Simultaneous Detection and Quantification of Adenine Nucleotides in Mammalian Cultured Cells by HPLC

TL;DR

This paper introduces a new HPLC method to accurately measure adenine nucleotides in mammalian cells, which is important for understanding cellular metabolism and signaling.

Contribution

A novel RP-HPLC method with UV–vis detection for simultaneous quantification of multiple adenine nucleotides in cultured cells.

Findings

The developed RP-HPLC method is sensitive and capable of detecting physiological concentrations of adenine nucleotides.

The method successfully detects changes in ATP, AMP, and NAD+ levels caused by pharmacological treatments.

The approach provides a reliable and accessible alternative for quantifying adenine nucleotides in biological samples.

Abstract

Adenine nucleotides,
including ATP, ADP, ADP-ribose, AMP, NAD+, and NADH, play
central roles in cellular homeostasis and
are involved in multiple metabolic and signaling pathways. Owing to
their broad functional relevance in cell biology, the accurate quantification
of these metabolites is essential for diverse research areas such
as bioenergetics, cell signaling, and cancer biology. Several analytical
methods have been described for the measurement of adenine nucleotides,
ranging from enzymatic assays to mass spectrometry-based approaches.
In this study, we developed a reverse-phase high-performance liquid
chromatography (RP-HPLC) method with UV–vis detection that
enables the simultaneous quantification of ATP, ADP, ADP-ribose, AMP,
NAD+, and NADH. This method is simple, sensitive within
the physiological concentration range of all analytes, and capable
of detecting biologically relevant changes in ATP, AMP, and NAD+ levels induced by pharmacological treatments. Therefore,
it presents an accessible and reliable alternative for the quantification
of these nucleotides in biological samples.

Figures

Chemical structure of the six adenine nucleotides analyzed here.

Schematic representation of the experimental workflow.

Separation profile of standards and linearity of the standard curves. (A) Chromatogram of standard analytes at 5 μg/mL concentration. ATP, ADP, ADPr, AMP, and NAD+ were run together, presented as a “Mix” curve, and NADH was run separately (red). NAD+ formed due to NADH degradation is indicated by the green arrow. (B) Standard curves for each analyte obtained in the conditions shown in (A). Equations are represented as y = ax + b, and both axes for linear regressions are represented on a logarithmic scale.

Simultaneous quantification of six adenine nucleotides in biological samples. (A) representative chromatograms for each treatment, where NT: nontreated control; DMSO: 0.01% DMSO for 24 h; FK866:10 nM FK866 for 24 h; FCCP: 20 μM FCCP for 2h; and O2D: 1 μM oligomycin +20 mM 2-deoxyglucose for 2h 30 min. (B–G) concentration of each adenine nucleotide, normalized by protein mass in the extracts (μM/ μg of total protein). Results are presented as the mean ± standard deviation of four independent biological replicates, performed in duplicate. Different biological replicates are represented as different geometrical forms, as indicated at the bottom of the figure.

Ratios of adenine nucleotides calculated from the data in Figure (B–G). (A–C) Usual ATP/ADP, ATP/AMP, and NAD+/NADH ratios. (D) Adenylate energy charge (AEC), calculated as (ATP + 0.5 ADP)/(ATP + ADP + AMP). Results are presented as mean ± standard deviation of four independent biological replicates, performed in duplicate. Different biological replicates are represented as different geometrical forms, indicated at the bottom of the figure.

Tables

1. Retention Time, Limit of Quantification (LOQ), and Estimation of Loss Due to the Sample Preparation Procedure for Each Analyte

	ATP	ADP	ADPr	AMP	NAD+	NADH
retention time (min)	7.5	8.4	9.1	10.7	11.5	13.7
LOQ (nM)	197	234	179	288	151	150
loss rate (%)	15.25	16.45	20.25	–18.70	15.29	30.27

Full text

Introduction

Adenine nucleotides play essential roles in biological systems, ranging from energy metabolism to cellular signaling. Adenosine triphosphate (ATP), through coupled reactions, is the primary driver of thermodynamically unfavorable processes in living systems and is therefore widely regarded as the “energy currency” of the cell. ?,? In addition to its bioenergetic function, ATP serves as the main phosphate donor in kinase-catalyzed phosphorylation reactions that regulate numerous signaling pathways. Both ATP and adenosine diphosphate (ADP) can also function as signaling molecules, notably in purinergic signaling and related pathways.? Adenosine monophosphate (AMP), generated from ADP hydrolysis, accumulates under conditions of high energy demand and acts as a central regulator of cellular energy sensing through activation of the AMP-activated protein kinase (AMPK) pathway.? AMPK, in turn, modulates a wide range of downstream processes, including glucose and lipid metabolism, cell growth, autophagy, and other cellular pathways.?

Oxidized nicotinamide adenine dinucleotide (NAD^+^) plays a pivotal role in cellular redox reactions, acting as a coenzyme in key metabolic pathways such as glycolysis, fatty acid oxidation, and the citric acid cycle. ?,? The intracellular ratio of NAD^+^ to its reduced form, NADH, is an important indicator of the cellular redox state and a critical regulator of multiple cellular responses. NAD­(P)H also represents a major source of reducing equivalents for antioxidant systems, including the glutathione peroxidase cycle, which relies on NADP^+^, generated by phosphorylation of NAD^+^,? to regenerate enzyme activity.?

Beyond their canonical bioenergetic functions, adenine nucleotides are also critical for genomic stability. In response to DNA damage, poly­(ADP-ribose) polymerase 1 (PARP1) utilizes NAD^+^ as a substrate to catalyze the formation of poly­(ADP-ribose) (PAR) chains, in which ADP-ribose (ADPr) moieties are covalently attached to target proteins, concomitant with the release of nicotinamide.? PARP1 activity is strongly induced by DNA damage, particularly single-strand breaks, leading to PARylation of histones surrounding the damage site and facilitating the recruitment of the DNA repair machinery. ?,? Following lesion resolution, PAR chains are degraded by hydrolases such as poly­(ADP-ribose) glycohydrolase (PARG), generating free ADPr monomers that can be recycled into NAD^+^ by the NAD salvage pathway. ?,?

Given their widespread and fundamental roles, the quantification of ATP, ADP, and AMP in biological samples has been pursued since the early 1950s, primarily using enzymatic and chromatographic approaches. Early methods relied on ion-exchange chromatography with borate complexes, requiring extensive sample processing to separate the nucleotides prior to analysis. ?,? In 1954, the first luminescence-based assay for ATP quantification was introduced, shortly after the discovery that ATP stimulates light emission in firefly lantern extracts.? This approach was later refined through the use of purified luciferin/luciferase systems and enzymatic conversion of ADP and AMP to ATP, forming the basis of most commercial luciferase-based ATP quantification kits available today.?

More recently, analytical techniques such as reverse-phase high-performance liquid chromatography (RP-HPLC) and liquid chromatography tandem mass spectrometry (LC-MS/MS) have been developed to enable the simultaneous quantification of ATP, ADP, and AMP in biological samples.? Although LC-MS/MS offers superior sensitivity and lower limits of quantification (LOQ), its use can be limited by challenges in chromatographic separation using MS-compatible mobile phases and by degradation of ATP and ADP to AMP in the ionization source, which can compromise accurate quantification of the latter,? despite successful implementation reported in the literature.?

The simultaneous determination of NAD^+^ and NADH also presents analytical challenges due to their distinct stability profiles across different pHs. NAD^+^ is more stable under acidic conditions, whereas NADH is stabilized at alkaline pH, likely owing to proton-catalyzed autoxidation of the reduced form.? Consequently, rapid metabolic quenching during sample preparation is critical for accurate measurements.? Nevertheless, NAD^+^ and NADH have been successfully quantified using colorimetric assays,? fluorescence spectrophotometry,? and by LC-MS/MS-based methods.?

In contrast, analytical methods for the quantification of free ADPr remain limited, as studies have focused on its polymerized form, PAR. In biological samples, PAR is predominantly measured using antibody-based approaches.? Measurements of free ADPr are comparatively scarce and have largely been restricted to in vitro assays of PARP1 activity using HPLC? or to LC-MS/MS-based quantification alongside NAD^+^ following acidic extraction from yeast? or mammalian cells.?

In the present study, we developed a method for the simultaneous detection of six metabolically relevant adenine nucleotidesATP, ADP, AMP, ADPr, NAD^+^, and NADH (Figure)in a single biological sample using RP-HPLC coupled to UV–vis detection. This approach reduces the number of experimental replicates required, significantly shortens sample preparation and analysis times, and improves overall experimental sustainability and usability. Moreover, HPLC represents a more cost-effective and accessible alternative to LC-MS/MS. Importantly, the method developed here achieves limits of detection compatible with physiological concentration ranges of all six metabolites in cultured human cells. Validation experiments using pharmacological agents known to alter concentrations of ATP, AMP, and NAD+ levels further demonstrated that this method reliably detects biologically meaningful variations in adenine nucleotide concentrations.

Chemical structure of the six adenine nucleotides analyzed here.

Materials and Methods

Cell Culture

and Treatments

MRC-5 normal human pulmonary fibroblasts were maintained in DMEM high glucose containing 10% fetal bovine serum +1% penicillin/streptomycin (complete medium) at 37 °C under a 5% CO_2_ atmosphere. Cultures were routinely subcultured using 0.5% trypsin when 80% confluence was reached. For the treatments, cells were seeded in 100 mm plates and incubated with 20 μM carbonyl cyanide p-trifluoromethoxy phenylhydrazone (FCCP) for 60 min; 20 mM 2-deoxyglucose (2-deoxyGlu) + 1 μM oligomycin for 2.5 h; or 10 nM FK866 for 24 h. All treatments were carried out in complete medium, while control cells received 0.01% DMSO as a vehicle control.

Nucleotide Extraction

Cells grown on 100 mm dishes were treated as described above and carefully washed with phosphate-buffered saline (PBS). After PBS removal, 1 mL of HPLC-grade methanol (Merck) containing 1 mM deferoxamine mesylate (Sigma-Aldrich) was added to the plate and homogenized with a cell scraper. Cell extracts were then transferred to a 2 mL microtube and incubated for two h at 25 °C with a 1300 rpm agitation in a Thermomixer C (Eppendorf). Subsequently, samples were centrifuged at 20,000g for 10 min at 4 °C. The supernatant, which contains all nucleotides, was stored at −80 °C. The samples were dried at 20 °C in a SpeedVac and then stored at −80 °C until the analysis. The pellet was stored at −20 °C for protein quantification. Figure depicts the experimental workflow for the extraction and analysis of samples.

Schematic representation of the experimental workflow.

Protein Purification and Quantification

The pellets obtained after separation of the supernatant for nucleotide quantification were used for protein determination. The pellets were suspended in 250 μL of freshly prepared extraction buffer (100 mM ammonium bicarbonate and 8 M urea) and mixed in a Thermomixer at 25 °C until completely dissolved. When pellet solubilization was not complete, samples were mixed at 1000 rpm at 55 °C in a Thermomixer for up to 1 h. Subsequently, 1 mL of ice-cold acetone was added to the samples, and the samples were vortexed thoroughly. The samples were kept at −20 °C overnight for protein precipitation, followed by centrifugation at 13,000g for 10 min at 4 °C. The supernatant was completely removed, and samples were air-dried in a fume hood. The dry pellets were resuspended in 200 μL of extraction buffer and used for protein quantification. Before quantification, samples were diluted so that the final concentration of urea did not exceed 3 M, to avoid interfering with the protein quantification methods. We used Bio-Rad Protein Assay Dye Reagent Concentrate (Bio-Rad) according to the manufacturer’s instructions, and bovine γ globulin (Bio-Rad) as the standard.

RP-HPLC Analysis

All analytes were analyzed in a single run using a Luna C18(2) 5 μm 100 Å 250 mm × 4.6 mm column (Phenomenex) with guard column AQ C18 4 mm × 3 mm ID (Phenomenex) in an SCL-40 HPLC System (Shimadzu). Two mobile phases were utilized: phase A being 25 mM diammonium hydrogen phosphate ((NH_4_)2_HPO_4) at pH 6,0, adjusted using 85% phosphoric acid, and phase B being HPLC-grade methanol. Column oven temperature was 25 °C, and the flow was 0,8 mL/min. Samples were separated using the following gradient: 0–10 min, 1% B; 10–15 min, 1%–20% B; 15–20 min, 20% B; 20–22 min, 20%–40% B; 22–27 min, 40% B; 27–30 min, 40%–1% B; and 30–40 min, 1% B. Analytes were monitored at 254 nm. It should be noted that the analytes of interest elute before the 20 min running time, but the 40 min run ensures that the column is properly cleaned before the next sample is injected (Supporting Figure S1).

Dried samples prepared as above were resuspended in 100 μL of phase A, agitated at 1000 rpm for 10 min at 37 °C, and centrifuged at room temperature for 10 min at 10,000g. The supernatant was transferred to a new tube, and 45 μL was injected per analysis. The same injection volume was also used to prepare the standard curves, which consisted of solutions in the 0.1 to 200 μg/mL concentration range for each analyte. The concentrations used for each analyte, in μM, are presented in Supporting Table 7. We note that to ensure the reproducibility of the standard curves, all standards must be freshly prepared immediately before the experiment, as we observed that ATP and NADH, shown in Supporting Figure S2, are rapidly degraded when stored at −20 °C. All reagents used here were acquired from Sigma-Aldrich.

Recovery Assay

for Estimation of Analyte Loss during the Experimental Procedure and Limit of Quantification Estimations

To test how the extraction procedure affected analyte recovery and estimate the loss of each metabolite during the nucleotide extraction process, samples containing all analytical standards were submitted to similar processing as the biological samples, and the peak areas obtained in RP-HPLC analysis were then compared to the area of control samples, which were not submitted to extraction. For this test, ATP, ADP, ADP-ribose, and AMP were tested together to account for the possibility of nucleotide hydrolysis into AMP during the extraction process. NAD^+^ and NADH were evaluated separately. Loss rates were calculated as the percentage of peak area lost after extraction and used as correction factors in the calculation of the concentration for the biological samples. The limit of quantification (LOQ) of each analyte was determined by visual inspection of the integrable peaks, with the lowest concentration with an integrable peak being considered the LOQ. The LOQ was estimated based on at least 3 independent experiments.

Statistical Analysis

Linear regressions were obtained using GraphPad Prism 10 and Origin 2024. To evaluate statistical significance, one-way ANOVA with a Tukey posthoc test for each metabolite was performed using GraphPad Prism 10.

Results and Discussion

Sample Preparation Procedure

Several methodologies have been reported for the preparation of samples for adenine nucleotide quantification, employing distinct strategies. ?−? ?,?,? A fundamental aspect shared among these approaches is metabolic quenching, which is essential to preserve nucleotide concentrations as close as possible to their intracellular levels. Acidification is a commonly used quenching strategy, as it rapidly inactivates ATPases and thereby preserves ATP levels. ?,? However, this approach is incompatible with experimental designs that require the simultaneous determination of NAD^+^ and NADH because NADH is spontaneously oxidized to NAD^+^ under acidic conditions,? leading to distortion of their relative concentration.

To overcome this limitation, metabolic quenching can alternatively be achieved using organic solvents, such as methanol or combinations of solvents, ?,? which was the strategy adopted in the present study. To preserve the cellular NAD^+^/NADH ratio during extraction, 100 mM deferoxamine mesylate, an iron chelating agent, was added to 100% methanol. This combination was found to effectively minimize metabolite loss during sample preparation (data not shown).

RP-HPLC Provides

a Complete Separation of Six Metabolites and Low LOQ

Many biologically relevant processes lead to changes in absolute concentrations or relative ratios of adenine nucleotides, underscoring the importance of accurately quantifying these species. Ideally, all analytes should be measured in the same samples to minimize artifacts arising from the sample handling and processing. Accordingly, we developed an HPLC-based protocol enabling the simultaneous determination of six adenine nucleotides with central roles in mammalian energy metabolism: ATP, ADP, AMP, NAD+, NADH, and ADP-ribose. As shown in Figure, the chromatography conditions established herein provided adequate resolution of all six analytes.

Separation profile of standards and linearity of the standard curves. (A) Chromatogram of standard analytes at 5 μg/mL concentration. ATP, ADP, ADPr, AMP, and NAD+ were run together, presented as a “Mix” curve, and NADH was run separately (red). NAD+ formed due to NADH degradation is indicated by the green arrow. (B) Standard curves for each analyte obtained in the conditions shown in (A). Equations are represented as y = ax + b, and both axes for linear regressions are represented on a logarithmic scale.

It should be noted that the standard chromatograms shown in FigureA were obtained by using solutions containing all nucleotides except NADH. NADH standards were analyzed separately to prevent interference with NAD^+^ quantification caused by artifactual oxidation of NADH present in mixed standards. Even under these conditions, a small peak corresponding to the NAD^+^ retention time was detected in the NADH standard chromatogram (Figure, green arrow and red line), illustrating the susceptibility of NADH to oxidation. Figure shows the individual standard curves for all six nucleotides.

Using these standards, we determined both the analyte-specific loss rates associated with experimental handling and the limits of quantification (LOQ) for each compound. ATP, ADP, ADP-ribose, and NAD^+^ exhibited comparable loss rates, with a mean loss of 16.82% (Table). This parameter is particularly relevant for biological samples, as extensive analyte loss during sample preparation may preclude quantification if postextraction concentrations fall below the LOQ. Taken together, the relatively low loss rate and LOQ for these nucleotides indicate that this method is both sensitive and well suited for biological applications.

1: Retention Time, Limit of Quantification (LOQ), and Estimation of Loss Due to the Sample Preparation Procedure for Each Analyte

The LOQ is a critical performance metric for analytical methods applied to biological samples, given that endogenous metabolite concentrations may approach the detection limits. As summarized in Table, the LOQ values obtained with the optimized HPLC method ranged from 150 to 288 nM, which is compatible with the analysis of biological samples, in which most of the metabolites are typically present in the millimolar range.?

In contrast, AMP and NADH displayed loss rates that differed from those of the other nucleotides (Table). In mixtures containing ATP, ADP, ADP-ribose, and AMP, higher than expected levels of AMP were detected after extraction, resulting in a negative apparent loss rate. This increase is likely attributable to spontaneous hydrolysis of ATP and ADP to AMP during sample processing. NADH exhibited a loss rate (30.27%) that was higher than that of the other metabolites, which can be partially explained by its oxidation to NAD^+^. Collectively, these findings highlight the importance of determining loss rates for each individual analyte and experimental protocol and emphasize that loss should not be considered an intrinsic property of the extraction method alone, but rather as a variable dependent on the chemical properties and reactivity of the metabolites analyzed.

Validation Experiments

with Cultured Human Cells Detect Expected Biological Responses

To validate the ability of the present method to detect biologically relevant changes in adenine nucleotide concentrations, MRC-5 cells were subjected to experimental conditions known to modulate the levels of the analytes of interest. Cells were treated with the nicotinamide phosphoribosyl transferase (NAMPT) inhibitor FK866,? which impairs the NAD^+^ salvage pathway and thereby reduces intracellular NAD^+^ levels. A second condition involved treatment with the oxidative phosphorylation uncoupler FCCP, which stimulates mitochondrial respiration and is expected to decrease NADH levels. A third condition consisted of combined treatment with oligomycin, an inhibitor of the ATP synthase, and 2-deoxyglucose, a potent inhibitor of glycolysis; together, these agents are expected to decrease ATP levels and concomitantly increase AMP concentration. Samples were processed as described above, and representative chromatograms of the treated cells are shown in FigureA. Visual inspection of peak intensities revealed clear treatment-dependent changes in metabolite levels, which are quantitatively presented in FigureB–G.

Simultaneous quantification of six adenine nucleotides in biological samples. (A) representative chromatograms for each treatment, where NT: nontreated control; DMSO: 0.01% DMSO for 24 h; FK866:10 nM FK866 for 24 h; FCCP: 20 μM FCCP for 2h; and O2D: 1 μM oligomycin +20 mM 2-deoxyglucose for 2h 30 min. (B–G) concentration of each adenine nucleotide, normalized by protein mass in the extracts (μM/ μg of total protein). Results are presented as the mean ± standard deviation of four independent biological replicates, performed in duplicate. Different biological replicates are represented as different geometrical forms, as indicated at the bottom of the figure.

As anticipated, FK866 treatment resulted in a significant reduction in NAD^+^ concentration (FigureF; p < 0.05 relative to both control conditions), accompanied by a modest, nonsignificant decrease in NADH levels (FigureG). Despite these changes, the NAD^+^/NADH ratio was not significantly altered following FK866 treatment (FigureJ), due to the concomitant decrease in both metabolites. These findings underscore the importance of quantifying the absolute concentration of individual nucleotides rather than relying solely on metabolite ratios.

The results obtained following combined oligomycin and 2-deoxyglucose (O2D) treatment further support the notion that ratios may fail to capture biologically meaningful alterations. Although statistically significant changes in ATP (FigureB) and AMP (FigureE) concentrations were observed, neither the ATP/ADP nor the ATP/AMP ratios differed significantly between the treated and control samples (FigureA,B). In the case of the ATP/AMP ratio (FigureB), relatively high sample variability likely contributed to the lack of statistical significance.

Ratios of adenine nucleotides calculated from the data in Figure (B–G). (A–C) Usual ATP/ADP, ATP/AMP, and NAD+/NADH ratios. (D) Adenylate energy charge (AEC), calculated as (ATP + 0.5 ADP)/(ATP + ADP + AMP). Results are presented as mean ± standard deviation of four independent biological replicates, performed in duplicate. Different biological replicates are represented as different geometrical forms, indicated at the bottom of the figure.

The adenylate energy charge (AEC), a well-established indicator of the cellular metabolic state and ATP-dependent processes, was also calculated for all conditions. AEC values range from 0 to 1,? and were determined using the equation (ATP

• 0.5 ADP)/(ATP + ADP + AMP). Across the conditions analyzed, AEC remained largely stable (FigureD), with statistically significant differences observed between the O2D-treated sample and all other conditions.

Conclusions

Given the central role of adenine nucleotides in biological processes, robust analytical methods are required for their accurate quantification in biological samples. Nonetheless, the applicability of several existing approaches is constrained by the high salt concentrations employed during sample preparation as well as the need. For specialized instrumentation, such as mass spectrometry, it is not readily available to many cell biology laboratories. In this context, we developed an improved method based on HPLC with a diode array detector (HPLC-DAD) that enables the simultaneous analysis of six adenine nucleotides in a single chromatographic run. The method relies on a simple and rapid sample preparation procedure that allows samples to be dried and stored, thereby decoupling biological experimentation from chromatographic analyses and permitting parallel processing, which enhances the reproducibility. Furthermore, the use of organic solvent extraction in combination with deferoxamine minimizes the artifactual degradation of analytes, thereby improving the accuracy of quantification. Simultaneous determination of all six adenine nucleotides within the same sample also contributes to reduced intersample variability and increased analytical precision.

We further demonstrated that this method is sufficiently sensitive to detect biologically relevant changes in nucleotide levels in cultured mammalian cells with average intracellular concentrations well above the limit of detection for each analyte. Experiments employing established pharmacological interventions targeting cellular bioenergetics underscore the importance of measuring absolute nucleotide concentrations, as biologically meaningful effects may not be apparent when only metabolite ratios are considered. Notably, although FK866 treatment did not significantly alter the NAD^+^/NADH ratio, absolute quantification revealed a concomitant decrease in both NAD^+^ and NADH levels, an effect that would have remained undetected if only ratios had been assessed.

Taken together, the simplicity, adaptability, and sensitivity of the sample preparation and chromatography procedures suggest that this method represents an accessible and reliable alternative to LC-MS/MS-based approaches for the quantitation of adenine nucleotides in biological samples of diverse origins.

Supplementary Material