Neutrophil extracellular traps characterize caseating granulomas

Abstract

Tuberculosis (TB) remains one of the top 10 causes of death worldwide and still poses a serious challenge to public health. Recent attention to neutrophils has uncovered unexplored areas demanding further investigation. Therefore, the aim of this study was to determine neutrophil activation and circulatory neutrophil extracellular trap (NET) formation in various types of TB. Sera from TB patients (n = 91) and healthy controls (NHD; n = 38) were analyzed for NE-DNA and MPO-DNA complexes, cell-free DNA (cfDNA), and protease activity (elastase). We show that these NET parameters were increased in TB sera. Importantly, NET formation and NE activity were elevated in TB patients with extensive tissue damage when compared to those with minor damage and in patients with relapse, compared to new cases. We discuss the importance of balancing NET formation to prevent tissue damage or even relapse and argue to analyze circulating NET parameters to monitor the risk of disease relapse. To investigate the tissues for NETs and to find the source of the circulating NET degradation products, we collected sections of granulomas in lung and lymph node biopsies. Samples from other diseases with granulomas, including sarcoidosis (SARC) and apical periodontitis (AP), served as controls. Whereas NET formation characterizes the caseating granulomas, both caseating and non-caseating granulomas harbor DNA with unusual conformation. As TB is associated with hypercoagulation and thromboembolism, we further imaged the pulmonary vessels of TB patients and detected vascular occlusions with neutrophil aggregates. This highlights the dual role of neutrophils in the pathology of TB.

Fig. 1. Circulating NET degradation products are associated with lung destruction and relapse in TB patients.

We detected NETs by the quantification of circulating NET degradation products and found that they are elevated in patients with TB when compared to NHD. a We quantified cfDNA in the sera by PicoGreen fluorescence, b MPO–DNA and c NE-DNA complexes by ELISA and d NE activity by the conversion of a specific fluorogenic substrate. e Summarizes mean values and ranges for all parameters. f–m Clinical parameters were associated with the results obtained by NET formation analyses of TB sera. f NE-DNA complexes, g MPO–DNA complexes, h NE activity and i neutrophil counts in the sera of TB patients, associated with the type of TB case: new (green) versus relapse (red). Note that NET formation is increased in the sera of TB patients in relapse. j NE-DNA complexes, k MPO–DNA complexes, l NE activity and m neutrophil counts in the sera of TB patients, grouped by the extent of tissue destruction: no destruction (green) or destruction (red). Note that NET formation, NE activity and neutrophil counts are increased in patients with tissue destruction. All statistical analyses were performed employing the Mann–Whitney test. Data are presented as mean ± standard deviation (s.d.). At least three technical replicates were obtained for each data set. NHD normal healthy donors, TB tuberculosis, MPO myeloperoxidase, NE neutrophil elastase, MFI mean fluorescence intensity, n.s. not significant.

Fig. 2. Caseating granulomas in TB and SARC contain NETs.

Various staining for NETs (NE, citH3, MPO) in a, b TB and c, d SARC. Upper panel: a lung granuloma and b lymph node granuloma from a representative TB patient. Lower panel: c lung granuloma and d lymph node granuloma from a representative SARC patient. All antigens (NE, citH3, MPO) are shown in green, and sorted according to their abundance from left to right (low to high). Note the difference between TB and SARC; MPO is most abundant in TB, citH3 in SARC. Mock staining without the primary antibody (wo1st) served as control. Hoechst33342 was used as a DNA dye (red). Arrows mark NET-rich regions; white squares mark NETs which have been further enlarged and are displayed to the right of the original images. We applied additional tonal corrections to the insets to improve visibility. Labeled and mock-stained images from the same tissue sample were co-processed. HE staining, purple; CD31 staining, brown. Bars represent 1 mm.

Fig. 3. Granulomas harbor the DNase-resistant Z-form DNA.

Detection of B and Z isoforms of DNA by IF in caseating granulomas in TB and caseating or non-caseating granulomas in SARC, as well as in parts without granuloma. Antibodies recognizing B- or Z-form DNA and the low molecular weight DNA stain DAPI are displayed in green and red, respectively; Mock staining without a primary antibody served as control. TB tuberculosis, SARC sarcoidosis. Bars represent 500 µm.

Fig. 4. Morphometry of various granulomas.

We performed morphometry analyses of all granulomas, and grouped them based on caseation, disease, and type of tissue. a–c Caseating versus non-caseating granulomas; d–f TB versus SARC; and g, h lungs versus lymph nodes. Note the further discrimination of TB and SARC in (c), and of caseating versus non-caseating granulomas in (g, h). Note that caseating granulomas are larger than the non-caseating ones and that NET markers and Z-form DNA are more abundant in TB and SARC granulomas, respectively. All statistical analyses were performed employing the Mann–Whitney test. Data are presented as mean ± standard deviation (s.d.).

Fig. 5. TB and SARC granulomas differ in B-form or Z-form DNA staining.

FIt-SNE plots obtained from the morphometric analysis of B-form DNA or Z-form DNA MFI and grouped according to the type of disease and the affected organ. For further differentiation of caseating and non-caseating granulomas see Supplementary Figs. S4 and 6. TB tuberculosis, SARC sarcoidosis.

Fig. 6. Caseating and non-caseating granulomas can be distinguished by staining for NET markers and DNA isoforms (B/Z).

FIt-SNE plots are calculated by morphometry of MPO, NE, B-form DNA, or Z-form DNA MFI values, grouped according to the subtypes of granuloma; caseating (cavitating) versus non-caseating.

Fig. 7. Most pulmonary vessels in the granulomatous regions of TB are occluded by seemingly viable neutrophils.

Native endogenous fluorescence (NEF) was used to detect pulmonary vascular occlusions in TB biopsies (n = 4). a One of the analyzed lung sections showing an open vessel (*), a partially occluded vessel (arrow), and occluded vessels (+). b The percentage of pulmonary vessels which are open (green), partially occluded (orange), or completely occluded (red). P = 0.0065 for partially occluded versus occluded. c The size of open, partially occluded, and occluded vessels (single dots represent individual blood vessels). d Staining for NET-associated markers (citH3, MPO, NE) in occlusions. Note that citH3 is absent and NE or MPO showed cytoplasmic appearances, typical for viable neutrophils. White square marks neutrophil aggregates which have been further enlarged to improve visibility. Statistical analyses were performed using the two-way ANOVA. NHD normal healthy donor.