Mycobacterium avium biofilm attenuates mononuclear phagocyte function by triggering hyperstimulation and apoptosis during early infection

Abstract

Mycobacterium avium subsp. hominissuis is an opportunistic human pathogen that has been shown to form biofilm in vitro and in vivo. Biofilm formation in vivo appears to be associated with infections in the respiratory tract of the host. The reasoning behind how M. avium subsp. hominissuis biofilm is allowed to establish and persist without being cleared by the innate immune system is currently unknown. To identify the mechanism responsible for this, we developed an in vitro model using THP-1 human mononuclear phagocytes cocultured with established M. avium subsp. hominissuis biofilm and surveyed various aspects of the interaction, including phagocyte stimulation and response, bacterial killing, and apoptosis. M. avium subsp. hominissuis biofilm triggered robust tumor necrosis factor alpha (TNF-α) release from THP-1 cells as well as superoxide and nitric oxide production. Surprisingly, the hyperstimulated phagocytes did not effectively eliminate the cells of the biofilm, even when prestimulated with gamma interferon (IFN-γ) or TNF-α or cocultured with natural killer cells (which have been shown to induce anti-M. avium subsp. hominissuis activity when added to THP-1 cells infected with planktonic M. avium subsp. hominissuis). Time-lapse microscopy and the TUNEL (terminal deoxynucleotidyltransferase-mediated dUTP-biotin nick end labeling) assay determined that contact with the M. avium subsp. hominissuis biofilm led to early, widespread onset of apoptosis, which is not seen until much later in planktonic M. avium subsp. hominissuis infection. Blocking TNF-α or TNF-R1 during interaction with the biofilm significantly reduced THP-1 apoptosis but did not lead to elimination of M. avium subsp. hominissuis. Our data collectively indicate that M. avium subsp. hominissuis biofilm induces TNF-α-driven hyperstimulation and apoptosis of surveilling phagocytes, which prevents clearance of the biofilm by cells of the innate immune system and allows the biofilm-associated infection to persist.

FIG 1

THP-1 cells are stimulated during exposure to M. avium subsp. hominissuis biofilms. (A) THP-1 cells were added on top of M. avium subsp. hominissuis strain A5 and 104 14-day-old biofilms (equal inoculum levels), and supernatant was collected at 0.5, 1, 2, 4, 24, and 48 h. TNF-α ELISA was carried out for each time point for each strain. (B) Filter-sterilized supernatants from 14-day-old biofilms were added at a 10% (vol/vol) concentration in RPMI 1640 plus 10% FBS with THP-1 cells for 4 h. Culture supernatant was then collected, and TNF-α ELISA was performed. Abbreviations: NS, not stimulated; 104 S, M. avium subsp. hominissuis 104 supernatant; A5 S, M. avium subsp. hominissuis A5 supernatant. (C) UV-killed M. avium subsp. hominissuis A5 biofilms (A5 BF UV) were compared side by side with non-UV-treated biofilms (A5 BF) and planktonic bacteria that were immobilized on the bottom surface of the plate via centrifugation (A5 plank) for TNF-α production by THP-1 cells 4 h after being placed on top of the bacteria. (D) THP-1 cells were placed on top of M. avium subsp. hominissuis A5 biofilm, and O₂⁻ was assessed spectrophotometrically as described in Materials and Methods. (E) THP-1 cells were placed on top of M. avium subsp. hominissuis A5 biofilm, and nitric oxide (NO) was assessed using the Griess reagent system. Abbreviations for panels D and E: neg, negative control for respective assay; THP-1, THP-1 cells placed in empty wells with no bacteria present; THP-1 A5, THP-1 cells placed in wells containing M. avium subsp. hominissuis A5 biofilm. Bars represent means ± standard deviations. *, P < 0.05.

FIG 2

The number of CFU of M. avium subsp. hominissuis A5 biofilm is unaffected by the addition of THP-1 cells, even when preactivated or cocultured with NK cells. (A) THP-1 cells were added on top of established M. avium subsp. hominissuis A5 biofilms for 24 and 72 h. In some instances, THP-1 cells were first prestimulated with 50 ng/ml of TNF-α or 100 ng/ml of IFN-γ. At each time point, wells were resuspended, diluted, and plated to obtain CFU/well of M. avium subsp. hominissuis A5. (B) Suspensions of THP-1 cells alone, NK cells alone, or a coculture of the two cell types were added on top of established M. avium subsp. hominissuis A5 biofilms for 24 and 48 h before CFU enumeration. (C) Coculture of NK cells and THP-1 cells infected with planktonic M. avium subsp. hominissuis demonstrating NK cell-assisted bacterial killing after 4 days. (D) Infection of THP-1 with planktonic M. avium subsp. hominissuis A5 and two M. avium subsp. hominissuis A5-derived biofilm-deficient mutants (6H9 and 5G4), demonstrating that planktonic M. avium subsp. hominissuis can grow inside THP-1 cells. For both panels A and B, all comparisons to the M. avium subsp. hominissuis A5 control (no host cells added) for each time point were not significant. Bars represent means ± standard deviations. **, P < 0.01.

FIG 3

THP-1 cells rapidly undergo apoptosis after exposure to M. avium subsp. hominissuis A5 biofilm. (A to F) Microscopy showing THP-1 cells at 5 min (A), 15 min (B), 25 min (C and D), and 40 min (E and F). (G) THP-1 cells were placed on top of established M. avium subsp. hominissuis A5 biofilm, and the TUNEL assay was performed at 0, 0.5, 24, and 48 h. In each field of view, apoptotic cells were counted microscopically and divided by the total number of counted cells to attain percentage of apoptosis. Fifty fields of view were counted for each time point and averaged. The bars represent the mean of each time point ± the standard deviation.

FIG 4

Blocking TNF-α reduces apoptosis of THP-1 cells during M. avium subsp. hominissuis A5 biofilm exposure. THP-1 cells were placed on top of established M. avium subsp. hominissuis A5 biofilms with or without anti-TNF-R1 (1:10 dilution of a hybridoma suspension) or anti-TNF-α (10 μg/ml). (A) TUNEL assay was performed at 2 and 6 h after exposure. The ratio of apoptotic to total cells was counted in 50 fields of view microscopically and was averaged for each time point. (B) THP-1 cells were lysed at 24 h after exposure, and wells were resuspended, diluted, and plated for CFU of M. avium subsp. hominissuis. Bars represent the mean ± standard deviation. *, P < 0.05.