ER-transiting bacterial toxins amplify STING innate immune responses and elicit ER stress

Abstract

The cGAS/STING sensor system drives innate immune responses to intracellular microbial double-stranded DNA (dsDNA) and bacterial cyclic nucleotide second messengers (e.g., c-di-AMP). STING-dependent cell-intrinsic responses can increase resistance to microbial infection and speed pathogen clearance. Correspondingly, STING activation and signaling are known to be targeted for suppression by effectors from several bacterial pathogens. Whether STING responses are also positively regulated through sensing of specific bacterial effectors is less clear. We find that STING activation through dsDNA, by its canonical ligand 2'-3' cGAMP, or the small molecule DMXAA is potentiated following intracellular delivery of the AB₅ toxin family member pertussis toxin from Bordetella pertussis or the B subunit of cholera toxin from Vibrio cholerae. Entry of pertussis toxin or cholera toxin B into mouse macrophages triggers markers of endoplasmic reticulum (ER) stress and enhances ligand-dependent STING responses at the level of STING receptor activation in a manner that is independent of toxin enzymatic activity. Our results provide an example in which STING responses integrate information about the presence of relevant ER-transiting bacterial toxins into the innate inflammatory response and may help to explain the in vivo adjuvant effects of catalytically inactive toxins.

Keywords: ADP ribosylation; GPCR; STING; cGAS; cholera toxin; endoplasmic reticulum; pertussis toxin; type I interferon.

Fig 1

Pertussis toxin intoxication amplifies STING-dependent innate immune responses to cyclic nucleotides and dsDNA. (A) Primary murine BMDMs were cultured in media alone or intoxicated overnight with 250 ng/mL of purified PTx. The following morning, media containing toxin were removed and replaced with fresh culture media. BMDMs were subsequently stimulated by the addition of 10 µg/mL cGAMP to culture media for 4 hrs. Total RNA was harvested, and the expression of indicated genes was quantified by qRT-PCR. ^* P = 0.00234 left panel; ^* P = 0.0115 right panel. (B) Primary BMDMs were treated with 250 ng/mL PTx alone for 6 hrs or pre-treated with 250 ng/mL of PTx for 1, 6, or 18 hrs followed by changing of culture media and immediate 4-hr treatment with 10 µg/mL cGAMP. Total RNA was harvested, and the expression of indicated genes was quantified by qRT-PCR. ^* P = 0.001. (C) RAW264.7 cells were intoxicated overnight with 250 ng/mL of purified pertussis toxin. The following morning, media containing toxin were changed, and cells were transfected with 5 µg total ISD dsDNA for 5 hrs. Total RNA was harvested, and the expression of indicated genes was quantified by qRT-PCR. ^* P = 0.00056 left panel; ^* P = 0.00033 right panel. n = 5. (D) Cell culture supernatants harvested at 4 hrs from macrophages treated as in (A) were analyzed for the expression of IFN-β protein by ELISA. ^* P = 0.000345 for cGAMP; ^* P = 0.0045 for PTx/cGAMP. (E) BMDMs were treated for 18 hrs with media alone, 10 ng/mL Escherichia coli lipopolysaccharide (LPS), 10 µg/mL cGAMP, or 250 ng/mL of purified pertussis toxin. Cells were lysed, and lysates were probed using antibodies against the indicated proteins by western blot. (F) BMDMs were treated for 18 hrs with media alone or 250 ng/mL of purified pertussis toxin and subsequently stimulated for 8 hrs with 10 µg/mL cGAMP.

Fig 2

Pertussis toxin intoxication increases STING-dependent innate signaling. (A–C) BMDMs were cultured in media alone or intoxicated overnight (18 hrs) with 250 ng/mL of purified pertussis toxin. The following morning, media containing toxin were removed and replaced with fresh culture media. BMDMs were stimulated by the addition of STING ligand DMXAA for indicated times. Whole-cell lysates were harvested and analyzed by immunoblot using antibodies directed against the indicated protein targets. Representative of n = 5.

Fig 3

Pertussis toxin increases hallmarks of STING receptor activation by DMXAA. (A) BMDMs were cultured in media alone or intoxicated overnight with 250 ng/mL of purified pertussis toxin. The following morning, media containing toxin were removed and replaced with fresh culture media. BMDMs were stimulated by the addition of STING ligand DMXAA for indicated times. Whole-cell lysates were analyzed by immunoblot analysis under non-reducing conditions using the indicated antibodies. (B) BMDMs were stimulated as in (A), and whole-cell lysates were analyzed under standard reducing conditions using antibodies against the indicated protein targets. Representative of n = 3.

Fig 4

Pertussis toxin potentiates STING responses independent of ADP-ribosylating activity. (A) BMDMs were treated overnight with media alone, 250 ng/mL of WT, or 250 ng/mL R9K;E129A mutant purified pertussis toxin. The following morning, media containing toxin were removed and replaced with fresh culture media. BMDMs were stimulated by the addition of 10 µg/mL cGAMP to culture media for 4 hrs. Total RNA was harvested, and the expression of indicated genes was quantified by qRT-PCR. ^* P < 0.01. (B) Cell culture supernatants harvested at 4 hrs from macrophages treated in (A) were analyzed for the expression of IFN-β protein by ELISA. ^* P < 0.05.

Fig 5

Pertussis toxin induces canonical markers of ER stress independent of catalytic activity, and ER stress contributes to the effect of PTx on STING. (A) BMDMs were treated overnight with media alone, 250 ng/mL of WT, or 250 ng/mL R9K;E129A mutant purified pertussis toxin. The following morning, media containing toxin were removed and replaced with fresh culture media for 4 hrs. Total RNA was then harvested, and the expression of Bip/GRP78 was quantified by qRT-PCR. (B) BMDMs were treated as in (A), and whole-cell lysates were harvested and analyzed by immunoblot for the indicated genes. Representative of n = 4 (C) BMDMs was cultured overnight with media alone or 250 ng/mL of WT PTx in the presence or absence of 100 µM TUDC. The following morning, media containing toxin was removed and replaced with fresh culture media. BMDMs were stimulated by the addition of 10 µg/mL cGAMP to culture media for 4 hrs. Total RNA was harvested, and the expression of indicated genes was quantified by qRT-PCR. ^* P < 0.01.

Fig 6

The catalytically inactive cholera toxin B subunit induces ER stress and primes STING responses. (A and B) BMDMs were cultured in media alone or intoxicated for 3 hrs with 10 µg of purified cholera toxin. BMDMs were subsequently stimulated by the addition of 10 µg/mL cGAMP to culture media for 4 hrs. Total RNA was harvested, and the expression of indicated genes was quantified by qRT-PCR. ^* P ≤ 0.05. (C) BMDMs were pre-treated with Forskolin for 3 hrs prior to stimulation with cGAMP for 4 hrs. Representative of n = 3. ^* P ≤ 0.05. (D) BMDMs were pre-treated with purified CTx B overnight prior to stimulation with 10 µg of cGAMP for 3 hrs. ^* P = 0.003. (E) BMDMs were treated with CTx B overnight as in (D). Whole-cell lysates were harvested and analyzed by immunoblot using antibodies directed against the indicated protein targets. Representative of n = 3. (F) BMDMs were cultured overnight with media alone or 20 ng of CTx B in the presence or absence of 100 µM TUDC. The following morning, media containing toxin were removed and replaced with fresh culture media. BMDMs were stimulated by the addition of 10 µg/mL cGAMP to culture media for 4 hrs. Total RNA was harvested, and the expression of indicated genes was quantified by qRT-PCR. ^* P < 0.05. N = 3. (G) Immortalized murine intestinal epithelial cells were treated overnight with media alone, 20 ng/mL CTx B, or 200 ng/mL PTx followed by stimulation with DMXAA for 4 hrs.