Identification and Distribution of the Dietary Antioxidant Ergothioneine in Humans and Animal Models Combining NMR, RANSY, and MS Methods

Abstract

Ergothioneine (ERG), a unique, naturally occurring antioxidant of dietary origin, is gaining increasing attention due to its crucial roles in human health and diseases. Despite its significance, ERG is rarely detected in biospecimens by mass spectrometry (MS) and, to date, had not been characterized by nuclear magnetic resonance (NMR) spectroscopy, two widely used analytical techniques in metabolomics. In this study, we investigated human plasma, whole blood (WB), and red blood cells (RBC), as well as mouse blood and tissues combining NMR, LC-MS and ratio analysis techniques. The results demonstrate the ability of simple 1D ¹H NMR to routinely identify and quantify ERG in various biological samples. The levels of ERG vary widely and depend on the type of biological sample, with human RBC exhibiting remarkably high concentrations, often exceeding 1.5 mM. The average levels of ERG in human plasma, WB, and RBC were in the ratios of 1:70:140, respectively. Moreover, ERG levels showed a linear correlation between plasma and WB (R² = 0.59), plasma and RBC (R² = 0.75), and WB and RBC (R² = 0.98). In mice, ERG levels exhibit a distinct whole-body distribution, with the average levels in mouse skeletal muscle, brain, heart, kidney, and liver in the ratios of 0:1:10:35:45, respectively. The demonstration of surprisingly high levels of ERG in biological samples using ¹H NMR opens new avenues for its analysis using both NMR and MS methods to explore its roles in human health and diseases, as a part of routine global or targeted metabolomics studies.

Graphical Abstract

INTRODUCTION

Ergothioneine (ERG) is a unique, naturally occurring sulfur-containing derivative of the amino acid histidine and is widely found in human and animal biospecimens, various foods, and other species.^1,2 Due to its potent antioxidant and anti-inflammatory properties,³ there is growing interest in understanding its roles in human health and diseases.⁴ Notably, the antioxidant capacity of ERG is considered to be comparable to that of glutathione (GSH), a key cellular antioxidant in humans.⁵ Human blood levels of ERG have been shown to decline with age, and its lower levels are associated with various health conditions,⁶ including cardiovascular disease,^2,7 cognitive impairment,^6,8–11 dementia,^8,12 Parkinson’s disease,¹³ and an increased risk of frailty.^8,14,15 Conversely, higher plasma levels of ERG have been linked to a reduced risk of mortality.²

ERG is synthesized by certain fungi and bacteria, with edible fungi such as mushrooms serving as particularly rich sources of this compound.^4,16–20 Although the human body cannot synthesize ERG, it is absorbed from the diet, distributed to various organs, and, interestingly, retained in the body for several months.³ Although ERG was first isolated from a fungus more than a century ago,²¹ research focused on this compound gained momentum only after the discovery of its specific transporter in 2005.²² Since then, studies have increasingly focused on its physiological roles in human health, disease prevention, and treatment.^4,23 It is now classified as a ‘longevity vitamin’²⁴ and has been deemed safe as a dietary supplement by both European^25,26 and United States²⁷ food safety agencies. Consequently, numerous methods for its large-scale production have been developed.²⁸

Despite the growing need for ERG measurement and significant advances in analytical methods in the last more than two decades, the literature on the analysis of the ubiquitous metabolite remains limited in the context of metabolomics studies. Notably, ERG is rarely detected by mass spectrometry (MS)^7–15 and, to date, had not been well characterized by NMR spectroscopy, which are the two primary analytical techniques in metabolomics. This highlights the persisting challenges in identifying unknown metabolites in biological samples. To address these challenges, the present study builds on our ongoing efforts to identify unknown metabolites and develop simple, reliable methods for their analysis in metabolomics applications.^29–34 We conducted a comprehensive investigation of human plasma, whole blood (WB), red blood cells (RBC), and mouse tissues, including heart, liver, kidney, brain, and skeletal muscle, using a combination of 1D and 2D NMR spectroscopy, MS techniques, and ratio analysis. Here, we demonstrate the detection, identification, and absolute quantitation of ERG in various biological specimens using a simple 1D ¹H NMR method. Our findings reveal a distinctive distribution pattern of ERG in both human and animal models. The study demonstrates that ERG is accessible for routine analysis using both NMR and MS techniques as part of metabolomics studies, facilitating investigations of its important roles in human health and diseases.

MATERIALS AND METHODS

Chemicals and solvents:

Methanol, chloroform, sodium phosphate monobasic (NaH₂PO₄), sodium phosphate dibasic (Na₂HPO₄), fumaric acid, maleic acid, and 3-(trimethylsilyl)propionic acid-2,2,3,3-d₄ sodium salt (TSP) were obtained from Sigma-Aldrich (St. Louis, MO). Deuterium oxide (D₂O) was obtained from Cambridge Isotope Laboratories, Inc. (Andover, MA). Deionized (DI) water was purified using an in-house Synergy Ultrapure Water System from Millipore (Billerica, MA). All chemicals were used with no further purification.

Preparation of phosphate buffer:

Buffer solution (100 mM; pH 7.4) was prepared by dissolving 1124 mg anhydrous disodium hydrogen phosphate (Na₂HPO₄) and 250 mg anhydrous monosodium phosphate (NaH₂PO₄) in 100 g D₂O. TSP (50 or 124 μM). Fumaric acid (227.9 μM) or maleic acid (255.5 μM) were added as internal references for the quantitation of metabolites in biological mixtures.³⁵ The calculated pH of the buffer solution was 7.4 and the measured pH was 7.33. This buffer was used without further pH correction.

Human plasma, whole blood (WB), and red blood cells (RBC):

Human blood specimens from healthy individuals were collected in heparinized BD Vacutainer tubes (BioVision, CA). The biospecimen collection protocol was approved by the University of Washington IRB. Blood (35–40 mL) was collected from 25 overnight fasted healthy volunteers (15 male, 10 female; age range 21–63 years). None of the volunteers took ERG supplements. A portion of the blood was used for hematocrit measurements, and the remaining portion was aliquoted (200 μL or 1200 μL) into Eppendorf tubes (2 mL) as depicted in Figure S1. Briefly, the 1200 μL aliquots were placed on ice or the bench, at room temperature, and one aliquot from each volunteer was centrifuged at 2000×g and 4 °C for 15 min after different delay times varying from 0 hrs. to 144 hrs., and RBC and the supernatant plasma (both 200 μL aliquots) were transferred to fresh vials. The aliquots of WB (200 μL) were mixed with an 800 μL mixture of methanol and chloroform (1:1 v/v ratio), and the aliquots of RBC (200 μL) were mixed with a 900 μL mixture of water, methanol, and chloroform (0.5:2:2 v/v/v ratio). The aliquots of plasma, WB, and RBC were then stored at −80 °C until used for analysis.

Mouse tissue specimens:

Mouse tissue specimens were obtained with the approval of the Institutional Animal Care and Use Committee at the University of Washington. A total of 22 C57BL/6J mice, aged 3.5 to 6 months (Table S1), were used for identification and quantitation of the unknown metabolite. After anesthetizing each mouse, tissue specimens from the heart, kidney, brain, liver, and skeletal muscle were collected, snap frozen in liquid nitrogen, and stored at −80 °C until used for analysis.

Hematocrit measurements:

Approximately 70 μL human WB from each volunteer was loaded into open-ended Fisherbrand Hematocrit Capillary Tubes (Thermo Fisher Scientific, Catalog No. 22–362574; Cincinnati, OH). One end of the tube was sealed using Fisherbrand Hemato-Seal Capillary Tube Sealant (Catalog No. 02–678). The sealed tubes were then centrifuged in a DSC-100MH-2 Hematocrit Centrifuge (OpticsPlanet, Inc., Northbrook, IL) at 14,837×g for 5 mins. The length of the packed red blood cell column and the total sample column length were obtained using a Unico Microhematocrit CMH30-Reader (OpticsPlanet, Inc., Northbrook, IL). Hematocrit was calculated as the ratio of the packed red blood cell column length to the total sample column length, expressed as a percentage. All measurements were performed in triplicate for each sample to ensure precision and accuracy.

Aqueous metabolites extraction from plasma, WB, and RBC:

Frozen plasma samples were mixed with methanol (1:2 v/v/), vortexed for 30 s and stored at 20 °C for 20 min. The mixtures were then centrifuged at 13,400×g for 40 min to pellet proteins. The clear solutions were transferred to fresh vials and dried under a stream of nitrogen gas. The dried samples were mixed with 200 μL phosphate buffer in D₂O containing maleic acid and TSP and transferred to 3 mm NMR tubes for analysis.

Frozen WB and RBCs aliquots, which had been previously treated with organic solvents and stored at −80 °C, were vortexed for 2 min or until a homogenous mixture was achieved. The samples were then sonicated for 20 min at 4 °C and vortexed again for 30 s. The mixtures were then centrifuged at 13,400×g for 40 min to pellet proteins and cell debris. The clear supernatants were transferred to fresh vials and dried under nitrogen gas. The dried samples were mixed with 200 μL phosphate buffer in D₂O containing fumaric acid and TSP and transferred to 3 mm NMR tubes for analysis.

Aqueous metabolites extraction in mouse tissue:

Frozen tissue specimens (~80 to 100 mg) were mixed with a solution of deionized water and methanol (200 μL; 1:5 v/v, 4 °C) using lockable Eppendorf vials and then homogenized using disposable tissue pestles. A further 800 μL of cold water and methanol (1:5 v/v) was added, and each mixture was then vortexed and incubated on dry ice (−75 °C) for 30 min. Subsequently, the mixtures were sonicated in an ice bath for 10 min and centrifuged at 2000×g for 5 min at 4 °C. The soluble extracts were separated, frozen using dry ice, and lyophilized to dryness. The dried extracts were then mixed separately with a cold phosphate buffer (100 mM; pH = 7.4; 4 °C) in D₂O containing 50 μM TSP (210 μL for 3 mm NMR tube, 600 μL for 5 mm NMR tube) and transferred to NMR tubes for analysis.

NMR Spectroscopy:

NMR experiments were performed at 298 K on a Bruker Avance III 800 MHz spectrometer equipped with a cryogenically cooled probe and Z-gradients suitable for inverse detection. The NOESY pulse sequence with water suppression or the CPMG (Carr-Purcell-Meiboom-Gill) pulse sequence with residual water suppression using presaturation were used for ¹H 1D NMR experiments. Spectra were obtained using a 9615 Hz spectral width, 32,768 time-domain points, and 5 s recycle delay. For the CPMG experiment, a pulse train length of 128 or 256 ms was used. Separately, to confirm the newly identified metabolite peaks, ¹H spectra were obtained after the addition of a stock solution of ERG (1 to 10 μL; 1.0 or 50 mM). The raw data were Fourier transformed using a spectral size of 32,768 points after multiplying by an exponential window function with a line broadening of 0.5 Hz. Separately, ¹H 1D NMR spectra were obtained in duplicate for human plasma solutions containing ergothioneine at concentrations ranging from 0.1 μM to 100 μM (specifically, 0.1, 0.5, 1.0, 3.0, 10.0, 50.0, and 100.0 μM). The CPMG pulse sequence and parameters used for plasma, whole blood (WB), and red blood cell (RBC) spectra were used. Based on the relative peak area of ergothioneine, the limit of detection (LOD) and the limit of quantitation (LOQ) were determined.

To establish the identity of the unknown metabolite, ¹H-¹H homonuclear and ¹H-¹³C heteronuclear two-dimensional (2D) experiments were performed for a few representative plasma, WB, and RBC samples. 2D ¹H-¹H correlation spectroscopy (COSY) using the pulse sequence ‘cosygpqf’ and 2D ¹H-¹H total correlation spectroscopy (TOCSY) using the pulse sequence ‘mlevphpr’ were performed with or without suppression of the residual water signal by presaturation during the relaxation delay. A sweep width of 9615 Hz was used in both dimensions; 512 FIDs were obtained with t₁ increments, each with 2048 complex data points. The number of transients used was 16 for COSY and 40 for TOCSY and the relaxation delay was 1.0 s. The resulting 2D data were zero-filled to 4096 (for COSY) or 2048 ( for TOCSY) points in the t₂ dimension and 1024 points in the t₁ dimension. A 45° (for COSY) or 90° (for TOCSY) shifted squared sine-bell window function was applied to both dimensions before Fourier transformation.

2D ¹H-¹³C heteronuclear single quantum coherence (HSQC), and 2D ¹H-¹³C heteronuclear multiple bond correlation (HMBC) experiments were performed using the ‘hsqcetgpsisp2.2’ and ‘hmbcgplpndprdf’ pulse sequences, respectively. Spectral widths of 9615 Hz (¹H) and 50310 Hz (¹³C) for HSQC and 8012 Hz (¹H) and 50309 Hz (¹³C) for HMBC were used. FIDs were obtained with 256 or 400 t₁ increments, each with 2048 or 4096 complex data points for HSQC and HMBC, respectively. The number of transients used was 64 or 196 and the relaxation delay was 1.0 s. The obtained 2D data were zero-filled to 4096 and 1024 points in t₂ and t₁ dimensions; a 45° shifted squared sine-bell window function was applied to both dimensions before Fourier transformation.

All spectra were phase and baseline corrected and the chemical shift scales were calibrated based on the TSP signal for ¹H or ¹³C for both 1D and 2D spectra. Additionally, plasma, WB, and RBC ¹H NMR spectra obtained as part of our previous studies^36,37 were used to compare the levels of the identified metabolite. Bruker Topspin versions 4.1.4 and 3.6.5 software packages were used for NMR data acquisition, processing and analyses.

Liquid chromatography-mass spectrometry (LC-MS):

Extracts of RBC samples were subjected to global, high-resolution MS analysis as described previously.^38,39 Briefly, LC-MS analysis was performed using an Agilent 6546 Q-TOF MS system attached to an Agilent 1290 Infinity II liquid chromatography system. Metabolites were separated using hydrophilic interaction liquid chromatography (HILIC; WATERS XBridge BEH Amide; 15 cm × 2.1 mm; 2.5 μm) using mobile phases A (10 mM ammonium acetate and 0.2% acetic acid in 95% H₂O + 3% acetonitrile + 2% methanol) and B (10 mM ammonium acetate and 0.2% acetic acid in 5% H₂O + 93% acetonitrile + 2% methanol). High resolution mass spectral data were collected in positive electrospray ionization mode. The tentatively identified peak, based on the analysis of NMR spectra, was subjected to MS/MS analysis using different collision energies. The identified metabolite was further confirmed based on LC-MS and LC-MS/MS analysis of the standard compound.

Spectral analysis, metabolite identification, quantitation and data analysis.

Figure S2 shows the major steps followed in the identification of the unknown metabolite and its quantitation. We first used RANSY (Ratio Analysis NMR Spectroscopy)⁴⁰ for selective identification of peaks associated with the unknown metabolite peak at 6.8061 ppm. RANSY generates spectra for an individual metabolite by exploiting the fact that the peak ratios for any metabolite in the NMR spectrum are fixed and proportional to the relative numbers of magnetically distinct protons. When the peak ratios are divided by their coefficient of variation derived from a set of NMR spectra, the generation of an individual metabolite spectrum is enabled. For RANSY, peaks in 1D NMR spectra of the biospecimens were integrated using Bruker AMIX software. The data were then subjected to RANSY using an Excel template that we have developed based on the code reported previously.⁴⁰

Subsequently, identity of the unknown metabolite peaks in the NMR spectra were first tentatively established based on the comprehensive analysis of a combination of 1D and 2D (COSY, TOCSY, HSQC, and HMBC) NMR spectra. It was confirmed by MS and MS/MS analysis of extracts of human RBC. The identified compound was further confirmed by spiking experiments using the authentic compound. Absolute concentrations of the newly identified metabolite (ERG) in various human and animal biospecimens were determined using its NMR peak areas relative to the area of the internal standard (maleic acid or fumaric acid) by taking into consideration the concentration of the internal standard and the number of protons that ERG and internal standard peaks represented. Concentrations of ERG in different biospecimens were compared and correlated. Its levels in human plasma, whole blood, and RBCs were evaluated as a function of hematocrit, age, and gender. The levels of ERG in plasma separated from blood that stayed on ice or on the bench at room temperature for different duration were also evaluated. Microsoft Excel (Microsoft 365) was used to perform statistical linear correlations and to generate the box-and-whisker plots.

RESULTS AND DISCUSSION

¹H NMR spectra of human plasma, whole blood, and RBC as well as mouse specimens, including heart, kidney, brain, and liver, showed a peak at 6.8061 ppm, whose identity had not been established previously. In this study, based on a comprehensive analysis combining NMR, MS and RANSY methods, we identified the peak in the NMR spectra as arising from ergothioneine (ERG). This identification was conducted ab initio, without prior knowledge linking the observed NMR signals to ergothioneine. As an example, Figure 1 shows typical ¹H NMR spectra of plasma, WB, and RBC from a healthy donor, highlighting prominent peaks corresponding to the identified metabolite, ERG.

Figure 1.

Typical 800 MHz ¹H NMR spectra of solvent extracted (a) plasma; (b) whole blood; and (c) red blood cells from a healthy donor highlighting peaks from the newly identified metabolite, ergothioneine (ERG). ATP: Adenosine triphosphate; 2, 3-BPG: 2,3-Bisphosphoglycerate. The inset in (a) shows expanded region highlighting the ERG peak.

Peaks from the unknown metabolite were initially identified through analysis of 1D NMR spectra using RANSY method.⁴⁰ The well isolated peak at 6.8061 ppm was used as the driving peak. RANSY selectively identified three of the four sets of peaks from the unknown metabolite unambiguously (Figure S3). To assist with the identification, various 2D NMR spectra, including ¹H-¹H COSY, ¹H-¹H TOCSY, ¹H-¹³C HSQC, and ¹H-¹³C HMBC spectra, were analyzed following strategies described in our previous studies (Figures S4 and S5).^29,33 For these analyses, we utilized human RBCs, which exhibited higher intensity peaks for the unknown metabolite in the NMR spectra compared to other biospecimens.

However, a major challenge was that weak signals from the unknown metabolite were masked by the dominant glucose peaks, even in the RBC spectra, complicating the analysis. To overcome this, we leveraged the metabolic activity of RBCs in whole blood. Since glycolysis is the primary metabolic pathway in RBCs, these cells actively consume glucose, leading to its progressive depletion over time. This depletion occurred more rapidly when blood samples were stored at room temperature compared to on ice. After 144 hours of room temperature storage, glucose was nearly completely consumed by the metabolically active RBCs (Figure S6). Notably, the peaks corresponding to the unknown metabolite remained largely unchanged under these conditions. As a result, the NMR spectra of these samples displayed well-resolved signals for the unknown metabolite, free from interference by glucose. These glucose-depleted samples were subsequently used for the identification of the unknown metabolite using NMR.

Although analyses of 2D NMR spectra provided numerous hints about the structure of the unknown metabolite, the data were insufficient to establish the structure unambiguously. To address this challenge, we conducted mass spectrometry (MS) analysis on RBC samples. To enable correct identification of the mass peak from complex mass spectra, we filtered RBC samples based on their ¹H 1D NMR spectra and selected samples that exhibited vastly different intensity for the known metabolite. Specifically, two samples were selected such that, the concentration of the unknown metabolite in one sample was higher (by nearly a factor of 6) than the other sample (Figure S7). Liquid chromatography-mass spectrometry (LC-MS) data were acquired in positive ionization mode, as the metabolite was predicted to carry a positive charge due to the trimethylamine group, inferred from its characteristic ¹H/¹³C NMR chemical shifts and long range ¹H-¹³C couplings (Figure S5). The mass spectral data for both samples were compared, and a mass-to-charge ratio (m/z) peak exhibiting nearly a six-fold intensity difference between the two samples was identified (Figure S8; Table S2). This peak, with an m/z value of 230.0959, was further investigated, and the predicted structures for this mass were retrieved through a database search (HMDB).⁴¹ These structures were then compared to the tentative structure proposed from the NMR data. The unidentified metabolite peaks in NMR spectra were thus identified to be from ERG. To further confirm the identity of the metabolite, we rerecorded ¹H NMR spectra after spiking with a commercially obtained ERG compound and compared LC-MS and LC-MS/MS spectra of an RBC sample and the standard compound at varying collision energies (Figures S9–S12). These results confirmed the identity of the metabolite as ERG. The ¹H and ¹³C NMR chemical shifts for ERG, derived from 1D and 2D NMR spectra, are listed in Table S3. A reverse literature search following the identification of ERG by NMR uncovered several LC-MS-based metabolomics studies involving this metabolite.^7–15

An important characteristic of ERG for its routine analysis using ¹H NMR is the singlet peak at 6.8061 ppm, which originates from the hydrogen atom in the imidazole ring moiety (H #5 of ERG shown in Figure 1). This hydrogen has no adjacent hydrogens, preventing spin-spin couplings that would otherwise cause peak multiplicity. Furthermore, this peak is isolated and does not overlap with signals from other metabolites in any of the biospecimens analyzed, including human plasma, WB, and RBC (Figure 2). As a result, the 6.8061 ppm singlet provides excellent resolution and sensitivity, making it the optimal choice for accurate measurement of ERG in routine ¹H NMR based metabolomics studies. While the singlet peak at 3.2794 ppm, arising from the trimethyl group of the betaine moiety in ERG (H #7 of ERG shown in Figure 1), offers nearly an order of magnitude higher sensitivity than the 6.8061 ppm singlet, it is less ideal for the accurate analysis of the metabolite. This is because, the 3.2794 ppm peak often overlaps with peaks from other metabolites including the prominent glucose signals (Figure 1). The limit of detection (LOD), measured on an 800 MHz instrument equipped with a cryoprobe, was ~46 ng (1.0 μM in 200 μL solution, 3 mm NMR tube) for the peak at 6.8061 ppm. The corresponding limit of quantitation (LOQ) was 152 ng (3.3 μM). These values, however, depend on a number of parameters including magnetic field strength, probe sensitivity, and the number of transients. Figure S13 presents a representative plot of peak area versus ERG concentration. Due to spectral overlap, the peak at 3.2794 ppm did not improve the LOD or LOQ and could not be reliably integrated. Therefore, as discussed above, this peak is not ideal for accurate quantitation of ERG.

Figure 2.

Overlaid portions of ¹H NMR spectra of solvent extracted (a) plasma (n=150), (b) whole blood (n=128), and (c) red blood cells (n=316), highlighting the characteristic singlet peak from ERG at 6.8061 ppm. NADH: nicotinamide adenine dinucleotide (reduced); NADPH: nicotinamide adenine dinucleotide phosphate (reduced).

The concentrations of ERG in human plasma, WB, and RBC varied significantly between individuals (Figure 3). RBC exhibited surprisingly high levels of ERG, reaching up to ~1.7 mM, with average concentrations approximately 140-fold higher than those in plasma and about 2-fold higher than those in WB (Figure 4). Positive correlations were observed between ERG concentrations in plasma, WB, and RBC, with a coefficient of determination (R²) of 0.59 for the plasma - WB correlation, 0.75 for plasma - RBC, and 0.98 for RBC - WB (Figure 5). These findings for plasma and WB are consistent with previous reports using LC-MS⁹ and indicate that plasma ERG levels are reflective of its distribution in WB, RBC, and potentially throughout the body.

Figure 3.

Portions of ¹H NMR spectra of human red blood cells (RBC) from three healthy individuals (a) to (c) highlighting the massive variation in the levels of ERG among the individuals. Metabolites from the samples were solvent extracted prior to analysis.

Figure 4.

Box and whisker plots of ergothioneine (ERG) levels in human plasma (n=25), whole blood (WB, n=21), and red blood cells (RBC, n=25) from the same individuals measured by ¹H NMR at 800 MHz. An expanded view of the plot for plasma is shown in the inset for clarity. The average levels of ERG in plasma, WB, and RBC were in the ratios of 1:70:140, respectively.

Figure 5.

Correlations of ERG levels between human (a) plasma and WB (n=21); (b) plasma and RBC (n=25); and (c) RBC and WB (n=21). ERG levels were measured by ¹H NMR at 800 MHz. WB: whole blood; RBC: red blood cells.

Since ERG is not synthesized in the human body and none of the individuals in this study took ERG supplements, the detected ERG must be derived from dietary sources. The substantial variation in ERG concentrations across individuals reflects differences in dietary habits, particularly in the consumption of foods rich in ERG. This is significant because ERG is increasingly recognized for its roles in disease prevention, particularly through its antioxidant and anti-inflammatory properties.³ Consequently, the ability to routinely measure the absolute concentration of ERG in the widely used biospecimen, blood, using ¹H NMR spectroscopy provides an objective tool for investigating its effects on human health and diseases. However, it is important to note that ERG concentrations in plasma increase over time as plasma remains in contact with RBC following blood collection (Figure S14). This suggests that ERG leaks from RBC into plasma if the blood is not processed promptly. Therefore, to ensure accurate analysis of ERG, plasma should be separated from blood cells immediately after blood collection.

In plasma, WB, and RBC, average levels of ERG were higher in younger (21 – 40 yrs.) compared to older (41 – 63 yrs.) donors (Figure S15), in males compared to females (Figure S16), and in individuals with higher hematocrit compared to those with lower hematocrit (Figure S17). However, none of the differences reached statistical significance (Tables S4–S6). This lack of significance is not surprising, given the relatively small sample size; importantly, ERG is exclusively diet-derived and hence dietary intake is a major confounder that needs to be addressed for interpretation of such data. Due to the lack of dietary data, however, the dietary associations with age, sex, or hematocrit could not be assessed.

We investigated the whole-body distribution of ERG in mice by measuring its concentrations in various organs using ¹H NMR spectroscopy. The results revealed significantly different levels across tissues (Figure 6). The liver exhibited the highest ERG concentrations, while the brain showed the lowest, and ERG was undetectable in skeletal muscle tissue (see also Figure S18). The average concentrations in the tissues followed approximate ratios of 0:1:10:35:45 for skeletal muscle, brain, heart, kidney, and liver, respectively. These variations suggest differing levels of ERG transporter expression across tissues, with absence of ERG in skeletal muscle potentially indicating the lack of its transporter. The distribution of ERG as determined by NMR is noteworthy, particularly given the challenges associated with measuring ERG using earlier analytical methods, which lacked the specificity and sensitivity of current techniques.¹⁶ Historically, ERG levels in rat plasma, RBC, and various tissues were measured using column chromatography.⁴² However, the ERG distribution reported in the earlier study is inconsistent with the findings of the current study in both humans and mouse models (Figures 4 and 6).

Figure 6.

Box and whisker plots of concentrations of ERG in mouse heart (n=20), skeletal muscle (SM, n=15), kidney (n=16), brain (n=4), and liver (n=16) tissues obtained using ¹H NMR at 800 MHz. Note, ERG was not detected in skeletal muscle. The inset shows an expanded view of the plot for brain tissue, for clarity. The average levels of ERG in skeletal muscle, brain, heart, kidney and liver were in the ratios of 0:1:10:35:45, respectively.

Surprisingly, despite its high concentrations, ubiquitous presence in biological samples and permanent positive charge that facilitates easy ionization and detection by MS with high sensitivity, ERG is rarely identified in MS-based metabolomics studies^7–15 and had not been identified in NMR-based metabolomics studies. In the ¹H NMR spectra of live RBC, a peak at 3.28 ppm was first reported by Rabenstein and colleagues in 1977,⁴³ and the same group later identified it as ERG.⁴⁴ However, this peak overlaps with peaks from dominant RBC metabolites such as glucose and betaine (Figures 3 and S6). Despite this early detection, a comprehensive investigation detailing the detection, identification, and strategies for routine analysis of ERG in biological samples from NMR-based metabolomics perspective had not yet been reported. The demonstration that ERG can be routinely analyzed by ¹H NMR as part of metabolomics studies opens new avenues for exploring its roles in health and diseases. Additionally, the ability to analyze ERG by NMR paves the way for the development of new protocols for its routine analysis in MS based metabolomics studies.

CONCLUSIONS

In this study, we have described the detection, identification, and analysis of ergothioneine (ERG) in human plasma, whole blood, and red blood cells, as well as in the heart, kidney, brain, skeletal muscle, and liver tissues of mice using ¹H NMR spectroscopy. We relied on a combined NMR, MS and ratio analysis approach to identify the compound. ERG, a unique antioxidant of dietary origin, is increasingly recognized for its potential for preventing disease development. It is ubiquitous and present in high concentrations in biological samples, with its ¹H NMR peak intensity often surpassing that of other metabolites in blood. The positive correlation between plasma levels and those in whole blood and RBC suggests that plasma ERG levels may reflect its whole-body distribution, making it a valuable, routinely measurable biomarker for human health assessment.

A unique characteristic for its analysis using simple 1D ¹H NMR spectra is that it shows isolated peak at 6.8061 ppm, which does not overlap with other peaks, enabling its reliable measurement in routine metabolomics studies. Surprisingly, despite its ubiquity and high concentration in biological samples from both humans and animal models, ERG had not been identified in NMR-based metabolomics studies, one of the most widely used analytical methods. Even global analyses using mass spectrometry (MS) have often failed to annotate the metabolite, highlighting the challenges current methods face in identifying unknown or even less common metabolites. Our demonstration that ERG can be routinely measured using simple 1D ¹H NMR provides a foundation for incorporating its routine analysis in both NMR- and MS-based metabolomics studies.

Supplementary Material

SUPPORTING INFORMATION:

Mouse tissue specimens used in the study (Table S1); list of mass peaks for RBC samples from two donors (Table S2); ¹H and ¹³C NMR chemical shifts for ergothioneine measured in RBC (Table S3); ergothioneine levels in younger and older donors (Table S4); ergothioneine levels in male and female donors (Table S5); ergothioneine levels in lower and higher hematocrit levels (Table S6); schematic diagram of the blood collection and treatment protocol (Figure S1); flow diagram for the identification and quantitation of ergothioneine (Figure S2); ¹H 1D CPMG NMR and RANSY spectra of RBC (Figure S3); 2D ¹H-¹H TOCSY NMR spectra of RBC (Figure S4); overlay of 2D ¹H-¹³C HSQC and 2D ¹H-¹³C HMBC spectra of RBCs (Figure S5); ¹H NMR spectra of RBC separated from whole blood at 0 and 144 hrs. (Figure S6); ¹H NMR spectra of RBC from two donors highlighting ergothioneine peak (Figure S7); plot of p-value versus fold change for m/z ions detected in RBC and overlay of extracted ion chromatograms for RBCs from two donors (Figure S8); extracted ion chromatograms and MS/MS spectra for ergothioneine (Figure S9); ¹H NMR spectra of human biospecimens after spiking with ergothioneine (Figure S10); ¹H NMR spectra of mouse biospecimens after spiking with ergothioneine (Figure S11); ¹H NMR spectrum of ergothioneine standard (Figure S12); plot of ERG peak area (6.8061 ppm) versus its concentration (Figure S13); overlay of ¹H NMR spectra of plasma separated from blood at different time points (Figure S14); box and whisker plots of ergothioneine levels in young vs. old individuals (Figure S15); box and whisker plots of ergothioneine levels in male vs female (Figure S16); box and whisker plots of ergothioneine levels in low vs high hematocrit values (Figure S17); ¹H NMR spectra of heart, kidney, brain, liver, and skeletal muscle tissue from four mice highlighting ergothioneine peak (Figure S18).