NAD hydrolysis by the tuberculosis necrotizing toxin induces lethal oxidative stress in macrophages

Abstract

Mycobacterium tuberculosis (Mtb) kills infected macrophages through necroptosis, a programmed cell death that enhances mycobacterial replication and dissemination. The tuberculosis necrotizing toxin (TNT) is the major cytotoxicity factor of Mtb in macrophages and induces necroptosis by NAD⁺ hydrolysis. Here, we show that the catalytic activity of TNT triggers the production of reactive oxygen species (ROS) in Mtb-infected macrophages causing cell death and promoting mycobacterial replication. TNT induces ROS formation both by activating necroptosis and by a necroptosis-independent mechanism. Most of the detected ROS originate in mitochondria as a consequence of opening the mitochondrial permeability transition pore. However, a significant part of ROS is produced by mechanisms independent of TNT and necroptosis. Expressing only the tnt gene in Jurkat T-cells also induces lethal ROS formation indicating that these molecular mechanisms are not restricted to macrophages. Both the antioxidant N-acetyl-cysteine and replenishment of NAD⁺ by providing nicotinamide reduce ROS levels in Mtb-infected macrophages, protect them from cell death, and restrict mycobacterial replication. Our results indicate that a host-directed therapy combining replenishment of NAD⁺ with inhibition of necroptosis and/or antioxidants might improve the health status of TB patients and augment antibacterial TB chemotherapy.

Keywords: TNT; macrophages; necroptosis; nicotinamide adenine dinucleotide; reactive oxygen species.

Figure 1.. TNT-induced reactive oxygen species mediate macrophage cell death.

THP-1 macrophages were infected with Mtb strains at an MOI of 10:1 and treated with N-acetylcysteine (NAC, 100 μM) when indicated. ROS levels were measured with the fluorescent probe H₂DCFDA at (A) 4, 24 and 48 h or (B) 48 h post-infection. (C) Cell viability of infected macrophages was measured 48 h after infection as the total ATP content with a luminescent ATP detection assay kit. (D) The Mtb intracellular growth in infected macrophages was measured at 4 h and 48 h post-infection and expressed as CFU per ml. (E) Wt Mtb H37Rv was grown in Middlebrook 7H9 medium with 0.5% glycerol, 0.02% Tyloxapol and 10% OADC and supplemented with N-acetyl-cysteine (NAC, 100 μM). The OD₆₀₀ was measured at the indicated time points. Asterisks indicate significant differences (p-value<0.01, calculated using the Two-way ANOVA with Bonferroni’s correction) compared with the indicated conditions. Data are represented as mean ± SEM.

Figure 2.. TNT-induced ROS accumulation occurs downstream NAD⁺ depletion, necroptosis activation and mitochondrial damage.

THP-1 macrophages were treated with (A) nicotinamide (5 mM) or cyclosporin A (CsA, 5 μM) and (C) the RIPK3 inhibitor GSK´872 (10 μM) or the MLKL inhibitor necrosulfonamide (NSA, 10 μM), and infected with Mtb strains at an MOI of 10:1 for 48 h. ROS levels (measured with the fluorescent probe H₂DCFDA) were analyzed. (B) THP-1 macrophages were treated with necrostatin-1s (Nec-1s, 10 μM), GSK’872 (10 μM), necrosulfonamide (NSA, 10 μM) and/or TNF-a, cycloheximide, and zVAD-fmk (T/C/Z) (Cho et al., 2009), and cell viability was measured by trypan blue. Asterisks indicate significant differences (p-value<0.01, calculated using the One-way [B] or Two-way [A, C] ANOVA with Bonferroni’s correction) compared with the indicated conditions. Data are represented as mean ± SEM.

Figure 3.. TNT is sufficient to induce lethal oxidative stress in Jurkat-T cells.

Expression of TNT in the Jurkat-T cell line containing an integrated Tet-regulated TNT expression cassette (Jurkat 655-TNT) was induced with different doxycycline (Dox) concentrations at the indicated time points. When indicated, cells were treated with N-acetyl-cysteine (NAC, 1 mM), nicotinamide (NAM, 5 mM) or cyclosporin A (CsA, 5 μM). (A, C) Cell viability was measured by trypan blue staining. (B) ROS levels were measured with the fluorescent probe H₂DCFDA. (D) Western blot for (i) TNT (purified specific polyclonal antibody) and GAPDH (loading control) in the whole-cell lysates of Jurkat 655-TNT cells induced with 5 ng/mL doxycycline at the indicated time points, and for (ii) different amounts of purified WT TNT protein to determine the purified specific αTNT polyclonal antibody sensitivity. Asterisks indicate significant differences at 72 h (p-value<0.01, calculated using the One-way ANOVA with Bonferroni’s correction) compared with the indicated conditions. Data are represented as mean ± SEM.

Figure 4.. NAD⁺ depletion induces ROS in macrophages.

THP-1 macrophages were treated with the indicated concentrations of FK866 for 6 h and (A) the mitochondrial membrane potential (analyzed by using a JC-1 dye-based fluorescent probe) and (B) ROS levels (measured with the fluorescent probe H₂DCFDA) were analyzed. THP-1 macrophages were treated with CCCP (4 μM) and H₂O₂ (10 mM, 30 min) as positive controls for mitochondrial membrane depolarization and ROS detection, respectively. (C) THP-1 macrophages were treated with FK866 at 10 μM and/or N-acetyl-cysteine (NAC) at 1 mM for 6 h, and cell viability was measured by trypan blue. Asterisks indicate significant differences (p-value<0.01, calculated using the One-way ANOVA with Bonferroni’s correction) compared with the indicated conditions. Data are represented as mean ± SEM.

Figure 5.. ROS do not activate necroptosis in macrophages.

(A) THP-1 macrophages were treated with different menadione concentrations and cell viability was measured at the indicated time points by trypan blue. THP-1 macrophages were treated with menadione (Mena, 160 μM), N-acetyl-cysteine (NAC, 1 mM), necrostatin-1s (Nec-1s, 10 μM), GSK’872 (10 μM), necrosulfonamide (NSA, 10 μM) and cyclosporin A (CsA, 5 μM), and (B) ROS levels (measured with the fluorescent probe H₂DCFDA) and (C) cell viability (measured by trypan blue) were determined. Data are represented as mean ± SEM.

Figure 6.. Model of ROS production in macrophages infected with Mtb.

After phagocytosis, Mtb permeabilizes the phagosomal membrane enabling the translocation of TNT to the cytosol (Sun et al., 2015). TNT hydrolyzes the cytosolic NAD⁺ to nicotinamide (NAM) and ADP-ribose (ADPr), activating the necroptosis effectors RIPK3 and MLKL (Pajuelo et al., 2018). Activated RIPK3 migrates to the mitochondria and recruits Bcl-x_L (Zhao et al., 2017). The RIPK3-Bcl-x_L complex prevents caspase-8 activation and, together with oligomerized pMLKL, induces ROS production by opening the mitochondrial permeability transition (MPT) pore and other mechanisms (pathway 1a). The enzymatic activity of TNT also increases ROS production in a necroptosis-independent manner as a consequence of either NAD⁺ depletion or by signaling of the NAD⁺ hydrolysis products (pathway 1b). Additionally, necroptosis is activated by Mtb in a TNT-independent manner and induces ROS accumulation by the RIPK3/MLKL pathway (pathway 2a). Other uncharacterized Mtb factors also contribute to ROS production in macrophages infected with Mtb in a TNT- and necroptosis-independent manner (pathway 2b).