Glucose Triggers ATP Secretion from Bacteria in a Growth-Phase-Dependent Manner

Ippei Hironaka,a,b Tadayuki Iwase,a Shinya Sugimoto,a Ken-ichi Okuda,a Akiko Tajima,a Katsuhiko Yanaga,b Yoshimitsu Mizunoea

Department of Bacteriologya and Department of Surgery,b The Jikei University School of Medicine, Minato-ku, Japan

ATP modulates immune cell functions, and ATP derived from gut commensal bacteria promotes the differentiation of T helper17 (Th17) cells in the intestinal lamina propria. We recently reported that Enterococcus gallinarum, isolated from mice and hu-mans, secretes ATP. We have since found and characterized several ATP-secreting bacteria. Of the tested enterococci, Enterococ-cus mundtii secreted the greatest amount of ATP (>2 �M/108 cells) after overnight culture. Glucose, not amino acids and vita-mins, was essential for ATP secretion from E. mundtii. Analyses of energy-deprived cells demonstrated that glycolysis is themost important pathway for bacterial ATP secretion. Furthermore, exponential-phase E. mundtii and Enterococcus faecalis cellssecrete ATP more efficiently than stationary-phase cells. Other bacteria, including Pseudomonas aeruginosa, Escherichia coli,and Staphylococcus aureus, also secrete ATP in exponential but not stationary phase. These results suggest that various gut bac-teria, including commensals and pathogens, might secrete ATP at any growth phase and modulate immune cell function.

ATP is the source of energy within living cells and acts as anallosteric effector of numerous cell processes, such as active

transport, nucleic acid synthesis, muscle activity, and movement(1–5).

In mammalian cells, ATP is the neurotransmitter for puriner-gic signal transmission (6). ATP is stored in secretory vesicles andexocytosed, leading to various purinergic responses, such as cen-tral control of autonomic functions, pain and mechanosensorytransduction, neural-glial interactions, control of vessel tone andangiogenesis, and platelet aggregation through purinoceptors,thus establishing its role as a chemical transmitter (7–11).

ATP modulates immune cell function by activating the ATPreceptors P2X and P2Y (11–13). Furthermore, ATP has beenfound in the intestinal lumen, where it promotes differentiation ofT helper 17 (Th17) cells in the lamina propria (14). Germ-free(GF) mice exhibited much lower concentrations of fecal ATP,accompanied by fewer lamina propria Th17 cells, than specific-pathogen-free (SPF) mice. In addition, antibiotic-treated SPFmice showed marked reductions in the number of Th17 cells andin the concentration of fecal ATP. Systemic or rectal administra-tion of ATP markedly increases the number of lamina propriaTh17 cells in GF mice. Thus, ATP driven from commensal bacte-ria activates the lamina propria, leading to local differentiation ofTh17 cells. However, ATP-secreting bacteria have not been iden-tified in the intestinal lumen of SPF mice.

Recently, we reported that Enterococcus gallinarum, a vanco-mycin-resistant Gram-positive coccus isolated from SPF mice, se-cretes ATP; other tested bacteria, such as Enterococcus faecalis,Enterococcus faecium, Escherichia coli, and Staphylococcus aureus,did not (15). To our knowledge, this is the first report of theisolation and identification of ATP-secreting bacteria from thefeces of SPF mice. However, whether other enterococcal speciessecrete ATP and what culture conditions are optimal for ATPsecretion remain unclear.

In this study, we found that several enterococcal strains otherthan E. gallinarum secrete ATP; Enterococcus mundtii secretes thegreatest amount of ATP. We therefore used this strain to discoverthat glucose is an essential nutrient and that glycolysis is impor-tant for ATP secretion under the tested conditions. Furthermore,

biochemical analyses using energy-deprived cells revealed that ex-ponential-phase cells secreted more efficiently than stationary-phase cells. We also showed that several other bacterial speciessecrete ATP at exponential phase even though ATP secretion wasnot detected after overnight (O/N) culture.

MATERIALS AND METHODSBacterial strains and culture conditions. Bacterial strains used in thisstudy are described in Table 1. All strains were grown in brain heart infu-sion (BHI; BD, NJ) medium at 37°C O/N and then stored at �80°C in20% (wt/vol) glycerol. RPMI 1640 medium (Gibco, CA), nutrient broth(NB; BD, NJ), Luria-Bertani (LB) broth (LB; Merck, Germany), trypticsoy broth (TSB; BD, NJ), heart infusion (HI) broth (HI; BD, NJ), and BHImedium were used in this study.

Preparation of culture supernatant. All strains were precultured in 5ml of BHI medium for 16 h at 37°C under aerobic conditions with shak-ing. After cells reached stationary phase, the cultures were centrifuged at4,000 � g for 10 min, and the cell pellets were resuspended in RPMI 1640medium. The cell suspensions were then inoculated into RPMI 1640 me-dium to an optical density at 660 nm (OD660) of 0.1 and cultured for 16 hat 37°C under aerobic conditions with shaking. Growth was monitored bymeasuring the OD660 with a Taitec MiniPhoto 518R spectrophotometer.To investigate the effect of oxygen on ATP secretion, E. mundtii NBRC100490T was cultured anaerobically in an anaerobic jar (Mitsubishi GasChemical Company, Inc., Japan) at 37°C for 16 h with shaking. If re-quired, NB, LB broth, TSB, HI broth, BHI medium, and modified RPMI1640 medium (with no amino acids, vitamins, or glucose; pH 7.4) wereused instead of RPMI 1640 medium. To determine the effect of glucose onbacterial ATP secretion, 0.2% (wt/vol) or 1% (wt/vol) glucose was addedto NB, LB, TSB, HI, and BHI media. Cultures were centrifuged at 4,000 �g for 10 min at 4°C, and the supernatants were filtered using a 0.2-�m-pore-size membrane (Kanto Chemical Co., Inc., Japan) to completely

Received 14 December 2012 Accepted 23 January 2013

Published ahead of print 25 January 2013

Address correspondence to Yoshimitsu Mizunoe, mizunoe@jikei.ac.jp.

Copyright © 2013, American Society for Microbiology. All Rights Reserved.

doi:10.1128/AEM.03871-12

2328 aem.asm.org Applied and Environmental Microbiology p. 2328–2335 April 2013 Volume 79 Number 7

on June 13, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

remove residual cells. The filtered culture supernatants were used forquantification of extracellular ATP.

Quantification of extracellular and intracellular ATP. The filteredculture supernatant (100 �l) was mixed with an equal volume of BacTiter-

Glo ATP measurement reagent (Promega, Inc., WI). The biolumines-cence response in relative light units was detected (500 ms) with a lumi-nometer (Luminoskan Ascent; Thermo Fisher Scientific KK, Japan). ATPconcentration was determined using standard ATP (Sigma, MO) solu-tions. RPMI 1640 medium was used as the negative control. To measurethe concentration of intracellular ATP, bacterial cultures were directlymixed with BacTiter-Glo ATP measurement reagent. The cell lysis timefor these bacteria was empirically determined to be 5 min. The concentra-tion of intracellular ATP was calculated by subtracting the amount ofextracellular ATP from that of the uncentrifuged bacterial cultures. Bio-luminescence measurements for each sample were obtained in triplicate.Reconstituted BacTiter-Glo reagent has a minimum half-life of over 30min; reagent decay did not limit the detection of ATP in these experi-ments.

Preparation of energy-deprived cells and inhibition of glycolysis. Atmid-exponential (OD660 of 0.6) and stationary phases (cultivation for 16h), E. mundtii NBRC 100490T and E. faecalis CG110 cells grown in BHImedium were harvested and washed twice with phosphate-buffered saline(PBS) at 4°C. To deprive the cells of intracellular ATP, the suspensionswere incubated for 30 min in 0.5 mM dinitrophenol at 37°C and washedthree times with ice-cold PBS (17). After samples were subjected to cen-trifugation at 4,000 � g for 10 min at 4°C, the cell pellets were resuspendedin RPMI 1640 medium without glucose, and the suspensions were incu-bated at 37°C for 2 h in the presence or absence of 1% (wt/vol) glucose.Glycolysis inhibition was performed by adding 10 �M iodoacetic acid(IAA) to the energy-deprived cells of E. mundtii 100490T at 60 min afterthe addition of glucose. At specific time points (see Fig. 4), intracellularand extracellular ATP concentrations were measured as described above.

Bacterial viability assay. Exponential-phase energy-deprived cells ofE. mundtii NBRC 100490T were stained with a Live/Dead BacLight bacte-rial viability kit (Invitrogen, CA) and observed by fluorescence micros-copy (Eclipse E600; Nikon, Japan). Isopropyl alcohol-treated cells wereused as a control for dead cells.

Quantification of metabolites. Exponential-phase energy-deprivedcells of E. mundtii NBRC 100490T were suspended in RPMI 1640 mediumwithout glucose that was supplemented with 1% (wt/vol) glucose andincubated at 37°C for 10 min. After incubation, intracellular and extracel-lular concentrations of ATP, ADP, AMP, and NAD� were measured by

TABLE 1 Bacterial strains used in this study

Species Straina

Source orreference

Enterococcus asini NBRC 100681T NBRCEnterococcus avium NBRC 100477T NBRCEnterococcus camelliae NBRC 101868T NBRCEnterococcus canis NBRC 100695T NBRCEnterococcus cecorum NBRC 100674T NBRCEnterococcus dispar NBRC 100678T NBRCEnterococcus durans NBRC 100479T NBRCEnterococcus faecalis CG110 34Enterococcus faecium NBRC 100485T NBRCEnterococcus gallinarum Egm10 15Enterococcus gilvus NBRC 100696T NBRCEnterococcus hirae NBRC 3181T NBRCEnterococcus malodoratus NBRC 100489T NBRCEnterococcus moraviensis NBRC 100710T NBRCEnterococcus mundtii NBRC 100490T NBRCEnterococcus pallens NBRC 100697T NBRCEnterococcus phoeniculicola NBRC 100711T NBRCEnterococcus pseudoavium NBRC 100491T NBRCEnterococcus raffinosus NBRC 100492T NBRCEnterococcus saccharolyticus NBRC 100493T NBRCEnterococcus sulfureus NBRC 100680T NBRCEnterococcus thailandicus NBRC 101867T NBRCEnterococcus villorum NBRC 100699T NBRCPseudomonas aeruginosa Ishii This studyEscherichia coli MC4100 35Staphylococcus aureus JCM2874 JCMa The superscript “T” indicates the type strain of the species. NBRC, National Instituteof Technology and Evaluation Biological Resource Center, Japan; JCM, JapanCollection of Microorganisms.

FIG 1 ATP-secreting property of enterococcal strains. The indicated enterococcal strains were precultured to stationary phase, harvested, and resuspended inRPMI 1640 medium. The cell suspensions were subsequently inoculated into RPMI 1640 medium to an OD660 of 0.1 and cultured for 16 h at 37°C under aerobicconditions with shaking. Extracellular ATP was measured as described in Materials and Methods. Data are the means of triplicate experiments, and the error barsindicate standard deviations. n.d., not detected.

Growth-Phase-Dependent ATP Secretion by Bacteria

April 2013 Volume 79 Number 7 aem.asm.org 2329

on June 13, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

capillary electrophoresis time of flight mass spectrometry (CE-TOFMS)in Human Metabolome Technologies, Inc., Japan.

Statistical analysis. The data were analyzed with a two-tailed Stu-dent’s t test (Microsoft Excel 2007). A P value of �0.05 was consideredsignificant.

RESULTSATP secretion by enterococcal species. We recently reported thatE. gallinarum, a commensal enterococcal species in mice and hu-mans, secretes ATP, but other tested bacteria did not (15). In thisstudy, we further examined the ATP-secreting property of 22 en-terococcal species available from the National Institute of Tech-nology and Evaluation Biological Resource Center Japan (NBRC)or gifted from J. Nakayama (Kyushu University, Japan) (Table 1).As shown in Fig. 1, we identified seven new ATP-secreting entero-coccal strains: Enterococcus cecorum NBRC 100674T, E. faeciumNBRC 100485T, Enterococcus gilvus NBRC 100696T, E. mundtiiNBRC 100490T, Enterococcus saccharolyticus NBRC 100493T, En-terococcus sulfureus NBRC 100680T, and Enterococcus thailandicusNBRC 101867T. E. mundtii NBRC 100490T secreted the greatestamount of ATP (�2 �M/108 cells), much more than that pro-duced by a previously isolated E. gallinarum (�1 �M/108 cells).This strain was used for the remainder of our analyses.

Effect of culture medium on ATP secretion by E. mundtii. Inour previous study, only RPMI 1640 medium was used to analyze

bacterial ATP secretion (15). To assess the physiological signifi-cance of culture conditions for ATP secretion, we used variousmedia, such as NB, LB broth, TSB, HI medium, and BHI medium,which are generally used for bacterial cultivation, in addition toRPMI 1640 medium. After incubation of E. mundtii NBRC100490T at 37°C for 16 h, concentrations of extracellular ATP weremeasured. Notably, extracellular ATP was found only when thebacterium was cultured in RPMI 1640 medium (Fig. 2A), suggest-ing that the element or elements that exist in RPMI 1640 mediumbut do not exist in the other media are necessary for bacterial ATPsecretion. Since RPMI 1640 medium is a complete synthetic me-dium, all compositions have been defined, and certain nutrientscan easily be eliminated on demand. To determine the most im-portant factor for ATP secretion by E. mundtii NBRC 100490T, weused modified RPMI 1640 medium in which all components ofcommercial RPMI 1640 medium were mixed manually, andamino acids, vitamins, or glucose were omitted as needed. Thebacterial growth was impaired in all deletion groups; thereforeATP concentrations were expressed per 108 cells. Remarkably,the bacterium did not secrete ATP in the absence of glucose(Fig. 2B). Deletion of amino acids and vitamins yielded negli-gible and moderate effects, respectively. These results stronglysuggest that glucose in RPMI 1640 medium is critical for secre-tion of ATP. The glucose concentration in RPMI 1640 medium

FIG 2 Effect of culture medium on ATP secretion of E. mundtii. (A) E. mundtii NBRC 100490T was cultured under aerobic conditions with shaking for 16 h at37°C in the indicated media. (B) E. mundtii NBRC 100490T was cultured under aerobic conditions with shaking for 16 h at 37°C in modified RPMI 1640 mediumin which amino acids, vitamins, or glucose was omitted from the RPMI 1640 medium. (C) E. mundtii NBRC 100490T was cultured under aerobic conditions withshaking for 16 h at 37°C in the indicated media supplemented with 0.2% (wt/vol) and 1% (wt/vol) glucose. Extracellular ATP was measured as described in thelegend of Fig. 1. Data are the means of triplicate experiments, and error bars indicate standard deviations. n.d., not detected; n.s., not significant; �, P � 0.01; ��,P � 0.05.

Hironaka et al.

2330 aem.asm.org Applied and Environmental Microbiology

on June 13, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

is 0.2% (wt/vol) and is below the level of detection in the othertested media.

To investigate the effect of glucose on ATP secretion by E.mundtii NBRC 100490T more precisely, glucose was added to theindicated culture medium, and the amount of extracellular ATPsecreted by the bacterium was measured. Although effective glu-cose concentrations varied by medium, E. mundtii NBRC 100490T

secreted ATP in all tested culture media supplemented with 0.2%(wt/vol) (basal level in RPMI 1640 medium) and 1% (wt/vol)glucose (Fig. 2C). These results clearly indicate that glucose pro-motes bacterial ATP secretion.

Effect of oxygen on bacterial ATP secretion. Oxygen is one ofthe major factors influencing bacterial growth and ATP synthesis.To examine oxygen requirements for ATP secretion, E. mundtiiNBRC 100490T was cultured under aerobic and anaerobic condi-tions. As shown in Fig. 3, no significant difference in the amountsof extracellular ATP secreted by this strain was observed underthese conditions. In addition, culture ODs under anaerobic con-ditions were similar to those of cultures grown under aerobic con-ditions. These results indicate that oxygen is dispensable for ATPsecretion and growth of this strain. Although ATP is generallysynthesized by glycolysis, the tricarboxylic acid cycle (TCA) cycle,and electron transfer system, no lactic acid bacteria (LAB) havebeen reported to carry genes required for a complete TCA cycle(18). Furthermore, LAB are nonrespiring, fermenting, acid-pro-ducing bacteria that cannot perform respiratory metabolism ex-cept under specific conditions (19). This knowledge and our re-sults suggest that glycolysis is crucial for ATP secretion.

Glucose-dependent ATP secretion by energy-deprived cells.The use of growing cells is sometimes not effective for analyzingmetabolic pathways and biochemical reactions in vitro. In thesecases, resting cells and in vitro reconstituted model systems areuseful. In this study, we used energy-deprived cells prepared bytreatment with dinitrophenol, which disrupts membrane poten-tial and inhibits ATP synthesis driven by F1Fo ATPase (20). After

cells are incubated with dinitrophenol for several minutes, intra-cellular ATP is completely exhausted and converted to ADP,AMP, and adenosine (20).

By adding glucose to energy-deprived E. mundtii NBRC100490T and E. faecalis CG110 cells, time-dependent changes inintracellular and extracellular ATP were determined. After theaddition of glucose to energy-deprived E. mundtii NBRC 100490T

cells prepared at mid-exponential phase, a two-step elevation ofintracellular ATP was observed. The first increase occurred imme-diately and was maintained from 5 min to 15 min; a second in-crease occurred, and then the amount of ATP decreased graduallyafter reaching the maximum level (Fig. 4A). In contrast, extracel-lular ATP increased gradually from 5 min to the end of the exper-iment (120 min) (Fig. 4C).

When IAA, a glycolysis inhibitor, was added to the energy-deprived cells at 60 min, intracellular ATP decreased immediately,and extracellular ATP became constant (Fig. 4A and C). Theseresults indicate that ATP synthesis was completely arrested by IAA

FIG 3 Oxygen requirement for bacterial ATP secretion. E. mundtii NBRC100490T was cultured under aerobic or anaerobic conditions with shaking for16 h at 37°C in RPMI 1640 medium supplemented with 0.2% (wt/vol) or 1%(wt/vol) glucose. Extracellular ATP was measured as described in the legend ofFig. 1. Data are the means of triplicate experiments, and error bars indicatestandard deviations. n.d., not detected; n.s., not significant.

FIG 4 Time-dependent change of intracellular and extracellular ATP concen-trations in energy-deprived enterococcal cells. Energy-deprived E. mundtiiNBRC 100490T (�) and E. faecalis CG110 (Œ) cells at mid-exponential phase(OD660 of 0.6) (A, C, E, and G) and stationary phase (cultivation for 16 h at37°C) (B, D, F, and H) were prepared as described in Materials and Methods.Intracellular (A, B, E, and F) and extracellular (C, D, G, and H) ATP wasmeasured in the presence of 1% (wt/vol) glucose. The arrows represent thetime of addition of IAA. �, ATP concentrations in the presence of IAA. Dataare the means of triplicate experiments, and error bars indicate standard devi-ations.

Growth-Phase-Dependent ATP Secretion by Bacteria

April 2013 Volume 79 Number 7 aem.asm.org 2331

on June 13, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

inhibition of glycolysis and suggest that ATP produced by glycol-ysis from glucose might be secreted into the extracellular milieu.We also used energy-deprived cells prepared at stationary phaseand investigated intracellular and extracellular ATP levels. Eleva-tion of intracellular ATP in stationary-phase cells was observedimmediately after the addition of glucose, and, subsequently, thelevel was higher than that in exponential-phase cells (Fig. 4B);however, stationary-phase cells secreted less ATP than exponen-tial-phase cells (Fig. 4D). Addition of IAA to energy-deprived cellsprepared at stationary phase showed that intracellular ATP de-creased more rapidly than it did in exponential-phase cells (Fig.4B), and extracellular ATP decreased gradually (Fig. 4D). Intra-cellular and extracellular ATP did not increase without glucosesupplementation in exponential- and stationary-phase cells (datanot shown), indicating again that glucose is essential for ATP syn-thesis and secretion.

E. faecalis CG110, an ATP-secretion-negative strain, was usedin the first trial using O/N cultures (Fig. 1); intracellular ATPexhibited a two-step increase (Fig. 4E and F), and exponential-phase cells secreted much more ATP than stationary-phase cells,as is the case in E. mundtii NBRC 100490T (Fig. 4G and H).

In these experiments, we determined bacterial CFU counts at 0h and 2 h after the addition of glucose and found that the bacterialCFU counts remained constant (data not shown). We also ana-lyzed the viability of exponential-phase energy-deprived E.mundtii NBRC 100490T cells by Live/Dead staining and observedvery few dead cells (Fig. 5). These results suggest that cell integritywas maintained and that extracellular ATP is not due to bacteri-olysis. In addition, to confirm whether there exist any metabolitesother than ATP, such as ADP, AMP, and NAD�, CE-TOFMS wasperformed (Fig. 6). No other metabolites were detected in the

extracellular milieu although intracellular concentrations of ADPand AMP were approximately two-thirds the ATP concentration,and NAD� was approximately 1.5 times the ATP concentration.These results might eliminate the possibility that extracellularATP production was simply due to bacteriolysis.

Bacteria generally secrete ATP during growth. Bacterial ATPsecretion has been examined using O/N culture supernatants, butwe showed that exponential-phase cells secrete much more ATPthan stationary-phase cells by using energy-deprived cells, as de-scribed above. We assumed that enterococcal species and bacteriabelonging to other genera might secrete ATP during growth. Here,we monitored bacterial growth and extracellular ATP concentra-

FIG 5 Fluorescence microscopy images of energy-deprived E. mundtii NBRC 100490T cells stained for cell viability. Exponential-phase energy-deprived cells ofE. mundtii NBRC 100490T were untreated (A and B) or treated with 70% isopropyl alcohol (C and D). Bacterial cells were stained with a Live/Dead BacLightbacterial viability kit. Live cells and dead cells are stained in green and red, respectively. Scale bar, 5 �m.

FIG 6 Metabolite concentrations of E. mundtii NBRC 100490T. Exponential-phase energy-deprived cells of E. mundtii NBRC 100490T were suspended inRPMI 1640 medium supplemented with 1% (wt/vol) glucose and incubated at37°C for 10 min. After incubation, intracellular (A) and extracellular (B) con-centrations of the indicated metabolites were measured by CE-TOFMS. n.d.,not detected (�0.01 �M).

Hironaka et al.

2332 aem.asm.org Applied and Environmental Microbiology

on June 13, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

tions of E. mundtii NBRC 100490T, E. faecalis CG110, E. faeciumNBRC 100485T, Pseudomonas aeruginosa Ishii, E. coli MC4100,and S. aureus JCM2874 (Fig. 7). As expected, E. mundtii NBRC100490T cells secrete large amounts of ATP, from exponentialphase to early stationary phase. From late exponential phase toearly stationary phase, surprisingly, E. faecium NBRC 100485T

secreted much more ATP than did E. mundtii NBRC 100490T

although at late stationary phase, the extracellular ATP concentra-tion in E. mundtii NBRC 100490T was higher. In the case of E.faecalis CG110, a low but significant amount of extracellular ATPwas detected at exponential phase although it was not detected inO/N cultures. P. aeruginosa, E. coli, and S. aureus also secreted ATPonly at exponential phase. In addition, we measured extracellularATP in each bacterial strain cultured in RPMI 1640 medium with-out glucose and detected no extracellular ATP at all tested timepoints (Table 2). These results suggest that a variety of bacteriahave the potential to secrete ATP in a growth-phase-dependentmanner in the presence of glucose.

DISCUSSION

ATP derived from commensal bacteria drives differentiation ofintestinal Th17 cells, and administration of ATP exacerbates a

T-cell-mediated colitis model with enhanced Th17 cells (14).Thus, clarifying bacterial ATP secretion mechanisms will provideimportant insight into the cause of chronic ulcerative colitis. Wereported that E. gallinarum, isolated from SPF mice, secretes ATP(15). In this study, we found seven new enterococcal species thatsecrete ATP (e.g., E. mundtii secretes a large amount of ATP) (Fig.1). E. mundtii is a Gram-positive coccus, lactic acid bacteriumisolated from cow teats, the hands of milkers, plants, and soil (21).The species has rarely been isolated from human sources, andthere is little evidence of virulence in humans (22, 23). However,the other ATP-secreting enterococcus, E. faecium, is important fornosocomial infections (24, 25). E. gilvus was isolated from the bileof a patient suffering from cholecystitis (pathogenic role is un-clear) (26), and E. cecorum is associated with endocarditis (27).Therefore, ATP-secreting enterococci might be considered clini-cally problematic pathogens.

Furthermore, we found that glucose is an essential nutrient forbacterial ATP secretion (Fig. 2). Luminal glucose concentration inthe gut depends on the time of day and the segment of the gut;concentrations in the small intestine ranged from 0.2 to 48 mM inrats fed regular chow (28). Thus, our tested glucose concentra-

FIG 7 Extracellular ATP concentrations during bacterial growth. Growth (�) of and extracellular ATP concentrations (Œ) in E. mundtii NBRC 100490T, E.faecalis CG110, E. faecium NBRC 100485T, P. aeruginosa Ishii, E. coli MC4100, and S. aureus JCM2874 are shown. Data are the means of triplicate experiments,and error bars indicate standard deviations. (Insets) Enlarged view of the period 0 to 10 h.

TABLE 2 Extracellular ATP concentrations in bacterial culture supernatants at various time points

Organism

ATP concn (�M/108 cells) by treatment and time point (h)a

� Glucose � Glucose

0 4 8 0 4 8

E. mundtii 0.05 � 0.079 0.25 � 0.135 0.55 � 0.046 ND ND NDE. faecalis 0.01 � 0.018 0.079 � 0.137 ND ND ND NDE. faecium 0.051 � 0.015 0.524 � 0.007 2.576 � 0.065 ND ND NDP. aeruginosa 0.053 � 0.055 0.051 � 0.027 ND ND ND NDE. coli 0.255 � 0.013 0.249 � 0.003 0.119 � 0.007 ND ND NDS. aureus ND 0.970 � 0.285 0.273 � 0.222 ND ND NDa Cells were grown in RPMI 1640 medium with (�) or without (�) glucose supplementation. The data are presented as the means � standard deviations of three experiments. ND,not detected.

Growth-Phase-Dependent ATP Secretion by Bacteria

April 2013 Volume 79 Number 7 aem.asm.org 2333

on June 13, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

tions (0, 0.2, and 1% [wt/vol]) were consistent with normal phys-iological conditions in the gut. These findings might suggest thatluminal glucose is sufficient for ATP secretion by gut microbes. Inaddition, E. mundtii NBRC 100490T secreted ATP under aerobicand anaerobic conditions in the presence of glucose (Fig. 3), andinhibition of glycolysis suppressed ATP secretion (Fig. 4). These find-ings indicate that glycolysis is crucial for bacterial ATP secretion.

Kinetic analyses of ATP secretion in energy-deprived entero-cocci demonstrated that intracellular ATP immediately increasedafter the addition of glucose, and subsequently extracellular ATPincreased (Fig. 4). Furthermore, exponential-phase cells secretemore efficiently than stationary-phase cells (Fig. 4). There are twoexplanations for this phenomenon. First, the stringent response isassociated with the reduction of ATP secretion in stationaryphase. During the stringent response, cells attenuate rRNA syn-thesis, induce and repress metabolic pathways in accordance withtheir physiological needs, and induce many stationary-phase sur-vival genes (29). Thus, stationary-phase cells may use ATP tomaintain cellular functions and prepare for regrowth, so they se-crete less ATP than exponential-phase cells. Second, some ATPexport machinery may be expressed less at stationary than at ex-ponential phase.

In addition, we confirmed cell integrity and checked metabo-lite leakage of exponential-phase energy-deprived E. mundtii100490T cells (Fig. 5 and 6). These results indicate that extracellu-lar ATP is not simply due to bacteriolysis and that bacteria secreteATP spontaneously.

Our finding that exponential-phase cells secrete much higherconcentrations of ATP motivated us to reexamine whether bacte-ria other than enterococci secrete ATP during growth. Interest-ingly, all tested bacteria, including P. aeruginosa, E. coli, and S.aureus that are regarded as commensal bacteria or infectious mi-crobes, secreted ATP at exponential phase even though extracel-lular ATP was not detected in O/N culture supernatants (Fig. 7).Therefore, various bacterial species may secrete ATP at anygrowth stage. Although ATP secretion varied, nanomolaramounts of ATP near the mammalian cell surface are sufficient toelicit functional changes such as neutrophil chemotaxis and air-way epithelial cell volume regulation (16, 30–32). Accordingly, allbacteria tested in this study may have the ability to modulatemammalian cellular functions. Our results support the hypothesisthat commensal bacterium-derived ATP activates specific den-dritic cells in the lamina propria to induce inflammatory cyto-kines, leading to local differentiation of Th17 cells, thereby exac-erbating colitis (14).

Several mechanisms, such as ATP binding cassette transport-ers, connexin hemichannels, stretch-activated channels, cargo-vesicle trafficking, and exocytotic granule secretion (4, 33), areprobably involved in ATP secretion from bacteria. Our resultshave potentially important implications for clarifying the rolesand the mechanisms of bacterial ATP secretion. The entire mech-anism of bacterial ATP-secreting systems is an interesting topic forfuture studies, which will provide further information about ATP-mediated bacteria-bacteria communication and bacteria-host in-teractions.

ACKNOWLEDGMENTS

This work was supported by grants-in-aid for scientific research (C)23590521 from the Ministry of Education, Science, Culture and Sports ofJapan and The Jikei University Graduate Research Fund.

We thank Jiro Nakayama for the gift of E. faecalis CG110 and SatomiYamada for technical support.

REFERENCES1. Brown GC. 1992. Control of respiration and ATP synthesis in mamma-

lian mitochondria and cells. Biochem. J. 284:1–13.2. Burnstock G, Campbell G, Satchell D, Smythe A. 1970. Evidence that

adenosine triphosphate or a related nucleotide is the transmitter sub-stance released by non-adrenergic inhibitory nerves in the gut. Br. J. Phar-macol. 40:668 – 688.

3. Jouaville LS, Pinton P, Bastianutto C, Rutter GA, Rizzuto R. 1999.Regulation of mitochondrial ATP synthesis by calcium: evidence for along-term metabolic priming. Proc. Natl. Acad. Sci. U. S. A. 96:13807–13812.

4. Lundin A, Thore A. 1975. Comparison of methods for extract of bacterialadenine nucleotides determined by firefly assay. Appl. Microbiol. 30:713–721.

5. Perriman R, Barta I, Voeltz GK, Abelson J, Ares M, Jr. 2003. ATPrequirement for Prp5p function is determined by Cus2p and the structureof U2 small nuclear RNA. Proc. Natl. Acad. Sci. U. S. A. 100:13857–13862.

6. Burnstock G. 2006. Historical review: ATP as a neurotransmitter. TrendsPharmacol. Sci. 27:166 –176.

7. Burnstock G. 2007. Physiology and pathophysiology of purinergic neu-rotransmission. Physiol. Rev. 87:659 –797.

8. Johnson RG, Jr. 1988. Accumulation of biological amines into chromaf-fin granules: a model for hormone and neurotransmitter transport.Physiol. Rev. 68:232–307.

9. Pankratov Y, Lalo U, Verkhratsky A, North RA. 2006. Vesicular releaseof ATP at central synapses. Pflugers Arch. 452:589 –597.

10. Sawada K, Echigo N, Juge N, Miyaji T, Otsuka M, Omote H, YamamotoA, Moriyama Y. 2008. Identification of a vesicular nucleotide transporter.Proc. Natl. Acad. Sci. U. S. A. 105:5683–5686.

11. Schnurr M, Toy T, Shin A, Waqner M, Cebon J, Maraskovsky E. 2005.Extracellular nucleotide signaling by P2 receptors inhibits IL-12 and en-hances IL-23 expression in human dendritic cells: a novel role for thecAMP pathway. Blood 105:1582–1589.

12. Khakh BS, North RA. 2006. P2X receptors as cell-surface ATP sensors inhealth and disease. Nature 442:527–532.

13. North RA. 2002. Molecular physiology of P2X receptors. Physiol. Rev.82:1013–1067.

14. Atarashi K, Nishimura J, Shima T, Umesaki Y, Yamamoto M, OnoueM, Yagita H, Ishii N, Evans R, Honda K, Takeda K. 2008. ATP driveslamina propria Th17 cell differentiation. Nature 455:808 – 812.

15. Iwase T, Shinji H, Tajima A, Sato F, Tamura T, Iwamoto T, Yoneda M,Mizunoe Y. 2010. Isolation and identification of ATP-secreting bacteriafrom mice and humans. J. Clin. Microbiol. 48:1949 –1951.

16. Okada SF, Nicholas RA, Kreda SM, Lazarowski ER, Boucher RC. 2006.Physiological regulation of ATP release at the apical surface of humanairway epithelia. J. Biol. Chem. 281:22992–23002.

17. Okuda K, Aso Y, Nakayama J, Sonomoto K. 2008. Cooperative transportbetween NukFEG and NukH in immunity against the lantibiotic nukacinISK-1 produced by Staphylococcus warneri ISK-1. J. Bacteriol. 190:356 –362.

18. Hoskins J, Alborn WE, Jr, Arnold J, Blaszczak LC, Burgett S, DeHoffBS, Estrem ST, Fritz L, Fu DJ, Fuller W, Geringer C, Gilmour R, GlassJS, Khoja H, Kraft AR, Lagace RE, LeBlanc DJ, Lee LN, Lefkowitz EJ, LuJ, Matsushima P, McAhren SM, McHenney M, McLeaster K, MundyCW, Nicas TI, Norris FH, O’Gara M, Peery RB, Robertson GT, RockeyP, Sun PM, Winkler ME, Yang Y, Young-Bellido M, Zhao G, Zook CA,Baltz RH, Jaskunas SR, Rosteck PR, Jr, Skatrud PL, Glass JI. 2001.Genome of the bacterium Streptococcus pneumoniae strain R6. J. Bacteriol.183:5709 –5717.

19. Yamamoto Y, Poyart C, Trieu-Cuot P, Lamberet G, Gruss A, Gaudu P.2005. Respiration metabolism of group B Streptococcus is activated byenvironmental haem and quinone and contributes to virulence. Mol. Mi-crobiol. 56:525–534.

20. Chen F, Cushion MT. 1994. Use of an ATP bioluminescent assay toevaluate viability of Pneumocystis carinii from rats. J. Clin. Microbiol. 32:2791–2800.

21. Collins MD, Farrow JAE, Jones D. 1986. Enterococcus mundtii sp. nov.Int. J. Syst. Bacteriol. 36:8 –12.

22. Higashide T, Takahashi M, Kobayashi A, Ohkubo S, Sakurai M, Shirao

Hironaka et al.

2334 aem.asm.org Applied and Environmental Microbiology

on June 13, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

Y, Tamura T, Sugiyama K. 2005. Endophthalmitis caused by Enterococcusmundtii. J. Clin. Microbiol. 43:1475–1476.

23. Kaufhold A, Ferrieri P. 1991. Isolation of Enterococcus mundtii fromnormally sterile body sites in two patients. J. Clin. Microbiol. 29:1075–1077.

24. Murray BE. 1990. The life and times of the enterococcus. Clin. Microbiol.Rev. 3:46 – 63.

25. Willems RJ, Hanage WP, Bessen DE, Feil EJ. 2011. Population biologyof Gram-positive pathogens: high-risk clones for dissemination of antibi-otic resistance. FEMS Microbiol. Rev. 35:872–900.

26. Tyrrell GJ, Turnbull L, Teixeira LM, Lefebvre J, Carvalho MDGS,Facklam RR, Lovgren M. 2002. Enterococcus gilvus sp. nov. and Entero-coccus pallens sp. nov. isolated from human clinical specimens. J. Clin.Microbiol. 40:1140 –1145.

27. Ahmed FZ, Baig MW, Gascoyne-Binzi D, Sandoe JA. 2011. Enterococcuscecorum aortic valve endocarditis. Diagn. Microbiol. Infect. Dis. 70:525–527.

28. Ferraris RP, Yasharpour S, Lloyd KC, Mirzayan R, Diamond JM. 1990.Luminal glucose concentrations in the gut under normal conditions. Am.J. Physiol. 259:G822–G837.

29. Chang DE, Smalley DJ, Conway T. 2002. Gene expression profiling ofEscherichia coli growth transitions: an expanded stringent response model.Mol. Microbiol. 45:289 –306.

30. Chen Y, Corriden R, Inoue Y, Yip L, Hashiguchi N, Zinkernagel A,Nizet V, Insel PA, Junger WG. 2006. ATP release guides neutrophilchemotaxis via P2Y2 and A3 receptors. Science 314:1792–1795.

31. Corriden R, Insel PA. 2010. Basal release of ATP: an autocrine-paracrinemechanism for cell regulation. Sci. Signal. 104:re1. doi:10.1126/scisignal.3104re1.

32. Praetorius HA, Leipziger J. 2009. ATP release from non-excitable cells.Purinergic Signal. 5:433– 446.

33. Lazarowski ER, Boucher RC, Harden TK. 2003. Mechanism of release ofnucleotides and integration of their action as P2X- and P2Y-receptor ac-tivating molecules. Mol. Pharmacol. 64:785–795.

34. Gawron-Burke C, Clewell DB. 1982. A transposon in Streptococcus faeca-lis with fertility properties. Nature 300:281–284.

35. Casadaban MJ. 1976. Transposition and fusion of the lac genes to selectedpromoters in Escherichia coli using bacteriophage lambda and Mu. J. Mol.Biol. 104:541–555.

Growth-Phase-Dependent ATP Secretion by Bacteria

April 2013 Volume 79 Number 7 aem.asm.org 2335

on June 13, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from