared with untreated macrophages. These findings implicate Cpn60.2 as a macrophage-associated adhesin.

A similar masking strategy was used by Abou-Zeid et al (17) to identify the BCG85 complex of Mycobacterium bovis BCG, which is also present in M. tuberculosis as the antigen 85 complex, as a fibronectin-binding adhesin. In this report, the authors showed that M. bovis BCG that was pre-incubated with anti-BCG85 antibodies demonstrated a diminished capacity to adhere to fibronectin compared with unmasked M. bovis BCG. They also found that fibronectin that was pre-incubated with purified BCG85A and BCG85B proteins inhibited the adhesion of M. bovis BCG to fibronectin.

ECM protein binding assays

In addition to their ability to bind to host cells, M. tuberculosis also adheres to ECM proteins, particularly collagens, fibronectin, and laminin. Therefore, numerous studies have identified M. tuberculosis adhesins based on their ability to bind to ECM proteins. Kumar et al (18) used an image-based enzyme-linked immunosorbent assay (ELISA) to demonstrate the binding of the recombinant Rv0309, Rv2599, and Rv3717 proteins to immobilized ECM proteins (fibronectin, laminin, and/or collagen). Detection of protein binding was achieved using horseradish peroxidase conjugated anti-6x histidine antibodies and ortho-phenylenediamine dihydro-chloride.

Alteri et al (19) used flow cytometry to measure the binding of wild-type and MTP-deficient M. tuberculosis strains to ECM proteins, by the detection of anti-ECM antibody complexes, using Alexa Fluor conjugate. The binding of MTP to laminin was confirmed using a sandwich ELISA. In this experiment, the authors used varying concentrations of laminin to demonstrate that immobilized purified MTP protein bound to this ECM protein. Bound laminin was detected using anti-laminin antibodies and IgG peroxidase conjugate. Similarly, using flow cytometry and ELISA, Pethe et al (20) showed that the Mycobacterium smegmatis laminin-binding protein (which is also expressed by M. tuberculosis) binds to laminin.

Using ELISA and reverse ELISA, Kinhikar et al (21) demonstrated that purified recombinant malate synthase bound to fibronectin and laminin. This was confirmed by fractioning the recombinant protein using sodium dodecyl sulphate polyacrylamide gel electrophoresis, treating the Western blots with these ECM proteins, and detecting ECM protein binding using anti-ECM and secondary antibodies. These researchers also showed that recombinant malate synthase adhered to epithelial cells. This was done by detecting bound malate synthase using anti-malate synthase antibodies and flow cytometry. Using a similar immunoblotting strategy to Kinhikar et al (21), Espitia et al (22) and Xolalpa et al (23) found that the fractioned recombinant Rv1759c PE-PGRS protein and glutamine synthetase A1 bound to fibronectin, respectively.

Microsphere adhesion assays

Although microspheres have generally been used in M. tuberculosis phagocytosis assays, they also have application in the study of bacterial adhesion. Esparza et al (24) used 1 μm green fluorescent polystyrene microbeads coated with purified PstS-1 ABC phosphate transporter protein to interact with macrophages adhering to glass coverslips. Non-adherent microbeads were removed by washing and adherent microbeads were identified by analyzing randomly selected cells by confocal microscopy. Similarly, Diaz-Silvestre et al (25) used this technique to confirm the role of the 19-kDa antigen (LpqH glycolipoprotein) as an adhesin.

Chitale et al (26) used 0.3 and 1.1 μm latex beads coated with recombinant Mycobacterium cell entry-1 (Mce1) protein to infect HeLa cells cultured on glass coverslips. Bead association was assessed by light and transmission electron microscopy. The use of 1 μm fluorescent latex beads has also been described, where fluorescence emission gives an indication of bead association (26).

M. tuberculosis auto-aggregation assays

Apart from their function in bacteria-eukaryotic cell interactions, adhesins also mediate cell-to-cell contact between bacterial cells and are, thus, key contributors to bacterial aggregation and biofilm formation. Menozzi et al (27) reported that incorporating purified HBHA to M. tuberculosis cultures resulted in an aggregation of the bacilli, in a dose-dependent manner. Ramsugit et al (28) showed that a MTP-deficient M. tuberculosis strain displayed a significant reduction in bacterial aggregation and pellicle formation in detergent-free Sauton’s media compared with the wild-type strain. Both HBHA and MTP were subsequently shown to also function as adhesins in the host-pathogen interaction (12-14). The identification of adhesins that promote M. tuberculosis auto-aggregation could, therefore, enable the identification of potential adhesins that are associated with host interaction.

In silico analyses

Bioinformatics analyses have led to the identification of numerous potential M. tuberculosis adhesins. Kumar et al (18) screened all of the M. tuberculosis H37Rv proteins through several bioinformatics tools and identified 20 novel putative adhesins (Table 2). This was achieved using SPAAN adhesin predictor (29) at a threshold of Pad > 0.65, subcellular localization prediction algorithms [LOCtree (30), PSORTb (31), and SubLoc (32)], and removing potential human homologues that were identified by BLAST. This was followed by an argumentation-based approach to process ‘claims’ and ‘attacks’ from different algorithms. In addition, further filtering criteria for globular and low-molecular-weight proteins that were not previously reported in the literature were applied. Three of the prospective adhesins identified in that study (Rv0309, Rv2599, and Rv3717) were subsequently experimentally confirmed to bind to ECM proteins (18). This validates the usefulness of computational methods to identify M. tuberculosis adhesins.

The MycobacRV Database of Mycobacterial Vaccine Candidates provides information regarding possible adhesins of 22 human pathogenic mycobacterial species and their potential as vaccine candidates (33). This database uses SPAAN (29) at Pad>0.6 and PSORTb (34) at a very stringent cut-off to identify adhesin and adhesin-like proteins. Using the MycobacRV web server (http://mycobacteriarv.igib.res.in), 37 novel prospec-tive adhesins were identified in M. tuberculosis H37Rv (Table 2).

Conclusion

TB remains a major challenge. Despite global efforts to improve our understanding of M. tuberculosis pathogenesis, insufficient knowledge has been gained on the M. tuberculosis adherence mechanisms, the first (and arguably the most critical) step in the pathogenesis of this bacterium. Several M. tuberculosis adhesins have been identified; however, computational analyses have predicted a large number of additional adhesin-encoding genes in this pathogen’s genome. Several of these genes encode proteins belonging to the PE-PGRS and Pro-Pro-Glu (PPE) protein families. It, therefore, seems likely that key information regarding this organism’s adherence mechanisms could be obtained by the study of the adhesin capability of proteins belonging to these families.

With several methods available for the study of M. tuberculosis adhesion, as highlighted in this review, identifying novel M. tuberculosis adhesins is possible and should be explored further. Advances in gene disruption technologies open up the possibility of screening mutant libraries to identify M. tuberculosis genes involved in the pathogen’s adherence. Adapting the adhesion methodologies to enable high-throughput screening of libraries of potential anti-adhesion therapeutics is needed and may have value in the control of this ravaging disease.