Infection-induced bystander-apoptosis of monocytes is TNF-alpha-mediated

Abstract

Phagocytosis induced cell death (PICD) is crucial for controlling phagocyte effector cells, such as monocytes, at sites of infection, and essentially contributes to termination of inflammation. Here we tested the hypothesis, that during PICD bystander apoptosis of non-phagocyting monocytes occurs, that apoptosis induction is mediated via tumor necrosis factor-alpha (TNF-α and that TNF-α secretion and -signalling is causal. Monocytes were infected with Escherichia coli (E. coli), expressing green fluorescent protein (GFP), or a pH-sensitive Eos-fluorescent protein (EOS-FP). Monocyte phenotype, phagocytic activity, apoptosis, TNF-receptor (TNFR)-1, -2-expression and TNF-α production were analyzed. Apoptosis occured in phagocyting and non-phagocyting, bystander monocytes. Bacterial transport to the phagolysosome was no prerequisite for apoptosis induction, and desensitized monocytes from PICD, as confirmed by EOS-FP expressing E. coli. Co-cultivation with non-infected carboxyfluorescein-succinimidyl-ester- (CFSE-) labelled monocytes resulted in significant apoptotic cell death of non-infected bystander monocytes. This process required protein de-novo synthesis and still occurred in a diminished way in the absence of cell-cell contact. E. coli induced a robust TNF-α production, leading to TNF-mediated apoptosis in monocytes. Neutralization with an anti-TNF-α antibody reduced monocyte bystander apoptosis significantly. In contrast to TNFR2, the pro-apoptotic TNFR1 was down-regulated on the monocyte surface, internalized 30 min. p.i. and led to apoptosis predominantly in monocytes without phagocyting bacteria by themselves. Our results suggest, that apoptosis of bystander monocytes occurs after infection with E. coli via internalization of TNFR1, and indicate a relevant role for TNF-α. Modifying monocyte apoptosis in sepsis may be a future therapeutic option.

Figure 1. Apoptosis occurs after infection with E. coli-GFP 4 h p.i. in phagocyting and non-phagocyting monocytes.

PBMC were infected with E. coli-GFP for 4 hours and free bacteria were removed. When indicated, CytoD was supplemented 30 minutes prior to infection. Monocyte apoptosis was detected by annexin V (A; n = 5), DNA strand breaks (B; n = 7), and hypodiploidity (Fig. 1C, n = 14, * p<0.05, ** p<0.01). In the same experimental setup, CD14+ monocytes were gated for the absence or presence of GFP-fluorescence and analyzed for their percentage of hypodiploid DNA (Fig. 1D; n = 14, * p<0.05).

Figure 2. Transport to the phagolysosome is not obligatory for monocyte apoptosis.

The gating strategy and nomenclature is shown by dot-plot analysis of monocytes, infected with EOS-FP E.coli for the indicated time intervals (A). Bound bacteria exhibit a red fluorescence, in the FL-2 channel (red bacteria in the micrograph, gated BI). Bacteria, transported to the phagolysosome, exhibit a green fluorescence in the FL-1 channel (green bacteria in the micrograph, gate P). Monocytes without bacterial contact were gated in NC. The chart (B) summarizes the proceeding the phagocytic process within the indicated time intervals. Details are given in the supplement. (C): Assessment of monocyte apoptosis by hypodiploid DNA content in total (second column) and in the three subfractions described (n = 7; ** = p<0.001).

Figure 3. Bystander apoptosis occurs without cellular contact to the phagocyting monocyte.

Monocytes were infected with E. coli-GFP and GFP –positive and -negative cells were separated by FACS sorting. Apoptosis was assessed 24 hours p.i. by quantification of hypodiploid DNA (Fig. 3A; n = 3, *p<0.05). Monocytes were infected with E.coli-GFP for one hour and free bacteria were removed. CFSE labelled monocytes of the same donor were co-incubated for 4 hours in a concentration of 10∶1. Un-infected, unlabelled monocytes served as controls. Apoptosis was assessed 4 hours p.i. in GFP+ and CFSE+ monocytes (B; n = 5). Monocytes were infected with E. coli-GFP as described (i.e. 1 hour w/o antibiotics and 3–21 hours with gentamycin) in a transwell setup for the time intervals indicated (“cis” chamber, C, sketch to the right); non-infected cells served as controls. Monocytes from the same donor were co-cultivated in compartments, separated by teflon membranes (“trans” chamber). Apoptosis was detected by assessment of hypodiploid DNA-content (C, n = 5; * p<0.05, ** = p<0.001). In the same experimental setup, a CFU-assay was performed by lysis of cells (Fig. 3D, n = 5; * p<0.05, ** p<0.001).

Figure 4. Infection with E. coli induces TNF-α production and monocyte apoptosis in phagocyting and non-phagocyting cells.

Monocytes pre-incubated with Cyto D or CHX were infected with E. coli-GFP for 4 hours and TNF-α production was detected in culture supernatants by ELISA (A; n = 8, * p<0.05.) In the same experimental setup, monocyte apoptosis was quantified by TUNEL-staining after treatment with TNF-α neutralizing anti-TNF-α antibody, or E. coli, or combinations (B, left;; n = 9, * p<0.05), or IL-6 (B, right; n = 4, * p<0.05). In the same setup, intracellular TNF-α production of monocytes was analyzed by FACS-stain (C, left) and gated for GFP-positivity (C; n = 5, * p<0.05, ** p<0.001).

Figure 5. Monocyte apoptosis with regard to bacterial phagocytosis and TNF-α production.

Monocytes were infected with E.coli-GFP for four hours and free bacteria were removed. Four hours p.i. hypodiploid DNA content was quantified and analyzed for absence or presence of intracellular viable bacteria (GFP-negative or positive, respectively), and TNF-α content (A; n = 7, * p<0.05, ** p<0.001). Monocytes were infected with E.coli-EOS-FP for one hour and free bacteria were removed. 4 hours p.i. we quantified hypodiploid DNA content in TNF-α -positive and –negative monocytes in the BI- and P-gate (B; n = 5, * p<0.05, ** p<0.001).

Figure 6. TNFR1 is internalized after incubation with TNF-α and after infection with E. coli and leads to monocyte apoptosis.

Monocytes were infected with E.coli-GFP for one hour and free bacteria were removed. 4 hours p.i. TNFR1 and -2 expression were analyzed by FACS (A; n = 9, **p<0.01). Analysis of TNFR1 internalization following TNF-α administration after 30 minutes; one group received CytoD (B; n = 7, p<0.05). In the described experimental setup, monocyte apoptosis was analyzed via Vybrant DNA-content and distinguished by internalization of TNFR1 (C, * p<0.05, ** = p<0.001). One group was cultivated with CytoD. TNFR1 internalization 30 min p.i. Internalization was quantified for GFP-negative and –positive monocytes (D; n = 5,* p<0.05, ** p<0.001.