Measuring NAD+ by HPLC and Mass Spectrometry: What the Analytical Literature Reports

Quantifying NAD+ is described in the method literature as an analytical problem rather than a routine assay. A review of NAD+ metabolome ("NADome") measurement notes that most pyridine nucleotides auto-oxidise and show differential chemical stability across biological matrices, so that reported concentrations depend heavily on extraction procedure, choice of internal standard, and analyte stability during handling [1]. The same review reports that specialised HPLC-UV, NMR, capillary zone electrophoresis, and colorimetric enzymatic assays are inexpensive and widely available but were judged to lack the specificity and sensitivity required for quantification in human biological samples, with liquid chromatography–mass spectrometry presented as the alternative for blood, cerebrospinal fluid, and urine [1]. Separation chemistry is the main point of divergence between published methods, because NAD+ and its phosphorylated relatives are highly polar and are poorly retained on conventional reversed-phase stationary phases without ion-pairing reagents [2]. One tandem LC-MS/MS approach separated metabolites on an amino column using a dual HILIC–reversed-phase gradient with heated electrospray ionisation in mixed-polarity multiple reaction monitoring mode, and quantified 17 metabolites of the NAD+ metabolome in U251 human astroglioma cell extracts in vitro, with additional measurements in murine oocytes under different culture conditions [3]. A zwitterionic HILIC tandem mass spectrometry method applied isotope dilution to compensate for loss of these labile metabolites and was validated across low, medium, and high biomass samples of two in vitro model systems, Escherichia coli and the human JJN-3 cell line, with sample preparation that avoided solvent removal [2]. A HILIC-mode UHPLC-MS method quantified NAD, NADH, NADP, NADPH, FAD, AMP, ADP, and ATP in a single extraction from Saccharomyces cerevisiae cultures in vitro, reporting calibration linearity of r² > 0.98 from 0.1 to 100 µmol and a lowest limit of detection of 0.055 nmol [4]. Ion-pairing reversed-phase methods form a second family. An ion-pair reversed-phase UHPLC-MS/MS method using dibutylammonium acetate as a volatile ion-pairing reagent quantified NAD, NADH, NADP, NADPH, FMN, FAD, and several coenzyme A species in Saccharomyces cerevisiae cell extracts in vitro, using uniformly ¹³C-labelled cell extract as internal standard for isotope dilution mass spectrometry and reporting limits of detection and quantification, repeatability, and intermediate precision [5]. A later ion-pair reversed-phase UHPLC-MS method built on BEH C18 columns modified with hybrid surface technology reported capacity factors above 2, retention time reproducibility within 0.09 min across six injections, tailing factors below 2, and limits of detection under 238 fmol on column for phosphorylated metabolites including NAD-based cofactors, with application to liver tissue, human urine and plasma, cells, and faeces [6]. An extension of that approach quantified 125 phosphorylated analytes in a single run, including NAD and NADP, reporting limits of detection below 0.95 pmol, linearity R² > 0.99, recoveries of 80–120%, and coefficients of variation under 20% for both precision and inter-technician consistency, and was applied to human urine, plasma, cells, and faeces and to rabbit liver tissue [7]. Some methods target NAD+ as one node in a broader redox panel. An integrated LC-MS/MS strategy on a QTRAP platform quantified 23 metabolites spanning NAD+, FAD, GSSG, and ATP pathways, and was applied to plasma from humans, hamsters, and mice, with the authors reporting that hamster profiles resembled human profiles more closely than mouse profiles did [8]. The same work used dithiothreitol and hydrogen peroxide treatment of A549 and HeLa cells in vitro as a method-validation challenge, reporting that dithiothreitol moderately increased and hydrogen peroxide substantially decreased most analytes [8]. Non-mass-spectrometric options remain in use for specific questions. A reverse-phase HPLC method with UV-vis detection was developed for the simultaneous quantification of ATP, ADP, ADP-ribose, AMP, NAD+, and NADH in mammalian cultured cells in vitro, and was reported to detect changes in ATP, AMP, and NAD+ levels induced by pharmacological treatment within the physiological concentration range of the analytes [9]. For the upstream tryptophan–NAD pathway, an HPLC method with fluorescence detection based on enzymatic conversion of quinolinic acid to nicotinic acid mononucleotide and then to a BODIPY-labelled deamido-NAD reported a limit of detection of 5.0 nmol/L, at least 30-fold lower than published HPLC quantitation limits for quinolinic acid, with results in human cerebrospinal fluid samples correlating with a GC/MS method [10]. Separately, an HPLC–electrospray ionisation mass spectrometry method in selected ion monitoring mode measured five urinary acids of the tryptophan–NAD pathway using deuterated picolinic and nicotinic acid internal standards and mixed-mode solid-phase extraction, reporting relatively uniform levels among rats and larger variation among human urine samples [11]. Applications of these platforms illustrate how the analytical choice constrains interpretation. In an in vitro metabolomics comparison of breast cancer and breast epithelial cell lines using UPLC-ESI-Q-TOF and HPLC-ESI-QqQ multiple reaction monitoring, the triple-negative MDA-MB-231 line showed elevated nicotinamide, 1-ribosyl-nicotinamide, and NAD+ relative to the other lines examined [12]. At the human population level, a study using a rigorously validated ultra-high-performance liquid chromatography coupled to high-resolution mass spectrometry system designed to account for real-world analytical variability quantified NAD+ across seven independent human cohorts and reported that whole-blood NAD+ levels remained stable with age and across lifestyle interventions, while changing in response to nicotinamide riboside supplementation; the authors concluded that this challenges the utility of blood NAD+ as a biomarker of ageing or lifestyle factors [13].