Tuberculosis Bacteria’s Secreted Proteins Emerge as New Drug Targets

Tuberculosis remains one of humanity’s deadliest infectious diseases, and the bacterium responsible, Mycobacterium tuberculosis, owes much of its lethal success to a toolkit of proteins it exports beyond its own cell walls. A new review published in Frontiers in Tuberculosis by Anthony Musaazi Kaweesa and Divya Tiwari of the Blizard Institute at Queen Mary University of London brings together, for the first time in a single framework, the major secreted proteins of M. tuberculosis and evaluates how feasible it really is to turn them into drug targets. Their conclusion is both encouraging and sobering: secreted virulence factors are chemically accessible in ways that intracellular targets are not, yet most inhibitor programs targeting them remain stuck at the preclinical stage, hampered by low potency, off-target activity, and built-in redundancy among the proteins themselves.

The strategic logic behind targeting secreted proteins is compelling. Many classic tuberculosis drugs must cross the bacterium’s notoriously robust and impermeable cell wall to reach their cytosolic targets, and those that do make it inside can still be neutralized by efflux pumps or other xenobiotic defenses. Secreted proteins, by contrast, sit outside or on the surface of this barrier, at the interface between pathogen and host. A drug acting on them does not necessarily need to penetrate the cell wall at all, which may in principle bypass a whole class of resistance mechanisms. Moreover, disabling virulence factors rather than killing the bacterium outright could reduce the selective pressure that drives antibiotic resistance, a persistent threat in tuberculosis therapy, where current multi-drug regimens must be taken for months and adherence failures fuel drug-resistant strains.

The review begins with the machinery itself. M. tuberculosis exports its proteins through three principal systems: the ESX, or Type VII, secretion systems, of which there are five, designated ESX-1 through ESX-5; the highly conserved general secretory, or Sec, pathway; and the twin-arginine translocation, or Tat, pathway. ESX-1 is the best characterized and secretes a stable heterodimer of EsxA, also known as the 6-kDa early secretory antigenic target ESAT-6, and EsxB, or CFP-10. The ESX-1 system is encoded in the region of difference 1 locus, which is absent from all strains of the BCG vaccine, and its loss is thought to have contributed to the attenuation of the vaccine strain, underscoring just how central this system is to virulence. The Sec pathway uses ATPases SecA1 and SecA2 with the SecYEG channel to export unfolded proteins, while the Tat pathway moves folded proteins through TatA, TatB, and TatC. ESX-3 and ESX-5 have also been implicated in virulence, whereas ESX-2 and ESX-4 remain poorly understood.

EsxA and EsxB top the list of druggable virulence factors. EsxA, a small 6-kDa helical protein, dissociates from its 10-kDa partner under acidic conditions and forms membrane-permeabilizing structures that rupture the phagosome, allowing the bacterium to escape into the host cell’s cytosol. EsxA also dampens pro-inflammatory cytokine production by binding toll-like receptor 2 and promotes apoptosis by upregulating caspases. EsxB is less understood but stabilizes EsxA and interacts with the EccCb1 ATPase of the secretion apparatus. Inhibiting this axis has required creative chemistry because protein-protein interactions are notoriously difficult to disrupt with conventional small molecules. Nevertheless, proof-of-concept exists: the small molecule IMB-BZ blocks EsxB’s interaction with EccCb1 and, in Mycobacterium marinum infection models, reduced bacterial cytolytic activity to levels comparable to an ESX-1-deficient strain while rescuing phagosome maturation. HCL2, a peptide derived from human cytochrome c oxidase subunit 3, binds EsxA and disrupts the EsxA-EsxB complex, and an antibody-derived ESAT-6-specific nanobody expressed in the cytoplasm of host cells has blocked M. tuberculosis growth in macrophages. Indirect approaches are also viable: ethoxzolamide, an FDA-approved carbonic anhydrase inhibitor, downregulates the PhoPR regulon and suppresses ESX-1 secretion, producing a phenotype resembling the attenuated PhoPR mutant in macrophages and mice.

The antigen 85 complex represents a more enzymatically tractable target. Comprising the homologous proteins Ag85A, Ag85B, and Ag85C, secreted via the Sec pathway in a roughly 2:3:1 ratio, this complex acts as a mycolyl-transferase that attaches mycolic acids to arabinogalactan and catalyzes the biosynthesis of trehalose dimycolate, the cord factor critical for the bacterium’s cord-like appearance and survival. The complex also binds fibronectin, facilitating host cell attachment and invasion. Inhibitor development began with trehalose analogues that mimicked natural substrates, but these showed limited potency and whole-cell activity. More promising are ebselen and its derivatives, which allosterically inhibit the enzyme through covalent modification of Cys209 on Ag85C, and cyclophostin and cyclipostin analogues that covalently target the catalytic Ser124. The cyclipostin analogues have inhibited intracellular growth of M. tuberculosis with limited host-cell cytotoxicity, while ebselen’s activity in macrophage models still requires validation. Notably, ebselen has been shown to potentiate the activity of streptomycin, suggesting cell-envelope targeting can synergize with existing antibiotics. Functional redundancy among the three isoforms is a challenge, but their highly conserved active sites offer a rationale for multi-isoform inhibitors that raise the mutational threshold for resistance.

Among the phosphatases, SapM, PtpA, and PtpB each sabotage the phagosome in distinct ways. SapM, a 28-kDa lipid phosphatase, strips phosphatidylinositol 3-phosphate from the phagosomal membrane, halting phagosome maturation and phagosome-lysosome fusion; mutants lacking SapM fail to arrest maturation and grow poorly in macrophages and guinea pig tissues. Competitive inhibitor 2-phospho-L-ascorbic acid showed limited potency, but newer uncompetitive inhibitors, including Tyrphostin 51 and Tyrphostin Ag183 with IC50 values of 6.3 and 8.2 micromolar respectively, sustain inhibition even at high substrate concentrations. PtpA dephosphorylates host VPS33B and excludes the vacuolar H+-ATPase to block acidification, while also suppressing JNK and p38 MAPK signaling. Unfortunately, PtpA closely resembles human protein tyrosine phosphatases such as PTP1B, so most reported inhibitors, from stevastelins to prodigiosins and chalcones, lack specificity, though some low-micromolar chalcones did inhibit intracellular growth. PtpB, by contrast, has low similarity to human PTP1B and is therefore a more attractive target; inhibitors including I-A09, L01Z08, Kuwanon G, and Fusarielin M have all reduced bacterial burden in infected macrophages, and phosphatase inhibitors have been shown to augment the bactericidal activity of the standard drug regimen.

Protein kinase G, or PknG, presents a different risk profile. This 82-kDa serine/threonine kinase blocks phagosome-lysosome fusion and promotes survival under hypoxic stress, but it shares high homology with eukaryotic kinases, raising the specter of off-target effects on host signal transduction. The landmark inhibitor AX20017, however, achieved high selectivity for mycobacterial PknG over 28 archetypal human kinases, thanks to unique amino acid side chains within its narrow binding pocket, and restored lysosomal localization of bacteria in a dose-dependent manner. Later inhibitors, including AZD7762, R406, NU-6027, sclerotiorin, RO9021, and natural flavanones, showed activity against intracellular mycobacteria, though AZD7762’s mild macrophage cytotoxicity and the unvalidated efficacy of the newest compounds illustrate how selective this field still is. The toxin channel CpnT, whose C-terminal tuberculosis necrotizing toxin hydrolyzes NAD+ to kill macrophages, and the acyl carrier protein AcpM, which suppresses apoptosis, reactive oxygen species, and autophagy while also serving fatty acid biosynthesis, currently have no direct inhibitors reported, though the endogenous antitoxin IFT offers a template for future molecules.

Lipoproteins and heat shock proteins round out the review. Lipoproteins such as LprA and LpqH act as TLR2 agonists driving TNF and IL-12 production, while LprG, LppX, and LpqN support lipid transport and cell wall homeostasis; because they are non-redundant, immunogenic, and often essential for intracellular survival, disrupting their maturation or export could simultaneously compromise bacterial viability and restore immune clearance. Among the heat shock proteins, HspX, part of the DosR dormancy regulon, drives immunosuppressive M2 macrophage polarization, and a human monoclonal IgA antibody called 2E9 against HspX has reduced lung bacterial load through enhanced phagocytosis. Hsp70, a key chaperone maintaining proteostasis during oxidative stress, hypoxia, and nutrient deprivation, is targeted by telaprevir, a repurposed inhibitor that enhanced aminoglycoside efficacy and reduced rifampicin resistance, hinting that destabilizing stress-response networks could synergize bactericidally with existing drugs.

The authors are candid about the obstacles. Some secreted proteins are also immunodominant antigens, so inhibiting them risks compromising protective host immune responses, and standard growth-inhibition screens are poorly suited to targets whose main contribution is virulence rather than in vitro growth. They propose prioritizing dual-function targets like the antigen 85 complex, which is both an immunodominant T-cell antigen and essential for cord factor biosynthesis, and point to CRISPRi-mediated gene silencing in Mycobacterium abscessus as a model for validating target vulnerability during infection. With multiple new drugs now advancing through the tuberculosis pipeline and clinical trials testing entirely new regimens, the authors argue this is an opportune moment to develop anti-virulence agents as adjuncts to standard chemotherapy. If potency, selectivity, and delivery challenges can be overcome, the proteins M. tuberculosis painstakingly exports to disarm its host may become the very handle by which modern medicine finally pulls the pathogen out of its hiding places.

Subject of Research: Secreted virulence proteins of Mycobacterium tuberculosis as therapeutic targets in tuberculosis drug discovery

Article Title: Secreted proteins of Mycobacterium tuberculosis as therapeutic targets for drug discovery

Article References: Secreted proteins of Mycobacterium tuberculosis as therapeutic targets for drug discovery. (n.d.). https://doi.org/10.3389/ftubr.2026.1736139

Image Credits: AI Generated

DOI: 10.3389/ftubr.2026.1736139

Keywords: tuberculosis, Mycobacterium tuberculosis, secreted proteins, ESX-1, ESAT-6, antigen 85, virulence factors, drug discovery, phosphatases, protein kinase G, heat shock proteins, antimicrobial resistance