Abstract

   Bacterial membrane vesicles (BMVs) are secreted by many pathogenic
   bacteria and known to stimulate various host responses upon infection,
   thereby contributing to the pathogenicity of bacterial pathogens like
   Pseudomonas aeruginosa. While the effects of BMVs on host immune
   responses are well studied, little is known about their impact on cell
   metabolism and mitochondrial respiration. Here, we show that P.
   aeruginosa BMVs (i) reprogram cell metabolism of human lung cells, (ii)
   negatively affect mitochondrial respiration by (iii) specifically
   inhibiting complex III of the electron transport chain, leading to (iv)
   the activation of adenosine monophosphate-activated protein kinase
   (AMPK) signaling, which in turn results in (v) AMPK-dependent
   inhibition of global protein synthesis.

   Keywords: bacterial membrane vesicles (BMVs), Pseudomonas aeruginosa,
   metabolism, electron transport chain, AMPK
     __________________________________________________________________

Significance Statement.

   The study reveals that Pseudomonas aeruginosa bacterial membrane
   vesicles (BMVs) significantly disrupt human lung cell metabolism and
   mitochondrial function. Specifically, BMVs inhibit complex III of the
   electron transport chain, leading to impaired mitochondrial
   respiration. This disruption activates adenosine
   monophosphate-activated protein kinase (AMPK) signaling, which
   subsequently results in an AMPK-dependent inhibition of global protein
   synthesis. These findings elucidate a previously uncharacterized
   mechanism by which BMVs contribute to bacterial pathogenicity.

Introduction

   The opportunistic pathogen Pseudomonas aeruginosa causes milder local
   to severe systemic infections. These infections by often
   multiantibiotic resistant bacteria are particularly prevalent in
   immunocompromised patients, emphasizing its importance as a significant
   health concern in global health care ([38]1, [39]2). During infection,
   P. aeruginosa secretes bacterial membrane vesicles (BMVs) containing
   metabolites, nucleic acids, and proteins, including virulence factors,
   which are delivered to the host cell during infection and play a
   pivotal role in its pathogenicity ([40]3–6). BMVs are well known for
   triggering host immune responses, capable of inducing the release of
   proinflammatory cytokines like interleukin-8 ([41]7–9). Notably, BMVs
   can also activate the adenosine monophosphate-activated protein kinase
   (AMPK) within the host cell, leading to autophagy induction ([42]10).
   However, how the activation of AMPK in response to BMVs is regulated
   remains unclear.

   Given that AMPK serves as a metabolic sensor, activated by
   mitochondrial dysfunction ([43]11, [44]12), and considering evidence of
   BMVs inhibiting mitochondrial activity in macrophages ([45]13), it is
   conceivable that mitochondrial dysfunction might be the primary event
   facilitating BMV-induced AMPK activation. Some bacterial pathogens are
   known to affect mitochondrial function by specifically inhibiting
   protein complexes of the electron transport chain (ETC), which is
   essential for oxidative phosphorylation and ATP generation ([46]14).
   For P. aeruginosa, extracellular secreted factors such as
   1-hydroxyphenazine, pyocyanin, and exotoxin A have been described as
   possible inhibitors of the ETC, leading to reduced mitochondrial
   respiration ([47]15–18). However, the functional basis of mitochondrial
   dysfunction caused by BMVs is still unknown.

   Here, we show that treatment of human lung cells with BMVs isolated
   from the pathogenic P. aeruginosa strain PA14 leads to metabolic
   reprogramming and impaired mitochondrial respiration by specifically
   inhibiting complex III of the ETC, resulting in mitochondrial
   dysfunction. Moreover, this event activates AMPK signaling, leading to
   AMPK-dependent inhibition of global protein synthesis in the host cell.

Materials and methods

Bacterial strains and isolation of BMVs

   The bacterial strains P. aeruginosa PA14 ([48]19) and Pseudomonas
   putida KT2440 ([49]20) were used in this study. Both strains were
   cultivated in lysogeny broth (LB) (Roth, X968.2) in flasks with
   baffles. BMVs were isolated as described before ([50]21). Briefly,
   bacterial cultures were inoculated with an optical density measured at
   a wavelength of 600 nm (OD[600]) of 0.05 and grown at 37 °C (P.
   aeruginosa) or 30 °C (P. putida) at 160 rpm until they reached the
   early-stationary phase of growth. BMVs were isolated from culture media
   after removing bacterial cells by centrifugation at 4 °C and 8,000×g
   for 30 min and filtration through 0.22-µM PES filters (Corning,
   431097). Supernatants were concentrated by ultrafiltration using
   Vivaspin 20 PES ultrafiltration units with a molecular weight cutoff of
   100 kDa (Sartorius, VS2041). BMVs were isolated by ultracentrifugation
   at 4 °C and 150,000×g for 2 h. Isolated vesicles were resuspended in 1
   × PBS (Gibco, 18912-014) and sterile-filtered. BMVs were quantified by
   using the membrane lipid dye FM4-64 (Invitrogen, [51]T13320) to
   calculate the vesicle load per µL (VL/µL) as described before ([52]21).
   Vesicle samples were stored at 4 °C.

Cell culture

   The cell lines A549 (DSMZ, ACC 107), HCC44 (DSMZ, ACC 534), and human
   bronchial epithelial cells (HBEpC, PromoCell, C-12640) were used in
   this study. A549 cells were cultivated in Dulbecco's modified Eagle
   medium (DMEM) (Gibco, 11965-092) containing 25 mM glucose and 4 mM
   glutamine. HCC44 cells were cultivated in RPMI 1640 medium (Gibco,
   21875-034) containing 11 mM glucose and 2 mM glutamine. Both growth
   media were supplemented with 10% FBS (Bio&SELL, FBS.SAM.0500). HBEpC
   cells were cultivated in Airway Epithelial Cell Growth Medium
   supplemented with Growth Medium SupplementMix (PromoCell, C-21060). All
   cell lines were incubated in a humidified atmosphere with 5% CO[2] at
   37 °C. Cell detachment was performed with 0.05% trypsin–EDTA (Gibco,
   25300-054) for A549 and HCC44 cells or accutase solution
   (Sigma-Aldrich, A6964) for HBEpC cells.

Analysis of cell growth and viability

   To analyze cell growth of the proliferating cell lines A549 and HCC44,
   cell confluence was measured in 6-well plates (Greiner Bio-One, 657160)
   at 37 °C with a microplate reader (Tecan Spark). For the analysis of
   cell viability of nonproliferating HBEpC cells, the PrestoBlue cell
   viability reagent (Invitrogen, A13261) was used according to the user
   manual. Fluorescence signals were measured by using an excitation
   wavelength of 560 nm and an emission wavelength of 590 nm.

Stable isotope labeling

   For stable isotope labeling, cells were seeded in 6-well plates and
   incubated at 37 °C and 5% CO[2] overnight. The next day, cells were
   washed with 1 × PBS and treated as indicated in the figure legends in
   the following labeling media: A549 cells were cultivated in DMEM
   (Gibco, A14430-01) supplemented with 25 mM unlabeled glucose or
   [U-^13C[6]]-glucose and 4 mM [U-^13C[5]]-glutamine or unlabeled
   glutamine and 10% dialyzed FBS (dFBS). HCC44 cells were cultivated in
   SILAC RPMI 1640 medium (Gibco, A24942-01) supplemented with 11.1 mM
   unlabeled glucose or [U-^13C[6]]-glucose, 2.05 mM [U-^13C[5]]-glutamine
   or unlabeled glutamine, 0.22 mM lysine, 1.15 mM arginine, and 10% dFBS.
   HBEpC cells were cultivated in DMEM (Gibco, A14430-01) supplemented
   with 25 mM unlabeled glucose or [U-^13C[6]]-glucose, 4 mM
   [U-^13C[5]]-glutamine or unlabeled glutamine, and 2% growth medium
   SupplementMix (PromoCell, C-39165).

Metabolite extraction and gas chromatography coupled with mass spectrometry
analysis

   For gas chromatography coupled with mass spectrometry (GC/MS) analysis
   of cellular metabolites, cells were seeded in 6-well plates and
   incubated at 37 °C and 5% CO[2] overnight. The next day, cells were
   washed with 1 × PBS and treated as indicated in the figure legends.
   Treated cells were extracted as described by Sapcariu et al. ([53]22).
   The upper polar phase was transferred in GC/MS vials and dried in a
   speedvac at 4 °C overnight. Dried samples were derivatized with 15 µL
   of 2% (w/v) methoxyamine hydrochloride solved in pyridine by shaking at
   40 °C for 90 min and additional 15 µL of
   N-methyl-N-(trimethylsilyl)trifluoroacetamide (MSTFA) by shaking at 40
   °C for 30 min or N-tert-butyldimethylsilyl-N-methyltrifluoroacetamide
   (MTBSTFA) by shaking at 55 °C for 60 min.

   The derivatized samples (1 µL) were injected into a SSL injector in
   splitless mode and heated up to 270 °C. GC/MS measurements were taken
   with an Agilent Technologies 7890B GC system including a 30-m
   Phenomenex ZB-35 and a 5-m Duraguard capillary column, connected to an
   Agilent Technologies 5977B MSD, under electron ionization at 70 eV. The
   MS source temperature was held at 230 °C and the quadrupole temperature
   at 150 °C. Helium was used as a carrier gas with a flow rate of
   1 mL/min. The temperature profile of the GC oven depended on the used
   measuring method. For the measurement of MTBSTFA-derivatized samples,
   the GC oven temperature was held at 100 °C for 2 min, then increased up
   to 300 °C at 10 °C/min, and held at 300 °C for 4 min. For the
   measurement of MSTFA-derivatized samples, the GC oven temperature was
   held at 80 °C for 6 min, then increased up to 300 °C at 6 °C/min, held
   at 300 °C for 10 min, raised to 325 °C at 10 °C/min, and held at 325 °C
   for 4 min. The total abundances of metabolites and distributions of
   mass isotopomers were calculated by the integration of mass fragments
   and corrected for natural isotope abundances by using the software
   MetaboliteDetector as previously described ([54]23).

Quantitative PCR analysis

   RNA was isolated from the interphase of extracted cells as described
   before ([55]22) by using the NucleoSpin RNA kit (MACHEREY-NAGEL,
   740955.50). Isolated RNA was converted to cDNA by using the
   High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems,
   4368813). RNA and cDNA concentrations were measured with a microplate
   reader (Tecan Spark). For qPCR analysis, TaqMan gene expression assays
   (Applied Biosystems) for the housekeeping gene 18S (Hs99999901 s1) and
   target genes HMGCR (Hs00168352 m1) and SREBF2 (Hs01081784 m1) were
   used, together with the iTaq Universal Probes Supermix (Bio-Rad,
   1725132) and the QuantStudio 5 Real-Time PCR system (Applied
   Biosystems). Data were analyzed with the QuantStudio design and
   analysis software.

Protein extraction and immunoblotting

   For protein analysis, cells were plated in 10-cm (A549 and HCC44 cells)
   or 6-cm (HBEpC cells) cell culture dishes (Greiner Bio-One) in growth
   medium and incubated at 37 °C and 5% CO[2] overnight. The next day,
   cells were treated as indicated in the figure legends. After treatment,
   cells were lysed with the M-PER extraction reagent (Thermo Scientific,
   78501) containing 1 × Halt protease and phosphatase inhibitors (Thermo
   Scientific, 78441), mixed at 4 °C and 1,400 rpm for 10 min, and
   centrifuged at 4 °C and 14,000×g for 10 min. Supernatants containing
   protein were stored at −20 °C. Protein quantification was performed by
   using the Pierce BCA protein assay kit (Thermo Scientific, 23227)
   according to the user manual. Samples were loaded with a 5 × Laemmli
   buffer, incubated at 95 °C for 5 min, and centrifuged at 16,000×g for
   1 min. Twenty-five micrograms of total protein were separated on 4–20%
   precast SDS–PAGE gels (Bio-Rad). Depending on the protein size, a
   prestained protein marker covering the range of 10–180 kDa (Thermo
   Scientific, 26616) or 43–315 kDa (Cell Signaling Technology, 12949) was
   used.

   Proteins were transferred onto a 0.45-µm PVDF membrane (Carl Roth,
   T830.1) by using the Trans-Blot SD Semi-Dry Transfer Cell (Bio-Rad).
   The membrane was blocked with 5% BSA (w/v) (Biomol, 01400.100) in
   Tris-buffered saline with 0.1% Tween 20 (TBS-T) for 1 h. Primary
   antibodies diluted in 1 × TBS-T with 5% BSA (Table [56]S1) were added
   to the membranes and incubated at room temperature for 1 h. Membranes
   were then incubated with secondary antibodies diluted in 1 × TBS-T with
   5% BSA (Table [57]S2) at room temperature for 1 h. For signal
   detection, the Immobilon Classico Western HRP substrate (Millipore,
   WBLUC0500) was used and imaged with a Bio-Rad ChemiDoc imaging system.
   Band intensities were analyzed by using the ImageJ 1.53k software.

Respirometry

   Cell respiration was measured by using an Agilent Seahorse XFe96
   Analyzer together with Seahorse XFe96 Extracellular Flux Assay Kits.
   Cells were seeded in Seahorse XF96 V3 PS Cell Culture Microplates
   (Agilent, 101085-004) and cultured in growth medium at 37 °C and 5%
   CO[2] overnight. At the same time, the sensor cartridge (Agilent,
   103792-100) was incubated with sterile water together with the Seahorse
   XF Calibrant Solution (Agilent, 100840-000) at 37 °C. The next day, the
   medium was replaced with Seahorse XF DMEM medium, pH 7.4 (Agilent,
   103575-100), supplemented with 10 mM glucose, 2 mM glutamine, and 10%
   dFBs and incubated at 37 °C for 60 min. For sensor cartridge
   calibration, the water was replaced with Seahorse XF Calibrant Solution
   and incubated at 37 °C for 45 min. To assess mitochondrial function,
   the oxygen consumption rate (OCR) (pmol O[2]/min) was measured and
   normalized to basal respiration. Mitochondrial respiratory chain
   deficiencies were analyzed based on the study of Jaber et al. ([58]24).

Results and discussion

P. aeruginosa BMVs inhibit proliferation and reduce viability of human lung
cells

   BMVs from several bacterial species are reported to inhibit the
   proliferation of different host cell types ([59]21, [60]25–27). Thus,
   we aimed to explore the impact of P. aeruginosa BMVs on the
   proliferation of human lung cells in initial experiments. To this end,
   we treated A549 and HCC44 lung cancer cells with BMVs isolated from the
   pathogenic P. aeruginosa strain PA14 and monitored their cell
   confluence over 72 h. Moreover, we analyzed the effects of PA14 BMVs on
   the viability of primary bronchial epithelial cells (HBEpC) by
   fluorescently labeling living cells, as these cells are
   nonproliferating. For all cell lines, we observed a decrease in cell
   confluence (Fig. [61]1A) or viability (Fig. [62]S1A) after vesicle
   treatment, highlighting the pathogenic potential of PA14 BMVs. In
   contrast, treatment with BMVs isolated from the nonpathogenic strain P.
   putida KT2440 resulted in a significantly weaker effect (Fig. [63]S1A
   and [64]B). This suggests that the specific cargo of the pathogenic
   PA14 BMVs was responsible for the host reaction, rather than BMVs in
   general. To additionally exclude that the effects on cell confluence
   are not mediated by potential contaminants from vesicle isolation or
   components of the media and solutions used, we treated A549 cells with
   a LB medium concentrate prepared under identical conditions as BMVs.
   Neither the LB concentrate nor P. putida KT2440 BMVs significantly
   affected A549 cell confluence compared with PBS-treated cells,
   suggesting that the specific cargo of the pathogenic PA14 BMVs was
   responsible for the host reaction (Fig. [65]S1C). However, the precise
   molecular mechanism of the observed antiproliferative effect remains
   unknown.

Fig. 1.

   [66]Fig. 1.
   [67]Open in a new tab

   Metabolic reprogramming in host cells after PA14 BMV treatment. A)
   Confluence of A549 and HCC44 cells after treatment with 25 VL/mL BMVs
   for 72 h. Data were obtained from three replicates. B) Glucose uptake
   and lactate secretion of A549, HCC44, and HBEpC cells after treatment
   with 25 VL/mL BMVs for 24 h. Data were obtained from three replicates.
   C) Signal intensities (AU) of malate, aspartate, and glutamate in A549
   and HBEpC cells after vesicle treatment (25 VL/mL for 24 h). Data were
   obtained from 5 to 6 (A549) or 2 to 3 (HBEpC) replicates and normalized
   to cell confluence. D) Schematic overview of the incorporation of a
   [U-^13C[6]]-glucose tracer into metabolites of the TCA cycle. Created
   with [68]BioRender.com. E) Malate MIDs after [U-^13C[6]]-glucose
   labeling of BMV-treated cells. A549 and HCC44 cells were treated for
   24 h and HBEpC cells for 6 h with 25 VL/mL. Data were obtained from two
   or three replicates. F) Schematic overview of the incorporation of a
   [U-^13C[5]]-glutamine tracer into metabolites of the TCA cycle. Created
   with [69]BioRender.com. G) Malate MIDs after [U-^13C[5]]-glutamine
   labeling of BMV-treated cells. A549 and HCC44 cells were treated for
   24 h and HBEpC cells for 6 h with 25 VL/mL. Data were obtained from
   three replicates. All bar plots in this figure are depicted as mean ±
   SEM. For the comparison between control (Ctrl) and BMV-treated cells
   for each cell line, statistical significance was analyzed using
   unpaired t test (n.s. = not significant, *P < 0.05, **P < 0.01, ***P <
   0.001).

P. aeruginosa BMVs induce rapid metabolic reprogramming by modulating TCA
cycle activity

   BMVs have been recognized for modulating the host immune response
   during P. aeruginosa infection ([70]7, [71]8). However, the underlying
   molecular dynamics remain elusive. Given the close link between
   cellular functions and cell metabolism, investigating the metabolic
   responses of the host cell induced by PA14 BMVs would be revealing for
   the understanding of the host–pathogen interaction of P. aeruginosa.
   Recently, we developed a method for the isolation and quantification of
   BMVs in order to analyze vesicle-induced metabolic changes in mammalian
   cell cultures ([72]21). To ascertain the potential metabolic effects
   induced by BMVs, we initially examined changes in glucose uptake and
   lactate secretion of vesicle-treated lung cancer cells as well as
   primary bronchial epithelial cells. We observed an increase in glucose
   uptake and lactate secretion for all tested cell lines (Fig. [73]1B),
   indicating increased glycolytic activity. This phenomenon is known for
   various bacterial and viral infections, especially in immune cells,
   presumably to provide biosynthetic intermediates for the synthesis of
   nucleotides, amino acids, and lipids to support host cell proliferation
   ([74]28–30).

   To further investigate BMV-induced metabolic shifts in host cells, we
   employed GC/MS to analyze BMV-treated A549, HCC44, and HBEpC cells. Our
   results indicate broad metabolic reprogramming of both lung cancer and
   primary lung cells. Specifically, we observed decreased levels of TCA
   cycle-associated metabolites such as malate, aspartate, and glutamate
   in cells treated with PA14 BMVs (Fig. [75]1C). Analogous to cell
   growth, these effects were reduced when cells were treated with BMVs of
   the nonpathogenic P. putida KT2440 strain (Fig. [76]S1D and [77]E).
   Conversely, the levels of most amino acids were increased in A549 cells
   after BMV treatment, presumably due to autophagy induction, as
   previously reported ([78]21).

   To understand the metabolic pathways contributing to these
   observations, we used stable isotope-assisted metabolomics. By feeding
   cells with [U-^13C[6]]-glucose and [U-^13C[5]]-glutamine during vesicle
   exposure, we noted considerable shifts in the mass isotopomer
   distribution (MID) of TCA cycle-associated metabolites, exemplified by
   malate. As glucose can enter the TCA cycle via acetyl-CoA, the use of
   the [U-^13C[6]]-glucose tracer results in the formation of M2 citrate
   (Fig. [79]1D) and, for the oxidative TCA cycle flux, in M2 malate.
   Interestingly, the formation of M2 malate was decreased in all tested
   cell lines after vesicle treatment (Fig. [80]1E). For A549 and HCC44
   cells, we also observed an increased formation of M3 malate from
   [U-^13C[6]]-glucose, involving the conversion of fully labeled pyruvate
   to oxaloacetate by carboxylation via pyruvate carboxylase (PC). In this
   step, free unlabeled CO[2] gets incorporated into oxaloacetate
   resulting in M3 malate (Fig. [81]1D). The increased formation of M3
   malate in A549 and HCC44 cells after BMV treatment (Fig. [82]1E)
   suggests a shift towards increased PC activity, presumably to
   compensate for the reduced oxidative TCA cycle flux. However, we did
   not observe this effect in HBEpC cells, indicating an inactive PC and
   no possibility to evade the affected flux.

   Using [U-^13C[5]]-glutamine as a substrate, active oxidative TCA cycle
   flux results in M4 malate (Fig. [83]1F). After BMV treatment, we
   observed decreased formation of M4 malate from [U-^13C[5]]-glutamine
   for all tested cell lines (Fig. [84]1G), confirming the ^13C-glucose
   results. Moreover, the formation of M3 malate increased in A549 and
   HCC44 cells (Fig. [85]1G), explained by a shift to the reductive TCA
   cycle flux ([86]31), resulting in M5 citrate that can be converted to
   M2 acetyl-CoA and M3 oxaloacetate via ATP-citrate lyase (ACL, Fig.
   [87]1F). Similar to the glucose labeling results, we did not observe
   the increased formation of M3 malate in HBEpC cells (Fig. [88]1G).
   Deregulation of enzymes such as PC and ACL, associated with the
   reductive TCA cycle flux, is known for various pathological conditions
   like cancer, but also in infection ([89]32, [90]33). Our results
   suggest that the shift towards a higher activity of these enzymes after
   BMV treatment is specific for lung cancer cell metabolism and does not
   occur in primary lung cells. To investigate whether these effects are
   caused by specific virulence agents of PA14 BMVs or vesicles in
   general, we analyzed potential effects of BMVs isolated from the
   nonvirulent P. putida KT2440 strain. We observed much smaller effects
   when cells were treated with these vesicles, which further suggests
   that the effects are due to specific factors contained in BMVs isolated
   from the pathogenic P. aeruginosa PA14 strain (Fig. [91]S1F and [92]G).

   Given that P. aeruginosa BMVs contain lipopolysaccharide (LPS) on their
   surface, which is known to induce a potent immune response in the host
   ([93]8), we analyzed whether LPS mediates similar effects on cell
   metabolism in A549 cells.

   To that end, we performed [U-^13C[6]]-glucose and [U-^13C[5]]-glutamine
   labeling during treatment with 1 µg/mL LPS. We observed no metabolic
   changes, shown by the similar malate MIDs after LPS treatment compared
   with the control cells (Fig. [94]S2), excluding LPS as the responsible
   factor of BMV-mediated metabolic reprogramming.

   To analyze how fast the observed metabolic changes appear after BMV
   treatment, we analyzed the effects after different treatment times (15,
   60, and 240 min). We observed rapid metabolic reprogramming following
   BMV treatment, with decreased M2 and increased M3 malate already
   observable after 15 min in A549 cells (Fig. [95]S3A). Moreover, the
   cellular malate levels were reduced by 20% at this time (Fig. [96]S3B).
   Our results indicate a rapid PA14 vesicle-driven impact on cell
   metabolism, evidenced by the decreased oxidative TCA cycle flux
   following BMV treatment. The swift onset of this effect suggests that
   the observed metabolic changes are not due to prior cellular
   alterations, such as changes at gene expression level. Instead, this
   indicates the presence of a vesicle-associated factor that triggers an
   immediate effect on cellular respiration.

P. aeruginosa BMVs impair mitochondrial respiration of human lung cells by
inhibiting ETC complex III

   The aforementioned metabolic changes indicate a reduced oxidative TCA
   cycle flux after BMV treatment, resulting in decreased generation of
   NADH and FADH[2], which are essential for mitochondrial respiration and
   ATP production. Deo et al. ([97]13) have previously demonstrated the
   inhibitory effects of P. aeruginosa BMVs on mitochondrial activity in
   macrophages, leading to mitochondrial apoptosis and inflammation.
   Consistent with these results, we observed a significantly higher
   NADH/NAD^+ ratio and decreased ATP levels in A549 and HCC44 cells after
   BMV treatment (Fig. [98]2A and B). To determine whether P. aeruginosa
   BMVs alter mitochondrial respiration in human lung cells, we analyzed
   changes of cellular respiration by measuring the OCR in BMV-infected
   A549 cells. We observed a decreased OCR immediately after vesicle
   treatment (Fig. [99]2C), indicating a rapid effect on mitochondrial
   respiration (<7 min), which might explain the swift metabolic
   adaptation discussed above.

Fig. 2.

   [100]Fig. 2.
   [101]Open in a new tab

   PA14 BMVs induce mitochondrial dysfunction in host cells. A) NADH and
   NAD^+ levels of A549 cells after treatment with 25 VL/mL BMVs and 1 mM
   metformin (Met) for 24 h measured as luminescence intensities (RLU).
   Data were obtained from 15 replicates. B) ATP levels of A549 and HCC44
   cells after BMV treatment (25 VL/mL for 24 h) measured as luminescence
   intensities (RLU). Data were obtained from 10 replicates. C) OCR of
   A549 cells after BMV treatment. 25 VL/mL BMVs were added at the
   depicted time point. Data were obtained from 15 replicates. D)
   Respirometry test for complex III activity of the ETC of A459 cells
   after BMV treatment (25 VL/mL). Isolation of complex III was performed
   by the addition of 2 µM rotenone (Rot), 40 µM malonate (Malo), 2 mM
   itaconate (Ita), 25 µg/mL digitonin (Dig), and 1 mM duroquinol (DQ) at
   the depicted time points. Data were obtained from 30 replicates. All
   OCR values in this figure were normalized to the baseline and depicted
   as mean ± SEM. E) Location of electron transport deficiencies by
   functionally isolating cytochrome c and the complexes of the ETC to
   test their activity after vesicle treatment. Rotenone, malonate, and
   itaconate or AA are added to inhibit complex I, II, or III, which
   interrupts the electron transfer to the following complexes. The
   electron transport can be restored by the addition of the electron
   donor duroquinol for complex III or TMPD and its reducing agent
   ascorbate for cytochrome c. PA14 BMVs are able to specifically inhibit
   complex III of the ETC. All bar plots in this figure are depicted as
   mean ± SEM. Statistical significance was analyzed using unpaired
   one-way ANOVA followed by Tukey's multiple comparison test (A) or using
   unpaired t test (B) (n.s. = not significant, ***P < 0.001).

   Mitochondrial respiration relies on oxidative phosphorylation, which
   requires the ETC located in the inner mitochondrial membrane and
   composed of complexes I, II, III, and IV, as well as proteins like
   cytochrome c. It is well known that some bacterial pathogens are able
   to inhibit complexes of the ETC ([102]14). For P. aeruginosa, the
   extracellular secreted factors exotoxin A, 1-hydroxyphenazine, and
   pyocyanin have been reported to affect mitochondrial respiration,
   presumably by disrupting the ETC ([103]15–18). To determine whether P.
   aeruginosa BMVs specifically impair one of the ETC complexes, leading
   to the observed respiratory deficit, we followed the protocol by Jaber
   et al. ([104]24). This protocol involves a stepwise series of
   experiments to map the location of electron transport deficiencies,
   beginning with cytochrome c. If BMVs cause a cytochrome c deficit, then
   adding exogenous cytochrome c should rescue respiration. Because
   cytochrome c cannot cross the plasma membrane, we selectively
   permeabilized the plasma membrane using digitonin. However, we did not
   observe a rescue effect in the OCR of BMV-treated A549 cells
   supplemented with cytochrome c (Fig. [105]S4A), suggesting that a
   deficit in another component of the ETC is limiting respiration.

   After excluding cytochrome c deficiency, we functionally isolated the
   individual complexes of the ETC by selectively limiting electron entry
   to each complex and determining whether the deficiency can still be
   observed. Since complexes I to IV function in series, a dysfunctional
   complex would affect subsequent complexes. For this reason, we started
   with complex IV.

   To functionally isolate complex IV, we inhibited complex III with
   antimycin A (AA) (Fig. [106]2E), which interrupts electron transfer to
   complex IV, leading to a decreased OCR (Fig. [107]S4B). We then added
   N,N,N′,N′-tetramethyl-p-phenylenediamine (TMPD) and its reducing agent
   ascorbate to restore electron transfer to complex IV, bypassing complex
   III (Fig. [108]2E). We observed that complex IV activity was not
   affected by BMVs (Fig. [109]S4B), leaving complexes I to III as
   potential targets.

   Next, we analyzed complex III activity by inhibiting complex I with
   rotenone and complex II with malonate and itaconate (Fig. [110]2E)
   ([111]34, [112]35). To ensure electron transfer to complex III, we
   added the electron donor duroquinol, which recovers respiration by
   feeding electrons into complex III (Fig. [113]2E). Interestingly, we
   observed impaired complex III activity immediately after vesicle
   addition, indicating specific inhibition of this complex by PA14 BMVs
   (Fig. [114]2D). Moreover, we observed impairment of complexes I and II,
   presumably as a consequence of complex III inhibition (Fig. [115]S4C
   and [116]D).

   Overall, these findings suggest that PA14 BMVs specifically target
   complex III of the ETC, leading to a cascade of inhibition that affects
   complexes I and II, thereby impairing overall mitochondrial
   respiration.

P. aeruginosa BMVs suppress cholesterol biosynthesis and global protein
synthesis while inducing inflammation

   Since BMVs are able to induce drastic effects on host cell metabolism,
   we assumed that other general cellular processes would also be
   impaired. To gain an overview of altered cellular functions, we
   performed RNA-sequencing analysis followed by pathway enrichment
   analysis (Fig. [117]S5). We identified the suppression of cholesterol
   biosynthesis as an important target pathway in A549 cells treated with
   PA14 BMVs, as all participating genes were downregulated (Fig.
   [118]3A). Suppression of cholesterol biosynthesis is well documented in
   viral infections ([119]36), but less explored in bacterial infections.
   For further validation, we analyzed the expression of selected
   downregulated genes by using quantitative PCR (Figs. [120]3B and
   [121]S6). We observed a decreased expression of
   3-hydroxy-3-methylglutaryl-coenzyme A reductase (HMGCR) and sterol
   regulatory element binding transcription factor 2 (SREBF2) after 24 h
   of BMV exposure. HMGCR is the rate-limiting enzyme of the cholesterol
   synthesis pathway ([122]37), and its expression can be activated by
   SREBF2 which regulates various key enzymes associated with cholesterol
   and fatty acid synthesis ([123]38, [124]39).

Fig. 3.

   [125]Fig. 3.
   [126]Open in a new tab

   PA14 BMVs affect transcription and translation of the host cell. A)
   RNA-sequencing analysis of A549 cells treated with PA14 BMVs (PA) for
   24 h. B) Relative gene expression of HMGCR and SREBF2 in A549 cells
   after treatment with 25 VL/mL BMVs for 24 h. Data were obtained from
   three replicates. C) IL-8 secretion of A549 and HBEpC cells after BMV
   treatment (25 VL/mL for 24 h). Data were obtained from three
   replicates. D) Global protein synthesis of A549 cells after treatment
   with 50 µg/mL CHX and 25 VL/mL BMVs for 30 min (+30 min pretreatment).
   Data were obtained from four or five replicates. All bar plots in this
   figure are depicted as mean ± SEM. Statistical significance was
   analyzed using unpaired t test (B, C) or using unpaired one-way ANOVA
   followed by Tukey's multiple comparison test (D) (*P < 0.05, **P <
   0.01, ***P < 0.001).

   We also observed an upregulation of markers associated with
   inflammation (TNF signaling) and disease (lung fibrosis) after BMV
   treatment (Fig. [127]3A). For example, we determined a higher
   expression of interleukin-8 (IL-8, CXCL8), which plays a key role in
   acute inflammation ([128]40). Bauman and Kuehn have previously shown
   that vesicles isolated from the P. aeruginosa PAO1 strain induce IL-8
   activation in lung epithelial cells ([129]7). To confirm this result
   using PA14 BMVs, we exposed lung cancer cells (A549) and primary lung
   cells (HBEpC) to BMVs for 24 h and measured their IL-8 secretion (Fig.
   [130]3C). We observed an increased IL-8 secretion for both cell lines,
   with a more pronounced effect in A549 cells, which could be explained
   by the general importance of IL-8 in cancer progression ([131]41).
   Next, we analyzed whether vesicles affect not only transcription, but
   also translation in the host cell, since there are known bacterial
   regulators of protein translation, such as the exotoxin A of P.
   aeruginosa ([132]42). Moreover, inhibition of protein synthesis by BMVs
   has already been described in macrophages due to mitochondrial stress
   ([133]13). To confirm similar changes in human lung cells, we labeled
   newly translated proteins of A549 cells with a fluorescent dye and
   analyzed their signal intensities after treatment with the translation
   inhibitor cycloheximide (CHX) and PA14 BMVs. We observed a decreased
   fluorescence signal for both treatments already after 30 min,
   indicating the inhibition of global protein synthesis by PA14 BMVs in
   A549 cells (Fig. [134]3D).

P. aeruginosa BMVs activate AMPK signaling through mitochondrial dysfunction,
leading to global protein synthesis inhibition

   We showed that PA14 BMVs are able to affect general cellular functions
   of the host cell, suggesting the activation of a global signaling
   pathway inside the cell. The AMPK is known to function as a metabolic
   sensor that can be activated e.g. by mitochondrial dysfunction, sensing
   decreased ATP levels caused by ETC inhibition ([135]43). Moreover,
   mitochondria-localized AMPK is known to enable mitochondrial function
   ([136]44). Losier et al. ([137]10) demonstrated that AMPK is stimulated
   by the detection of BMVs during infection with the pathogen Salmonella
   enterica serovar Typhimurium, resulting in autophagy induction.
   Interestingly, AMPK activation is known to downregulate HMGCR, which in
   turn suppresses cholesterol synthesis ([138]45), aligning with our
   result that HMGCR gene expression is decreased after BMV treatment
   (Fig. [139]3A and B).

   To confirm whether AMPK signaling is activated by PA14 BMVs, we treated
   A549 cells with P. aeruginosa BMVs and analyzed the activation of the
   AMPK target enzyme acetyl-CoA carboxylase 1 (ACC1, Fig. [140]4A) by
   determining its phosphorylation via western blot (WB) analysis. When
   phosphorylated, ACC1 inhibits the conversion of acetyl-CoA to
   malonyl-CoA, which is needed for fatty acid synthesis ([141]46). We
   observed a general decrease in ACC1 protein levels after vesicle
   treatment, while the p-ACC1/ACC1 ratio increased in a
   concentration-dependent manner, indicating the activation of AMPK by
   PA14 BMVs (Fig. [142]4B). This result also suggests a regulation of
   ACC1 protein translation by BMVs, aligning with our observation that
   PA14 BMVs inhibit global protein synthesis in the host cell (Fig.
   [143]3D). Since AMPK signaling can regulate protein translation via the
   eukaryotic elongation factor 2 (eEF2), we analyzed its activation by WB
   analysis and observed a highly increased p-eEF2/eEF2 ratio after BMV
   treatment (Fig. [144]4C). As eEF2 acts as a negative regulator of
   protein synthesis ([145]47), its activation indicates an inhibition of
   global protein synthesis by PA14 BMVs. We observed the activation of
   eEF2 in lung cancer cells (Fig. [146]4C and D) as well as in primary
   lung cells (Fig. [147]4E), with a higher phosphorylation increase in
   cancer cells.

Fig. 4.

   [148]Fig. 4.
   [149]Open in a new tab

   Activation of AMPK signaling in host cells by PA14 BMVs. A) Overview of
   the AMPK signaling targets ACC1 and eEF2 and their activation. B)
   Western blot (WB) analysis of p-ACC1 (Ser79) in A549 cells after
   vesicle treatment (25 and 50 VL/mL) for 24 h. Data were obtained from
   two independent experiments. C) WB analysis of p-eEF2 (Thr56) in A549
   cells after vesicle treatment (25 and 50 VL/mL) for 24 h. Data were
   obtained from two independent experiments. D) WB analysis of p-eEF2
   (Thr56) in HCC44 cells after vesicle treatment (25 and 50 VL/mL) for
   24 h. E) WB analysis of p-eEF2 (Thr56) in HBEpC cells after vesicle
   treatment (25 and 50 VL/mL) for 24 h. All bar plots in this figure are
   depicted as mean (±SEM).

   Taken together, our results suggest that PA14 BMVs inhibit global
   protein synthesis in both lung cancer and primary lung cells, with a
   more pronounced effect in cancer cells.

   Since it has also been shown that BMVs of Salmonella enterica serovar
   Typhimurium inhibit mTOR signaling ([150]10), and given that eEF2 can
   also be regulated by mTOR, we analyzed the activity of the upstream
   kinase eEF2K, using an antibody for the mTOR-specific phosphorylation
   site Ser366 ([151]47, [152]48). However, we did not observe any change
   in phosphorylation at this site (Fig. [153]S7), suggesting that the
   vesicle-driven effects on protein synthesis are regulated independently
   of mTOR. To confirm that translation is inhibited via AMPK signaling,
   we treated BMV-infected A549 cells with the AMPK inhibitor compound C
   (CC) and analyzed the activity of ACC1 and eEF2. We discovered that
   AMPK inhibition suppressed the vesicle-mediated activation of both
   targets (Fig. [154]5A and B), connecting the inhibition of protein
   synthesis by BMVs to the activation of AMPK signaling.

Fig. 5.

   [155]Fig. 5.
   [156]Open in a new tab

   AMPK-mediated inhibition of global protein synthesis through ETC
   inhibition after PA14 BMV treatment. A) WB analysis of p-ACC1 (Ser79)
   in A549 cells after treatment with 25 VL/mL BMVs and 10 µM CC for 24 h.
   B) WB analysis of p-eEF2 (Thr56) in A549 cells after treatment with
   25 VL/mL BMVs and 10 µM CC for 24 h. C) Global protein synthesis of
   A549 cells after treatment with 25 VL/mL BMVs or 1 µM AA for 30 min
   (+30 min pretreatment). Data were obtained from five replicates. D)
   Overview of the effects of PA14 BMVs on the host cell and their
   underlying signaling pathway. Vesicles specifically inhibit complex III
   of the ETC resulting in mitochondrial dysfunction which activates the
   AMPK signaling pathway leading to an AMPK-dependent inhibition of
   protein synthesis via the translation regulator eEF2. Mitochondrial
   dysfunction and AMPK activation both induce a metabolic reprogramming
   of the host cell. Created with [157]BioRender.com. All bar plots in
   this figure are depicted as mean (±SEM). Statistical significance was
   analyzed using unpaired one-way ANOVA followed by Tukey's multiple
   comparison test (C) (n.s. = not significant, **P < 0.01).

   As mentioned, AMPK is known to sense mitochondrial dysfunction caused
   by ETC inhibition. Toyama et al. ([158]43) have already shown that AA,
   a commonly used inhibitor of complex III of the ETC, can rapidly
   activate AMPK, leading to mitochondrial fragmentation. A similar
   connection between mitochondrial dysfunction and AMPK activation is
   observed with the common type 2 diabetes drug metformin. Metformin
   inhibits complex I of the ETC and activates AMPK signaling, resulting
   in decreased cell proliferation and protein synthesis ([159]49).

   Since BMVs and AA both target complex III of the ETC, we wanted to test
   whether this impairment results in an inhibition of host cell protein
   synthesis. To that end, we treated A549 cells with PA14 BMVs and AA and
   analyzed global protein synthesis. We observed a similar inhibitory
   effect on global protein synthesis for both treatments (Fig. [160]5C).

   Taken together, our results suggest that PA14 BMVs mediate
   mitochondrial dysfunction in the host cell due to the specific
   inhibition of complex III of the ETC, which then activates AMPK
   signaling, leading to AMPK-dependent inhibition of protein synthesis
   (Fig. [161]5D).

Conclusion

   In this study, we investigated the effects of P. aeruginosa PA14 BMVs
   on cell metabolism and mitochondrial respiration in human lung cells.
   In summary, we observed cellular metabolic reprogramming following
   vesicle treatment, particularly affecting TCA cycle-associated
   metabolites. In agreement with recent investigations ([162]13), we
   identified vesicle-driven mitochondrial dysfunction, but more
   specifically, inhibition of complex III of the mitochondrial ETC by
   PA14 BMVs. We further demonstrated that BMV-mediated activation of the
   metabolic sensor AMPK ([163]10) is presumably a consequence of impaired
   complex III function, and identified AMPK-dependent inhibition of
   global protein synthesis in the host cell. However, the findings
   presented in this study are limited by the reliance on in vitro
   experiments and the use of a single P. aeruginosa PA14 laboratory
   strain, which does not reflect the genetic diversity and virulence
   profiles of clinical isolates. To enhance the biological relevance and
   translational potential of the results, the use of primary cells from
   multiple donors and in vivo models should be considered in further
   studies. Despite these limitations, our study builds upon previously
   reported effects of BMVs on host cells, advancing this knowledge by
   uncovering a link between mitochondrial dysfunction and AMPK activation
   mediated by BMVs ([164]10, [165]13).

Supplementary Material

   pgaf248_Supplementary_Data
   [166]pgaf248_supplementary_data.docx^ (1.9MB, docx)

Contributor Information

   Julia Müller, Braunschweig Integrated Centre of Systems Biology
   (BRICS), Technische Universität Braunschweig, Rebenring 56, 38106
   Braunschweig, Germany; Department of Bioinformatics and Biochemistry,
   Technische Universität Braunschweig, Rebenring 56, 38106 Braunschweig,
   Germany.

   Marcel Kretschmer, Braunschweig Integrated Centre of Systems Biology
   (BRICS), Technische Universität Braunschweig, Rebenring 56, 38106
   Braunschweig, Germany; Department of Bioinformatics and Biochemistry,
   Technische Universität Braunschweig, Rebenring 56, 38106 Braunschweig,
   Germany.

   Elise Opitsch, Braunschweig Integrated Centre of Systems Biology
   (BRICS), Technische Universität Braunschweig, Rebenring 56, 38106
   Braunschweig, Germany; Department of Bioinformatics and Biochemistry,
   Technische Universität Braunschweig, Rebenring 56, 38106 Braunschweig,
   Germany.

   Svea Holland, Braunschweig Integrated Centre of Systems Biology
   (BRICS), Technische Universität Braunschweig, Rebenring 56, 38106
   Braunschweig, Germany; Leibniz Institute DSMZ, German Collection of
   Microorganisms and Cell Cultures GmbH, Inhoffenstraße 7B, 38124
   Braunschweig, Germany.

   José Manuel Borrero-de Acuña, Departamento de Microbiología, Facultad
   de Biología, Universidad de Sevilla, Av. de la Reina Mercedes 6, 41012
   Sevilla, Spain.

   Dieter Jahn, Braunschweig Integrated Centre of Systems Biology (BRICS),
   Technische Universität Braunschweig, Rebenring 56, 38106 Braunschweig,
   Germany; Institute for Microbiology, Technische Universität
   Braunschweig, Rebenring 56, 38106 Braunschweig, Germany.

   Meina Neumann-Schaal, Braunschweig Integrated Centre of Systems Biology
   (BRICS), Technische Universität Braunschweig, Rebenring 56, 38106
   Braunschweig, Germany; Leibniz Institute DSMZ, German Collection of
   Microorganisms and Cell Cultures GmbH, Inhoffenstraße 7B, 38124
   Braunschweig, Germany.

   Andre Wegner, Braunschweig Integrated Centre of Systems Biology
   (BRICS), Technische Universität Braunschweig, Rebenring 56, 38106
   Braunschweig, Germany; Department of Bioinformatics and Biochemistry,
   Technische Universität Braunschweig, Rebenring 56, 38106 Braunschweig,
   Germany.

Supplementary Material

   Supplementary material is available at [167]PNAS Nexus online.

Funding

   Work in A.W.'s laboratory was supported by the Ministry of Science and
   Culture (MWK) of Lower Saxony, Germany (SMART BIOTECS alliance between
   the Technische Universität Braunschweig and the Leibniz Universität
   Hannover) and the German Federal Ministry of Education and Research
   (BMBF, PeriNAA-01ZX1916B).

Author Contributions

   Julia Müller (Data curation, Formal Analysis, Investigation,
   Visualization, Writing—original draft, Writing—review & editing),
   Marcel Kretschmer (Data curation, Formal Analysis, Investigation,
   Writing—review & editing), Elise Opitsch (Investigation), Svea Holland
   (Investigation, Writing—review & editing), José Manuel Borrero-de Acuña
   (Conceptualization, Writing—review & editing), Dieter Jahn
   (Conceptualization, Writing—review & editing), Meina Neumann-Schaal
   (Conceptualization, Supervision, Writing—review & editing), and Andre
   Wegner (Conceptualization, Funding acquisition, Supervision,
   Visualization, Writing—original draft, Writing—review & editing).

Preprints

   A preprint of this article is published at
   [168]https://doi.org/10.1101/2024.06.17.599317.

Data Availability

   All data are included in the manuscript and [169]Supplementary
   material. The RNA-Seq dataset (RNA-Seq PA14 A549) is available at
   [170]https://doi.org/10.15490/FAIRDOMHUB.1.STUDY.1393.2.

References