Intramyocellular diacylglycerol concentrations and [U-¹³C]palmitate isotopic enrichment measured by LC/MS/MS.

Diacylglycerols (DAG) are important lipid metabolites thought to induce muscle insulin resistance when present in excess; they can be synthesized de novo from plasma free fatty acids (FFA) or generated by hydrolysis of preexisting intracellular lipids. We present a new method to simultaneously measure intramyocellular concentrations of and the incorporation of [U-13C]palmitate from an intravenous infusion into individual DAG species. DAG were extracted from pulverized muscle samples using isopropanol:water:ethyl acetate (35:5:60; v:v:v). Chromatographic separation was conducted on reverse-phase column in binary gradient using 1.5 mM ammonium formate, 0.1% formic acid in water as solvent A, and 2 mM ammonium formate, 0.15% formic acid in methanol as solvent B. We used UPLC-ESI+-MS/MS in the multiple reaction monitoring (MRM) mode to separate the ions of interest from sample. Because DAG are a neutral lipid class, they were monitored as an ammonium adduct [M+NH4]+. To measure isotopic enrichment (for 13C16:0/16:0-DAG and 13C16:0/C18:1-DAG), we monitored the basic ions as [M+2+NH4]+ and the enriched compounds as [M+16+NH4]+. We were able to measure concentration and enrichment using 20 mg of skeletal muscle samples obtained from rats receiving a continuous infusion of [U-13C]palmitate. Applying this protocol to biological muscle samples proves that the method is sensitive, accurate, and efficient.

Recently, more attention has been focused on DAG because of their purported role to induce muscle and liver insulin resistance when present in excess (1)(2)(3)(4). Because of the biological functions of DAG, especially DAG species containing saturated fatty acids, we endeavored to develop a method to precisely and rapidly measure muscle DAG.
Most methods for DAG measurement quantify total DAG content. One of the fi rst and common quantitative analysis of DAG was performed by enzymatic analysis using diacylglycerol kinase assay ( 5 ) using labeled ATP 32 . A nonradioactive technique for total DAG quantifi cation uses HPLC analysis following derivatization ( 6 ). The composition of DAG fatty acids can be determined by hydrolysis of DAG and analysis of the liberated and derivatized fatty acid by the means of gas-liquid chromatography with fl ame ionization detector (GLC-FID). Typically, DAG are separated from an extract of total lipids using thin-layer chromatography (TLC) or HPLC. TLC separation requires long processing times, leads to loss of analyte, and is susceptible to sample contamination. Development of mass spectrometry methods has facilitated analysis of DAG by reducing the analysis time, the amounts of consumables, and the likelihood of sample contamination. Diacylglycerols and triacylglycerols are classes of neutral lipids without any permanent charge; therefore, they are usually analyzed by ESI/MS as a sodium [M + Na] + , lithium [M + Li] + , or ammonium adduct that is created spontaneously by adding the appropriate salt to the solvent system ( 7,8 ) or after the derivatization procedure ( 9,10 ).
Because the potential contribution of circulating FFA to DAG synthesis is unknown, we developed a new method for measuring both concentration and isotopic enrichment of DAG in tissue from animals infused intravenously with uniformly labeled [U- 13 C]palmitate. Our method displays excellent inter-and intraassay reproducibility, low Abstract Diacylglycerols (DAG) are important lipid metabolites thought to induce muscle insulin resistance when present in excess; they can be synthesized de novo from plasma free fatty acids (FFA) or generated by hydrolysis of preexisting intracellular lipids. We present a new method to simultaneously measure intramyocellular concentrations of and the incorporation of [U- 13  We were able to measure concentration and enrichment using 20 mg of skeletal muscle samples obtained from rats receiving a continuous infusion of [U-concentrated albumin solution (essentially fatty acid-free, Sigma-Aldrich) with a ratio of 1 g of albumin per 10 mg tracer. The solution was heated to 60°C, stirred until clear, and sonicated for 1 min with probe sonicator. Despite the use of fatty acid-free albumin, residual unlabeled palmitate decreased fi nal isotopic enrichment of the infusates to 86.65% (± 0.18% SD, n = 6 independent preparations). The solution was diluted with sterile phosphate buffered saline to a fi nal concentration of 1 mM 13 C 16 -palmitate, taking into account both the tracer isotopic purity and the presence of preexisting albumin-bound palmitate. This concentration was used to calculate infusion rates. The total palmitate concentration in infusate (M+16 palmitate and all other sources) was 1.134 mM. The infusate was passed through 0.2 m sterile syringe fi lter, divided into aliquots, and stored at Ϫ 20°C until use. Plasma palmitate concentration and isotopic enrichment in both the infusate and plasma samples was measured according to Persson et al. ( 14 ).

Isotope infusion and tissue collection
The food was withdrawn from animals 6 h prior to infusion. Animals received pentobarbital anesthesia (40 mg/kg of body weight) and were placed on a heating blanket. The tracer was delivered into proximal dorsal tail vein using a Model NE-1000 syringe pump (New Era Syringe Pumps, Farmingdale, NY) via 0.5 mm id tubing (Rotiolab, Dreieich, Germany) and 25 gauge needle at a constant rate of 50 nmol·min Ϫ 1 ·kg Ϫ 1 body weight for 2 h.
A priming dose of 500 nmol/kg was given in the fi rst 10 s of infusion to prime the palmitate pool and accelerate isotopic equilibration. Every 15 min, a 50 l blood sample was taken from saphenous vein into a heparinized Microvette capillary tube (Stardstedt, Numbrecht, Germany). Plasma was obtained by centrifugation for analysis of palmitate concentration and enrichment. To prevent vascular volume overload , the infused volume did not exceeded 17% of total plasma volume [calculated according to Lee et al. ( 15 )] and was less than estimated 2 h urine output according to the values for Wistar rats from Rat Phenome Database ( 16 ). The tracer infusion had no effect on the level of plasma FFA. The pre-and postinfusion concentration of total plasma FFA and FFA palmitate did not differ signifi cantly (274 µmol/l versus 271 µmol/l for total FFA; 64 µmol/l versus 64 µmol/l for palmitate). The mean plasma palmitate tracer enrichment at plateau was 0.107 ± 0.005 molar percentage excess (MPE, n = 6). At the end of 2 h infusion, the last blood sample was collected from inferior vena cava, and rats were sacrifi ced by heart incision. Samples of longissimus thoracis muscle were collected and immediately frozen in liquid nitrogen for analysis of DAG concentration and enrichment.

Liquid chromatography condition
The chromatographic separation was performed using an Agilent 1290 Infi nity Ultra Performance Liquid Chromatography (UPLC). The analytical column was a reverse-phase Zorbax SB-C8 day-to-day variation, and can detect concentration and enrichment in as little as 20 mg of muscle. Measuring both concentration and isotopic enrichment of DAG and other lipid groups ( 11,12 ) can help to elucidate the contribution of plasma FFA to intramyocellular fatty acid-containing compounds. We believe that this method can be used to elucidate the dynamics of the intramyocellular DAG pool and its relationship to plasma FFA in physiological and pathological states.

Extraction of diacylglycerols
Muscle samples were pulverized and weighed, and 20 mg of each was taken for analysis. To these samples we added 50 ng of the internal standard (ISTD) (1,3 dipentadecanoyl-sn -glycerol) as well as 200 µl of homogenization buffer consisting of 0.25 M sucrose, 20 mM KCl, 50 mM Tris, and 0.5 mM EDTA, pH 7.4. After homogenization, 1.5 ml of the extraction mixture isopropanol: water:ethyl acetate, (35:5:60; v:v:v) was added to each sample. The samples were then vortexed, sonicated, and centrifuged for 10 min at 4,000 rpm (Sorvall Legend RT). The supernatant was transferred to a new tube, and the pellet was reextracted using the same extraction mixture. After centrifugation, supernatants were combined and evaporated under nitrogen. The dried sample was reconstituted in 100 µl of LC solvent B for LC/ MS/MS analysis. To prevent sample contamination, all the glassware used for extraction of lipids, including glass pipette tips and vials, was disposable. All solvents, including water, were of LC/ MS grade.

Animals and isotope infusion
Male Wistar rats weighing approximately 350 g were acclimatized for one week prior to the experiment by housing them in controlled environment under ambient temperature and humidity in 12 h light/dark cycle and feeding them standard laboratory rat chow. Experiments were approved by Institutional Animal Care and Use Committee of Medical University of Bialystok.

Preparation of standard solutions
The ISTD solution and stock solution of DAG standards for standard curve preparation were prepared in ethanol. Each 50 µl of the ISTD contained 50 ng of 1,3 15/15-DAG. Each 10 µl of the standard curve stock solution contained 50 ng of the particular DAG. After preparation of the six-point standard curve (representative plot is presented in Fig. 1 ), each aliquot was spiked with 50 µl of ISTD solution, and all aliquots underwent the extraction procedure. For enrichment analysis, a ten-point standard curve was prepared by mixing constant amounts of unlabeled standards and different amounts of labeled standards for each point ( Fig. 2 ).

Method validation
Recovery and effi ciency.

Mass spectrometry
The DAG species were analyzed as the ammonium adducts, without derivatization, using electrospray ionization conditions in the positive ion mode. To generate this charged adduct species, a 2 mM ammonium formate in solvent A and 1.5 mM ammonium formate in solvent B were used. The following conditions were employed: the drying gas temperature was 300°C, the drying gas fl ow was 5 l/min, the spray voltage was 3,500 V, sheath gas was 11 L/min, and transfer capillary was 250°C.    After chromatographic separation, the eluate was directed to the ionization source without derivatization. Because neutral lipids, such as DAG and triacylglycerols, lack a permanent charge, we analyzed DAG as the ammonium adduct [M+NH 4 ] + (the masses for precursor and product ions are provided in Table 1 ). Using the sample preparation and analytical conditions described above, we obtained clean total ion chromatogram in MRM mode ( Fig. 3 ). Enriched peaks from skeletal muscle and corresponding base peaks [M+2+NH 4 ] + are presented in Fig. 4 . Our approach contrasts to published mass spectrometry methods for DAG measurement by applying triple quadrupole mass spectrometry equipment designed specifi cally for quantitative analysis. Most LC/MS methods for DAG quantifi cation use ion trap mass spectrometers, more suitable for identifi cation then for quantifi cation, with a derivatization step before analysis ( 9,10 ). A method for DAG quantifi cation without derivatization prior to analysis has been described using a direct infusion ionization source operating in full-scan mode ( 18 ), but without chromatographic separation. The principle advantages of our method are simple sample preparation without derivatization, short run time (10 min), and accurate and precise analysis of both concentration and enrichment based on mass transition (MRM mode) and retention time. The method requires as little as 20 mg of muscle.

Standard curves for enrichment estimation
The standard curve samples contained [U- 13

Method validation
Recovery and effi ciency. The average calculated recovery of extracted standards ranged from 86 to 93%. In the biological matrix, the extraction recovery ranged from 82 to 89%.
Reproducibility. We found the concentration and iso topic enrichment method's reproducibility, both sample-to-sample and day-to-day, to be excellent ( Table 3 , concentration;  Table 4 , enrichment), indicating this method is reliable for both isotopic enrichment and concentration measurement in a biological matrix.
Sensitivity. Using the sample preparation and extraction procedure described above, we could reliably quantitate DAG concentration and isotopic enrichment using 20 mg of tissue. This amount of sample was suffi cient to precisely establish both values. The tissue mass equivalent used for each injection was comparable to only 0.6 mg of the original muscle sample. The measurement of DAG was prepared by dissolving DAG standards in 100 µl of solvent B and directly analyzing the standard solution by LC/MS/MS, while another 100 l solution was dried under gentle N 2 fl ow, extracted as described for the samples, and redissolved in 100 l of solvent B for LC/MS/MS analysis. The peak areas for DAGs [M + NH 4 ] + ions were integrated and compared. This experiment was performed three times, and the average recovery was calculated.
To estimate recovery from the biological matrix, we prepared homogenate from muscle samples that was kept at room temperature for 24 h and then incubated at 45°C for 1 h. A portion of the homogenate was extracted to ensure there was no detectable DAG. The remaining homogenate was divided into three aliquots to which 100 l of the above DAG standard mixture was added, followed by the extraction and LC/MS/MS analysis procedures. To measure the extraction recovery from sample containing the biological matrix and standard mixture, we compared the DAG peak areas of both sample types.

Reproducibility.
To test the sample-to-sample reproducibility, three rat longissimus thoracis muscle samples were taken after [U-13 C 16 ]palmitate infusion. Samples were analyzed fi ve times by LC/MS.MS, using 20 mg of tissue for each analysis. To study day-to-day reproducibility, the samples were analyzed on three different days.

RESULTS AND DISCUSSION
We used UPLC (1290 Infi nity, Agilent) connected to an Agilent 6460 triple quadrupole mass spectrometer.   Values are given as mean MPE ± SD. Values in parentheses represent day-to-day coeffi cient of variation (CV%) for the respective samples and DAG molecular species. concentration alone could be done with a smaller sample; however, any further decrease of sample requirement would introduce variation in estimation of sample weight and isotopic enrichment.

Biological application
Animals with greater plasma palmitate enrichment had greater tissue DAG enrichment ( Fig. 5A ). The muscle DAG enrichments were compared with the corresponding plasma palmitate enrichments. There was a close correlation between the enrichment of plasma FFA and tissue DAG in rat muscle samples ( Fig. 5B ). Enrichment of both molecular species of DAG displayed a linear relationship with plasma tracer enrichment. We also noticed differences between labeling effi ciency of dipalmitoyl-DAG and palmitoyl/oleoyl-DAG (curve slope and mean enrichment value under the same tracer pool), which suggests differences in synthesis or degradation of intramyocellular DAG containing saturated and monounsaturated fatty acids. These data suggest to us that plasma FFA are an important source of fatty acids for intramyocellular, de novo DAG synthesis. It also suggests that the individual intramyocellular DAG species differ in respective intracellular turnover.

CONCLUSION
We found that this UPLC-ESI/MS/MS in the MRM mode is a rapid, simple, and reliable method to simultaneously detect low levels of stable isotopic tracer enrichments ( 13 C16:0/16:0 DAG and 13 C16:0/18:1DAG) and DAG concentrations in small samples of muscle tissue. Remarkably, as little as 3 l of lipid extract, equivalent to 0.6 mg of original sample weight, can be reliably analyzed for DAG concentration and enrichment with excellent quantifi cation accuracy and sample-to-sample and dayto-day reproducibility. One limitation to this method is the limited availability of DAG species standards and enriched standards, which can be quite costly to purchase.