Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2024 Aug 20:13:RP96678.
doi: 10.7554/eLife.96678.

Released bacterial ATP shapes local and systemic inflammation during abdominal sepsis

Daniel Spari  1   2 , Annina Schmid  1   2 , Daniel Sanchez-Taltavull  1   2 , Shaira Murugan  1   2 , Keely Keller  1   2 , Nadia Ennaciri  1   2 , Lilian Salm  1   2 , Deborah Stroka  1   2 , Guido Beldi  1   2
Affiliations

Released bacterial ATP shapes local and systemic inflammation during abdominal sepsis

Daniel Spari et al. Elife. .

Abstract

Sepsis causes millions of deaths per year worldwide and is a current global health priority declared by the WHO. Sepsis-related deaths are a result of dysregulated inflammatory immune responses indicating the need to develop strategies to target inflammation. An important mediator of inflammation is extracellular adenosine triphosphate (ATP) that is released by inflamed host cells and tissues, and also by bacteria in a strain-specific and growth-dependent manner. Here, we investigated the mechanisms by which bacteria release ATP. Using genetic mutant strains of Escherichia coli (E. coli), we demonstrate that ATP release is dependent on ATP synthase within the inner bacterial membrane. In addition, impaired integrity of the outer bacterial membrane notably contributes to ATP release and is associated with bacterial death. In a mouse model of abdominal sepsis, local effects of bacterial ATP were analyzed using a transformed E. coli bearing an arabinose-inducible periplasmic apyrase hydrolyzing ATP to be released. Abrogating bacterial ATP release shows that bacterial ATP suppresses local immune responses, resulting in reduced neutrophil counts and impaired survival. In addition, bacterial ATP has systemic effects via its transport in outer membrane vesicles (OMV). ATP-loaded OMV are quickly distributed throughout the body and upregulated expression of genes activating degranulation in neutrophils, potentially contributing to the exacerbation of sepsis severity. This study reveals mechanisms of bacterial ATP release and its local and systemic roles in sepsis pathogenesis.

Keywords: ATP; E. coli; OMV; bacterial ATP; immunology; infectious disease; inflammation; microbiology; mouse; sepsis.

Plain language summary

Sepsis is a severe condition often caused by the body’s immune system overreacting to bacterial infections. This can lead to excessive inflammation which damages organs and requires urgent medical care. With sepsis claiming millions of lives each year, new and improved ways to treat this condition are urgently needed. One potential strategy for treating sepsis is to target the underlying mechanisms controlling inflammation. Inflamed and dying cells release molecules called ATP (the energy carrier of all living cells), which strongly influence the immune system, including during sepsis. In the early stages of an infection, ATP acts as a danger signal warning the body that something is wrong. However, over time, it can worsen infections by disturbing the immune response. Similar to human cells, bacteria release their own ATP, which can have different impacts depending on the type of bacteria and where they are located in the body. However, it is not well understood how bacterial ATP influences severe infections like sepsis. To investigate this question, Spari et al analysed how ATP is released from Escherichia coli, a type of bacteria that causes severe infections. This revealed that the bacteria secrete ATP directly in to their environment and via small membrane-bound structures called vesicles. Spari et al. then probed a mouse model of abdominal sepsis which had been infected with E. coli that release either normal or low levels of ATP. They found that the ATP released from E. coli impaired the mice’s survival and lowered the number of neutrophils (immune cells which are important for defending against bacteria) at the site of the infection. The ATP secreted via vesicles also altered the role of neutrophils but in more distant regions, and it is possible that these changes may be contributing to the severity of sepsis. These findings provide a better understanding of how ATP released from bacteria impacts the immune system during sepsis. While further investigation is needed, these findings may offer new therapeutic targets for treating sepsis.

PubMed Disclaimer

Conflict of interest statement

DS, AS, DS, SM, KK, NE, LS, DS, GB No competing interests declared

Figures

Figure 1.
Figure 1.. Sepsis-associated bacteria release adenosine triphosphate (ATP) in a growth-dependent manner.
(A) Experimental approach to isolate and cultivate sepsis-associated bacteria from abdominal fluid of patients with abdominal sepsis. (B) Bacterial species identified by whole 16S-rRNA Sanger sequencing from abdominal fluid of patients with abdominal sepsis. Three colonies out of 25 could not be identified. (C) Measurement of released ATP (M) and growth (OD600) over time (hours) from the four sepsis-associated bacteria E. coli, K. pneumoniae, E. faecalis, and S. aureus isolated from patients. N=2 independent bacteria cultures. Means and standard deviations are shown. (D) Area under the curve (AUC) of released ATP over time (M*hours) of the previously assessed bacteria (cumulative ATP). One-way ANOVA, N=2 independent bacteria cultures. Means and individual values are shown. (E) Experimental approach to isolate and cultivate sepsis-associated bacteria from abdominal fluid of mice with abdominal sepsis. (F) Bacterial species identified by whole 16S-rRNA Sanger sequencing from abdominal fluid of mice with abdominal sepsis. Seven colonies out of 25 could not be identified. (G) Measurement of released ATP (M) and growth (OD600) over time (hours) from the three sepsis-associated bacteria E. coli, E. faecalis, and S. aureus isolated from mice. N=2 independent bacteria cultures. Means and standard deviations are shown. (H) AUC of released ATP over time (M*hours) of the previously assessed bacteria (cumulative ATP). One-way ANOVA, N=2 independent bacteria cultures. Means and individual values are shown.
Figure 1—figure supplement 1.
Figure 1—figure supplement 1.. Experimental approach to measure released bacterial adenosine triphosphate (ATP) and growth over time.
Figure 2.
Figure 2.. Adenosine triphosphate (ATP) release is dependent on ATP synthesis.
(A) Illustration depicting the location of ATP synthase and cytochrome bo3 oxidase in gramneg bacteria. (B) Measurement of released ATP (M) and growth (OD600) over time (hours) from cytochrome bo3 oxidase (cyo) and ATP synthase (atp) mutants. The parental strain (PS) was added as a control. N=2 independent bacteria cultures. Means and standard deviations are shown. (C) Area under the curve (AUC) of released ATP over time (M*hours) of the previously assessed bacteria (cumulative ATP) is shown individually in the left panel. N=2 independent bacteria cultures. Means and individual values are shown. Means of grouped cyo and atp mutants are compared in the right panel. t-Test. Means and individual values are shown. (D) Cumulative ATP (M*hours) and cumulative growth (OD600*hours) of all assessed cyo and atp mutants and the PS were plotted against each other. Pearson’s correlation (r) and coefficient of determination (R2) of the applied linear model are depicted. 95% confidence level is shown by the black dashed lines.
Figure 2—figure supplement 1.
Figure 2—figure supplement 1.. Peak ATP (M) and peak growth (OD600*hours) of all assessed cyo and atp mutants and the PS were plotted against each other.
Pearson’s correlation (r) and coefficient of determination (R2) of the applied linear model are depicted. 95% confidence level is shown by the black dashed lines.
Figure 3.
Figure 3.. Outer membrane integrity and bacterial death determine bacterial adenosine triphosphate (ATP) release during growth.
(A) Illustration depicting the location of outer membrane porins in gramneg bacteria. (B) Measurement of released ATP (M) and growth (OD600) over time (hours) from outer membrane porin mutants. The parental strain (PS) and the PS supplemented with either 1 mM Ca2+ or 0.5 mM EDTA were added as controls. N=2 independent bacteria cultures. Means and standard deviations are shown. The red line marks the individual peak of ATP release and growth (OD600) at that time point. (C) Area under the curve (AUC) of released ATP over time (M*hours) of the previously assessed bacteria (cumulative ATP). One-way ANOVA, N=2 independent bacteria cultures. Means and individual values are shown. (D) ATP concentration (M) and growth (OD600) at the individual peak of ATP release of all assessed outer membrane porin mutants, the PS, and the PS+Ca2+ (no peak for the EDTA control) were plotted against each other. Pearson’s correlation (r) and coefficient of determination (R2) of the applied linear model are depicted. 95% confidence level is shown by the black dashed lines. (E) Gating strategy to identify added counting beads, live, injured, and dead bacteria. (F) Quantitative assessment of injured and dead bacteria, as identified by flow cytometry after 4 hr of culturing (ATP peak) of the PS, ΔompF and ΔompC. One-way ANOVA followed by Tukey post hoc test, N=4 independent bacteria cultures. Means and individual values are shown. (G) ATP concentration (M) after 4 hr of culturing (ATP peak) of the PS, ΔompF and ΔompC. One-way ANOVA followed by Tukey post hoc test, N=2 independent bacteria cultures. Means and individual values are shown.
Figure 4.
Figure 4.. Bacterial adenosine triphosphate (ATP) reduces neutrophil counts and impairs sepsis outcome in vivo.
(A) Experimental approach to determine the local role of bacterial ATP in vivo, intraabdominal (i.a.) injecting parental strain (PS)+pEMPTY or PS+pAPY. (B) Measurement of released ATP (M) in bacteria culture supernatant immediately before bacteria were i.a. injected. t-Test, N=2 independent bacteria cultures. Means and individual values are shown. (C) Measurement of ATP (M) in abdominal fluid from mice 4 hr after i.a. injection of bacteria. t-Test, n=5 animals per group of N=2 independent experiments. Means and individual values are shown. (D) Quantitative assessment of colony forming units in abdominal fluid and (E) blood from mice 4 hr after i.a. injection of bacteria. Wilcoxon rank-sum test, n=5 animals per group of N=2 independent experiments. Means and individual values are shown. No growth for controls was detected. (F) Kaplan-Meier curves of mice after i.a. injection of bacteria. Log-rank test, n=10 animals per group. (G) Heatmap showing surface marker expression (x-axis), which was used to characterize the different immune cell populations (y-axis). (H) Concatenated (n=5 animals for each treatment group, n=3 animals for control group of N=2 independent experiments) and down-sampled images of immune cell populations characterized in the abdominal cavity 4 hr after sham treatment or i.a. injection of bacteria. (I) Abundance of neutrophils, small peritoneal macrophages (SPM), and CX3CR1pos monocytes in abdominal fluid from mice 4 hr after sham treatment or i.a. injection of bacteria. One-way ANOVA followed by Tukey post hoc test, n=5 animals for each treatment group, n=3 animals for control group of N=2 independent experiments. Means and individual values are shown.
Figure 4—figure supplement 1.
Figure 4—figure supplement 1.. Immune cell characterization 8 hr after intraabdominal (i.a.) injection of bacteria.
(A) Measurement of released adenosine triphosphate (ATP) (M) and growth (OD600) over time (hours) from parental strain (PS)+pEMPTY and PS+pAPY. n=2 measurements of N=3 independent bacteria cultures. Means and standard deviations are shown. (B) Area under the curve (AUC) of released ATP over time (M*hours) of the previously assessed bacteria (cumulative ATP). t-Test, n=2 measurements of N=3 independent bacteria cultures. Means and individual values are shown. (C) Measurement of ATP (M) in abdominal fluid from mice 8 hr after i.a. injection of bacteria. t-Test, n=5 animals per group of N=2 independent experiments. Means and individual values are shown. (D) Quantitative assessment of colony forming units in abdominal fluid and (E) blood from mice 8 hr after i.a. injection of bacteria. Wilcoxon rank-sum test, n=5 animals per group of N=2 independent experiments. Means and individual values are shown. No growth for controls was detected. (F) Concatenated (n=5 animals for each treatment group, n=3 animals for control group of N=2 independent experiments) and down-sampled images of immune cell populations characterized in the abdominal cavity 8 hr after sham treatment or i.a. injection of bacteria. (G) Abundance of neutrophils, small peritoneal macrophages (SPM), and CX3CR1pos monocytes in abdominal fluid from mice 8 hr after sham treatment or i.a. injection of bacteria. One-way ANOVA followed by Tukey post hoc test, n=5 animals for each treatment group, n=3 animals for control group of N=2 independent experiments. Means and individual values are shown.
Figure 5.
Figure 5.. Outer membrane vesicles (OMV) contain adenosine triphosphate (ATP) and can be exploited as a model to assess the systemic relevance of bacterial ATP.
(A) Illustration depicting the location of assessed proteins that lead to a hypervesiculation phenotype if knocked out in the gramneg bacterium E. coli. (B) Relative amount of OMV compared to the parental strain (PS) isolated from growth cultures of the assessed hypervesiculation mutants after 5 hr (exponential growth phase) and O/N (stationary phase). n=2 measurements of N=3 independent bacteria cultures. Means and individual values are shown. (C) Absolute quantification of ATP in OMV isolated from growth cultures of the PS, ΔnlpI and ΔtolB at their individual peak of ATP release and after 24 hr. n=2 measurements of N=3 independent bacteria cultures. Means and individual values are shown. (D) Amount of protein (BCA assay) detected in different fractions after density gradient ultracentrifugation. n=2 measurements of the different fractions. 20 µl of E. coli growth culture and 20 µl of each fraction were then characterized by Coomassie blue staining and specific detection of outer membrane ompF and cytoplasmic ftsZ. (E) Characterization of OMV by nanoparticle tracking analysis (n=5 measurements per sample) and electron microscopy (representative image) before and after electroporation. (F) Absolute quantification of ATP in OMV, which were loaded using different strategies. Columns 2–5: different concentrations of ATP incubated for 1 hr at 37°C (passive filling). Columns 6–12: different voltages with fixed settings for resistance (100 Ω) and capacitance (50 µF). N=2–9 independent experiments. Means and standard deviations are shown. (G) Relative quantification of ATP in OMV over 24 hr at 37°C after electroporation (0 hr=100%). n=2 measurements of N=3 independent experiments. Means and individual values are shown.
Figure 5—figure supplement 1.
Figure 5—figure supplement 1.. Adenosine triphosphate (ATP) measurement of the parental strain (PS), ΔnlpI as well as ΔtolB and outer membrane vesicle (OMV) collection and characterization.
(A) Measurement of released ATP (M) and growth (OD600) over time (hours) from PS, ΔnlpI and ΔtolB. OMV collection time points are marked in purple. n=2 measurements of N=3 independent bacteria cultures. Means and standard deviations are shown. (B) OMV before and after density gradient ultracentrifugation for 16 hr at 150,000×g. (C) Statistical parameters of OMV before electroporation as well as ATP-loaded and empty OMV after electroporation. (D) Relative quantification of ATP in OMV 16 hr at 4°C after electroporation (0 hr=100%). n=2 measurements of N=3 independent experiments. Means and individual values are shown.
Figure 6.
Figure 6.. Bacterial adenosine triphosphate (ATP) within outer membrane vesicles (OMV) upregulates lysosome-related pathways and degranulation processes in neutrophils.
(A) Experimental approach to determine the systemic role of bacterial ATP in vivo, intraabdominal (i.a.) injecting ATP-loaded or empty OMV. (B) Representative microscopic images of cells from the abdominal cavity 1 hr after i.a. injection of either ATP-loaded or empty OMV. OMV: DiI, Nucleus: DAPI, Neutrophils: Ly-6G-FITC. (C) Cells from remote organs were isolated 1 hr after i.a. injection of either ATP-loaded or empty OMV. OMV were mainly taken up by neutrophils (except in the spleen, ratio ≈ 1). t-Test with Benjamini-Hochberg correction, n=5 animals per group of N=2 independent experiments. Means and individual values are shown. (D) Representative microscopic image of pulmonary neutrophils 1 hr after i.a. injection of either ATP-loaded or empty OMV. OMV co-localize with the endolysosomal compartment. OMV: DiI, Endolysosomal system: LysoTracker Deep Red, Neutrophils: Ly-6G-FITC. (E) Pulmonary neutrophils were isolated 1 hr after i.a. injection of ATPγs-loaded or empty OMV, bead-sorted, and RNA sequencing was done. Principal component analysis shows significantly different clustering between neutrophils that took up ATPγs-loaded (NA) or empty OMV (NE). PERMANOVA, n=6 animals in the NE group, n=5 animals in the NA group. Ellipses represent 95% confidence level. (F) Volcano plot of RNA sequencing results shows an upregulation of genes mainly in the NA group. Genes classified in either lysosome (LYSO) or neutrophil degranulation pathways (NDG) or both, which were mentioned in the text, were highlighted. (G) Heatmap of the lysosome pathway (LYSO) showing the gene expression per sample. (H) Heatmap of the neutrophil degranulation pathway (NDG) showing the gene expression per sample.
Figure 6—figure supplement 1.
Figure 6—figure supplement 1.. Uptake of outer membrane vesicles (OMV) by neutrophils.
Representative images of OMV uptake by neutrophils in the abdominal cavity 1 hr after intraabdominal (i.a.) injection additionally assessed using flow cytometry (ImageStream).
Figure 6—figure supplement 2.
Figure 6—figure supplement 2.. Characterization of local immune response in the abdominal cavity.
(A) Gating strategy to identify large peritoneal macrophages (LPM), small peritoneal macrophages (SPM), and neutrophils in abdominal fluid. (B) Abundance of OMVpos/(OMVpos+OMVneg) LPM, SPM, and neutrophils 1 hr after intraabdominal (i.a.) injection of adenosine triphosphate (ATP)-loaded or empty OMV. t-Test with Benjamini-Hochberg correction, n=5 animals per group of N=2 independent experiments. Means and individual values are shown. (C) Abundance of LPM, SPM, and neutrophils 1 hr after sham treatment or i.a. injection of either ATP-loaded or empty OMV. One-way ANOVA, n=5 animals for each treatment group, n=3 animals for control group of N=2 independent experiments. Means and individual values are shown.
Figure 6—figure supplement 3.
Figure 6—figure supplement 3.. Assessment of outer membrane vesicle (OMV) uptake by immune cells in remote organs.
(A) Gating strategy to identify total OMVpos cells and specifically OMVpos neutrophils in blood and remote organs (lung, liver, kidney, and spleen). (B) Fraction of OMVpos/(OMVpos+OMVneg) neutrophils 1 hr after intraabdominal (i.a.) injection of adenosine triphosphate (ATP)-loaded or empty OMV. t-Test, n=5 animals per group of N=2 independent experiments. Means and individual values are shown. (C) Fraction of % of CD45poslive cells 1 hr after intraabdominal (i.a.) injection of adenosine triphosphate (ATP)-loaded or empty OMV including control without OMV injection. t-Test, n=5 animals per group of N=2 independent experiments. Means and individual values are shown.
Figure 6—figure supplement 4.
Figure 6—figure supplement 4.. Assessment of the purity of bead-sorted pulmonary neutrophils.
Pulmonary neutrophils were isolated 1 hr after i.a. injection of ATPγs-loaded or empty outer membrane vesicle (OMV), bead-sorted and assessed for purity by flow cytometry. A representative image is shown.
Figure 6—figure supplement 5.
Figure 6—figure supplement 5.. List of significantly different pathways after enrichment analysis of RNA sequencing results.
Pulmonary neutrophils were isolated 1 hr after intraabdominal (i.a.) injection of ATPγs-loaded or empty outer membrane vesicle (OMV), bead-sorted, and RNA sequencing was done. This resulted in these significantly different pathways between the groups after enrichment analysis. DESeq, n=6 animals in the NE group, n=5 animals in the NA group.
Author response image 1.
Author response image 1.
Author response image 2.
Author response image 2.
See this image and copyright information in PMC

Update of

  • doi: 10.1101/2024.03.07.583973
  • doi: 10.7554/eLife.96678.1
  • doi: 10.7554/eLife.96678.2

Similar articles

See all similar articles

Cited by

References

    1. Alvarez CL, Corradi G, Lauri N, Marginedas-Freixa I, Leal Denis MF, Enrique N, Mate SM, Milesi V, Ostuni MA, Herlax V, Schwarzbaum PJ. Dynamic regulation of extracellular ATP in Escherichia coli. The Biochemical Journal. 2017;474:1395–1416. doi: 10.1042/BCJ20160879. - DOI - PubMed
    1. Araujo MEG, Liebscher G, Hess MW, Huber LA. Lysosomal size matters. Traffic. 2020;21:60–75. doi: 10.1111/tra.12714. - DOI - PMC - PubMed
    1. Arbizu PM. pairwiseAdonis. cb190f7GitHub. 2023 https://github.com/pmartinezarbizu/pairwiseAdonis
    1. Atarashi K, Nishimura J, Shima T, Umesaki Y, Yamamoto M, Onoue M, Yagita H, Ishii N, Evans R, Honda K, Takeda K. ATP drives lamina propria T(H)17 cell differentiation. Nature. 2008;455:808–812. doi: 10.1038/nature07240. - DOI - PubMed
    1. Baba T, Ara T, Hasegawa M, Takai Y, Okumura Y, Baba M, Datsenko KA, Tomita M, Wanner BL, Mori H. Construction of Escherichia coli K-12 in-frame, single-gene knockout mutants: the Keio collection. Molecular Systems Biology. 2006;2:2006.0008. doi: 10.1038/msb4100050. - DOI - PMC - PubMed

Substances

Associated data