BacFlash signals acid-resistance gene expression in bacteria

Wu, Di; Qi, Wenfeng; Nie, Wei; Lu, Zhengyuan; Ye, Yongxin; Li, Jinghang; Sun, Tao; Zhu, Yufei; Cheng, Heping; Wang, Xianhua

doi:10.1038/s41422-020-00431-3

Article
Published: 06 November 2020

BacFlash signals acid-resistance gene expression in bacteria

Di Wu¹^na1,
Wenfeng Qi¹^na1,
Wei Nie^1,2^na1,
Zhengyuan Lu ORCID: orcid.org/0000-0002-4262-1193¹,
Yongxin Ye¹,
Jinghang Li¹,
Tao Sun¹,
Yufei Zhu^1,2,
Heping Cheng ORCID: orcid.org/0000-0002-9604-6702^1,2 &
…
Xianhua Wang^1,2

Cell Research volume 31, pages 703–712 (2021)Cite this article

1751 Accesses
4 Citations
1 Altmetric
Metrics details

Subjects

A Correction to this article was published on 11 December 2023

This article has been updated

Abstract

Intracellular pH (pH_i) homeostasis is crucial for cellular functions and signal transduction across all kingdoms of life. In particular, bacterial pH_i homeostasis is important for physiology, ecology, and pathogenesis. Here we report an exquisite bacterial acid-resistance (AR) mechanism in which proton leak elicits a pre-emptive AR response. A single bacterial cell undergoes quantal electrochemical excitation, termed “BacFlash”, which consists of membrane depolarization, transient pH_i rise, and bursting production of reactive oxygen species. BacFlash ignition is dictated by acid stress in the form of proton leak across the plasma membrane and the rate of BacFlash occurrence is reversely correlated with the pH_i buffering capacity. Through genome-wide screening, we further identify the ATP synthase F_o complex subunit a as the putative proton sensor for BacFlash biogenesis. Importantly, persistent BacFlash hyperactivity activates transcription of a panel of key AR genes and predisposes the cells to survive imminent extreme acid stress. These findings demonstrate a prototypical coupling between electrochemical excitation and nucleoid gene expression in prokaryotes.

You have full access to this article via your institution.

Download PDF

The nucleotide pGpp acts as a third alarmone in Bacillus, with functions distinct from those of (p)ppGpp

Article Open access 23 October 2020

A meet-up of two second messengers: the c-di-AMP receptor DarB controls (p)ppGpp synthesis in Bacillus subtilis

Article Open access 22 February 2021

Autoinducer-2 and bile salts induce c-di-GMP synthesis to repress the T3SS via a T3SS chaperone

Article Open access 05 November 2022

Introduction

Intracellular pH (pH_i) homeostasis is of paramount importance across all kingdoms of life because most proteins require an optimal pH_i to maintain their structures and functions.¹ In particular, single-cell organisms have developed sophisticated pH_i homeostatic systems to cope with acid or alkaline challenges from fluctuating and unpredictable environments. For instance, Escherichia coli (E. coli), an intestinal bacterium prefers a neutral environmental pH (pH_e) and uses the heavy pH_i buffering provided by weak acids and bases, such as phosphate groups, ammonia, and the side-chains of amino acids, as the first line of defense to protect itself against rapid pH_i swings induced by pH_e changes.^2,3 Sustained pH_e acidification activates programmed acid-resistance (AR) responses, including the up-regulation of transporters catalyzing active proton efflux, such as proton-pumping electron transport chain (ETC) complexes, proton-coupled ATPase, and cation-proton antiporters.^4,5,6,7 Meanwhile, metabolic remodeling gears up to consume excess intracellular protons by enlisting the glutamate-dependent AR system involving the decarboxylases GadA and GadB to decarboxylate glutamate, and the antiporter GadC to export the resultant γ-aminobutyric acid.^8,9,10 In addition, periplasmic chaperon proteins such as HdeA and HdeB contribute to AR by preventing acid-induced protein aggregation and facilitating protein refolding after pH_e neutralization.¹¹

Proton signaling is also highly conserved and crucial in cellular bioenergetics and signal transduction. In prokaryotes, as well as eukaryotic mitochondria, respiration-coupled electron transfer and proton transport develop a proton electrochemical potential, which acts as a source of free energy for ATP synthesis.¹² Fed by the electron donors NADH and FADH₂, electron flow through the ETC in the eukaryotic mitochondrial inner membrane or prokaryotic plasma membrane drives outwardly proton pumping, building up the proton motive force (Δμ_H+), which comprises both the membrane potential (ΔΨ) and the proton gradient components (ΔpH).¹² The re-entry of protons at F₁F_o-ATP synthase harvests the free energy of Δμ_H+ for the synthesis of ATP. The plasma membrane of bacterial cells is not only an active site for proton signaling, but also the first line of defense against acid stress. It has been shown that, under extremely low pH_e, the bacterial cells flip electric potential to slow proton movement into the cell through repulsion.¹³ Whether and how the bacterial plasma membrane can sense acid stress and then relay the signal to express the downstream protective genes remains unclear.

In the present study, we discover a novel form of electrochemical excitation in single E. coli cells, termed “BacFlash”, which composes a transient pH_i rise, membrane depolarization, and bursting production of reactive oxygen species (ROS). Importantly, we demonstrate that the BacFlash activity senses acid stress in the form of proton leak with the ATP synthase F_o complex subunit a as the putative proton sensor. Moreover, we show that persistent BacFlash hyperactivity serves as an early warning signal of imminent acid stress and activates the transcription of a panel of key AR genes at neutral pH_e, and thereby protects the cells against subsequent acid stress. These findings illustrate a prototypical coupling between electrochemical excitation and selective gene expression in prokaryotes.

Results

BacFlash: electrochemical excitation in single E. coli cells

To investigate bacterial pH_i homeostasis and signaling, we expressed the genetically-encoded pH indicator ecliptic pHluorin¹⁴ in BW25113 cells, a common E. coli strain (Fig. 1a; Supplementary information, Fig. S1a). Confocal imaging visualized pulsatile increases of fluorescence occurring randomly among cells bathed in M9 minimal medium (Fig. 1a, b; Supplementary information, Fig. S1a and Video S1), indicative of spontaneous pH_i transients in these bacterial cells. Individual pH_i transients exhibited a full duration at half maximum of 8.4 ± 0.3 s and reached a peak of 18% ± 1% increase in fluorescence (Supplementary information, Fig. S1b), the latter corresponding to an alkalization of ~0.49 pH units based on in situ calibration (pKa ~ 7.26) (Supplementary information, Fig. S2). Similar pH_i transients were observed with pHtomato, another genetically-encoded pH indicator with a pKa ~ 7.8¹⁵ (Fig. 1c). Similar events were detected in BW25113 cells bathed in LB medium (Supplementary information, Fig. S3). To further delineate this phenomenon, we stained E. coli cells with tetramethyl rhodamine methyl ester (TMRM) to measure membrane potential, or 7-dichlorodihydrofluorescein diacetate (DCF) to detect ROS. The staining appeared to be heterogeneous among the intact cells (Supplementary information, Fig. S4), due perhaps to the strong efflux activity of the bacteria.^16,17 Nonetheless, we acquired dual indicator measurements in the portion of successfully stained cells, and found that pH_i transients were accompanied by downward deflections of the TMRM signals (Fig. 1c) and by stepwise increases in the DCF signals (Fig. 1c). Interestingly, robust pH_i transients persisted in spheroplasts whose cell walls were removed by lysozyme and EDTA treatment (Supplementary information, Fig. S5), indicating that neither the cell wall nor the periplasm of the bacterium is required for the biogenesis of pH_i transients. Spheroplasts displayed much improved dye staining, with about 80% cells stained by TMRM and 65% cells stained by DCF. In these dye-stained cells, we showed that almost all pH_i transients reported by pHluorin (97%) were accompanied with transient decrease of TMRM, and 67% pH_i transients reported by pHtomato cooccurred with stepwise increase of DCF signals (Supplementary information, Fig. S6). Overall, these results substantiate that pH_i rise, membrane depolarization, and ROS burst all occur concurrently and represent different aspects of a new type of electrochemical excitation in a single bacterial cell. We named it bacterial flash (BacFlash), in analogy to its eukaryotic counterpart called mitochondrial flash (mitoflash)^18,19,20,21 (see Discussion).

**Fig. 1: Spontaneous electrochemical excitation termed “BacFlash” in single bacterial cells.**

The rate of occurrence of BacFlashes was ~5 events/1000 μm²/100 s in BW25113 cells (Fig. 1d). BacFlashes with nearly identical frequency and properties were also detected in three other E. coli strains, BL21(DE3), DH5α, and Top10 (Fig. 1d; Supplementary information, Fig. S1). In colonies growing on agarose pads, BacFlashes were confined to individual cells (Supplementary information, Fig. S7), with no sign of intercellular propagation. During cell division, however, we observed synchronous BacFlashes in both daughter cells until their complete separation (Supplementary information, Fig. S8). In filamentous E. coli cells produced by cephalexin treatment to prohibit division,^22,23 BacFlashes were synchronized along the entire length of 20–100 μm (Supplementary information, Fig. S9 and Video S2). Inhibiting respiration with KCN and dissipation of proton electrochemical gradients with the ionophore carbonyl cyanide 4-(trifluoromethoxy) phenylhydrazone (FCCP) both markedly repressed BacFlash activity (Fig. 1e), indicating a tight link between BacFlash biogenesis and respiration in the bacteria. Altogether, we conclude that BacFlashes represent a fundamental activity that is intertwined with core physiology of E. coli cells.

Buffering pH_i quenches BacFlashes

In search of physiological regulators of BacFlashes, we serendipitously found a striking 40-fold increase in BacFlash frequency when ammonium (20 mM) was removed from the M9 minimal medium (Fig. 2a, b; Supplementary information, Video S3). The BacFlash response to ammonium removal was fully reversible (Fig. 2a) and was common to all four strains examined (Fig. 2b). These hyperactive BacFlashes were sensitive to inhibition of respiration (Supplementary information, Fig. S10b), as was the case in the presence of ammonium (Fig. 1e). In the absence of ammonium, BacFlash amplitude was moderately increased and the duration remained unaltered, while basal pH_i was slightly decreased (with F₀ decreased by ~10% corresponding to ~0.2 units) (Supplementary information, Figs. S2 and S10a). That is, ammonium acts as a powerful inhibitor of the biogenesis of BacFlashes, operating mainly in a frequency-modulatory mode.

**Fig. 2: Transmembrane proton leak triggers BacFlashes.**

Because ammonium participates in bacterial nitrogen metabolism, we tested whether the ammonium inhibition of BacFlashes could be caused by altered nitrogen source availability. We found that replacing ammonium with equimolar glycine or glutamate in M9 minimal medium failed to prevent the hyperactivation of BacFlashes induced by ammonium removal (Supplementary information, Fig. S11a). Further, in the TCE-Glu (PKUW151) strain, which uses glutamate as the preferred nitrogen source,²⁴ glutamate did not suppress the hyperactive BacFlashes either (Fig. 2c). In 13 mutants with ammonia metabolic gene knockout, not only was basal BacFlash activity unaltered in the presence of ammonium, but also their responses to ammonium removal remained intact (Supplementary information, Fig. S11b). These three lines of evidence exclude bacterial nitrogen metabolism as the mechanism underlying the ammonium-induced inhibition of BacFlashes.

Considering that ammonium can enter the cytosol and act as a weak base (pKa ~ 9.33),⁶ we then hypothesized that the inhibition of BacFlashes by ammonium might reflect its pH_i buffering capacity. In this regard, methylammonium (CH₃NH₃⁺), which can also enter the cytosol as a pH_i buffer (pKa ~ 10.73) but does not participate in bacterial nitrogen metabolism,^25,26 showed a concentration-dependent suppression of BacFlashes with an IC₅₀ of 7.91 mM (Fig. 2d), while quantitative analysis of ammonium inhibition of BacFlashes yielded an IC₅₀ of 1.56 mM (Fig. 2d). The higher IC₅₀ value for methylammonium over ammonium is consistent with its relatively lower pH buffering capacity in the physiological pH range (7.4–7.8 in E. coli^6,27). When present at high concentrations, both pH buffers completely abolished BacFlash activity, indicating that pH_i buffering quenches BacFlash biogenesis. Moreover, the inverse correlation between BacFlash frequency and pH_i buffering capacity suggests possible involvement of BacFlashes in signaling AR response.

Free protons trigger BacFlashes

Microscopically, increasing pH_i buffering would shorten the lifespan of a free proton and the distance it traverses before it is re-bound. That pH_i buffers repress BacFlash activity in a concentration-dependent manner suggests that free protons might serve as the predominant trigger for BacFlash biogenesis in E. coli cells. Since basal pH_i decreased only mildly (~0.2 pH units) (Supplementary information, Fig. S10a) even when BacFlash frequency swung over orders of magnitude (Fig. 2a, b), it is also clear that global pH_i is unlikely to be the most critical determinant of BacFlash ignition; instead, free protons that trigger BacFlashes must be highly localized to microdomains, presumably at putative proton-sensing sites. To test the hypothesis that local free protons trigger BacFlashes, we treated E. coli cells with nigericin or monensin, which are antibiotics mediating electroneutral H⁺/K⁺ or H⁺/Na⁺ antiport and can deliver nascent free protons at the inner surface of the plasma membrane without perturbing the membrane potential.²¹ We showed that, starting from a moderate basal frequency set by 2 mM ammonium, nigericin and monensin effectively elicited BacFlashes in a concentration-dependent manner, reaching 6.8- and 5.2-fold increases at 100 μM nigericin and monensin in intact E. coli cells, respectively (Fig. 2e, f). Again, pH_i buffering with 20 mM ammonium profoundly inhibited the BacFlash activity induced by these proton antiporters (Fig. 2f). In the absence of cell walls, these proton antiporters appeared to be much more efficacious, with similar BacFlash-triggering effects achieved at about two orders of magnitude lower concentrations in spheroplasts than intact cells (Supplementary information, Fig. S12). Together with BacFlash-quenching effect of pH_i buffers, these results substantiate the idea that free protons near the inner surface of the plasma membrane act as the predominant trigger of BacFlashes.

The ATP synthase F_o complex subunit a serves as a putative sensor for BacFlash biogenesis

Next, we sought to identify putative proton sensors responsible for BacFlash ignition by undertaking a genome-wide screening of the Keio collection of E. coli knockouts.²⁸ The screening was performed in ammonium-free M9 minimal medium allowing for high BacFlash activity (Fig. 3a), and a computer algorithm modified from FlashSniper²⁹ was used to automatically detect BacFlashes in massive image datasets (see Materials and methods). Two criteria were set for a putative proton sensor-knockout mutant: it would display a dramatically lower BacFlash frequency and would no longer respond to antiporter-introduced proton trigger. Out of 3985 non-essential knockout mutants in the collection, 3281 mutants successfully expressed the indicator pHluorin and were used for three rounds of screening (see Materials and methods). At a cut-off value of 70% decrease in BacFlash frequency, we identified 128 candidates in the first round (Supplementary information, Fig. S13a), confirmed 34 candidates in the second round (Supplementary information, Fig. S13b), and finally pinpointed three candidate genes, atpB, cyaA, and argO (Fig. 3b). We then applied nigericin (100 μM) to further assess the mutants for their responsiveness to artificially-introduced proton trigger. Strikingly, we found that atpB-knockout completely abolished the nigericin-induced BacFlash activity (Fig. 3c). In contrast, nigericin still significantly enhanced the BacFlash activity in the cyaA- and argO-knockout mutants, albeit at smaller magnitudes (Fig. 3c). These results indicate that only atpB satisfies both criteria for a putative proton sensor, whereas cyaA and argO appear to play regulatory, rather than obligatory, roles in BacFlash biogenesis.

**Fig. 3: Identification of putative proton sensor *atpB* for BacFlash ignition.**

AtpB encodes subunit a of the F_o sector of the membrane-bound ATP synthase which interacts with subunit c to form a ring-like structure obligatory to both the ATP synthase and ATPase activities.^30,31 Interestingly, subunit c deficiency had little effect on BacFlash frequency (fold change of 1.07 ± 0.08 compared to the wild-type). Because both subunit a and c are essential for ATP synthase and ATPase activity, these results indicate that the atpB-mediated regulation of BacFlash biogenesis depends on neither ATP synthase/ATPase activity nor ATP synthesis-coupled proton influx.

Hyperactive BacFlashes time-dependently activate AR gene expression and protect the cells against acid stress

During a BacFlash, transient pH_i rise, which is likely through enhancing ETC complex-mediated proton efflux, accounts for an immediate relief of acid stress and the transient membrane depolarization can also slow down the proton leak, suggesting that BacFlashes may constitute a novel AR mechanism. The question arising is whether and how the BacFlash mechanism is integrated into the sophisticated AR system developed in the single-cell organisms. To further appraise possible role of BacFlashes in bacterial AR response, we performed RNA-sequencing analysis under conditions of either hyperactive or hypoactive BacFlash activity. In particular, we investigated whether BacFlashes can regulate the expression of AR genes of different classes. Among 4040 E. coli genes analyzed, only 52 (~1.3%) were up-regulated and 47 (~1.2%) were down-regulated in cells displaying hyperactive BacFlashes compared with their counterparts in which BacFlash activity was inhibited by either 20 mM ammonium or 20 mM methylammonium (Supplementary information, Table S1). Bioinformatic analysis revealed that the genes up-regulated by ≥5 folds in hyperactive BacFlash group were highly enriched in multiple pH response pathways (Supplementary information, Fig. S14). In particular, the AR genes of multiple classes, including those in glutamate metabolism (the structural components gadA, gadB, and gadC and the central regulator gadE), the AR membrane protein (hdeD), and AR chaperones (hdeA and hdeB), were enriched (Fig. 4a). Using the lacZ assay, we further assessed the kinetics of the promoter activation of these induced AR genes.³² At low BacFlash activity with either 20 mM ammonium or methylammonium (Supplementary information, Fig. S15), the promoters of these AR genes all remained quiescent (Fig. 4b). After switching to ammonium-free M9 minimal medium, persistent hyperactivity of BacFlashes time-dependently activated these AR genes with a latency of ~1.5 h and a half-time of activation of 2.4–3.0 h in wild-type cells (Fig. 4b; Supplementary information, Fig. S15). Furthermore, in the atpB-knockout mutant, which lacks the putative proton sensor for BacFlash biogenesis and displays low BacFlash activities (Fig. 3b, c), none of the aforementioned AR genes was activated by switching from ammonium- or methylammonium-containing to ammonium-free M9 minimal medium (Fig. 4b). Functionally, we found that hyperactive BacFlashes induced with ammonium removal conferred the wild-type bacterial cells higher survival rates in response to acid stress (Fig. 4c), however, the atpB-knockout mutant showed high vulnerability to acid stress under all experimental conditions tested (Fig. 4c). It is noteworthy that this newly-identified AR mechanism operated at neutral pH_e (7.4) and is accompanied by only a small change in pH_i (~0.2 units) (Supplementary information, Fig. S10a), in sharp contrast to the AR response induced by an extremely acidic pH_e (2.0–4.0).^8,9,10 These results indicate that the BacFlash-activated AR gene expression serves as an early-warning signal, preconditioning bacterial cells to better survive imminent acid stress. Thus, the newly identified BacFlash-mediated AR mechanism senses an imbalance between proton leak and pH_i buffering, and preemptively activates the expression of a panel of key AR genes prior to overt acidosis.

**Fig. 4: Hyperactive BacFlashes time-dependently activate the expression of a panel of key AR genes.**

Discussion

We have shown that single E. coli cells undergo quantal and stochastic electrochemical excitation in the form of BacFlashes, which can be directly visualized by different indicators for pH_i, membrane potential, or ROS. Multiple processes take place simultaneously in a BacFlash: sudden membrane depolarization likely involves the gating of one or a few membrane channels, transient pH_i rise may reflect accelerated proton efflux by ETC operating at depolarized membrane potential, and ROS burst may arise from massive diversion of electrons from the ETC to the reaction with oxygen. Recently, transient membrane depolarization has been reported in single E. coli cells expressing an optical voltage sensor proteorhodopsin.^17,33 However, this type of membrane depolarization differs from BacFlashes in several important ways. First, in contrast to conscious pH_i rise in a BacFlash, no pH_i changes¹⁷ or a very small pH_i decrease (< 0.1 pH units)³³ was detected among these previous reported events. Second, increasing laser illumination augments the frequency and abbreviates the duration of the proteorhodopsin-reported events,¹⁷ but exerts little effects on the frequency as well as time course of BacFlashes (Supplementary information, Fig. S16). Thus, BacFlashes and the proteorhodopsin-reported events are mechanistically different and BacFlashes represent a previously unappreciated type of prokaryotic cellular dynamics.

On the other hand, BacFlashes bear remarkable resemblances to the eukaryotic mitoflashes that we and others reported previously.^{18,19,20,21,34} Spontaneous mitoflash in respiring mitochondria, also comprises an action potential-like mitochondrial depolarization, a transient rise in matrix pH, and bursting production of ROS. Mitoflashes of similar properties prevail among all eukaryotes examined^18,19,20,21 and even in isolated mitochondria,^35,36,37 indicating that it is a highly conserved and autonomous property of the organelle. Transient opening of mitochondrial permeability transition pore (mPTP), which is known as a megachannel in mitochondrial inner membrane,³⁸ has been reported to be tightly coupled with mitoflash generation, evidenced by precipitous dissipation of mitochondrial membrane potential and irreversible loss of small solutes from the matrix in a mitoflash.^18,19 We hypothesize that similar megachannel may exist in bacterial plasma membrane and its transient and reversible opening might account for transient depolarization of membrane potential in a BacFlash. Although the molecular identity of mPTP is still in debate, it has been proposed that ATP synthase might be one of the mPTP candidates.^38,39 In this regard, the present finding that subunit a of ATP synthase acts as a putative proton sensor for BacFlash generation is in line with an intimate relationship between mPTP and ATP synthase. Since the mitochondrion is thought to have originated from an alpha-proteobacterial ancestor through endosymbiosis,⁴⁰ it is conceivable that electrochemical excitation in the form of BacFlash might represent an evolutionarily ancient physiological activity in prokaryotes that has been essentially conserved in eukaryotes, assuming the form of mitoflash.

It has been shown that mitoflashes fulfill an elemental signaling role in a cell type-specific and context-sensitive manner, regulating mammalian cardiac ATP homeostasis,⁴¹ participating in skeletal muscle metabolism,⁴² and signaling structural long-term potentiation of neuronal synapses.⁴³ In the present study, we find that BacFlash biogenesis confers a new AR mechanism on bacterial cells by acting as an early warning signal. Differing from previously characterized AR mechanisms, the BacFlash-mediated AR appears to specifically respond to a new form of acid stress-sustained proton leak that is unbalanced by intracellular pH buffers. First, an inverse relationship between BacFlash frequency and pH_i buffering capacity was uncovered. BacFlash displays low frequency under normal culture medium containing heavy pH buffers, however, when changing the pH_i buffering capacity by varying the concentration of ammonium or methylammonium with the pH_e unchanged, the BacFlash frequency varied over orders of magnitude and vanished at high pH_i buffer capacity, even though the pH_i changed only mildly (~0.2 units). This testifies a high signal-to-background contrast and high signal-to-noise ratio for BacFlash frequency to decode proton leak, a predictor of imminent acid stress. Second, proton leak induced by the electroneutral H⁺/K⁺ antiporter nigericin or the H⁺/Na⁺ antiporter monensin ignited hyperactive BacFlashes, and this trigger effect can also be abolished by pH_i buffering. More importantly, through genome-wide screening of the Keio collection of E. coli knockouts, we showed that the ATP synthase F_o complex subunit a encoded by atpB meets the two criteria we set for a putative proton sensor for BacFlash generation: it is obligatory for both spontaneous and proton antiporter-induced BacFlashes. It should be noted that the atpB-mediated proton ignition of BacFlashes depends on neither ATP synthase/ATPase activity nor ATP synthesis-coupled proton influx because the F_oF₁-ATP synthase subunit c deficiency had little effect on BacFlash frequency. These results are in general agreement with the finding in mammalian cells that proton leak triggers mitoflashes over a nanometer scale.²¹

It appears that this newly identified AR mechanism is well integrated with the sophisticated bacterial AR system for both short-term and long-term responses. BacFlashes would slow down the proton leak owing to the membrane depolarization and enhance proton efflux as manifested by the pH_i rise, both of which would immediately relieve acid stress. More importantly, BacFlashes relay the acid stress signal to nucleoid gene expression, activating the transcription of multiple classes of AR genes in a coordinated, time-dependent manner. These AR genes include the key components of glutamate metabolism (gadA, gadB, gadC, and gadE), the AR membrane protein (hdeD), and AR chaperones (hdeA and hdeB). The BacFlash-activated expression of these AR genes effectively preconditions the bacterial cells to resist and survive subsequent acid stress. In addition to the immediate and long-term responses, our data also suggest an indirect mode of action for the BacFlash-mediated AR mechanism. In the genome-wide screen, we also identified two other BacFlash regulators that partake the AR response: cyaA and argO. Specifically, cyaA encodes an adenylate cyclase catalyzing the synthesis of the second messenger cyclic AMP, and its deficiency results in pH-sensitive growth impairment in E. coli.⁴⁴ ArgO encodes an arginine transporter mediating arginine export, further supporting the role of BacFlash biogenesis in pH_i homeostasis. Overall, we have demonstrated that the BacFlash AR mechanism operates at neutral pH_e (7.4), in sharp contrast to the AR response induced by an extremely acidic pH_e (2.0–4.0),^8,9,10 and teleologically, serves as an early warning signal to preemptively activate an AR response prior to overt cellular acidosis.

Materials and methods

Reagents

Nigericin, monensin, DCF, FCCP, lysozyme, methylamine hydrochloride (CH₃NH₂·HCl), sodium dodecyl sulfate, Na₂CO₃, CaCl₂, Na₂HPO₄, NaH₂PO₄, KH₂PO₄, KCl, NaCl, MgSO₄, MgCl₂, NH₄Cl, ampicillin, kanamycin, and chloramphenicol were from Sigma-Aldrich (Missouri, USA). 2-Mercaptoethanol was from Merck Millipore (Darmstadt, Germany). Polylysine was from Solarbio (Beijing, China). l-glutamic acid monosodium salt, d-glucose, and sucrose (RNA-free) were from VWR Life Science AMRESCO (Pennsylvania, USA). l-glutamine was from BioDee Biotechnology (Beijing, China). CCCP and 2-nitrophenyl β-d-galactopyranoside were from MedChemExpress (New Jersey, USA). Isopropyl β-d-thiogalactoside (IPTG) was from Thermo Fisher Scientific (Massachusetts, USA).

E. coli strains

The E. coli strains used in this study are listed in Supplementary information, Table S2. BL21(DE3) and BW25113 are in regular use in our laboratory. Top10 and DH5α were bought from TransGen Biotech (Beijing, China). PKUW13 and PKUW151 were gifted by Prof. Yi-Ping Wang of Peking University. Glycerol stocks were maintained at −80 °C, and individual colonies were picked for each experiment. The single gene-knockout mutants were from the Keio collection.²⁸ Each strain carries a non-essential gene replaced with a kanamycin resistance cassette.

Construction and expression of pHluorin and pHtomato

The genes for super-ecliptic pHluorin (GenBank: AF058694.2)¹⁴ and pHtomato (GenBank: JQ966306.1)¹⁵ were cloned into pTricHisA (Thermo Fisher Scientific, Massachusetts, USA) using the BamHI and HindIII restriction sites. The plasmid containing pHluorin or pHtomato was transfected into E. coli cells. A colony of cells from a fresh solid Luria-Bertani (LB) plate was inoculated into LB medium for overnight growth at 37 °C with shaking at 250 rpm for pre-culture. The next day, the pre-culture was diluted 1:100 into fresh LB medium as the experimental culture. Cells were grown to the early-log phase (OD₆₀₀ = 0.3–0.4), 0.2 mM IPTG was added, and growth was allowed to continue further. The cells were harvested after 1 h for further experimentation.

Preparation of filamentous E. coli and spheroplasts

Filamentous E. coli and spheroplasts were prepared from E. coli strain BW25113 as described previously.²² Briefly, BW25113 cells were transformed with pHluorin, and a colony of cells was inoculated into 5 mL LB medium for overnight growth at 37 °C with shaking at 250 rpm, then diluted 1:100 in the same medium and grown to OD₆₀₀ 0.5–0.7. This culture (3 mL) was then diluted 1:10 in LB medium and cephalexin was added to 60 μg/mL. The culture was then shaken for 2–2.5 h at 37 °C until single-cell filaments reached sufficient length (20–150 μm).

These filamentous cells were harvested by centrifugation at 1500 × g for 2 min, and the pellet was rinsed without resuspension by the gentle addition of 1 mL 0.8 M sucrose solution and then allowed to rest at room temperature (22–26 °C) for 1 min. The supernatant was then removed with a Pasteur pipette and discarded, and the pellet was re-suspended in 2.5 mL 0.8 M sucrose. To remove the cell wall, the following reagents were added in order: 150 μL of 1 M Tris-HCl (pH 7.8); 120 μL of lysozyme (5 mg/mL); 30 μL of DNase I (5 mg/mL); and 120 μL of 0.125 M Na-EDTA (pH 8.0). This mixture was incubated at room temperature for 4–8 min to hydrolyze the peptidoglycan layer. At the end of this incubation, 1 mL of a solution containing 20 mM MgCl₂ (to remove the EDTA and activate the DNase), 0.7 M sucrose, and 10 mM Tris-HCl (pH 7.8) was added slowly (over 1 min) while stirring, and the mixture was incubated at room temperature for four more minutes. The mixture containing spheroplasts was either used immediately or kept at −20 °C.

Confocal imaging of BacFlashes

E. coli cells grown in LB medium at 37 °C were harvested by centrifugation. The harvested cells were placed on 1% agarose pads (low melting-point agarose was added to the M9 minimal medium above to a final concentration of 1% weight/volume then heated until the agarose dissolved) or polylysine-coated glass slides (a 0.01% weight/volume solution of polylysine was dropped on the glass slides and dried in the hood) for confocal imaging. For polylysine adhesion, the E. coli cells were left to adhere to these slides put in 6-well plate for 5 min in M9 minimal medium and centrifuged at 4000 rpm for 15 min, followed by washing three times with M9 minimal medium to remove non-adhered cells. All experiments were carried out at room temperature, except for monitoring growth on the agarose pad, which was performed at 37 °C.

An inverted confocal microscope (Zeiss LSM 710) with a 40×, 1.3 NA oil-immersion objective was used for imaging. When imaging pHluorin alone, images were captured by excitation at 488 nm and the emission was collected at > 505 nm. For pHtomato imaging, images were captured by exciting at 543 nm and collecting the emission at > 550 nm. For TMRM measurements, the indicator (100 nM) was loaded at 37 °C for 20 min. Images were captured by exciting at 543 nm and collecting the emission at > 550 nm. To acquire pHluorin and TMRM signals simultaneously, images were acquired by alternating excitation at 488 and 543 nm and emission collection at 500–530 and > 560 nm, respectively. For DCF measurements, the indicator (25 μM) was loaded at 37 °C for 20 min followed by washing three times with M9 minimal medium. For staining of the spheroplasts, the DCF (25 μM) was kept in the bath solution during imaging. The DCF fluorescence imaging was performed by exciting at 488 nm and collecting emissions at > 500 nm, with a low laser intensity to minimize photochemical reactions of DCF. To record DCF and pHtomato simultaneously, images were acquired by alternating excitation at 488 and 543 nm and emission collection at 500–530 nm and > 560 nm, respectively.

In typical time-lapse recordings of BacFlashes, 100 frames of 512 × 512 pixels were collected at 1.58 μs/pixel in bidirectional scanning mode. The frame rate was 1 frame/s, and the axial resolution was set to 1.0 μm. All cells were imaged within 1 h of treatment, and all experiments were performed at room temperature unless specified otherwise.

In situ calibration of pHluorin

In pHluorin-expressing BW25113 cells, 200 μM CCCP was used to break down pH gradients across the cell membrane and was present throughout all experimental conditions. The first recording of pHluorin fluorescence was made in M9 minimal medium at pH_e 8.0. After changing the medium to the designated experimental pH_e, the pHluorin fluorescence was recorded and then normalized to the pH_e 8.0 recording. The averaged normalized fluorescence values at different pH_e values were fitted to a Hill curve to obtain the pKa and the Hill coefficient of pHluorin in situ.

RNA sequencing analysis

BW25113 cells at log phase (OD₆₀₀ = 0.12) in M9 minimal medium were centrifuged and re-suspended in ammonium-free M9 minimal medium supplemented with 20 mM sucrose, 20 mM ammonium, or 20 mM methylammonium. After treatment for 5 h at 37 °C with shaking at 250 rpm, total RNA was extracted using an RNAprep pure Cell/Bacteria kit (Tiangen, Beijing, China). RNA sequencing was then performed on an Illumina Hiseq 2500 platform (San Diego, California, USA), and 101-bp paired-end reads were generated for each treatment regimen.

The BW25113 genome and annotation files were downloaded from GenBank for the following analysis.⁴⁵ To obtain the genes differentially expressed between cells with different levels of BacFlash activity, the reads were first preprocessed to evaluate the data quality and rRNA rate and then aligned to the E. coli genome with Bowtie2.⁴⁶ The reads with high mapping quality (≥ 30) were normalized to calculate gene expression levels using FPKM (Fragments per Kilobase per Million mapped reads). Specific regions were used to represent gene loci to exclude the influence of overlapped regions in the gene expression calculation. Cuffdiff was used to obtain differentially expressed genes with cutoffs of fold-changes > 2 or < 0.5 and P values < 0.05. Two data sets from two independent biological replicates were merged to obtain the final up- or down-regulated genes. The genes up-regulated by ≥ 5 folds were bioinformatically analyzed using KOBAS3.0. The RNA sequencing data are available at https://www.ncbi.nlm.nih.gov/sra/PRJNA656061 and the NCBI accession number is PRJNA656061.

Promoter activity assay

The promoter regions of the target genes, as described previously⁴⁷ (Supplementary information, Table S3), were amplified from MG1655 genomic DNA using High-Fidelity DNA polymerase (NEB, Massachusetts, USA) and inserted into the pMP190 plasmid (gifted by Prof. Yi-Ping Wang, Peking University) using the XbaI and KpnI restriction sites. The inserted sequence was checked by sequencing. The BW25113 cells transformed by promoter-containing plasmid were cultured to the log phase (OD₆₀₀ = 0.4) in M9 minimal medium, centrifuged, and then re-suspended in ammonium-free M9 minimal medium supplemented with 20 mM sucrose, 20 mM ammonium, or 20 mM methylammonium. At different time points (0.1, 1.5, 3, and 5 h), the cells (~5 mL of cell culture) were collected and frozen on dry ice. The promoter activity was assessed as described previously.⁴⁸ Briefly, 200 μL ortho-nitrophenyl-β-galactoside was added to 1 mL cell lysate, and the reaction was stopped within 5 min by adding 0.5 mL 1 M Na₂CO₃. A colorimetric assay at 420 nm was performed using an ultraviolet spectrophotometer (Shimadzu, Kyoto, Japan), and the promoter activity was calculated as OD₄₂₀/min/mL/OD₆₀₀ as follows:

$${{{\mathrm{Activity}}}} = \frac{{{{{\mathrm{OD}}}}_{420}}}{{{{{\mathrm{OD}}}}_{600} \times {{{\mathrm{Time}}}}\left( {{{{\mathrm{min}}}}} \right) \times {{{\mathrm{Volum}}}}\left( {{\upmu {{{\mathrm{L}}}}}} \right)}} \times 1000$$

Acid resistance assay

The acid resistance assay was performed after overnight growth (18 h) of BW25113 cells in the LB medium. Briefly, the overnight culture was diluted 1:10 into M9 minimal medium and cultured for 2 h at 37 °C with shaking. The cells were then centrifuged and re-suspended in ammonium-free M9 minimal medium supplemented with 20 mM sucrose, 20 mM ammonium, or 20 mM methylammonium and cultured for 5 h at 37 °C with shaking. After adjusting cell densities according to their OD₆₀₀ value, the cells were suspended in M9 medium without ammonium, pH 2.0, and incubated for 0, 2, and 4 h, separately, at 37 °C with shaking. Then the cells were diluted 1:5000 in LB medium, plated to LB agar, and incubated at 37 °C for 15 h before colony counting. The survival rates were reported by normalizing the colony numbers after 2 or 4 h acid challenge to those at 0 h in each group.

Genome-wide screening of BacFlash regulator genes

The single gene-knockout mutants (3985 nonessential knockout mutants) used for genome-wide screening were from the Keio collection.²⁸ Briefly, each strain was transformed with pHluorin-containing plasmid and induced to express pHluorin by 0.2 mM IPTG. After 1.5 h of induction, the E. coli cells were centrifuged onto polylysine-coated slides in M9 minimal medium for confocal imaging. For each strain, the images were first captured in 20 mM ammonium-containing M9 minimal medium and then in ammonium-free M9 minimal medium. The image data were obtained in at least 3 regions from each condition. In the first-round screening, mutants showing ≥ 70% decrease of BacFlash frequency were considered as candidates for putative proton sensor genes. These candidates were then examined in a second-round screen using the same protocols and with ≥ 50% decrease of BacFlash frequency. A total of 3281, 128, and 34 mutants were used in the first, second, and third rounds of screening, respectively, and > 15,000 time-lapse image stacks of 512 × 512 × 100 pixels were generated.

Automated BacFlash detector

The automated detector of BacFlashes was modified from our previous Flashsniper for mitoflashes.²⁹ The overall detection scheme was divided into three steps: (1) image pre-processing to de-noise and eliminate motion artifacts, (2) cell edge-extraction by segmentation, and (3) event detection and parametric extraction.

Image pre-processing

We used a 5-kernel Gaussian-blur for image de-noising, and motion artifacts were eliminated through image registration. The Fourier-Mellin transformation⁴⁹ was used to estimate the motion parameters between the reference and target images. Considering that movement within an image stack was often slow and continuous, we fitted the frame-by-frame estimated parameters with a quadratic function over the entire stack. The inter-frame movements were then corrected accordingly, using the fitted parameters.

Cell edge-extraction

A comprehensive thresholding method combining global and local thresholds was used for segmentation. The global threshold was determined by Otsu thresholding.⁵⁰ The local threshold for each pixel was the average value of its local neighbors within a defined window centered at that pixel^51,52:

$$T_{ij} = \frac{1}{{MN}}\mathop {\sum}\limits_{m = - M/2}^{M/2} {\mathop {\sum}\limits_{n = - N/2}^{N/2} {f\left( {i + m,j + n} \right)} }$$

where f(i, j) is the intensity at pixel (i, j), and T_ij is the corresponding threshold. MN is the size of the neighborhood window. When the pixel intensity was higher than the global threshold, its threshold was set to the global threshold, otherwise it was set to the local threshold of the corresponding pixel. Then a binary image was computed with the calculated threshold:

$$T_{final}^{ij} = \left\{ {\begin{array}{*{20}{c}} {T_{otsu},f\left( {i,j} \right) \, < \, T_{otsu}} \\ {T_{ij},otherwise} \end{array}} \right.$$

where T_otsu is the global Otsu threshold and T_ij is the local threshold at pixel (i, j).

To segment overlapped cells, the binary image was further processed by distance transformation and gray reversal.⁵³ After basic smoothing, watershed segmentation was implemented on the reversed distance transform image.^54,55,56 The resultant watersheds were used to split the overlapped regions. To avoid over-segmentation, neighboring sub-regions were merged if their temporal correlation coefficient ρ was > 0.5.

Event detection and parametric extraction

We obtained the time course of pHluorin or pHtomatto fluorescence signals for each segmented E. coli cell. An algorithm for BacFlash event detection started with the location of a local peak in the signal. An event was detected when the peak was larger than its previous m points with a threshold T₁ and n following points with a threshold T₂ per the following formula:

$$\begin{array}{l}f_i\, > \max \left( {f_{i - m}:f_{i - 1}} \right) + T_1\;\& \\ f_i\, > \max \left( {f_{i + 1}:f_{i + n}} \right) + T_2\end{array}$$

The window size was m + n. We set m = 3, n = 3 and T₁ = 0.01 × maxima of the curve. Once identified, the event parameters basal fluorescence (F₀), peak (ΔF/F₀), full duration at half maximum, and rise time were measured based on the time courses of BacFlashes. The BacFlash frequency was indicated by quantifying the flash number in 1000 µm² of cell area in 100 s.

Statistics

Data are expressed as the means ± SEM. When appropriate, Student’s t-test was applied to determine the statistical significance of differences. In the promoter activity assay, one-way ANOVA was used to determine the significance of differences between three or more group sets. A two-tailed P < 0.05 was considered statistically significant.

Change history

11 December 2023
A Correction to this paper has been published: https://doi.org/10.1038/s41422-023-00900-5

References

Talley, K. & Alexov, E. On the pH-optimum of activity and stability of proteins. Proteins 78, 2699–2706 (2010).
Article PubMed PubMed Central CAS Google Scholar
Rius, N., Sole, M., Francia, A. & Loren, J. G. Buffering capacity and H+ membrane conductance of gram-negative bacteria. FEMS Microbiol. Lett. 130, 103–110 (1995).
PubMed CAS Google Scholar
Casey, J. R., Grinstein, S. & Orlowski, J. Sensors and regulators of intracellular pH. Nat. Rev. Mol. Cell Biol. 11, 50–61 (2010).
Article PubMed CAS Google Scholar
Foster, J. W. Escherichia coli acid resistance: tales of an amateur acidophile. Nat. Rev. Microbiol. 2, 898–907 (2004).
Article PubMed CAS Google Scholar
Slonczewski, J. L., Fujisawa, M., Dopson, M. & Krulwich, T. A. Cytoplasmic pH measurement and homeostasis in bacteria and archaea. Adv. Microb. Physiol. 55, 1–79, 317 (2009).
Article PubMed Google Scholar
Krulwich, T. A., Sachs, G. & Padan, E. Molecular aspects of bacterial pH sensing and homeostasis. Nat. Rev. Microbiol. 9, 330–343 (2011).
Article PubMed PubMed Central CAS Google Scholar
Kanjee, U. & Houry, W. A. Mechanisms of acid resistance in Escherichia coli. Annu. Rev. Microbiol. 67, 65–81 (2013).
Article PubMed CAS Google Scholar
Smith, D. K., Kassam, T., Singh, B. & Elliott, J. F. Escherichia coli has two homologous glutamate decarboxylase genes that map to distinct loci. J. Bacteriol. 174, 5820–5826 (1992).
Article PubMed PubMed Central CAS Google Scholar
De Biase, D., Tramonti, A., Bossa, F. & Visca, P. The response to stationary-phase stress conditions in Escherichia coli: role and regulation of the glutamic acid decarboxylase system. Mol. Microbiol. 32, 1198–1211 (1999).
Article PubMed Google Scholar
Hersh, B. M., Farooq, F. T., Barstad, D. N., Blankenhorn, D. L. & Slonczewski, J. L. A glutamate-dependent acid resistance gene in Escherichia coli. J. Bacteriol. 178, 3978–3981 (1996).
Article PubMed PubMed Central CAS Google Scholar
Hong, W., Wu, Y. E., Fu, X. & Chang, Z. Chaperone-dependent mechanisms for acid resistance in enteric bacteria. Trends Microbiol. 20, 328–335 (2012).
Article PubMed CAS Google Scholar
Nicholls, D. G. & Ferguson, S. L. Bioenergetics. (Academic Press, London, 2002).
Richard, H. & Foster, J. W. Escherichia coli glutamate- and arginine-dependent acid resistance systems increase internal pH and reverse transmembrane potential. J. Bacteriol. 186, 6032–6041 (2004).
Article PubMed PubMed Central CAS Google Scholar
Miesenbock, G., De Angelis, D. A. & Rothman, J. E. Visualizing secretion and synaptic transmission with pH-sensitive green fluorescent proteins. Nature 394, 192–195 (1998).
Article PubMed CAS Google Scholar
Li, Y. & Tsien, R. W. pHTomato, a red, genetically encoded indicator that enables multiplex interrogation of synaptic activity. Nat. Neurosci. 15, 1047–1053 (2012).
Article PubMed PubMed Central CAS Google Scholar
Bohnert, J. A., Karamian, B. & Nikaido, H. Optimized Nile Red efflux assay of AcrAB-TolC multidrug efflux system shows competition between substrates. Antimicrob. Agents Chemother. 54, 3770–3775 (2010).
Article PubMed PubMed Central CAS Google Scholar
Kralj, J. M., Hochbaum, D. R., Douglass, A. D. & Cohen, A. E. Electrical spiking in Escherichia coli probed with a fluorescent voltage-indicating protein. Science 333, 345–348 (2011).
Article PubMed CAS Google Scholar
Wang, W. et al. Superoxide flashes in single mitochondria. Cell 134, 279–290 (2008).
Article PubMed PubMed Central CAS Google Scholar
Hou, T., Wang, X., Ma, Q. & Cheng, H. Mitochondrial flashes: new insights into mitochondrial ROS signalling and beyond. J. Physiol. 592, 3703–3713 (2014).
Article PubMed PubMed Central CAS Google Scholar
Wang, W. et al. Mitochondrial flash: integrative reactive oxygen species and pH signals in cell and organelle biology. Antioxid. Redox Signal 25, 534–549 (2016).
Article PubMed PubMed Central CAS Google Scholar
Wang, X. et al. Protons trigger mitochondrial flashes. Biophys. J. 111, 386–394 (2016).
Article PubMed PubMed Central CAS Google Scholar
Martinac, B., Buechner, M., Delcour, A. H., Adler, J. & Kung, C. Pressure-sensitive ion channel in Escherichia coli. Proc. Natl. Acad. Sci. USA 84, 2297–2301 (1987).
Article PubMed PubMed Central CAS Google Scholar
Perozo, E. Gating prokaryotic mechanosensitive channels. Nat. Rev. Mol. Cell Biol. 7, 109–119 (2006).
Article PubMed CAS Google Scholar
Wang, J., Yan, D., Dixon, R. & Wang, Y. P. Deciphering the principles of bacterial nitrogen dietary preferences: a strategy for nutrient containment. MBio 7, e00792–16 (2016).
Article PubMed PubMed Central CAS Google Scholar
Yoch, D. C., Zhang, Z. M. & Claybrook, D. L. Methylamine metabolism and its role in nitrogenase “switch off” in Rhodopseudomonas capsulata. Arch. Microbiol. 134, 45–48 (1983).
Article PubMed CAS Google Scholar
Jayakumar, A., Epstein, W. & Barnes, E. M. Jr Characterization of ammonium (methylammonium)/potassium antiport in Escherichia coli. J. Biol. Chem. 260, 7528–7532 (1985).
Article PubMed CAS Google Scholar
Lund, P., Tramonti, A. & De Biase, D. Coping with low pH: molecular strategies in neutralophilic bacteria. FEMS Microbiol. Rev. 38, 1091–1125 (2014).
Article PubMed CAS Google Scholar
Baba, T. et al. Construction of Escherichia coli K-12 in-frame, single-gene knockout mutants: the Keio collection. Mol. Syst. Biol. 2, 2006.0008 (2006).
Article PubMed PubMed Central Google Scholar
Li, K. et al. Superoxide flashes reveal novel properties of mitochondrial reactive oxygen species excitability in cardiomyocytes. Biophys. J. 102, 1011–1021 (2012).
Article PubMed PubMed Central CAS Google Scholar
Fillingame, R. H. & Steed, P. R. Half channels mediating H(+) transport and the mechanism of gating in the Fo sector of Escherichia coli F1Fo ATP synthase. Biochim. Biophys. Acta. 1837, 1063–1068 (2014).
Article PubMed CAS Google Scholar
Allegretti, M. et al. Horizontal membrane-intrinsic alpha-helices in the stator a-subunit of an F-type ATP synthase. Nature 521, 237–240 (2015).
Article PubMed CAS Google Scholar
Miller, J. H. Experiments in Molecular Genetics. (Cold Spring Harbor Laboratory Press, New York, 1972).
Bruni, G. N., Weekley, R. A., Dodd, B. J. T. & Kralj, J. M. Voltage-gated calcium flux mediates Escherichia coli mechanosensation. Proc. Natl. Acad. Sci. USA 114, 9445–9450 (2017).
Article PubMed PubMed Central CAS Google Scholar
Breckwoldt, M. O. et al. Multiparametric optical analysis of mitochondrial redox signals during neuronal physiology and pathology in vivo. Nat. Med. 20, 555–560 (2014).
Article PubMed CAS Google Scholar
Zhang, X. et al. Superoxide constitutes a major signal of mitochondrial superoxide flash. Life Sci. 93, 178–186 (2013).
Article PubMed CAS Google Scholar
Wei-LaPierre, L. et al. Respective contribution of mitochondrial superoxide and pH to mitochondria-targeted circularly permuted yellow fluorescent protein (mt-cpYFP) flash activity. J. Biol. Chem. 288, 10567–10577 (2013).
Article PubMed PubMed Central CAS Google Scholar
Feng, G. et al. Mitoflash biogenesis and its role in the autoregulation of mitochondrial proton electrochemical potential. J. Gen. Physiol. 151, 727–737 (2019).
Article PubMed PubMed Central CAS Google Scholar
Bernardi, P., Rasola, A., Forte, M. & Lippe, G. The mitochondrial permeability transition pore: channel formation by F-ATP synthase, integration in signal transduction, and role in pathophysiology. Physiol. Rev. 95, 1111–1155 (2015).
Article PubMed PubMed Central Google Scholar
Carraro, M., Carrer, A., Urbani, A. & Bernardi, P. Molecular nature and regulation of the mitochondrial permeability transition pore(s), drug target(s) in cardioprotection. J. Mol. Cell Cardiol. 144, 76–86 (2020).
Article PubMed CAS Google Scholar
Archibald, J. M. endosymbiosis and eukaryotic cell evolution. Curr. Biol. 25, R911–R921 (2015).
Article PubMed CAS Google Scholar
Wang, X. et al. Mitochondrial flashes regulate ATP homeostasis in the heart. Elife 6, e23908 (2017).
Article PubMed PubMed Central Google Scholar
Fang, H. et al. Imaging superoxide flash and metabolism-coupled mitochondrial permeability transition in living animals. Cell Res. 21, 1295–1304 (2011).
Article PubMed PubMed Central CAS Google Scholar
Fu, Z. X. et al. Dendritic mitoflash as a putative signal for stabilizing long-term synaptic plasticity. Nat. Commun. 8, 31 (2017).
Article PubMed PubMed Central Google Scholar
Ahmad, D. & Newman, E. B. A deficiency in cyclic AMP results in pH-sensitive growth of Escherichia coli K-12. J. Bacteriol. 170, 3443–3447 (1988).
Article PubMed PubMed Central CAS Google Scholar
Grenier, F., Matteau, D., Baby, V. & Rodrigue, S. Complete genome sequence of Escherichia coli BW25113. Genome Announc. 2, e01038–14 (2014).
Article PubMed PubMed Central Google Scholar
Kulahoglu, C. & Brautigam, A. Quantitative transcriptome analysis using RNA-seq. Methods Mol. Biol. 1158, 71–91 (2014).
Article PubMed Google Scholar
Burton, N. A., Johnson, M. D., Antczak, P., Robinson, A. & Lund, P. A. Novel aspects of the acid response network of E. coli K-12 are revealed by a study of transcriptional dynamics. J. Mol. Biol. 401, 726–742 (2010).
Article PubMed CAS Google Scholar
You, C. et al. Coordination of bacterial proteome with metabolism by cyclic AMP signalling. Nature 500, 301–306 (2013).
Article PubMed PubMed Central CAS Google Scholar
Reddy, B. S. & Chatterji, B. N. An FFT-based technique for translation, rotation, and scale-invariant image registration. IEEE Trans Image Process 5, 1266–1271 (1996).
Article PubMed CAS Google Scholar
Otsu, N. A Threshold Selection Method from Gray-Level Histograms. IEEE Trans. Syst. Man Cybern. 9, 62–66 (1979).
Article Google Scholar
Ahmad, M. B. & Tae-Sun, C. Local threshold and Boolean function based edge detection. IEEE Trans. Consumer Electronics 45, 674–679 (1999).
Article Google Scholar
Shafait, F., Keysers, D. & Breuel, T. M. In Electronic Imaging. 6 (SPIE) (2008).
Borgefors, G. Distance transformations in digital images. Comput. Vis. Gr. Image Proc. 34, 344–371 (1986).
Article Google Scholar
Vincent, L. & Soille, P. Watersheds in digital spaces: an efficient algorithm based on immersion simulations. IEEE Trans. Pattern Anal. Mach. Intell. 13, 583–598 (1991).
Article Google Scholar
Haris, K., Efstratiadis, S. N. & Maglaveras, N. In Proceedings 1998 International Conference on Image Processing. ICIP98 (Cat. No.98CB36269). 338–342 vol. 333.
Bleau, A. & Leon, L. J. Watershed-based segmentation and region merging. Comput. Vis. Image Underst. 77, 317–370 (2000).
Article Google Scholar

Download references

Acknowledgements

We thank Dr. Yi-Ping Wang for the gifts of pMP190 plasmid and PKUW13 and PKUW151 strains, Dr. Liangyi Chen for the gift of pHluorin plasmid, Dr. Chuanyun Li and Dr. Wanqiu Ding for bioinformatics analysis, Mr. Dapeng Zhang for helping with automated BacFlash detection, Dr. Zhenzhen Rao for ANOVA analysis, Dr. Peng Yu and Mrs. Zhengling Yang for technical assistance, and Drs. Conghui You, Shiqiang Wang, and Peng Zou for valuable comments. This work was supported by the National Key Basic Research Program of China (2017YFA0504000 and 2016YFA0500403) and the National Science Foundation of China (31670039, 81827809, 31821091 and 31971158).

Author information

These authors contributed equally: Di Wu, Wenfeng Qi, Wei Nie

Authors and Affiliations

State Key Laboratory of Membrane Biology, Beijing Key Laboratory of Cardiometabolic Molecular Medicine, Peking-Tsinghua Center for Life Sciences, Institute of Molecular Medicine, Peking University, Beijing, 100871, China
Di Wu, Wenfeng Qi, Wei Nie, Zhengyuan Lu, Yongxin Ye, Jinghang Li, Tao Sun, Yufei Zhu, Heping Cheng & Xianhua Wang
Research Unit of Mitochondria in Brain Diseases, Chinese Academy of Medical Sciences, PKU-Nanjing Institute of Translational Medicine, Nanjing, Jiangsu, China
Wei Nie, Yufei Zhu, Heping Cheng & Xianhua Wang

Authors

Di Wu
View author publications
You can also search for this author in PubMed Google Scholar
Wenfeng Qi
View author publications
You can also search for this author in PubMed Google Scholar
Wei Nie
View author publications
You can also search for this author in PubMed Google Scholar
Zhengyuan Lu
View author publications
You can also search for this author in PubMed Google Scholar
Yongxin Ye
View author publications
You can also search for this author in PubMed Google Scholar
Jinghang Li
View author publications
You can also search for this author in PubMed Google Scholar
Tao Sun
View author publications
You can also search for this author in PubMed Google Scholar
Yufei Zhu
View author publications
You can also search for this author in PubMed Google Scholar
Heping Cheng
View author publications
You can also search for this author in PubMed Google Scholar
Xianhua Wang
View author publications
You can also search for this author in PubMed Google Scholar

Contributions

D.W., W.Q., W.N., Z.L., and Y.Z. performed experiments. Y.Y., J.L., and T.S. developed algorithms and software for automated BacFlash detection. X.W. and H.C. conceived and supervised the project and wrote the manuscript with inputs from all co-authors.

Corresponding authors

Correspondence to Heping Cheng or Xianhua Wang.

Ethics declarations

Competing interests

The authors declare no competing interests.

Additional information

The original online version of this article was revised: in the 'Acknowledgements' section of this article, one of the grant numbers relating to the National Science Foundation of China was incorrectly given as 8182780030 and should have been 81827809.

Supplementary information

Supplementary Figure S1

Supplementary Figure S2

Supplementary Figure S3

Supplementary Figure S4

Supplementary Figure S5

Supplementary Figure S6

Supplementary Figure S7

Supplementary Figure S8

Supplementary Figure S9

Supplementary Figure S10

Supplementary Figure S11

Supplementary Figure S12

Supplementary Figure S13

Supplementary Figure S14

Supplementary Figure S15

Supplementary Figure S16

Supplementary Table S1

Supplementary Table S2

Supplementary Table S3

Supplementary Video S1

Supplementary Video S2

Supplementary Video S3

Supplementary information, Video legends

Rights and permissions

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and permissions

About this article

Cite this article

Wu, D., Qi, W., Nie, W. et al. BacFlash signals acid-resistance gene expression in bacteria. Cell Res 31, 703–712 (2021). https://doi.org/10.1038/s41422-020-00431-3

Download citation

Received: 14 January 2020
Accepted: 14 October 2020
Published: 06 November 2020
Issue Date: June 2021
DOI: https://doi.org/10.1038/s41422-020-00431-3