In all living organisms, adenosine triphosphate (ATP) and NAD(H) represent universal molecular currencies for energy and redox state, respectively, and are thus widely applicable molecular proxies for an organism’s viability and activity. To this end, corresponding luciferase-based assays in combination with a microplate reader were established with the marine model bacterium Phaeobacter inhibens DSM 17395 (Escherichia coli K12 served as reference). Grey multiwell plates best balanced sensitivity and crosstalk, and optimal incubation times were 5 min and 30 min for the ATP and NAD(H) assay, respectively, together allowing limits of detection of 0.042, 0.470 and 0.710 nM for ATP, NAD+, and NADH, respectively. Quenching of bacterial cell samples involved Tris-EDTA-DTAB and bicarbonate base-DTAB for ATP and NAD(H) assays, respectively. The ATP and NAD(H) yields determined for P. inhibens DSM 17395 at ¼ ODmax were found to reside well within the range previously reported for E. coli and other bacteria, e.g., 3.28 µmol ATP (g cellsdry)−1. Thus, the here described methods for luciferase-based determination of ATP/NAD(H) pools open a promising approach to investigate energy and redox states in marine (environmental) bacteria.

© 2022 The Author(s). Published by S. Karger AG, Basel

Introduction

ATP (adenosine triphosphate), the chemical energy currency, and NAD(H) (nicotinamide adenine dinucleotide), a central hydrogen-transferring cofactor in energy metabolism, are used by all living organisms and have been investigated for a long time, representing hallmarks of biochemistry. ATP was discovered in the late 1930s [e.g., Lohmann, 1929], its role in linking energy-consuming and -producing reactions recognised in the early 1950s [Lipmann, 1941], and its formation by ion-pumping ATP synthases elucidated thereafter [Skou and Esmann, 1992; Boyer, 1993; Abrahams et al., 1994]. NADH was discovered at the beginning of the 20th century [Harden and Young, 1906], hydrid transfer during redox reactions indicated in the 1940s [Warburg and Christian, 1936] and its role in oxidative phosphorylation elucidated in the 1950s [Friedkin and Lehninger, 1949]. Chemotrophic (micro)organisms generate ATP by substrate level phosphorylation (e.g., pyruvate kinase in EMP pathway) and/or electron transport phosphorylation. NADH, provided by a multitude of catabolic oxidation reactions, serves as primary electron donor for the respiratory chain, which generates the transmembrane electrochemical gradient [Mitchell, 1961], in turn driving ATP synthase [Thauer et al., 1977; Steigmiller et al., 2008; Sazanov, 2014; Kühlbrandt, 2019].

Microbial life, as reflected, e.g., by maintenance, adaptation, or growth, is tightly connected with the cellular energy (ATP) and redox (NAD(H)) states. Conversely, knowledge of intracellular ATP/NAD(H) content represents a quantifiable molecular proxy of viability and metabolic activity. This adds to our understanding of resource allocation between assimilation and dissimilation, and of growth efficiency during various conditions of growth [Bremer and Dennis, 1987; del Giorgio and Cole, 1998; Russel and Cook, 1995]. Thus, in combination with habitat-mimicking cultivation regimes, reliable determination of intracellular ATP/NAD(H) concentrations could provide a valuable estimate of viability and activity of a given study organism under natural settings.

In accord with the aforementioned general relevance of ATP and NAD(H), a wide range of analytical methods has been established for measuring concentrations of these key metabolites. The methodological approaches include (i) high-performance liquid chromatography (HPLC) [Ritov et al., 2004; Olafsson et al., 2017], (ii) spectrophotometry [Wagner and Scott, 1994; Brown et al., 2020], (iii) electrochemical biosensors/assays [Bergel et al., 1989; Deng et al., 2009] and (iv) in situ fluorescence detection [Baraghis et al., 2011; Wu et al., 2020]. Compared to these methods, luciferase-based assays combine high sensitivity and large dynamic range with ease of use and suitability for high-throughput applications by means of microplate readers equipped with the respective optics [e.g., Fan and Wood, 2007; Inglese et al., 2007; Necchi et al., 2017].

Luciferase (EC 1.13.12.7), member of the adenylate-forming enzyme superfamily [Gulick, 2009], oxidatively decarboxylates luciferin (LH2) in an ATP-dependent manner to oxyluciferin (OxyLH2) with concomitant emission of yellow-green light. The substrate LH2 is composed of a benzothiazole and a thiazoline-carboxylic moiety, which are covalently linked via a C2′−C2 bond (Fig. 1a). The current view of the multi-step mechanism underlying the luciferase-based bioluminescence is schemed in Figure 1a and involves the following steps [Day et al., 2004; Marques and Esteves da Silva, 2009; Thorne et al., 2010; Hosseinkhani et al., 2011; Sundlov et al., 2012]: (i) in an initial nucleophilic displacement reaction, the carboxyl group at C4 of LH2 attacks the α-phosphorous of Mg2+-ATP forming enzyme-bound luciferyl-adenylate (LH2-AMP) and releasing Mg2+-PPi (Fig. 1a, top). (ii) Following base-catalysed H+-removal at C4 of LH2-AMP, oxidation by molecular oxygen (O2) yields the key LH2-AMP hydroperoxide intermediate (dioxetanone) and release of AMP (Fig. 1a, middle and bottom). (iii) Finally, decarboxylation at C4 of dioxetanone forms oxyluciferin with concomitant emission of light (Fig. 1a, bottom). The emission maximum (550–560 nm) apparently depends on the dissociation state of the hydroxyl group at C5′ combined with the oxidation/dissociation state of the oxygen at C4.

Fig. 1.

Biochemical principles of luciferase-catalysed bioluminescence. a Proposed multi-step mechanism of luciferase-catalysed reaction involving ATP-dependent activation of luciferin (LH2) to LH2-AMP followed by oxidative conversion to dioxetanone and final decarboxylation to oxyluciferin with concomitant emission of light (adapted from Marques and Esteves da Silva [2009] and Thorne et al. [2010]); note that the first two reactions are reversible and that at C4 of oxyluciferin, a keto-enol-tautomery is assumed. This reaction principle was applied in the assay for quantification of ATP. b Diaphorase-catalysed cycling reaction involves NADH-dependent reductive intramolecular lactonisation of a “trimethyl lock quinone” (e.g., quinone-trimethyl lock-C2-diamine-2-cyanobenzothiazole carbamate) followed by rapid luciferin formation (adapted from Zhou et al. [2014]). The latter then fuels luciferase-catalysed light emission (see above). This reaction principle was applied in the assay for quantification of NAD+/NADH.

To exploit luciferase-based assays also for the quantification of NAD(P)+/NAD(P)H, an upstream NAD(P)H-dependent diaphorase (EC 1.6.5.2) cycling reaction is employed [Zhou et al., 2014]. While the chemical composition of commercial assays is unknown (proprietary), the cycling reaction should proceed via the following principal steps. Diaphorase catalyses the NADH-dependent intramolecular lactonisation of a “trimethyl lock quinone” substrate yielding 6-hydroxyl cyanobenzothiazole next to the diamide and quinone moieties. Coupling with lactate-oxidising lactate dehydrogenase (EC 1.1.1.27) refurnishes NADH, thereby serving an amplifying function in the assay (Fig. 1b, top). Finally, excess of cysteine in the assay drives formation of luciferin from 6-hydroxyl cyanobenzothiazole under release of ammonium (Fig. 1b, bottom). Emission of light is achieved via the aforementioned luciferase-based reaction.

The aim of the present study was to establish and evaluate luciferase-based ATP/NAD(H) assays in multiwell plate format using a monochromator-based microplate reader for application with marine (environmental) bacteria. For this purpose, the aerobic heterotrophic Phaeobacter inhibens DSM 17395 was selected, since it is genome-sequenced [Thole et al., 2012], well-studied on the physiological and proteogenomic level in our laboratory [Drüppel et al., 2014; Trautwein et al., 2018; Wiegmann et al., 2014; Wünsch et al., 2019, 2020] and a member of the alphaproteobacterial Roseobacter group, which is widespread in pelagic oceanic water bodies [Buchan et al., 2005; Luo and Moran, 2014; Simon et al., 2017]. Facultative anaerobic, intestinal Escherichia coli K12 [Blattner et al., 1997; Hayashi et al., 2006; Keseler et al., 2017], belonging to the gammaproteobacterial family Enterobacteriaceae [Adeolu et al., 2016], was used as reference. To put the determined ATP/NAD(H) profiles into a more general physiological context, they were integrated with corresponding growth stoichiometric data.

Methods/DesignStrains, Media and General Cultivation Conditions

E. coli K12 MG1655 (DSM 18039) [Guyer et al., 1981] and P. inhibens (DSM 17395; originally deposited as Phaeobacter gallaeciensis) [Buddruhs et al., 2013] were obtained from the German Collection of Microorganisms and Cell Cultures (DSMZ; Braunschweig, Germany).

Routine cultivation, substrate adaptation of bacterial strains and main growth experiments were performed as follows: (i) E. coli K12 in 250 mL defined mineral M9-medium [Sambrook and Russell, 2001] containing 6 mM glucose (1,000-mL Erlenmeyer flasks, 37°C, 100 rpm); (ii) P. inhibens DSM 17395 in 250 mL defined marine mineral medium [Zech et al., 2009] containing 15 mM glucose (1,000-mL Erlenmeyer flasks, 28°C, 100 rpm). Prior to any of the experiments described below, each of the two strains was adapted to the cultivation conditions described above over five passages, starting from glycerol stocks in each case. Main cultures were then inoculated from actively growing pre-cultures. Organic substrates were provided from sterile stock solutions. Purity of the cultures was confirmed by microscopic examination (Axiostar; Zeiss AG, Göttingen, Germany). All chemicals used were of analytical grade.

Evaluation of Quenching Methods

E. coli cultures were used to determine the most suitable quenching method for the ATP assay (online suppl. Fig. S1; see www.karger.com/doi/10.1159/000522414 for all online suppl. material). Per tested centrifugation temperature, three replicate cultures (each derived from an independent glycerol stock) were used. Each individual culture was harvested at an optical density (OD) of ∼0.2. Two aliquots à 1 mL were immediately shock-frozen in liquid N2. Ten aliquots à 1 mL were centrifuged (20,800 × g, 5 min, 14°C or room temperature), the obtained cell pellets resuspended in the various quenching buffers, followed by shock freezing in liquid N2; the supernatant from 1 replicate was retained and also shock-frozen in liquid N2. All generated samples were stored at −80°C. The tested quenching buffers were: 1× PBS buffer; 6 M guanidine HCl in 1× PBS; 0.2 M NaOH base buffer with 1% (v/v) dodecyltrimethylammonium bromide (DTAB); bicarbonate base buffer with 1% (v/v) DTAB; and Tris (pH 7.75) with 2 mM EDTA and 1% (v/v) DTAB.

Growth Experiments, Sampling and Quenching

Growth experiments were conducted as described in the above section Strains, Media and General Cultivation Conditions. For determination of OD, cell counting and substrate depletion, 1-mL samples were retrieved at short intervals; for ATP/NAD(H) analyses 1.8/0.9-mL samples at ¼ ODmax, ½ ODmax, ¾ ODmax and ODmax.

Quenching of samples for ATP/NAD(H) analyses involved the following, swiftly performed steps. At first, the retrieved culture broth was centrifuged (20,800 × g, 14°C, 5 min). Then, the supernatants were decanted and shock-frozen in liquid N2, while the resultant cell pellets were resuspended on ice in Tris-EDTA with 1% (v/v) DTAB for the ATP assay or in bicarbonate base buffer with 1% (v/v) DTAB for the NAD(H) assay. These quenched samples were immediately shock-frozen in liquid N2 prior to storage at −80°C. The quenching procedure was performed using gloves, cleaned pipets, sterile pipets tips and micro reaction cups.

Monitoring Growth

OD was determined at 600 nm using a UVmini-1240 spectrophotometer (Shimadzu, Duisburg, Germany). Total cell counts (TCC) were determined with a Thoma counting chamber (Glaswarenfabrik Karl Hecht GmbH & Co. KG, Sondheim vor der Rhön, Germany). For determination of cellular dry weight (CDW) duplicate cultures per strain were harvested at ½ ODmax and ODmax by centrifugation (11,300 × g, 20 min, 4°C) and cell pellets were washed twice in cold 50 mM ammonium acetate. Resuspended cells (300 µL) were then transferred into pre-dried and weighed 1.5-mL tubes. CDW was determined by gravimetric analysis of tubes after drying to a constant weight at 60°C (the term cellsdry refers to the mass of dried cells). For determination of cellular protein content, duplicate cultures per strain were harvested at ¼ ODmax, ½ ODmax, ¾ ODmax and ODmax by centrifugation (11,300 × g, 20 min, 4°C) and washing of cell pellets twice in cold 0.9% NaCl (E. coli K12) or cold 3.7% NaCl (P. inhibens DSM 17395), followed by shock freezing in liquid N2 and storage at −80°C. Protein content was determined after alkaline hydrolysis (0.5 M NaOH containing 2% (w/v) SDS, 60°C, 10 min) using the RCDC assay (Bio-Rad Laboratories, Inc., Hercules, CA, USA) according to the manufacturer’s instructions. By correlating TCC and CDW with the respective OD, the course of TCC and CDW across the entire growth was estimated (online suppl. Fig. S2), in order to calculate qx/x and yields also for ¼ ODmax and ¾ ODmax.

The depletion of glucose during growth of P. inhibens DSM 17395 was determined by HPLC using an Ultimate3000 system (ThermoFisher, Germering, Bavaria, Germany) equipped with an RI-detector (Shodex RI-101; Showa Denko GmbH, München, Germany) and operated at 75°C. Separation was achieved on an Eurokat column (300 × 8 mm, 10-µm bead size; Knauer, Berlin, Germany), using 5 mM H2SO4 as eluent administered at 1.2 mL min‒1. The retention time was at 5.6 min and the dynamic range from 25 µM to 10 mM. Due to superposing background signals in the HPLC separation profile, the depletion of glucose during growth of E. coli K12 had to be determined by means of an enzymatic assay (Glucose Assay Kit ab65333; Abcam, Cambridge, UK) according to the manufacturer’s instructions, with the reaction conducted at 37°C and 400 rpm in the dark. Fluorescent recording (excitation: 535 nm; emission: 587 nm) was achieved by using a microplate reader (CLARIOstarPlus; BMG Labtech, Ortenberg, Germany).

Luciferase-Based Assays

To examine the suitability of different types of multiwell plates, the following commercial products were tested: white (Cat. # 655074; Greiner, Kremsmünster, Austria), grey (Cat. # 6002350; PerkinElmer, Waltham, MA, USA), black (Cat. # 655077; Greiner). Generally, sterile pipet tips and gloves were used. Samples and respective standards were always measured in triplicates.

ATP Assays

The ATP assays for growth experiments were conducted using the BacTiter-GloTM Assay (Cat. # G8231; Promega, Fitchburg, WI, USA) following the manufacturer’s instructions. First, the BacTiter-GloTM Buffer and BacTiter-GloTM Substrate were thawed and equilibrated to room temperature prior to use. Then, the BacTiter-GloTM Reagent was reconstituted by adding exactly 10 mL of the BacTiter-GloTM Buffer to the BacTiter-GloTM Substrate and gently mixing by inverting the solution for ∼1 min. The reagent was then equilibrated at room temperature for 15–120 min prior to use. The equilibration time depends on the strived for sensitivity of the assay, as trace amounts of ATP might be introduced through the manufacturing process and need to “burn off” to lower the background luminescence signal. The BacTiter-GloTM Reagent is then stable for 8 h at room temperature and might be stored longer at colder temperatures (see manufacturer’s technical bulletin). The quenched samples were thawed on ice, diluted in Tris (pH 7.75) with 2 mM EDTA (dilution buffer ATP; DB ATP) and equilibrated to room temperature, shortly before assaying. Then, 25 µL of the diluted samples and ATP standard series (0.1–1,000 nM ATP in Tris-EDTA; 100 mM ATP-stock, Cat. # R0441; Thermo ScientificTM, Waltham, MA, USA) were administered to the multiwell plate, followed by adding 25 µL BacTiter-GloTM Reagent to each used well and immediate mounting of the plate into the CLARIOstar® Plus reader (procedure and exemplary layout see online suppl. Fig. S3). Before recording of the luminescence signal at 545–550 nm, the plate was mixed in orbital shaker mode (10 s, 400 rpm) and incubated for 5 min at 25°C. For assessing plate types and quenching methods, the dilution buffer may vary, depending on the quenching buffer used (online suppl. Fig. S1).

NAD(H) Assays

The NAD(H) assays were conducted using the NAD/NADH-GloTM Assay (Cat. # G9071; Promega), following the manufacturer’s instructions. Initially, NAD/NADH-GloTM Buffer and NAD/NADH-GloTM Substrate were thawed and equilibrated to room temperature. Then, the Luciferin Detection Reagent was reconstituted by adding the whole content of the NAD/NADH-GloTM Buffer to the NAD/NADH-GloTM Substrate and gentle mixing by inverting the solution for ∼1 min. This reagent is stable for 24 h at room temperature and might be stored longer at colder temperatures (see manufacturer’s technical bulletin). Shortly before assaying, the NAD/NADH-GloTM Detection Reagent is reconstituted (stable <6 h) by adding per mL: 5 µL Reductase, 5 µL Reductase Substrate, 5 µL NAD reconstituted Cycling Enzyme, and 25 µL NAD Cycling Substrate. The reconstituted NAD/NADH-GloTM Detection Reagent is mixed by inverting 5 times and equilibrated to room temperature. Prior to use, 2 mM NAD+ (10 mg per vial; Cat. # N8285; Sigma-AldrichTM) and 0.2 mM NADH (0.2 mg per vial; Cat. # N6660; Sigma-AldrichTM) stocks were freshly prepared on ice by adding 7.5 mL or 1.4 mL of 0.05 M Tris buffer (pH 8.00) to the respective glass vials. To separately detect NAD+ or NADH with the assay, a pretreatment of the quenched samples was necessary. This pretreatment uses the selective stability of either reduced or oxidised form of the nicotinamide adenine dinucleotide coenzyme under low or high pH conditions while exposed to heat [Passonneau and Lowry, 1993]. On the one hand, NAD+ is rather stable under acidic conditions, while NADH decomposes, involving hydration, epimerization and cyclisation. On the other hand, NADH is rather stable at basic conditions, while NAD+ decomposes, e.g., via cleavage of the nicotinamide-ribose bond [Chenault and Whitesides, 1987]. This process is accelerated at higher temperatures [Passonneau and Lowry, 1993]. Therefore, the quenched samples were thawed on ice, mixed via vortexing, and 100-µL samples were pipetted into a new micro-reaction cup for each dinucleotide and its respective pretreatment. Samples intended for assaying NAD+ were acidified with 100 µL 0.4 M HCl, mixed and incubated for 15 min at 60°C. Samples destined for NADH assaying were incubated for 15 min at 60°C (already basic bicarbonate base buffer). After letting both sample types cool at room temperature for 10 min, NAD+ samples were neutralised with 100 µL 0.5 M Trizma®, while NADH samples were neutralised with 200 µL HCl/Trizma® solution (1:2 dilution of 0.4 M HCl and 0.5 M Trizma®). Both sample types were then diluted using a 1:3 bicarbonate base buffer-HCl/Trizma® solution (dilution buffer NADH; DB NADH). 25 µL of diluted sample and the respective standard series (10–400 nM NAD+/NADH) were then administered to a multiwell plate, 25 µL NAD/NADH-GloTM Detection Reagent was added to each occupied well, and the plate was immediately mounted into the CLARIOstar® Plus reader (procedure and exemplary layout see online suppl. Fig. S4). The plate was then mixed on orbital shaker mode (10 s, 400 rpm) and incubated for 30 min before recording the luminescence signal.

Discussion/ConclusionEstablishment of Luciferase-Based Assays

The luciferase-based assays were established on the basis of the ATP assay and involved evaluation of the suitability of different multiwell plate types, length of incubation time, quenching, and calibration methods. In case of quenching experiments, cultures of E. coli K12 were used.

Multiwell Plate Types

To assess horizontal crosstalk between adjacent wells (online suppl. Fig. S5A), multiwell plates with different colouring and design were compared. (i) White multiwell plates are usually applied for luciferase-based assays, since this colouring enhances the luciferase signal. (ii) Black multiwell plates are used in particular for fluorescent assays, producing signal strengths higher than luciferase assays. Since particles in the well walls block well-to-well transmission, the inter-well crosstalk is minimised, resulting in a better signal-to-noise ratio. (iii) Grey half-area multiwell plates are characterised by a crosstalk-minimising design (reduced well volume, increased inter-well distance, black spiked white colouring). Independent of the multiwell plate type tested, crosstalk due to light bleeding from neighbouring wells towards the overlying optics is minimised by the perforated aperture spoon implemented in the multiwell plate reader used here (online suppl. Fig. S5A). This aperture is automatically positioned between the movable optical detector and the specific well to be analysed, thereby excluding the undesirable bleeding signals. For each of the above-mentioned multiwell plate types, two different approaches were used to quantify crosstalk as described in the following.

Initially, a simple plate layout (online suppl. Fig. S5B) was used to measure crosstalk against a blank background when applying high ATP concentrations in the luciferase-based assay. According to this given plate setup, the CLARIOstar® software calculates a plate-/method-specific factor to level out the occurring horizontal crosstalk. The determined crosstalk factors were 0.05353% for the white plates, 0.00026% for the black plates, and 0.01277% for the grey plates.

Then, an ATP concentration series was used (considering the specific crosstalk factors) to assess the interplay between crosstalk and sensitivity when neighbouring wells harbour markedly different ATP amounts (online suppl. Fig. S5C). The 7-point concentration series was adjusted to the different multiwell plate types: 0.01–1,000 nM ATP for white and grey plates versus 0.05–1,000 nM ATP for the black plates. While the overall ATP concentration-dependent course of luciferase signals (luminescence) was similar for all three multiwell plate types (online suppl. Fig. S5D), sensitivity decreased from white via grey to black plates. Furthermore, wells containing very low ATP concentrations and neighbouring wells with high ATP concentrations exhibited ∼10-fold increased luminescence due to crosstalk (online suppl. Fig. S5E).

Taken together, these experiments indicate the following two points: first, the grey multiwell plates were found to provide the best compromise between sensitivity and crosstalk for the luciferase-based assays. An additional advantage of these plates is the reduced assay/sample volume. Second, the general setup of multiwell plates needs to avoid direct proximity between samples with markedly differing concentrations of ATP.

Incubation Time

To ensure the optimal incubation time prior to analysing the luciferase-based assays with the microplate reader, three time series with respective standards and buffer system (ATP in Tris-EDTA and NAD(H) in bicarbonate base-HCl/Trizma®) were conducted (online suppl. Fig. S6). To best cover for high and low concentrations, the incubation time for the ATP and NAD(H) assays was set to 5 and 30 min, respectively.

Quenching Methods

Culture samples reserved for luciferase-based assays have to be pre-treated by centrifugation and decanting of supernatants followed by quenching of the cells. This is essential, since (i) media components, especially high salt concentrations, might inhibit the luciferase reaction, resulting in lower sensitivity [Abushaban et al., 2017]; and (ii) immediate processing and analyses of samples concurrent with ongoing cultivations (∼8 h to over 200 h) are rather impracticable. The suitability of different quenching methods for determination of ATP was evaluated using E. coli K12 cultures harvested at OD600 ∼0.2, involving centrifugation at either 14°C or room temperature (Table 1 and online suppl. Figs. S1 and S7). Culture samples were applied instantaneously (“fresh cells”) to the ATP assay or following shock freezing in liquid N2 (“shock-frozen cells”), and culture supernatants served as references to elucidate potential analyte losses due to harvesting and quenching, respectively. The tested quenching methods comprised PBS, guanidine HCl, NaOH-DTAB, bicarbonate base-DTAB, and Tris-EDTA-DTAB [e.g., Hoffner et al., 1999; Squirrell et al., 2002; Pinu et al., 2017; Sobol et al., 2021], most of which contain the anionic surfactant DTAB [Sachin et al., 2018]. In case of the latter three buffers, additional experiments with a heating step were conducted, which, however, revealed that this measure had no effect on the determined ATP concentrations. Application of guanidine HCl did not allow to detect ATP at all. While the centrifugation temperature has significant albeit opposing effects with NaOH-DTAB and bicarbonate base-DTAB, nothing like that was observed with PBS and Tris-EDTA-DTAB. Analyses of the culture supernatant indicated that cells are apparently getting less leaky during centrifugation at 14°C as compared to room temperature. Thus, in summary, using Tris-EDTA-DTAB for quenching of ATP samples followed by centrifugation at 14°C were chosen as optimal conditions for all further experiments. In case of NAD(H) samples, the quenching method with bicarbonate base-DTAB was applied as described in the technical manual of the NAD/NADH-GloTM Assay and cell harvesting was conducted at the above-mentioned centrifugation temperature (14°C). Note that the buffers for preparation of standards and for sample dilutions were devoid of DTAB, in order to reduce any interfering effect with the luciferase-based assays.

Table 1.

ATP concentration depending on the quenching method applied prior to luciferase-based ATP assay

Calibration

In order to determine assay-specific parameters, 10-point calibrations were conducted for the ATP and the NAD(H) assays, applying the buffer conditions established in the above section Quenching Methods (Fig. 2). For calibration, an ATP standard series (0, 0.1, 0.125, 0.25, 0.5, 1, 10, 100, 500, and 1,000 nM ATP in Tris-EDTA) and NAD+/NADH standard series (0, 10, 12.5, 15, 20, 50, 75, 100, 200, and 400 nM NAD+ or NADH in bicarbonate base-HCl/Trizma®) were measured 16 times across two 96-well plates, applying the respective assay conditions. The limits of detection (LOD), identification (LOI) and quantification (LOQ) were calculated according to the method of blank determination [Magnusson and Örnemark, 2014]. For the ATP assay, the LOD was 0.042 nM, the LOQ was 0.141 nM and the linearity ranged to at least 1,000 nM; for the NAD+ assay, the LOD was 0.470 nM, the LOQ was 1.580 nM and the linearity ranged to at least 400 nM; for the NADH assay, the LOD was 0.710 nM, the LOQ was 2.360 nM and the linearity ranged to at least 400 nM. The high R2 (>0.99) depicted in each plot underlines the linearity and confidence of each calibration. The procedural standard deviations were 5.19 nM for ATP, 4.61 nM for NAD+, and 0.81 nM for NADH.

Fig. 2.

Quantitative determination of ATP (a), NAD+ (b) and NADH (c) using Ultra-GloTM luciferase-based assays. For the 10-point calibrations, the applied standards covered the following concentration range: 0–1,000 nM for ATP and 0–400 nM for NAD+ and NADH. For each standard concentration, 16 replicates (orange dots) were dispensed at even distances across two 96-well plates and measured at 545–550 nm. Black dots: mean of respective standard concentrations. Diagonal lines: dashed black, calibration curve; dashed blue, 95% confidence interval. The equation of the respective linear regression and the corresponding R2 are indicated in each plot. The limits of detection (LOD), identification (LOI) and quantification (LOQ) are indicated with grey dashed lines in the zoom-ins.

Application for E. coli K12 versus Marine P. inhibens DSM 17395

The established luciferase-based assays for ATP and NAD(H) were then applied to the standard bacterium E. coli K12 and the marine bacterium P. inhibens DSM 17395. The aim was to profile ATP/NAD(H) contents across growth stages (¼ ODmax, ½ ODmax, ¾ ODmax and ODmax) of batch cultures typically used for OMICS-based studies. Furthermore, these profiles should be correlated with growth stoichiometric parameters (OD, CDW, TCC, substrate consumption and biomass-specific growth rate) determined in the same growth experiments. Cultivation was performed under oxic conditions in shake flasks in defined mineral medium provided with glucose as single source of organic carbon and energy. The resultant data are compiled in Figure 3 and Table 2.

Table 2.

Yields and concentrations depending on type of normalizationa

Fig. 3.

Growth behaviour of the standard bacterium E. coli K12 (a) and the marine model bacterium P. inhibens DSM 17395 (b) during batch cultivation in defined mineral medium containing glucose as sole source of carbon and energy. Basic growth parameters are determined by measurement of optical density (OD600) (black, filled dots), concentration of cellular dry weight (CDW) (black, open bars) and glucose consumption (converted to carbon; red, filled dots). The growth energetics are assessed by the biomass-specific growth rate (qX/X) (black, open dots) and the biomass demand for ATP (YATP/X) (brown, filled dots), NAD+ (YNAD+/X) (green, filled dots) and NADH (YNADH/X) (green, filled dots).

Escherichia coli K12

The growth behaviour of E. coli strains in batch cultivations with defined mineral media (e.g., M9) and glucose as sole source of carbon and energy is probably one of the most studied and best understood cultivation processes. Therefore, we used E. coli K12 as a reference for comparison with marine P. inhibens DSM 17395. Growth of E. coli K12 is characterised by exponential growth until transition into a stationary phase due to carbon limitation. Even though exponential growth is generally regarded as ideal, unhindered growth (µmax) can result in the well-known effect of acetate overflow metabolism, which can be avoided by reducing µ in fed-batch cultivations [e.g., Basan et al., 2015; Bernal et al., 2016; Enjalbert et al., 2017]. This overflow effect underlies a complex, multifactorial, not well-understood process involving transcriptional and posttranscriptional regulation, posttranslational modification of proteins, etc. Additionally, high concentrations of primary electron donors (NADH), observed at high growth rates, were previously reported to reduce the flux through the TCA cycle by allosterically inhibiting both citrate synthase and isocitrate dehydrogenase leading to acetate accumulation [Yao et al., 2016]. In the present study, the growth behaviour of E. coli K12 was as outlined above, with exponential growth abruptly transiting into a stationary phase upon glucose depletion (Fig. 3a, top). ATP yields decreased linearly from ¼ ODmax to ¾ ODmax, even though cells were still growing exponentially, albeit at a somewhat reduced rate at ¾ ODmax (Fig. 3a, middle). One may speculate that this decline reflects inefficient and incomplete substrate catabolism as well as complex adaptations for transiting into the stationary phase [Yoshida et al., 2018]. This is accompanied by slightly rising NAD+ yields and sinking intracellular NADH levels (Fig. 3a, bottom). One explanation might be the aforementioned overflow metabolism, which has recently been attributed to higher cellular demand of NAD+ relative to ATP [Luengo et al., 2021].

Phaeobacter inhibens DSM 17395

The growth behaviour of P. inhibens DSM 17395 in standard sea water medium supplied with glucose is characterised by a short exponential growth phase followed by a long linear growth phase (Fig. 3b, top and middle). Growth is possibly limited due to N-limitation and inhibited by produced TDA, which dissipates the proton motive force [Trautwein et al., 2016, 2018; Wilson et al., 2016]. Additionally, the property of P. inhibens DSM 17395 to form aggregates [e.g., Drüppel et al., 2014] may increase metabolic constraints on individual cells in their centre, due to substrate deficiencies (e.g., C-source and oxygen). In accordance, the biomass-specific growth rate and the ATP yields declined in a somewhat exponential manner, implicating multiple possibly intertwined limitations and inhibitions. Likewise, the NAD+ yield declined in a similar manner (Fig. 3b, bottom), which might indicate a nitrogen limitation (among other factors) and constrain the profile recorded for E. coli K12. Such an interpretation would agree with the influence of intracellular nitrogen levels (2-oxoglutarate/glutamine ratio) on NAD+ production, involving the glutamine-dependent NAD+ synthetase (NadE), as proposed by Santos et al. [2020].

Comparison to Literature Data

To place the ATP and NAD(H) profiles determined in the present study in a larger context, the literature was searched for corresponding reports on bacteria. Intracellular concentrations of ATP or NAD(H) are typically given in the units µM or mM, but when estimations for the biovolume of a cell are not available, cellular, biomass and/or protein yields are also eligible to compare the energy states between different strains and growth phases [e.g., Neuhard and Nygaard, 1987; Lasko and Wang, 1996; Tsaplina et al., 2007; Hammes et al., 2010]. For E. coli, a broad range of intracellular ATP concentrations has been reported, ranging from ∼0.5 mM to ∼9 mM, depending on media composition, growth substrates and growth phase [e.g., Neuhard and Nygaard, 1987; Lasko and Wang, 1996; Yaginuma et al., 2014; Deng et al., 2021]. The same holds true for NAD+ and NADH yields ranging from ∼0.3 to ∼4 µmol NAD+ (g cellsdry)−1 or from ∼0.03 to ∼1.25 µmol NADH (g cellsdry)−1 [London and Knight, 1966; Sanchez et al., 2005; Vemuri et al., 2006]. London and Knight [1966] also report an extremely wide range of NAD+ yields in various bacterial strains upon growth under anoxic versus oxic conditions, ranging from 0.19 µmol NAD+ (g cellsdry)−1 in Pseudomonas fluorescens to 10.59 µmol NAD+ (g cellsdry)−1 in Streptococcus faecalis. The present study revealed ATP and NAD(H) yields at ¼ ODmax (per g cellsdry, cell or mg protein) and their concentrations (Table 2) to fall well into the ranges of the aforementioned literature data, e.g., 4.68 and 3.28 µmol ATP (g cellsdry)−1 for E. coli K12 and P. inhibens DSM 17395, respectively. Notably, measured NADH values for P. inhibens DSM 17395 were quite low and near the detection limit, in contrast to E. coli K12.

Statement of Ethics

Ethical approval was not required since the study involved environmental bacteria.

Conflict of Interest Statement

The authors have no conflicts of interest to declare.

Funding Sources

This study was supported by the Deutsche Forschungsgemeinschaft (award SFB TRR 51).

Author Contributions

D.W. and R.R. conceived the study; A.W., D.W., K.K. and S.S. conducted the cultivation experiments; D.W. and S.S. established and conducted luciferase-based assays; S.S. determined substrate depletion profiles; D.W. analysed the data; R.R. wrote the manuscript with contributions from D.W. All authors have agreed to the final version of the manuscript.

Data Availability Statement

All data generated or analysed during this study are included in this article and its online supplementary material. Further enquiries can be directed to the corresponding author.

This article is licensed under the Creative Commons Attribution-NonCommercial 4.0 International License (CC BY-NC). Usage and distribution for commercial purposes requires written permission. Drug Dosage: The authors and the publisher have exerted every effort to ensure that drug selection and dosage set forth in this text are in accord with current recommendations and practice at the time of publication. However, in view of ongoing research, changes in government regulations, and the constant flow of information relating to drug therapy and drug reactions, the reader is urged to check the package insert for each drug for any changes in indications and dosage and for added warnings and precautions. This is particularly important when the recommended agent is a new and/or infrequently employed drug. Disclaimer: The statements, opinions and data contained in this publication are solely those of the individual authors and contributors and not of the publishers and the editor(s). The appearance of advertisements or/and product references in the publication is not a warranty, endorsement, or approval of the products or services advertised or of their effectiveness, quality or safety. The publisher and the editor(s) disclaim responsibility for any injury to persons or property resulting from any ideas, methods, instructions or products referred to in the content or advertisements.