Validation of a new procedure to determine plasma fatty acid concentration and isotopic enrichment

Assessment of free fatty acid (FFA) concentration and isotopic enrichment is useful for studies of FFA kinetics in vivo. A new procedure to recover the major FFA from plasma for concentration and isotopic enrichment measurements is described and validated. The procedure involves extraction of plasma lipids with hexane, methylation with iodomethane (CH3I) to form fatty acid methyl esters (FAME), and subsequent purification of FAME by solid phase extraction (SPE) chromatography. The new method was compared with a traditional method using thin-layer chromatography (TLC) to recover plasma FFA, with subsequent methylation by BF3/methanol. The TLC method was found to be less reliable than the new CH3I method because of contamination with extraneous fatty acids, chemical fractionation of FFA species, and incomplete recovery of FFA associated with TLC. In contrast, the CH3I/SPE method was free of contamination, did not exhibit chemical fractionation, and had higher recovery. The iodomethane reaction was specific for free fatty acids; no FAME were formed when esterified fatty acids (triglycerides, cholesteryl esters, phospholipids) were subjected to the methylation reaction. We conclude that the CH3I/SPE method provides rapid and convenient recovery of plasma fatty acids for quantification or GC/MS analysis as methyl esters, and is not subject to the problems of contamination, reduced recovery, and chemical fractionation associated with recovery of FFA by TLC.—Patterson, B. W., G. Zhao, N. Elias, D. L. Hachey, and S. Klein. Validation of a new procedure to determine plasma fatty acid concentration and isotopic enrichment. J. Lipid Res. 1999. 40: 2118–2124. Assessment of free fatty acid (FFA) concentration and isotopic enrichment is useful for studies of FFA kinetics in vivo. A new procedure to recover the major FFA from plasma for concentration and isotopic enrichment measurements is described and validated. The procedure involves extraction of plasma lipids with hexane, methylation with iodomethane (CH3I) to form fatty acid methyl esters (FAME), and subsequent purification of FAME by solid phase extraction (SPE) chromatography. The new method was compared with a traditional method using thin-layer chromatography (TLC) to recover plasma FFA, with subsequent methylation by BF3/methanol. The TLC method was found to be less reliable than the new CH3I method because of contamination with extraneous fatty acids, chemical fractionation of FFA species, and incomplete recovery of FFA associated with TLC. In contrast, the CH3I/SPE method was free of contamination, did not exhibit chemical fractionation, and had higher recovery. The iodomethane reaction was specific for free fatty acids; no FAME were formed when esterified fatty acids (triglycerides, cholesteryl esters, phospholipids) were subjected to the methylation reaction. We conclude that the CH3I/SPE method provides rapid and convenient recovery of plasma fatty acids for quantification or GC/MS analysis as methyl esters, and is not subject to the problems of contamination, reduced recovery, and chemical fractionation associated with recovery of FFA by TLC.—Patterson, B. W., G. Zhao, N. Elias, D. L. Hachey, and S. Klein. Validation of a new procedure to determine plasma fatty acid concentration and isotopic enrichment. J. Lipid Res. 1999. 40: 2118–2124. Accurate assessment of plasma free fatty acid (FFA) concentrations and isotopic enrichment is critical for evaluating fatty acid metabolic kinetics in vivo using stable isotope methodology. Although numerous methods have been reported for the measurement of FFA concentrations and enrichments, these methods are seldom critically cross-validated. Perhaps the most commonly used approach for measuring plasma FFA concentrations and fatty acid enrichment involves extracting the lipid component of plasma, separating FFA by thin-layer chromatography (TLC), and preparing fatty acid methyl ester (FAME) derivatives using boron trifluoride (BF3) and methanol (1Morrison W.R. Smith L.M. Preparation of fatty acid methyl esters and dimethylacetals from lipids with boron fluoride–methanol.J. Lipid Res. 1964; 5: 600-608Google Scholar, 2Wolfe R.R. Radioactive and Stable Isotope Tracers in Biomedicine: Principles and Practice of Kinetic Analysis. Wiley-Liss, New York1992Google Scholar). An internal standard (e.g., heptadecanoic acid, C17:0) is used to quantify FFA concentration by gas chromatography (GC) using a flame ionization detector. The use of TLC requires time-consuming manual manipulations, including spotting samples on TLC plates and recovering TLC scrapings, which limit sample throughput. This approach assumes that recoveries of all fatty acid species are equivalent and that chemical fractionation does not occur. We developed an alternative approach for preparing FAME by using iodomethane (CH3I) and solid phase extraction (SPE) cartridges in an effort to decrease sample processing time and effort. In this report, we describe the CH3I/SPE procedure and identify systematic differences in the concentrations of major plasma FFAs and palmitate isotopic enrichment between this and the traditional TLC/BF3 procedure. By analysis of appropriate standards and biological samples, we provide evidence that the TLC/BF3 procedure introduces artifacts which decrease the accuracy of FFA concentration and palmitate enrichment measurements. In contrast, the new CH3I/SPE approach provides a more reliable analysis and permits a faster sample processing rate. Fatty acid concentration and enrichment were measured in biological plasma samples obtained from human subjects and in FAME and FFA standards. Plasma samples were obtained from subjects participating in two ongoing research protocols that focus on the hormonal regulation of whole-body lipolysis. These studies were approved by the Human Studies Committee of the Washington University School of Medicine, and informed consent of the participants was obtained. Samples were specifically chosen to include a broad range of plasma FFA concentration and enrichments. In both study protocols, subjects were infused with 2,2-2H2hexadecanoic acid (98 atom% 2H, Isotec, Miamisburg, OH) at a rate of 0.04 μmol/kg per min into an antecubital vein and blood samples were taken from a radial artery. Blood samples were collected in chilled tubes containing EDTA as an anticoagulant and placed immediately in ice. Plasma was obtained by refrigerated centrifugation within 20 min of blood drawing and stored at –70°C until subsequent analyses were performed. In one study, blood samples were obtained at regular time intervals during a pancreatic hormonal clamp with 4-stage epinephrine infusion. This study involved infusing somatostatin, insulin, and growth hormone to “clamp” the concentration of pancreatic hormones that affect lipolysis. Epinephrine was infused for 30 min at 0.00125 μg·kg fat free mass (FFM)-1·min-1, 0.005 μg·kg FFM-1·min-1, 0.0125 μg·kg FFM-1·min-1, and 0.025 μg·kg FFM-1·min-1 in 4 discrete stages, which caused a progressive step-wise increase in plasma FFA concentration and decrease in plasma palmitate enrichment. In the other study, blood samples were taken during an intravenous infusion of propranolol, a nonselective (both β1 and β2) β-adrenergic receptor antagonist, which caused a decrease in FFA concentration and an increase in palmitate enrichment. Blood samples from these two studies provided a 3-fold range of plasma FFA concentrations and a 5-fold range in palmitate isotopic enrichments. In addition, certain tests were performed on a larger pool of isotopically enriched plasma that was generated by combining aliquots from multiple plasma samples. FAME and FFA standards of known fatty acid profile were obtained from Nu-Chek-Prep, Inc. (Elysian, MN). The FAME standard (catalog # GLC-63) consisted of the methyl esters of C12:0 (2.0% by weight), C14:0 (2.0%), C14:1 (1.0%), C16:0 (22.0%), C16:1 (5.0%), C17:0 (10.0%), C18:0 (10.0%), C18:1 (32.0%), C18:2 (8.0%), C18:3 (4.0%), and C20:4 (4.0%). The fatty acid profile in this standard is similar to plasma FFA. The FFA standard (catalog # NIH-D) consisted of C14:0 (11.8%), C16:0 (23.6%), C16:1 (6.9%), C18:0 (13.1%), and C18:1 (44.6%). Sufficient 2H2palmitate was added to the FFA standard to provide a palmitate tracer:tracee ratio (TTR) of approximately 12.5%. Aliquots (250 μL) of plasma or water blanks were placed in 13 × 100 mm screw top tubes. Equal volumes (250 μL) of heptadecanoic acid (C17:0) internal standard (0.23 μmol/mL in heptane; stored at –20°C) and water were added. Samples were shaken on a platform vortexer for 3 min before adding 3 mL ice-cold acetone to precipitate plasma proteins. Samples were vortexed again for several seconds and placed at –20°C for 15 min. After centrifugation of precipitated proteins, the supernatant was poured into 16 mm × 125 mm screw-top tubes. Three-mL aliquots of hexane and water were added before securely capping the samples with Teflon-lined caps and shaking them gently in a horizontal platform shaker for 15 min. The samples were then centrifuged to separate the solvent and aqueous phases. The upper phase (hexane) was transferred into 13 × 100 mm screw-top tubes and dried in a SpeedVac centrifugal concentrator (Savant, Farmingdale, NY) or evaporated under nitrogen. A 0.25-mL aliquot of buffer (0.2 m dibasic potassium phosphate and 0.05 m tetrabutylammonium hydrogen sulfate, pH adjusted to 9.0 with tribasic potassium phosphate) and 0.25 mL iodomethane (Aldrich Chemical Co., Milwaukee, WI) in dichloromethane (1:10 vol:vol) were added, and samples were vortexed for 10 min to form FAME. Three mL of hexane was added and samples were vortexed for 15 min on a platform vortexer to extract FAME. After centrifugation to separate solvent phases, the upper layer (hexane) was transferred to 13 × 100 mm uncapped test tubes and dried in a SpeedVac centrifugal concentrator. Solid phase extraction cartridges (LC-Si, 3 mL size, catalog #505048; Supelco, Bellefonte, PA) were placed on a vacuum manifold and prepared by washing twice with 1.5-mL aliquots of hexane collected into waste tubes without allowing the cartridges to dry. Hexane (1.5 mL) was added to the dried samples and vortexed for 2 min. Samples were transferred to the SPE cartridges and rinsed slowly with 1.5 mL hexane. Waste tubes were replaced with 13 × 100 mm uncapped collection tubes, and FAME were eluted with 2 rinses (1.5 mL each) of 2% ethyl acetate in hexane. When analyzed by TLC, this fraction is free from tri-, di-, and mono-glycerides and phospholipids, but does contain cholesteryl esters. However, these esters do not interfere with subsequent GC/MS analysis of methyl palmitate or quantitative GC analysis of plasma FFA. Samples were dried in a SpeedVac concentrator. Heptane (100 μL) was added to each tube and samples were transferred to autosampler vials for subsequent quantitative GC and GC/MS analyses. This method has been previously described in detail (2Wolfe R.R. Radioactive and Stable Isotope Tracers in Biomedicine: Principles and Practice of Kinetic Analysis. Wiley-Liss, New York1992Google Scholar). Briefly, C17:0 internal standard was added to plasma or water blanks, lipids were extracted with hexane, dried, and FFA were separated by TLC on silica gel plates using a hexane–ether–formic acid solvent system. FFA were visualized by spraying with 0.01% rhodamine 6G. TLC scrapings containing FFA were extracted with chloroform–methanol 3:1, dried, and FAME were prepared by reaction with BF3 in methanol. Variations of the two basic procedures were used to validate the CH3I/SPE procedure or to delineate artifacts associated with the TLC/BF3 procedure. One test was used to evaluate the reliability of the solvent extraction procedures; lipids were extracted from plasma with chloroform–methanol (3Bligh E.G. Dyer W.J. A rapid method of total lipid extraction and purification.Can. J. Biochem. Physiol. 1959; 37: 911-917Google Scholar) rather than with hexane and the CH3I/SPE and TLC/BF3 procedures were conducted on the lipid extract. A second test was used to evaluate the effect of varying the size of the FFA band recovered from TLC on FFA concentration and enrichment. Eight identical aliquots of extracted plasma lipids were separated by TLC. After identification of bands by rhodamine, TLC scrapings were recovered which were approximately 7 mm wide (“narrow” cuts, which were entirely within the confines of the visible FFA band) or 12 mm wide (“broad” cuts which extended beyond the visible FFA band) (four lanes per test). A third test evaluated the specificity of the CH3I methylation procedure by adding approximately 1 mg of triglyceride, containing only C15:0 fatty acid (Nu-Chek-Prep), to 0.25-mL plasma samples before FFA analysis. A fourth test evaluated the importance of the order of FFA extraction and methylation procedures; plasma FFAs were methylated with iodomethane in situ within plasma before hexane extraction, rather than the normal CH3I/SPE procedure in which lipids were extracted with hexane before methylation. A fifth test evaluated methylation and TLC recovery procedures on FFA standards containing 2H2palmitate and C17:0 internal standard. FFA concentrations and palmitate TTR were measured on standards which were methylated with either BF3/methanol or CH3I, with or without separation by TLC or SPE. Post-TLC FFA methylation conditions evaluated included methylation with BF3/methanol or CH3I, either directly on TLC scrapings or after various solvent conditions which were used to extract FFA (chloroform–methanol 3:1 or 2:1 (vol:vol), single vs. multiple extractions). A final test evaluated the use of TLC rather than SPE for FAME recovery. Identical aliquots (100 μL) of FAME standards were placed directly in autosampler vials or dried in a SpeedVac concentrator, recovered from SPE or TLC, dried in a SpeedVac concentrator again, and with 100 for quantitative GC analysis. solvent extraction conditions (chloroform–methanol at 3:1 or (vol:vol), or were used for TLC recovery and compared with the normal SPE recovery procedure. FAME concentrations to the C17:0 methyl ester internal standard were and recovery of C17:0 methyl ester was by GC in recovered samples. GC analysis was by using a GC with a 30 m × mm Bellefonte, PA) and a flame ionization detector. was by using the FAME standard to of FAME to of a C17:0 internal standard. plasma only for major FFAs and are fatty acids that less than 2% of the total FFA concentration and are not The isotopic enrichment of methyl palmitate was measured by ionization GC/MS by using a with a 30 m × 0.25 mm with mass to of and the for and palmitate methyl were (2Wolfe R.R. Radioactive and Stable Isotope Tracers in Biomedicine: Principles and Practice of Kinetic Analysis. Wiley-Liss, New York1992Google Scholar, R.R. of methyl palmitate isotope by ionization gas Scholar). from chemical ionization of methyl palmitate R.R. of methyl palmitate isotope by ionization gas Scholar, G. S. accuracy and of gas measurements for metabolic Scholar) were by that similar were obtained for all samples measured within a single analysis. were obtained by measurement of appropriate isotopic enrichment standards. and were from Chemical are standard obtained by the TLC/BF3 and CH3I/SPE methods were compared by using a test for samples. A of was 1 the from an in vivo infusion study which caused a 3-fold range in plasma FFA concentrations were higher and palmitate isotopic were in samples analyzed using the TLC/BF3 procedure than the CH3I/SPE The in palmitate concentration between methods was by a in palmitate enrichment. The differences between methods in both concentration and enrichment with plasma FFA differences in measured FFA concentration between the CH3I/SPE and TLC/BF3 methods were for and but not for C16:1 and C18:1 FFA concentrations measured by TLC/BF3 were than measured by FFA with one were the measured by both FFA with two had concentrations measured by TLC/BF3 than by The concentration range measured for each FFA and the between the two methods C16:0 13 C16:1 C18:0 C18:1 C18:2 A analysis of the from 2 found that the fatty acids had and that were not from the for C18:2 was less than with an that was not than The fraction of total fatty acids as palmitate was higher by TLC/BF3 compared with CH3I/SPE samples that only the C17:0 internal standard were analyzed by both evaluated by the TLC/BF3 procedure the of of and C18:2 that and of the sample concentrations that were measured for these fatty acids; C16:1 and C18:1 were In contrast, blanks evaluated by the CH3I/SPE procedure did not of of these evaluate the effect of varying the size of the FFA band recovered from TLC, a pool of plasma lipids containing 2H2palmitate and C17:0 internal standard was prepared by combining chloroform–methanol 3:1 from several plasma samples. Identical aliquots from this pool were separated on a TLC and the FFA bands were recovered using or Identical were obtained using methylation with BF3/methanol or by extracting FFA with 3:1 by methylation with CH3I 2 for each the for both were for major plasma FFA are in TLC bands in higher concentrations than bands for all fatty acids with the of Although the and bands generated C16:0 was no in palmitate isotopic enrichment. GC/MS with to the of of other than in samples or blanks recovered from TLC not is the effect of TLC band size from of similar to which affect the quantitative GC of and bands of plasma FFA recovered from are 1 for FFA concentrations of identical aliquots of extracted plasma in a new are 1 for FFA concentrations of identical aliquots of extracted plasma The use of hexane for plasma lipid was critical to of plasma lipids when using the CH3I/SPE procedure. The use of extraction solvent concentrations for all FFA, and C18:2 concentration was higher and palmitate enrichment was when chloroform–methanol extraction was used compared with hexane extraction extraction of plasma with containing by CH3I methylation and in the of a of C15:0 methyl ester by ionization This was when the normal CH3I/SPE procedure with hexane extraction was used not of extraction procedures on plasma fatty acid concentrations by the CH3I/SPE are 1 for FFA concentrations Identical aliquots of plasma were by the normal CH3I/SPE method using hexane extraction, or by extracting plasma lipids with chloroform–methanol 3:1 with subsequent CH3I methylation and SPE. in a new are 1 for FFA concentrations Identical aliquots of plasma were by the normal CH3I/SPE method using hexane extraction, or by extracting plasma lipids with chloroform–methanol 3:1 with subsequent CH3I methylation and SPE. Identical were obtained the CH3I methylation reaction was performed on lipids that were extracted from plasma by hexane or on lipids in situ within plasma with subsequent hexane extraction measured FFA concentrations did not from the of A sample of the FFA standard with 2H2palmitate and C17:0 internal standard added was prepared and identical aliquots were subjected to that either did or did not TLC. which included TLC identical FFA concentrations and TTR methylation with BF3/methanol directly on TLC scrapings, or extraction procedures of TLC scrapings single or extraction with chloroform–methanol 3:1, or single extraction with chloroform–methanol by methylation with BF3/methanol. Three that did not TLC identical FFA concentrations and TTR methylation with BF3/methanol without methylation with CH3I without and methylation with CH3I by separation by SPE. by TLC in higher concentrations for C16:0 and C18:0 compared with the and and concentrations for C18:1 TLC and procedures the concentrations for C14:0 and The higher C16:0 concentration with TLC was associated with a 2H2palmitate isotopic enrichment vs. were performed to evaluate the specificity of FFA methylation with CH3I and subsequent FAME separation by SPE. with CH3I was specific for free fatty Fatty acid methyl esters were not generated when of triglyceride, cholesteryl or standards were to the CH3I reaction conditions not The FAME standard was used to evaluate fractionation and recovery of FAME by SPE compared with TLC. Identical aliquots of the FAME standard in hexane were either placed in autosampler vials or dried, recovered from SPE or TLC, and to a final with hexane. aliquot was then analyzed by quantitative FAME concentrations of all fatty acids were identical to the when the standard was recovered from SPE In to the contamination with FFA with the TLC blanks by using the SPE generated of FAME not of C17:0 methyl ester from SPE as by quantitative GC compared to a standard not subjected to SPE. In contrast, TLC in recovery of FAME compared to the C17:0 internal FAME recovery was The reduced recovery of FAME was for various extraction including chloroform–methanol 3:1 or and of C17:0 methyl ester by quantitative GC from TLC was reduced compared with no major differences were between extraction In this we describe a new procedure for measuring plasma free fatty acid concentrations by quantitative GC and fatty acid isotopic enrichment by This procedure involves the use of iodomethane to plasma FFA and SPE to separate FAME. of the was using FFA or FAME and to provide more reliable than a TLC/BF3 The procedures are to the major which of the plasma fatty acids; tests be to these studies to FFA The new CH3I/SPE method is specific for free fatty acids without from esterified fatty The of fatty acids with CH3I is more convenient than with the that CH3I is and does not rapid methylation of plasma fatty acids for gas Lipid Res. Scholar). However, CH3I is a and be used in a this approach was not by the two major problems associated with TLC separation of FFA, contamination of FFA and chemical fractionation of FFA. We contamination with FFA when TLC separation was This contamination was as FFA isotopic of and the of FFA when were Although of concentrations the accuracy of the TLC/BF3 method for FFA this contamination to enrichment measurements because of isotopic of FFA fractionation were when using TLC. chemical fractionation was by recovery of FFA to the C17:0 internal standard. This fractionation not be by varying solvent extraction This was not caused by the use of BF3/methanol as a methylation because similar problems were when FFA were extracted from TLC and then methylated with The second form of fractionation, fractionation, was on the that measured FFA concentrations on the ratio to the C17:0 internal was by the size of the TLC band used for FFA recovery. the TLC band size the recovery of FFA to the recovery of the C17:0 internal standard. The concentration of palmitate was by TLC size the isotopic enrichment was that the differences in concentration were to separation between C16:0 and C17:0 rather than contamination with extraneous the C16:0 with fractionation was for all FFA with the of fractionation because the size of the visualized FFA band for TLC scrapings is In to these problems of FFA separation by TLC, recovery of FAME by SPE using 2% ethyl acetate in hexane provided without chemical fractionation and was not subject to problems which in FFA measurements by the new procedure were and chloroform–methanol not be used to plasma because of fatty acids under these This in FFA concentrations as methyl and isotopic of palmitate because these methyl esters are formed before FAME are separated from other plasma the FAME were separated by SPE rather than TLC because we that TLC in FAME recovery and chemical fractionation which was as concentrations of FAME to the C17:0 methyl ester internal standard. The use of the TLC/BF3 procedure to plasma samples affect the accuracy of in vivo by using fatty acid contamination with FFA during TLC a of palmitate isotopic enrichment by approximately on the concentration of FFA in the sample A decrease in isotopic enrichment measured by TLC in an approximately of the palmitate rate of which is by the infusion rate of palmitate by the isotopic enrichment of and of on 1959; Scholar). FFA concentration measurements are to palmitate to total FFA which is by palmitate by the of total FFA as an increase in the of palmitate to total FFA from by to by in a decrease in total FFA However, these two associated with TLC of palmitate with of the to each that the effect on total FFA be systematic measurement are to affect the of from studies that are to differences or in FFA as during S. in lipid and during in J. Physiol. Scholar, S. R.R. and kinetics during in normal and Scholar) or S. R.R. during in and J. Physiol. Scholar, S. J. S. and during in and J. Physiol. Scholar). However, of measuring plasma FFA concentration and enrichment have a more in measurement of by D. G. S. of on lipid kinetics in and J. Physiol. systematic in in or of the In the CH3I/SPE method provides a and reliable method for plasma FFA concentration and enrichment. The method is not subject to the problems of contamination, reduced recovery, and fractionation associated with recovery of FFA by TLC. This was by and from the of fatty acid methyl esters free fatty acids gas chromatography gas rate of solid phase extraction layer chromatography to