The Bordetella effector protein BteA induces host cell death by disruption of calcium homeostasis

Abstract

Bordetella pertussis is the causative agent of whooping cough in humans, a disease that has recently experienced a resurgence. In contrast, Bordetella bronchiseptica infects the respiratory tract of various mammalian species, causing a range of symptoms from asymptomatic chronic carriage to acute illness. Both pathogens utilize type III secretion system (T3SS) to deliver the effector protein BteA into host cells. Once injected, BteA triggers a cascade of events leading to caspase 1-independent necrosis through a mechanism that remains incompletely understood. We demonstrate that BteA-induced cell death is characterized by the fragmentation of the cellular endoplasmic reticulum and mitochondria, the formation of necrotic balloon-like protrusions, and plasma membrane permeabilization. Importantly, genome-wide CRISPR-Cas9 screen targeting 19,050 genes failed to identify any host factors required for BteA cytotoxicity, suggesting that BteA does not require a single nonessential host factor for its cytotoxicity. We further reveal that BteA triggers a rapid and sustained influx of calcium ions, which is associated with organelle fragmentation and plasma membrane permeabilization. The sustained elevation of cytosolic Ca²⁺ levels results in mitochondrial calcium overload, mitochondrial swelling, cristolysis, and loss of mitochondrial membrane potential. Inhibition of calcium channels with 2-APB delays both the Ca²⁺ influx and BteA-induced cell death. Our findings indicate that BteA exploits essential host processes and/or redundant pathways to disrupt calcium homeostasis and mitochondrial function, ultimately leading to host cell death.IMPORTANCEThe respiratory pathogens Bordetella pertussis and Bordetella bronchiseptica exhibit cytotoxicity toward a variety of mammalian cells, which depends on the type III secretion effector BteA. Moreover, the increased virulence of B. bronchiseptica is associated with enhanced expression of T3SS and BteA. However, the molecular mechanism underlying BteA cytotoxicity is elusive. In this study, we performed a CRISPR-Cas9 screen, revealing that BteA-induced cell death depends on essential or redundant host processes. Additionally, we demonstrate that BteA disrupts calcium homeostasis, which leads to mitochondrial dysfunction and cell death. These findings contribute to closing the gap in our understanding of the signaling cascades targeted by BteA.

Keywords: Bordetella; calcium homeostasis; effector protein BteA; host cell death mechanism; type III secretion system (T3SS).

Fig 1

BteA-induced cell death is characterized by fragmentation of the endoplasmatic reticulum and mitochondrial networks. (A) Time lapse analysis of morphological changes and plasma membrane permeabilization in Hela cells. Hela cells were infected with B. bronchiseptica wild-type (BbWT) and BbΔbteA mutant (BbΔbteA), expressing the fluorescent protein mNeonGreen, at MOI of 10:1 in the presence of propidium iodide (5 µg/mL). A sequence of time lapse images is shown. Bright field, gray; bacteria, cyan; propidium iodide, magenta. Scale bar, 20 µm. Data are representative of three independent experiments. (B and C) Comparison of B. bronchiseptica and B. pertussis cytotoxicity against HeLa cells. HeLa cells were infected with B. bronchiseptica (B) or B. pertussis (C) wild-type strains and their mutant derivatives at the indicated MOI. Plasma membrane permeabilization was determined using the fluorescent DNA binding dye CellTox Green. For B. bronchiseptica, infections were conducted both in the presence (gly+) and absence of glycine (5 mM). Asterisks indicate a statistically significant difference (P < 0.05, unpaired two-tailed t-test) between BbWT and BbWT gly+ infection samples at 4 h. Data represent the mean ± SEM of a representative experiment out of 2 performed in technical triplicate. (D) Visualization of cellular structures. Hela cells were transfected to express fluorescent proteins tagged with localization signals for endoplasmatic reticulum (ER) and mitochondria (mito). One hour after infection with BbWT and BbΔbteA mutant at MOI 50:1, cells were fixed and analyzed by fluorescence imaging. ER, yellow; mitrochondria, magenta. Scale bar, 20 µm. The shown micrographs are representative of two independent experiments from which the organelle morphology was scored. Analysis was performed on at least 100 cells per experiment and condition and is plotted as morphology % ± SEM. Asterisks indicate statistically significant differences (P < 0.05, unpaired two-tailed t-test) between the % of fragmented ER or mito compared with uninfected cells.

Fig 2

BteA targets essential components of the host cell or genes with redundant functions. (A) Workflow for generating the HEK-Cas9 sublibraries A and B. Sublibraries A and B of the GeCKO v2 library were individually packaged into lentiviral particles and introduced into Cas9-expressing HEK cells. After selection with puromycin and expansion of both cell pools, genomic DNA was extracted, sgRNA was amplified and barcoded, and library complexity was verified by next-generation sequencing. (B) Verification of library complexity in the HEK-Cas9 sublibraries A and B. The number of targeted genes per detected sgRNA count is indicated. (C) Workflow for the enrichment of BteA-resistant cells. HEK-Cas9 sublibraries A and B were subjected to five rounds of selection with BbWT at MOI 100:1 for 5 h per round. Each selection resulted in the death of 95% of infected cells. To stop the infection, the medium containing BbWT was discarded and replaced with a fresh medium containing 100 µg/mL of gentamicin. In between selection rounds, the surviving HEK-Cas9 sublibraries A and B were expanded. (D) Susceptibility of selected HEK-Cas9 sublibraries A and B. Selected HEK-Cas9 sublibraries A and B, and parental HEK-Cas9 cells were infected with BbWT at the indicated MOI. The cytotoxicity was determined as lactate dehydrogenase (LDH) release 6 h post-infection. Susceptibility of selected HEK-Cas9 sublibraries A and B is not significantly different (P > 0.05, unpaired two-tailed t-test) compared with parental HEK-Cas9 using the corresponding MOI. Data represent the mean ± SEM of a representative experiment out of 2 performed in technical triplicate.

Fig 3

BteA disrupts cytosolic calcium homeostasis, leading to the permeabilization of cell plasma membrane. (A and B) Calcium influx precedes B. bronchiseptica-induced plasma membrane permeabilization. HeLa cells were infected with BbWT and BbΔbteA mutant at the indicated MOI. Calcium influx was monitored using cytosolic Ca²⁺ indicator Fluo-4/AM (A), whereas plasma membrane permeabilization was determined in parallel wells by fluorescent DNA binding dye CellTox Green (B). Data represent the mean ± SEM of a representative experiment out of 3 performed in technical duplicate. (C) Calcium imaging. HeLa cells loaded with cytosolic Ca²⁺ indicator Fluo-4/AM were infected with BbWT and BbΔbteA, expressing the fluorescent protein mScarlet (mSc), at MOI of 10:1. Sequence of time lapse images is shown. Bacteria, magenta; cytosolic Ca²⁺ indicator Fluo-4/AM, yellow. Scale bar, 20 µm. Data are representative of two independent experiments. (D and E) Treatment with 2-APB delays calcium influx and host cell death. HeLa cells were pre-incubated with 100 µM 2-APB (2-APB+) for 30 min, or left untreated, before being infected with BbWT or BbΔbteA derivative at MOI 25:1. Calcium influx was assessed using Fluo-4/AM Ca²⁺ indicator, (D) whereas plasma membrane permeabilization was determined in parallel wells by fluorescent DNA binding dye CellTox Green (E). Asterisks indicate a statistically significant difference (P < 0.05, unpaired two-tailed t-test) between BbWT and BbWT 2-APB+ infection samples at 1 h 30 min (maximum of BbWT Fluo-4/AM signal) (D) and at 4 h (E). Data represent the mean ± SEM of a representative experiment out of three performed in technical duplicate.

Fig 4

BteA-induced calcium influx correlates with ER and mitochondria fragmentation and yields pronounced elevation of mitochondrial calcium levels. (A) ER calcium imaging. Hela cells, transfected to express ER-targeted red Ca²⁺ sensor ER-LAR-Geco, were loaded with the cytosolic Ca²⁺ indicator Fluo-4/AM and infected with BbWT at MOI of 10:1. A sequence of time-lapse images is shown. Cytosolic Ca²⁺ indicator Fluo-4/AM, yellow; ER Ca²⁺ sensor ER-LAR-Geco, magenta. Scale bar, 20 µm. Data are representative of two independent experiments. The graph in the middle depicts the mean fluorescence intensity of the shown cell quantified over time. The right graph shows the relative fluorescence intensities of individual cells (n = 30) at the time of calcium influx. Asterisks indicate a statistically significant difference (P < 0.05, unpaired two-tailed t-test) between t-3 and t0. Cyto Ca²⁺, black asterisk; ER Ca²⁺, magenta asterisk. (B) Mitochondrial calcium imaging. Hela cells, transfected to express mitochondria-targeted red Ca²⁺ sensor mito-LAR-Geco, were loaded with the cytosolic Ca²⁺ indicator Fluo-4/AM, and infected with BbWT at MOI of 10:1. A sequence of time-lapse images is shown. Cytosolic Ca²⁺ indicator Fluo-4/AM, yellow; mitochondrial Ca²⁺ sensor mito-LAR-Geco, magenta. Scale bar, 20 µm. Data are representative of two independent experiments. The graph in the middle depicts the mean fluorescence intensity of the shown cell quantified over time. The right graph shows the relative fluorescence intensities (RFIs) of individual cells (n = 30) at the time of calcium influx. Asterisks indicate a statistically significant difference (P < 0.05, unpaired two-tailed t-test) between t-3 and t0. Cyto Ca²⁺, black asterisk; mito Ca²⁺, magenta asterisk.

Fig 5

BteA induces loss of mitochondrial membrane potential, mitochondrial swelling, and cristolysis. (A) BteA-induced calcium influx precedes loss of mitochondrial membrane potential. HeLa cells were loaded with the mitochondrial membrane potential indicator TMRM and the cytosolic Ca²⁺ indicator Fluo-4/AM, which was followed by infection with BbWT at MOI 10:1. Sequence of time lapse images is shown. Cytosolic Ca²⁺ indicator Fluo-4/AM, yellow; TMRM, magenta. Scale bar, 20 µm. Data are representative of two independent experiments. The graph in the middle depicts the mean fluorescence intensity of the shown cell (white arrow) quantified over time. The right graph shows the relative fluorescence intensities of individual cells (n = 30) at the time of calcium influx. Asterisks indicate statistically significant difference (P < 0.05, unpaired two-tailed t-test) between t-3 and t-0. Cyto Ca²⁺, black asterisk; TMRM, magenta asterisk. (B) Correlative light and electron microscopy of mitochondria in HeLa cells. HeLa cells, transfected to express mitochondria-targeted monomeric hyperfolder YFP fluorescent protein, were either infected with mScarlet-expressing BbWT at MOI 10:1 for 0.5 h or left untreated. Fluorescence imaging was performed during the fixation period to capture bacteria and mitochondrial structures. Samples were then prepared for electron microscopy and FIB-SEM imaging. Data visualization was processed in Amira 3D 2024.1 software. Mitochondria, magenta; bacteria, cyan. Scale bar FM, 10 µm; scale bar EM, 1 µm.

Fig 6

The schematic model of BteA-induced cell death. Upon delivery, BteA rapidly triggers a sustained elevation of cytosolic Ca²⁺ levels, which leads to mitochondrial Ca²⁺ overload. This disrupts mitochondrial membrane potential, causing swelling and cristae loss. The resulting mitochondrial dysfunction may activate the mitochondrial permeability transition pore (mPTP), driving plasma membrane blebbing and early cell damage. Inhibition of Ca²⁺ influx with the calcium channel modulator 2-APB delays these effects, underscoring the critical role of calcium dysregulation in BteA-induced cell death. Created in BioRender [Kamanova, J. (2024) BioRender.com/l36y378].