Journal of Chromatography & Separation Techniques
Open Access

Modification of a Targeted Metabolomics Method Using Hydrophilic Interaction Liquid Chromatography-High Resolution Mass Spectrometry to Measure Nine Purines in Commercial Dog Foods

Author info »

Introduction

Purines in the body, including nucleotides, nucleosides, and nucleobases, are derived from ingested foods or can be synthesized de-novo. In dogs, purines are usually metabolized to poorly water-soluble Uric Acid (UA) and further metabolized and excreted in urine as water-soluble allantoin. Urate uroliths are prone to form when excess uric acid is excreted in the urine of dogs with hepatic dysfunction or a genetic defect that impairs purine metabolism [1-4]. Preventative measures include feeding diets that manufacturers report to be low in purines or diets that contain raw materials with low purine content [5-9]. However, the purine content of commercial dog foods has not been well documented in peer-reviewed journals [10,11]. Therefore, a method that measures purines in different types of commercial diets would facilitate the selection of low-purine diets for individual patients to prevent urate urolithiasis.

Conventional methods for measuring purines in food ingredients use acid hydrolysis to break down all purines into four nucleobases [12,13]. Those methods quantify four nucleobases, which are then added together as total purine content, but do not differentiate between nucleotides, nucleosides, and nucleobases. Previous studies in people and rats have indicated that the different purine metabolites vary in their capacity to affect urinary UA concentrations [14-16]. Furthermore, purine metabolite content has been shown to change during storage when measured using thin-layer chromatography [17]. Ardente et al., showed that eight purine metabolite concentrations, quantified using LCMS/MS after aqueous extraction, differed between stored and fresh fish commonly consumed by dolphins [18]. They found differences in total and individual purine concentrations among fish, especially hypoxanthine concentrations, which may explain why dolphins fed by human beings are more likely than wild dolphins to develop urate uroliths.

This report describes how the purine analysis previously developed by Ardente et al., was adapted and validated for the measurement of purines in commercial dog foods [19]. Some adaptation was necessary because dog foods present a complex matrix composed of many ingredients, including protein and fat from animal or vegetable sources, soluble and insoluble fibers, and additives, such as vitamins and minerals. Also, a more sensitive detection method was desired because purine concentrations in dog foods designed for preventing urate uroliths were expected to contain much lower concentrations of purines than were previously measured in fish. Hydrophilic Interaction Liquid Chromatography (HILIC) was utilized because highly selective polar separation of purines has been reported [20,21]. Tandem High-Resolution Mass Spectrometry (HRMS/MS) with wide-isolation quantitative strategies described by Rubio et al., were also utilized to allow faster scan rates and higher selectivity [22]. An internal standard was used instead of more time-consuming standard additions to reduce sample processing time.

Materials and Methods

Chemicals and solutions

Stock solutions including the mobile phase, internal standards, and external standards were all prepared freshly at least bi-weekly using high-performance-liquid-chromatography or liquid chromatography-mass spectrometry grade solvents, when available. Ammonium formate, ammonium hydroxide, acetonitrile, acetone, Ultrapure Water (H₂O), Hydrochloric Acid (HCl), Sodium Chloride (NaCl), Sodium Hydroxide (NaOH), methanol, formic acid, and hexane were purchased from Fisher Scientific company, LLC (Suwanee, GA, USA).

Chromatography solutions differed from those used by Ardente, et al., [19]. Buffer stock solution for the mobile phase was 100 mM ammonium formate solution in H₂O, adjusted to a pH 9.2 with 30% ammonium hydroxide in H₂O. The additive stock solution was 5 mM methylene diphosphonic acid (TCI America, Portland, or USA) in H₂O to help improve peak shapes. Mobile phase A was 100 mL of buffer stock solution mixed with 900 mL of acetonitrile, while mobile phase B was 100 mL of buffer stock solution mixed with 60 mL of acetonitrile and 840 mL of H₂O. Each 999 mL of mobile phase was spiked with 1 mL of 5 mM methylene diphosphonic acid stock solution. Each mobile phase was shaken well and sonicated for 15 minutes.

Purine external standard solutions were prepared for calibration of the chromatographic method. Adenine (Ade), Adenosine (Ado), Hypoxanthine (Hxa), Xanthine (Xan), Guanine (Gua), Guanosine (Guo), Inosine (Ino), and Uric Acid (UA) were purchased from Sigma Aldrich (St. Louis, MO, USA). Adenosine Monophosphate (AMP), Inosine Monophosphate (IMP), and Guanosine Monophosphate (GMP) were purchased from TCI America. Aqueous 10 mM stock solutions of each purine were prepared. Solubility varied among purines, so Gua, Hxa, Xan, and UA were dissolved in 0.1 N NaOH, while Ade, Ado, Ino, and Guo were dissolved in 0.1 N HCl, and AMP, IMP and GMP were dissolved in 0.9% NaCl. These initial stock solutions were then combined to make three mixtures each of 5 mL final volume: Mixture A contained 1 mL of 0.1 N NaOH and 1 mL of each of the 10 mM stock solutions of Gua, Hxa, Xan, and UA; mixture B contained 1 mL of 0.1 N HCl and 1 mL of each of the 10 mM stock solutions of Ade, Ado, Ino, and Guo; mixture C contained 2 mL of 0.9% NaCl and 1 mL of each of the 10 mM stock solutions of AMP, IMP and GMP. A final stock solution containing 500 μM of each purine standard was prepared by combining 2.5 mL each of mixtures A, B, and C and 2.5 mL of 0.9% NaCl, and stored at -20℃. An aliquot of the final stock solution was thawed and sonicated for 10 minutes and diluted with 0.9% NaCl before each chromatographic run to prepare external standard solutions containing 100 nM, 500 nM, 1, 5, 10, 50, 100, and 250 μM of each individual purine.

This method used an internal standard, 2,6-Diaminopurine Monohydrate (DAP; 97.0+%, Fisher Scientific) to compensate for variations in recoveries and strong matrix effects where standard additions had been used previously [19]. A solution containing 400 μg/mL of DAP in H₂O was added to all food samples before extraction. An internal standard solution of stable isotope-labeled purines was added to each sample solution after extraction in conjunction with the external standards to calibrate the amount of purine in the sample solution. The internal standard solution included Ade-¹⁵N₅, Xan-¹³C, ¹⁵N₂, Hxa-¹³C₂, ¹⁵N, Ado-¹³C₅, Guo-¹³C, ¹⁵N₂, UA-¹³C₃, AMP-¹³C₅ (Toronto Research Chemicals, Toronto, ON, Canada) and Ino-¹⁵N₄, and Gua-¹⁵N₅ (Cambridge Isotope Laboratories, Cambridge, MA, USA). Labeled IMP and GMP were not readily available at the time of this study. Therefore, labeled AMP-¹³C5 was used as an alternative because it has the closest chemical structure and retention time to IMP and GMP. Individual solutions for each internal standard were prepared to a concentration of 1 mg/mL: Ade-¹⁵N₅, Gua-¹³C₂, ¹⁵N, Xan-¹³C, ¹⁵N₂, Hxa-¹³C₂, ¹⁵N, and UA-¹³C₃, were prepared in 0.1N NaOH; Ado-¹³C₅, Guo-¹³C, ¹⁵N₂, Ino-¹⁵N₄, and AMP-¹³C₅ were prepared in 0.1 N HCl. These internal standard solutions were combined and diluted in methanol before each run to achieve a final concentration of 10 μg/mL of each stable isotope-labeled compound.

Food samples and processing

Two lots of canned dog food were selected, one from each of two different readily available commercial dog foods (Supplementary table). One dog food was an adult maintenance diet presumed to contain increased purines because it contained organ meats and a minimum of 10% as fed crude protein. The other was presumed to be low in purine because it is marketed as a urate-urolith prevention diet. The adult maintenance diet was used for Quality Control (QC) and to evaluate between-day and intra-day variability. Both diets were used to assess recovery.

The entire contents of 1 large can (369 grams) of each canned diet was individually ground through a 2 mm sieve (Wiley Mill, Thomas Scientific., Swedesboro, NJ, USA), then homogenized by hand in a polyethylene bag with zip closure, and numbered for identification. Authors who performed the chromatographic analysis of samples were blind as to which sample identification numbers corresponded to which diet or replicate. All samples were lyophilized to stable weight (Free Zone 2.5 Liter -50℃ Benchtop Freeze Dryers with 3-shelf drying compartment, Labconco, Kansas City, MO, USA). Dried samples were stored in polyethylene bags with zip closures, in a climate-controlled (25℃) room, then lyophilized again just before preparation for purine analysis to eliminate any moisture accumulated from ambient air. Duplicated QC food samples were prepared whenever purine extraction was performed and then analyzed with the samples extracted on the same day.

Purine extraction

Purine extraction was performed largely according to the method described previously, with the following modifications to accommodate the different nature of the samples and the large number of samples being analyzed [19]. Homogenization, sonication, and heating of samples were performed as described by Clariana et al., [23]. After cooling in an ice water bath for 10 minutes, the solution was centrifuged at 18,100 rpm for 18 min at 20℃ (Sorval RC5C, Marshall Scientific, Hampton, NH, USA). The supernatant was filtered (Grade P5, Fisherbrand Qualitative-Grade Filter Paper Circles, Fisher Scientific), and an aliquot was mixed with an equal amount of hexane using a vortex mixer (Fisher Scientific) for 30 seconds and then centrifuged at 6,500 rpm, for 7 min, at 20℃ (Sorval RC5C, Marshall Scientific). The aqueous (bottom) layer was frozen at -20℃ until final processing and analysis.

A 100 μL aliquot of thawed extract was mixed with 20 μL of 10 μg/mL stable isotope-labeled internal standard solution, and 500 μL of protein precipitation solution (0.015% ammonium hydroxide in 50 mL of acetonitrile: acetone 9:1), using a vortex mixer for 10 seconds. The samples were incubated at 4℃ for 10 minutes to allow further precipitation, then centrifuged at 14,000 rpm at 4℃ for 10 minutes. An aliquot of the supernatant was transferred to an autosampler vial from which, a 10 μl sample was analyzed using Hydrophilic Interaction Liquid Chromatography (HILIC): High-Resolution Tandem Mass Spectrometry (HRMS/MS). The order of extracted samples, blanks, external standards, and DAP injected was shuffled randomly (RAND function of Microsoft Office Excel 2019).

HILIC separation

Separation of eleven purine metabolites, Ade, Gua, Hxa, Xan, UA, Ado, Guo, Ino, AMP, IMP, and GMP was carried out using HILIC (Dionex Ultimate 3000 UHPLC, Thermo Fisher Scientific). Separation was achieved on an InfinityLab Poroshell 120 HILIC-Z column, 150 mm × 2.1 mm, 1.9 μm with a 2.1 mm × 5 mm, 2.7 μm guard column (Agilent, Santa Clara, CA, USA). The elution conditions for separation began with 100%mobile phase A for 0.5 minutes, then changed to 77% mobile phase A: 23% mobile phase B over 1 minute, then held for 2.5 minutes, before further change to 50%:50% mobile phase A:B over 0.5 minutes, then held for 1 minute, then being increased back to 100% A over 0.5 minute, and then held for 4 minutes.

Detection and quantification

Purine metabolites were identified by mass and retention time using mass spectrometry with a Thermo Q-Exactive hybrid orbitrap Mass Spectrometer (Thermo Fisher Scientific) with Heated Electrospray Ionization (H-ESI) operated in both HRMS and HRMS/MS modes. Purines were detected by the full scan (HRMS) initially, and then ions were selected in the time window specified for tandem Mass Spectrometry (HRMS/MS).

The H-ESI source was operated in negative-ion mode with the following settings: Spray voltage 2500 V, capillary temperature 350℃, 35 arbitrary units of sheath gas pressure, 1.0 arbitrary units of ion sweep gas pressure, 10 arbitrary units of auxiliary gas pressure, probe heater temperature 275℃, and S-lens at 40. Data was processed using computer software (Xcalibur Quan Browser 4.2, Thermo Fisher). The FDA Bioanalytical method validation guidance for industry to standardize analytical methods was followed [24]. The sensitivity was determined by the lowest nonzero external standard (50 nM, 100 nM, 500 nM, 1, 5, 10, 50, 100, and 250 μM of each individual purine) on the calibration curve. The peak area ratio of the external standards relative to labeled internal standards at the different concentrations of the external standards was used to create a calibration curve. The purine concentration of the sample (μM) was determined by substituting the peak area ratio of the sample to the internal standard into the equation obtained from the calibration curve. Recovery of DAP in individual samples was calculated by dividing the measured DAP concentration by the added DAP concentration. Purine concentration (μmol/100 g food DM) in each sample was calculated by adjusting for sample weight and dilution during sample preparation, and dividing the measured concentration in the extract by the recovery of DAP. Spreadsheet software was used for these calculations. (Microsoft Excel for Mac version 16.75.2, Microsoft Corporation, Redmond, WA, USA).

Method validation

The Coefficient of Variation (CV) for within-day (n=6, six replicate sub-samples and between-days (n=18, two duplicated samples for nine days) was calculated for the content of each purine metabolite (μmol/100 g dry matter) in the high purine QC diet. To evaluate whether the recovery of the internal standard, DAP, accurately represented the recovery of individual purines, increasing amounts of mixed purine standard stock solution (500 μM of each purine in H₂O), were mixed, before extraction and analysis, with decreasing amounts of H₂O, 0.6 g of lyophilized sample of either the ‘high purine’ or ‘low purine’ QC diet, and 0.5 mL of 400 μg/mL DAP in H₂O to a final volume of 20 mL containing either 0 (no addition), 2.5 μM, 25 μM, and 100 μM of each added purine. Recovery (%) was calculated as follows: Firstly, the measured amount of each purine standard in the samples was determined as the difference in measured amount of each purine in food samples with and without a known amount of purine standard added; then, the measured amount of each purine standard was divided by the known amount of purine added and multiplied by 100. Recoveries were averaged for the two diets at low, moderate, and high internal standard additions to assess the accuracy of the method at each concentration.

Results and Discussion

Purine metabolites were separated and quantified by mass and retention time by HILIC-HRMS/MS or HILIC-HRMS (Figure 1).

Figure 1: Extracted ion chromatograms of eleven targeted purine metabolites in the high purine dog food. Time (minutes) and relative abundance are represented along the xaxis and y-axis, respectively. The Retention Time (RT) of each purine is noted next to each metabolite peak.

Most purines were quantified from the HRMS/MS data, whereas the full scan data with HRMS were used for only Xan and UA because product ion intensity was low. Instrument settings used for parallel reaction monitoring are listed in Table 1. Additional steps were necessary after extraction to clean samples because dog foods are complex matrices typically containing grains, meat, animal fat, vegetable oil, fiber, and mineral and vitamin supplements. Adding a mixture of acetonitrile and methanol and centrifugation achieved efficient protein precipitation and removal. The mobile phase being basic (pH 9.2) facilitated deprotonation and ionization of the purines and improved separation and response in negative ion mode. Flushing the system and chelating metals are expected to minimize ion suppression because phosphorylated organic compounds can interact with metals from iron-derived columns or solvents, which may affect peak detection [25,26]. Adding methylene diphosphonic acid to a mobile phase for this purpose improved chromatographic performance in HILIC mode.

Table 1: Instrument settings for parallel reaction monitoring.

We used external standards and two internal standards (DAP and stable-isotope-labeled purines) to improve the accuracy of quantification. Comparable information measuring purines in dog foods was not available from previous studies. Our aqueous extraction might not release all the purine in the diets compared to acid hydrolysis. Nevertheless, recoveries have previously been found to be poor in our laboratory when acid hydrolysis was used instead of aqueous extraction at least in part because acid conditions affect the solubility of some purines, such as UA [23]. Each sample was freeze-dried because pet foods can vary widely in moisture content but freeze-drying may have affected purine extraction [27].

Diaminopurine (DAP) was used as an internal standard after caffeine proved unsuitable in negative mode [21]. Recovery of DAP was only about 60% after the filtration step used in sample preparation but DAP proved to be suitable as an internal standard because recovery was consistent and mimicked the recovery of individual purines at concentrations found in the commercial dog foods. The between-day CV of each purine metabolite concentration (μmol/100g food dry matter) in 18 high purine QC diet samples (2 duplicates, 9 days) averaged 15% (range: 11%-19%). The mean recovery of DAP in between-day QC samples was 59% with a CV of 12%. The within-day CV of each purine metabolite concentration (μmol/100g food dry matter) in six QC samples was 6% on average (range: 4-13%) (Table 2). The mean recovery rate of DAP was 64% with a CV of 3% in within-day QC samples.

Table 2: Purine metabolite concentrations (μmol/100 g DM) and assay variability of the high purine diet.

The concentration of most purines (μmol/100g DM) were between 4 and 56 in both diets but higher for XA, Hxa and Ino in the high purine diet and higher for UA in the low purine diet (131, 306, 99, and 584, respectively) (Table 3). Both IMP and GMP were below the Limit of Quantitation (LOQ) in both diets. Adding standards to measure recoveries increased the concentration in the diets by 8-333 μmol/100g DM.

Table 3: Mean purine concentrations (μmol/100 g DM) in two commercial canned dog foods.

Average recoveries for the two diets were within 100% ± 20% for Ade, Gua, XA, Hxa, Ado, Guo, Ino, and AMP for the range of purine concentrations in the food samples (Table 4). Average recoveries for the two diets were higher for Ado, Guo and UA (133%, 138% and 156%, respectively) at the highest added concentration. Average recoveries for the two diets were within 100% ± 20% for IMP and GMP at the highest added concentration and at the intermediate added concentration for GMP but was inconsistent at other added concentrations.

Table 4: Recoveries of purine metabolites added to two commercial foods using diaminopurine as an internal standard. 

Recoveries were within acceptable limits for most of the purines at all concentrations evaluated but were unacceptable at lower concentrations for IMP and GMP. Recoveries were slightly high at the highest concentrations for Guo, Ado, and UA in both diets. Changing ratio of solvent volume to sample weight before extraction would improve accuracy because recoveries were consistent at low concentrations for all purines except IMP and GMP. Such modifications may not be necessary to measure Guo and Ado, and possibly also UA, however, because the highest concentration added was much higher than that found in both foods for Guo and Ado, and in the high purine diet for UA. Why a ‘low purine’ diet should contain so much more UA than the ‘high purine’ diet and whether that is true of all low purine diets needs to be determined. It is unlikely that the analytical errors overestimated UA concentrations, because repeat analyses of additional subsamples of each lot showed reproducible results.

The nucleotides, IMP and GMP were below the limit of quantification in these foods. It is likely that concentrations of these nucleotides were indeed low because IMP is degraded during storage to Ino and then Hxa [17]. Ardente et al. detected lower concentrations of IMP and higher concentrations of Hxa in frozen fish compared to fresh fish using their more laborious method, likely due to degradation during storage [18]. Purine degradation may also occur with cooking [27-29]. This study did not measure XMP (xanthine monophosphate) and xanthosine because internal standards that are stable in basic solutions were unavailable at the time of the study. Nevertheless, XMP and its nucleoside are unlikely to be present in more than negligible quantities because XMP is an intermediate in the metabolism of IMP to GMP [30].

Conclusion

This report describes how a previously reported targeted metabolomics assay was adapted and validated to enable quantification of nine purine metabolites in commercial dog foods. Use of external standards, two internal standards (DAP and stable-isotope-labeled purines), a basic mobile phase with added methylene diphosphonic acid, and HILIC-HRMS/MS or HILIC-HRMS enabled accurate measurement of poorly soluble purines in foods with a complex matrix. These adaptations saved at least 9 hours of extraction time processing one sample by omitting the evaporation, reconstitution, and standard addition steps in the previously reported method. Between-day and within-day variation and recoveries were within acceptable limits for nine purines at all concentrations evaluated but recoveries were unacceptable at lower concentrations for IMP and GMP. Thus, a broader range of purine metabolites in dog foods was successfully recovered in a reproducible fashion than has been previously reported using conventional approaches. As a result, the optimal dietary composition to prevent canine urate urolith can be assessed in a more direct and efficient way.

Acknowledgement

This research and the first authors’ position were funded by Royal Canin SAS, Aimargues, France.

References

1. Case LC, Ling GV, Ruby LA, Johnson DL, Franti CE, Stevens F. Urolithiasis in dalmations: 275 cases (1981-1990). J Am Vet Med Assoc. 1993;203(1):96-100.

   [Google Scholar] [PubMed]

2. Caporali EH, Phillips H, Underwood L, Selmic LE. Risk factors for urolithiasis in dogs with congenital extrahepatic portosystemic shunts: 95 cases (1999–2013). J Am Vet Med Assoc. 2015;246(5):530-536.

   [Crossref] [Google Scholar] [PubMed]

3. Bartges JW, Osborne CA, Lulich JP, Kruger JM, Sanderson SL, Koehler LA, et al. Canine urate urolithiasis: Etiopathogenesis, diagnosis, and management. Vet Clin North Am Small Anim Pract. 1999;29(1):161-191.

   [Crossref] [Google Scholar] [PubMed]

4. Bartges JW, Osborne CA, Lulich JP, Unger LK, Koehler LA, Bird KA, et al. Prevalence of cystine and urate uroliths in bulldogs and urate uroliths in dalmatians. J American Vet Med Assoc. 1994;204(12):1914-1918.

   [Google Scholar] [PubMed]

5. Lulich JP, Berent AC, Adams LG, Westropp JL, Bartges JW, Osborne CA. Acvim small animal consensus recommendations on the treatment and prevention of uroliths in dogs and cats. J Vet Intern Med. 2016;30(5):1564-1574.

   [Crossref] [Google Scholar] [PubMed]

6. Bartges CJ, Osborne CA, Felice LJ, Brown C, Allen TA, Koehler L, et al. Influence of two amounts of dietary casein on uric acid, sodium urate, and ammonium urate urinary activity product ratios of healthy beagles. Am J Vet Res. 1995;56(7):893-897.

   [Google Scholar] [PubMed]

7. Bartges JW, Osborne CA, Felice LJ, Allen TA, Brown C, Unger LK, et al. Diet effect on activity product ratios of uric acid, sodium urate, and ammonium urate in urine formed by healthy beagles. Am J Vet Res. 1995;56(3):329-333.

   [Google Scholar] [PubMed]

8. Bartges J, Osborne CA, Felice LJ, Allen TA, Brown C, Koehler LA, et al. Influence of four diets containing approximately 11% protein (dry weight) on uric acid, sodium urate, and ammonium urate urine activity product ratios of healthy beagles. Am J Vet Res. 1995;56(1):60-65.

   [Google Scholar] [PubMed]

9. Bijster S, Nickel RF, Beynen AC. Comparison of the efficacy of two anti-uric acid diets in dalmatian dogs. Acta Vet Hung. 2001;49(3):295-300.

   [Crossref] [Google Scholar] [PubMed]

10. Westropp JL, Larsen JA, Johnson EG, Bannasch D, Fascetti AJ, Biourge V, et al. Evaluation of dogs with genetic hyperuricosuria and urate urolithiasis consuming a purine restricted diet: A pilot study. BMC Vet Res. 2016;13(1):45.

    [Crossref] [Google Scholar] [PubMed]

11. Malandain E, Caussé E, Tournier C, Aladenise S, Rigolet A, Ecochard C, et al. Quantification of End-Products of Purine Catabolism in Dogs Fed Diets Varying in Protein and Purine Content. J Vet Int Med. 2008;22(3):732.

    [Google Scholar]

12. Kaneko K, Takayanagi F, Fukuuchi T, Yamaoka N, Yasuda M, Mawatari KI, et al. Determination of total purine and purine base content of 80 food products to aid nutritional therapy for gout and hyperuricemia. Nucleosides, Nucleotides Nucleic Acids. 2020;39(10–12):1449-1457.

    [Crossref] [Google Scholar] [PubMed]

13. Qu X, Sui J, Mi N, Lin H. Determination of four different purines and their content change in seafood by high-performance liquid chromatography: Determination of purines in seafood. J Sci Food Agric. 2017;97(2):520-525.

    [Crossref] [Google Scholar] [PubMed]

14. Brulé D, Sarwar G, Savoie L. Changes in serum and urinary uric acid levels in normal human subjects fed purine-rich foods containing different amounts of adenine and hypoxanthine. J Am Coll Nutr. 1992;11(3):353-358.

    [Crossref] [Google Scholar] [PubMed]

15. Clifford AJ, Riumallo JA, Young VR, Scrimshaw NS. Effect of oral purines on serum and urinary uric acid of normal, hyperuricemic and gouty humans. J Nutr. 1976;106(3):428-434.

    [Crossref] [Google Scholar]

16. Clifford AJ, Story DL. Levels of purines in foods and their metabolic effects in rats. J Nutr. 1976;106(3):435-442.

    [Crossref] [Google Scholar]

17. Fraser DI, Dingle JR, Hines JA, Nowlan SC, Dyer WJ. Nucleotide degradation, monitored by thin-layer chromatography, and associated postmortem changes in relaxed cod muscle. J Fish Res Bd Can. 1967 1;24(8):1837-1841.

    [Crossref] [Google Scholar]

18. Ardente AJ, Garrett TJ, Colee J, Vagt BJ, Walsh MT, Wells RS, et al. Differences in purine metabolite concentrations in the diet of managed and free-ranging common bottlenose dolphins (Tursiops truncatus). Aquas Mamm. 2017;43(6):618-628.

    [Crossref] [Google Scholar]

19. Ardente A, Garrett T, Wells R, Michael W, Smith C, Colee J, et al. A targeted metabolomics assay to measure eight purines in the diet of common bottlenose dolphins, tursiops truncates. J Chromatogr Sep Tech. 2016;7(5).

    [Crossref] [Google Scholar] [PubMed]

20. Chen J, Hou W, Han B, Liu G, Gong J, Li Y, et al. Target-based metabolomics for the quantitative measurement of 37 pathway metabolites in rat brain and serum using hydrophilic interaction ultra-high-performance liquid chromatography-tandem mass spectrometry. Anal Bioanal Chem. 2016;408(10):2527-2542.

    [Crossref] [Google Scholar] [PubMed]

21. Harmon TA Utilizing novel separation and detection techniques to expand metabolomic coverage in bioanalytical liquid chromatography-mass spectrometry discovery and quantitation. Doctoral dissertation, University of Florida. 2022.
22. Rubio VY, Cagmat JG, Wang GP, Yost RA, Garrett TJ. Analysis of tryptophan metabolites in serum using wide-isolation strategies for UHPLC-HRMS/MS. Anal Chem. 2020;92(3):2550-2557.

    [Crossref] [Google Scholar] [PubMed]

23. Clariana M, Gratacós-Cubarsí M, Hortós M, García-Regueiro JA, Castellari M. Analysis of seven purines and pyrimidines in pork meat products by ultra-high performance liquid chromatography-tandem mass spectrometry. J Chromatogr A. 2010;1217(26):4294-4299.

    [Crossref] [Google Scholar] [PubMed]

24. US Department of Health and Human Services. Food and drug administration, Center for Drug Evaluation and Research (CDER), Center for Veterinary Medicine (CVM). 2001.

    [Google Scholar]

25. Lou SN, Chen HH, Hsu PY, Chang DH. Changes in purine content of tilapia surimi products during processing. Fish Sci. 2005;71:889-895.

    [Crossref] [Google Scholar]

26. SN L. Changes in purine content of Tilapia mossambica during storage, heating and drying. Food Sci Agric Chem. 2001;3:23-29.

    [Google Scholar]

27. Xiao L, Sha W, Tao C, Hou C, Xiao G, Ren J. Effect on purine releasement of Lentinus edodes by different food processing techniques. Food Chem X. 2022;13:100260.

    [Crossref] [Google Scholar] [PubMed]

28. Nelson DL, Lehninger AL, Cox MM. Lehninger Principles of Biochemistry. Macmillan; 2008.

    [Crossref] [Google Scholar]

29. Tuytten R, Lemiere F, Witters E, Van Dongen W, Slegers H, Newton RP, et al. Stainless steel electrospray probe: A dead end for phosphorylated organic compounds?. J Chromatogr A. 2006;1104(1-2):209-221.

    [Crossref] [Google Scholar] [PubMed]

30. Muguruma Y, Tsutsui H, Akatsu H, Inoue K. Comprehensive quantification of purine and pyrimidine metabolism in Alzheimer's disease postmortem cerebrospinal fluid by LC-MS/MS with metal‐free column. Biomed Chromatogr. 2020;34(2):4722.

    [Crossref] [Google Scholar] [PubMed]