Combining Host Genetics and Functional Analysis to Depict Inflammasome Contribution in Tuberculosis Susceptibility and Outcome in Endemic Areas

Abstract

The interplay between M. tuberculosis (Mtb) and humans is multifactorial. The susceptibility/resistance profile and the establishment of clinical tuberculosis (TB) still remains elusive. The gain-of-function variant rs10754558 in the NLRP3 gene (found in 30% of the world population) confers protection against the development of TB, indicating a prominent role played by NLRP3 inflammasome against Mtb. Through genotype-guided assays and various Mtb strains (BCG, H37Rv, Beijing-1471, MP287/03), we demonstrate that Mtb strains activate inflammasome according to the NLRP3/IL-1ß or NLRC4/IL18 preferential axis. NLRP3 and NLRC4 genetic variants contribute to the presentation of TB. For the first time, we have shown that loss-of-function variants in NLRC4 significantly contribute to the development of extra-pulmonary TB. The analysis of inflammasome activation in a cohort of TB patients and their "household contacts" (CNT) revealed that plasma IL-1ß/IFN-α ratio lets us distinguish patients from Mtb-exposed-but-healthy individuals from an endemic region. Moreover, NLRP3 inflammasome seemed "exhausted" in TB patients compared to CNT, indicating a more efficient activation of inflammasome in resistant individuals. These findings suggest that inflammasome genetics as well as virulence-dependent level of inflammasome activation contribute to the onset of a susceptible/resistant profile among Mtb-exposed individuals.

Keywords: NLRP3; Nlrc4; genetics; inflammasome; tuberculosis.

Figure 1

M. tuberculosis strain H37Rv induces inflammasome activation in human MDM and IL-1ß production plays a major role in bacterial restriction. The healthy donors' monocyte-derived macrophages (MDM; n = 14) were challenged with the M. tuberculosis (Mtb) strain H37Rv (MOI: 0.033) for 3 h, then washed and cultured for 24 h. The release of IL-1ß (n = 14) (A) and IL-18 (n = 8) (B) was measured in culture supernatants of untreated (UN) and infected (H37Rv) MDM. Cells were harvested for an analysis of caspase-1 activity by a FAM-FLICA® assay and flow cytometry from three independent experiments. The percentage of FAM-FLICA+ cells was reported for H37Rv-infected (H37Rv) and untreated (UN) MDM (C). Lactate Dehydrogenase (LDH) liberation was measured in culture supernatants in three independent experiments. Data are expressed as a percentage in respect to the positive control (Triton, 100% cell lysis) (D). The healthy donors' MDM (n = 5) were infected with Mtb H37Rv (MOI: 0.033) for 3 h in the presence of exogenous recombinant human IL-1ß (rh-IL-1ß) or parthenolide (PTD; 10 μM), then lysed. Serial dilutions were plated in a complete agar medium. Bacterial growth was evaluated after 21 days by counting colony formation units (CFU) (E). The healthy donors' MDM (n = 4) were infected with Mtb H37Rv (MOI: 0.033) for 3 h in the presence of cytochalasin D (Cyt D; 10 μM), then washed and cultured for 24 h. The release of IL-1ß was measured at the end of incubation in infected MDM (H37Rv) and infected MDM in the presence of a phagocytosis inhibitor (Cyt D) (F). Individual data are reported with their mean. The Mann–Whitney test was used to compare treated and untreated groups (A–C,F). The Kruskall–Wallis test followed by a Bonferroni post-test were used to compare H37Rv, rh-IL-1ß, and PTD groups (E). Differences with a p < 0.05 were considered statistically significant (*).

Figure 2

NLRP3 contributes to inflammasome activation by Mtb H37Rv in human MDM. The healthy donors' monocyte-derived macrophages (MDM; n = 4) were cultivated (0.08 × 10⁶) in 16-well chamber slides (Thermo Fisher Scientific), challenged with M. tuberculosis (Mtb) H37Rv (MOI: 0.033) for 3 h, then washed and cultured for 24 h. Mouse anti-human NLRP3 and a fluorescent secondary antibody (Alexa 488-conjugated goat-anti-mouse IgG1) were used to label NLRP3. DAPI was used to counterstain the nuclei. NLRP3+ speck formation in untreated (UN) and infected MDM (H37Rv) was analyzed using fluorescence microscopy. A representative experiment was reported. Arrows indicated NLRP3+ specks. (A). The number of NLRP3+ specks was manually counted for each independent experiment (B). The healthy donors' monocyte-derived macrophages (MDM; n = 6) were treated with 10 μM MCC-950 or 10 μM parthenolide (PTD) 1 h before infection with M. tuberculosis H37Rv (MOI: 0.033) for 3 h, then washed and cultured for 24 h. The release of IL-1ß (C), IL-18 (D), and TNF (E) was measured in culture supernatants. Individual data are reported with their mean. The Mann–Whitney test was used to compare treated and untreated groups (B). The Kruskall–Wallis test followed by a Bonferroni post-test were used to compare H37Rv, MCC-950 and PTD groups (C–E). Differences with a p < 0.05 were considered statistically significant (*). **p < 0.001.

Figure 3

TB patients present an impaired NLRP3 inflammasome response. IL-1ß (A) level were measured in plasma from recently diagnosed untreated TB patients (TB; n = 27) and some of their “household contacts” (CNT; n = 15) from an endemic TB area (Manaus, metropolitan region, AM) and from healthy donors in a non-endemic area (São Paulo metropolitan region, SP) (HD; n = 5). IL-18 (B), IFN-a (C), and TNF (D) dosage was executed in a group of samples (HD, n = 6; TB, n = 8; CNT, n = 6). (E) A Principal Components (PC) analysis was executed including the plasma level of IL-1ß, a mycobacteria count in a sputum analysis and a diagnosis of TB (TB or CNT). Monocyte-derived macrophages (MDM) were isolated from HD (n = 6), TB patients (n = 6), and CNT (n = 6) and challenged with 1 μg/mL of LPS for 24 h (LPS), or 1 μg/mL of LPS for 24 h plus 1 mM of ATP for another 15 min (LPS+ATP). Cells were lysed for RNA isolation and expression analysis. The release of IL-1ß was measured in culture supernatants. (F–H) Basal expression of NLRP3, IL1B, and IL18 genes were analyzed in untreated MDM using gene-specific Taqman® assays and qPCR. Row target gene expression was normalized by the expression of the GAPDH house-keeping gene (ΔCt). Relative expression was calculated as 2^−ΔCt. Data are expressed as arbitrary units (AU). (I) IL-1ß was measured in culture supernatants of treated MDM. (J–L) LPS-induced expression of NLRP3, IL1B, and IL18 genes were analyzed in treated MDM using gene-specific Taqman® assays and qPCR. Row target gene expression was normalized by the expression of GAPDH house-keeping gene (ΔCt). Modulation of gene expression was calculated as Fold Change (FC = 2 ^−−ΔΔC) in respect to untreated (UN) cells. Individual data are reported with their mean. The Mann–Whitney test was used to compare two groups. The Kruskall–Wallis test followed by a Bonferroni post-test were used to compare more than two groups. Differences with a p < 0.05 were considered statistically significant (*).

Figure 4

Cathepsins activity is the major mechanism for inflammasome activation by M. tuberculosis H37Rv in human MDM, triggering NLRP3 and NLRC4. The healthy donors' monocyte-derived macrophages (MDM; n = 6) were treated or not with Ca074-Me (Ca074; 10 μM) before the infection with M. tuberculosis H37Rv (MOI: 0.033) for 3 h, then washed and cultured for 24 h. ATP (1 mM) was eventually added for another 15 min in H37Rv-infected MDM. The stimulation of MDM using LPS (1 μg/mL) for 24 h with or without ATP (1 mM) was included as a positive control for NLRP3 inflammasome activation. The release of IL-1ß (A) and IL-18 (B) was measured in culture supernatants. Healthy donors' monocyte-derived macrophages (MDM; n = 4) were cultivated (0.08 × 10⁶) in 16-well chamber slides (Thermo Fisher Scientific), challenged with M. tuberculosis H37Rv (MOI: 0.033) for 3 h, then washed and cultured for 24 h. Rabbit anti-human NLRC4 and fluorescent secondary antibody (Alexa 647-conjugated goat-anti-rabbit IgG1) were used to label NLRC4. DAPI was used to counterstain the nuclei. NLRC4+ speck formation in untreated (UN) and infected MDM (H37Rv) was analyzed using fluorescence microscopy. The number of NLRPC+ specks was manually counted for each independent experiment (C). A representative experiment was reported. Arrows indicated NLRC4+ specks. (D) Individual data are reported with their mean. The Kruskall–Wallis test followed by a Bonferroni post-test were used to compare more than two groups (A,B). The Mann–Whitney test was used to compare treated and untreated groups (C). Differences with a p < 0.05 were considered statistically significant (*). **p < 0.001.

Figure 5

Genetic variants in NLRP3 and NLRC4 genes affect host response to M. tuberculosis. The healthy donors' monocyte-derived macrophages (MDM; n = 14) were treated with M. tuberculosis H37Rv (MOI: 0.033) for 3 h, then washed and cultured for 24 h. The IL-1ß and IL-18 level was measured in culture supernatants at the end of experiment (n = 14 and 8, respectively), and in the plasma of TB patients (n = 27 and 20, respectively). NLRP3 rs10754558 and NLRC4 rs479333 SNVs were genotyped in the genomic DNA of healthy donors and TB patients. IL-1ß and IL-18 release data were then grouped by individual NLRP3 rs10754558 genotype into homozygotes for the minor G allele (G/G) and the others (C/C+C/G), according to a recessive model of inheritance for the minor allele (A–C); or by individual NLRC4 rs479333 genotype into homozygotes for the minor C allele (C/C) and the others (G/G+G/C) according to a recessive model of inheritance for the minor allele (D–F). Individual data are reported with their mean. The Mann–Whitney test was used to compare the groups. Differences with a p < 0.05 were considered statistically significant (*).

Figure 6

Mtb strains of different virulence and host genetic background shaped the inflammasome response to mycobacteria. The healthy donors' monocyte-derived macrophages (MDM) were treated with 10 μM MCC-950 or 10 μM parthenolide (PTD) 1 h before infection with M. tuberculosis (MOI: 0.033) H37Rv (n = 14), Beijin-1470 (Bj-1470; n = 9), MP287/03 (MP; n = 6) and Bacillus Calmette Guerin (BCG; n = 6) for 3 h, then washed and cultured for 24 h. The release of IL-1ß (A), IL-18 (B), TNF (C), and LDH (D) was measured in culture supernatants. Dosage for IL-1ß was executed for all samples, while for IL-18 and TNF in six samples, for LDH in three sample. Cytokines and LDH release were compared among different Mtb strains. Cells were lysed for RNA isolation and gene expression analysis. NLRP3 rs10754558 and NLRC4 rs479333 SNVs were genotyped in the genomic DNA of healthy donors. The effect of MCC-950 and PTD on inflammasome activation in infected MDM was analyzed by the meaning of IL-1ß (E) and IL-18 (F) release. Mtb-induced expression of NLRP3, NLRC4, IL1B, and IL18 genes in H37Rv-, Bj-1470-, MP- and BCG-infected MDM was calculated as 2^−ΔΔCt (fold-change, FC) and reported as log₂FC in a heat-map graph (G). IL-1ß release data were grouped by individual NLRP3 rs10754558 genotype into homozygotes for the minor G allele (G/G) and the others (C/C+C/G), according to a recessive model of inheritance for the minor allele (H). IL-18 release data were grouped by NLRC4 rs479333 genotype into homozygotes for the minor C allele (C/C) and the others (G/G+G/C) according to a recessive model of inheritance for the minor allele (I). Individual data are reported with their mean. The Kruskall-Wallis test followed by a Bonferroni post-test were used to compare groups (A–F). The Mann-Whitney test was used to compare two groups (H,I). Differences with a p < 0.05 were considered statistically significant (*). **p < 0.001.

Figure 7

Host genetic background and M.tuberculosis (Mtb) determine the outcome of TB. Schematic representation of the main results of this study.