Measurement of Plasma Fatty Acids
Modification of the method of Lagerstedt et al. was used to measure total fatty acid concentrations in plasma. In brief, a 100µL plasma sample was spiked with 100µL of a mixture of 11 internal standards (fatty acids labeled with stable isotopes) to account for recovery. Esterified fatty acids were hydrolyzed from lipids such as triglycerides, phospholipids and cholesteryl esters using sequential treatment with acid then base. Following base hydrolysis, the samples were re-acidified and total fatty acids were hexane-extracted from the matrix along with internal standards. Extracts were derivatized with pentafluorobenzyl bromide in the presence of triethylamine to form pentafluorobenzyl (PFB) esters and were reconstituted in hexane.
PFB-fatty acid derivatives were injected onto a capillary gas chromatographic column to resolve individual cis-fatty acids of interest from other matrix constituents. Analytes were detected using electron capture negative-ion mass spectrometry. Twenty-four cis-fatty acids (6 saturated- (SFA), 7 monounsaturated- (MUFA), and 11 polyunsaturated- (PUFA)) were quantified using selected ion monitoring; for each fatty acid, recovery was estimated and results were adjusted using the most appropriate isotopically-labeled internal standard.
Separations were carried-out on a 0.25mm × 60m (0.25µm film thickness) cyanopropyl-methylpolysiloxane phase (50:50) capillary column at a constant flow rate of 2 mL/min helium carrier gas. Injection volume per sample was 1µL; split ratios varied from 50:1 to 150:1. Injector temperature was maintained at 240°C. The column heating oven started at 170°C then ramped 30°C/min to 230°C and held for 10 min. The temperature then ramped 2°C/min to 250ºC and held for 4.5 minutes. All ramping of temperatures was linear. The total run time was 26.5 minutes per sample.
Methane (reagent gas) flow rate was set at 40% of the total flow (actual flow: 2mL/min). The mass spectrometer source was set at 170°C and the quadrupole was set at 150°C.
The acquisition method operated in negative ion mode. A selected ion monitoring (SIM) mass was monitored for each analyte and isotopically labeled standard. The 35 SIM masses were divided into SIM segments (5-7 SIM masses per segment) based on retention time. In some segments enhanced detector sensitivity was used to quantitate low abundance fatty acids.
Fragmentation of PFB-fatty acid esters using negative chemical ionization (NCI) resulted in a reproducible loss of the PFB moiety giving a stable carboxylate anion (mass minus 1 amu). Filament voltage, inlet temperature, percent reagent gas flow, and source temperature were optimized to yield optimum intensity of selected compounds.
Internal standards were labeled with deuterium or carbon-13.
| Fatty Acid Internal Standard
|| SIM Mass
In the following table, SIM masses are grouped by the internal standard that was used to normalize recovery for the group of fatty acids. Lipid Name indicates the carbon number in the long chain: number of double bonds in the carbon chain and the n- position of the bond closest to the terminal carbon (farthest from the carbonyl carbon).
SIM masses and limits of detection (LOD in µmol/L) for 24 fatty acids
||Fatty Acid Code
Quantitation was accomplished by comparing the peak area of the analyte in the unknown with peak areas of known amounts of fatty acid based on calibration curves. Calibration curves were prepared for each assay using four levels (Cal 10 - Cal 40) per fatty acid. Calculations were corrected based on the peak area of the internal standard in the unknown compared with the peak area of the internal standard in the calibrator solution. Best fits of concentration versus response were obtained when 1/x2 was used for all fatty acids except palmitic which used a quadratic fit to calibrate the assay. Peak area ratios (quantitation m/z: internal standard m/z) were linearly related to concentration. Calibration curves were required to have r2 of at least 0.98 and calibrators were required to be within 15% of their nominal values. The overall average r2 (SD) of calibration curves, including all analytes in all assays, was 0.999 (0.002).
Blanks were analyzed concurrently with the samples and were used to determine if any analyte contamination had occurred. Blanks were required to be less than 20% of the lowest calibrator (Cal 10).
Concentrations of fatty acids in the four calibration standards
|Fatty Acid Code
||Calibrator Cal 40 (µmol/L)
||Calibrator Cal 30 (µmol/L)
||Calibrator Cal 20 (µmol/L)
||Calibrator Cal 10 (µmol/L)
The laboratory and method were certified according to the Clinical Laboratory Improvement Amendment (1988) guidelines
Units of plasma were obtained from Solomon Park Research Laboratories (Kirkland, WA) and were screened using the methodology described above to obtain the concentrations of each of 24 fatty acids per unit. All quality control pools were prepared by blending units to achieve as best as possible, representative low, medium and high concentrations. Aliquots were dispensed into 2mL cryovials and stored at -70°C.
Before each analytical run, calibration standards, bench QC materials and plasma samples were thawed from -70°C to room temperature. Calibrators and QC materials were prepared and analyzed in parallel with the unknown samples.
Typical assays consisted of 4 calibration samples, 1 blank, 3 bench quality control samples, 34 subject person’s plasma followed by 3 quality control samples. Typically, one in every 20 subject person’s plasma in the assay was actually a blind QC pool which was labeled to resemble a subject person specimen. Two blind QC pools were used throughout the period of analysis of the survey specimens. Their results were required to be within 3 standard deviations of the characterization means for each analyte. Blind QC pools were characterized in singlicate in >40 assays. Mean CV (SD) for all analytes in both blind QC pools was 9% (10%); excluding DE1_N, mean CV (SD) was 7% (3%).
Bench QC pools (open label) were named as follows:
Bench QC pools were characterized to determine the quality control limits by analyzing duplicate samples in 22 assays from 1/13/2010 to 3/24/2010. After establishment of the control limits of the pools, QC samples contained within each analytical run were evaluated for validity by use of a multi-rule quality control. Mean CV (SD) for all analytes in the three bench QC pools was 8% (10%); excluding DE1_N, mean CV (SD) was 6% (2%).