| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Research Unit, Departments of
1
Gastroenterology and
2
Biochemistry. Hospital Universitari Germans Trias i Pujol, 08916 Badalona, Catalonia, Spain.
aAddress correspondence to this author at: Department of Gastroenterology, Hospital Universitari Germans Trias i Pujol, Carretera del Canyet s/n, 08916 Badalona, Spain. Fax 34-93-497-8951; e-mail mgassull{at}ns.hugtip.scs.es.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterial and MethodsResultsDiscussionReferences |
|---|
Methods: Lipoproteins were separated by sequential ultracentrifugation. Free or esterified heptadecanoic acid was used as internal standard. Fatty acids were derivatized to trimethylsilyl (TMS) esters. GC separation was in isothermal mode at 210 °C for 27 min. For both TMS-palmitate and TMS-oleate, M and [M + 1] signals were simultaneously acquired with a dual acquisition program in single-ion monitoring mode. Calibration mixtures containing increasing amounts of labeled fatty acids were prepared gravimetrically to construct calibration curves for isotopic enrichment. Likewise, five calibration curves (for increasing concentrations) were constructed for each fatty acid; this allowed selection of the most appropriate curve for the concentration in a plasma sample.
Results: Oleic acid-TMS ester was clearly separated from that of its stereoisomer, elaidic acid. Within a 10-fold concentration range, the isotopic ratio was independent on the amount of the analyte in the sample, with a maximum uncertainty of 0.34% in terms of molar percent excess. In addition, the within- and between-day imprecision (CV) of the method was <1%.
Conclusion: Results obtained with this method are independent of concentration and sufficiently precise for tracking 1-13C-labeled palmitic and oleic acids in biological samples
|
Introduction |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsDiscussionReferences |
|---|
The most suitable methods for such studies are isotope-dilution mass spectrometry methods, including gas chromatography–mass spectrometry (GC/MS) 1 and isotope ratio mass spectrometry coupled with a combustion chamber furnace and GC (GC/C/IRMS) (4). Although GC/MS requires relatively high amounts of tracer for detection, it is often used for measurements in biological fluids because it uses very small amounts of sample and can be used with a wide range of derivatization procedures, with optimal GC and MS performances (5). IRMS is sensitive to much smaller amounts of tracer, but needs larger sample volumes than GC/MS, as well as samples with very high purity (i.e., containing a single molecular species), making it relatively unsuitable for measurements in complex biological specimens (6). GC/C/IRMS would be preferred in the event that a good GC separation of compounds is achieved, but GC/MS is easier to use and much less expensive than GC/C/IRMS and is available in many laboratories. Furthermore, GC/MS provides structural information on the analyte.
Measuring isotopic enrichment by GC/MS requires controlling for many potential sources of variability, including analyte concentration, the intensity of the ion to be monitored, the spectrometric and chromatographic conditions, and the design of GC/MS (7). Furthermore, GC/MS assessment of fatty acids has additional drawbacks. One major difficulty is that the isotope ratios obtained are mostly dependent on the ionic concentration in the system, particularly when methyl derivatives are used (8). Nevertheless, most authors use methyl esters as derivatized compounds (8)(9)(10)
To overcome these difficulties in the GC/MS assessment of stable-isotope-labeled long-chain fatty acids in biological specimens, we have developed a new analytical protocol that includes the use of trimethylsilyl (TMS) esters as derivatized compounds and the calculation of isotopic enrichment from calibration curves obtained from mixtures containing known isotopic enrichments.
|
Material and Methods |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsDiscussionReferences |
|---|
The derivatizing agent bis-N,O-trimethylsilyl trifluoroacetamide (BSTFA) + 5% trimethylsilane was purchased from Fluka. The remaining organic solvents (acetonitrile, chloroform, methane, n-hexane, methanol, and ethanol) and the silica thin-layer chromatography plates were purchased from Teknokroma.
tracer administration and sampling
Labeled fatty acids (1 g of [1-13C]palmitic acid + 1 g of [1-13C]oleic acid) were mixed with 10 g of sunflower oil. After a 12-h overnight fast, the 15 probands [9 males and 6 females; mean (± SD) age, 54.0 ± 14.6 years] received the mixture orally together with a standard meal [20 g of margarine, three toasted bread slices, and 250 mL of a liquid nutritional supplement (Meritene®; Novartis Consumer Health, S.A.)]. Venous plasma samples were obtained at baseline and 360 min after administration of the tracer.
sample preparation
Chylomicrons (ChMs), VLDL, LDL, and HDL were isolated from plasma by sequential ultracentrifugation according to the Havel method (11), adapted to the sample volume for this study. The variability of triglyceride and cholesterol measurements in the different lipoproteins was used as an estimate of the precision of the ultracentrifugation method. Triglycerides were measured at 524 nm by the reaction of glycerophosphate oxidase coupled with the common Trinder sequence. Cholesterol was measured at 524 nm by the reaction of cholesterol oxidase coupled with a Trinder sequence.
Heptadecanoic acid (15 µL of a 1 g/L solution) was added to 400 µL of plasma as internal standard for free fatty acids, and known amounts of 10 g/L solutions of glyceryl triheptadecanoate, phosphatidylcholyl diheptadecanoate, and cholesteryl heptadecanoate were added to each lipoprotein fraction as internal standards for triglycerides, phospholipids, and cholesteryl esters, respectively.
Lipids were extracted with an chloroform–methanol mixture, according to the method of Bligh and Dyer (12). Free fatty acids, phospholipids, triglycerides, and cholesteryl esters were then separated by thin-layer chromatography on silica gel plates with hexane–diisopropyl ether–acetic acid (80:20:1 by volume) as mobile phase. Lipid fractions were recovered from plates and reextracted with chloroform–methanol (2:1 by volume). The chloroform layer was dried and hydrolyzed to fatty acids in alcoholic KOH at 70 °C for 2 h. Fatty acids were converted into their nonionic forms with HCl, reextracted in hexane, dried, and then derivatized to TMS esters with BSTFA + 5% trimethylsilane for 30 min at 70 °C. TMS esters were then dried and redissolved in 1 mL of hexane, and 1 µL of this extract was injected into the gas chromatograph-mass spectrometer by an AS 2000 autosampler (Thermo Quest).
gc/ms analysis
GC/MS measurements were performed in a MD 800 quadrupole mass spectrometer (Thermo Quest) operating in positive electronic impact set to 100 µA, connected to a CG 8060 gas chromatograph (Thermo Quest) equipped with a J&W DB-1 (60 m x 0.25 mm; 0.25-µm film thickness) column (Cromlab S.A.) with helium as the carrier gas. Injection was performed in splitless mode at 300 °C. TMS esters were separated at constant pressure (175 kPa) with the following oven program: (a) 165 °C for 1 min; (b) increase at a rate of 10 °C/min up to 210 °C; (c) hold at 210 °C for 27 min; (d) increase at a rate of 40 °C/min up to 315 °C; and (e) hold at 315 °C for 5 min. Although an oven temperature below the boiling point of the compounds would probably provide better resolution, we chose 165 °C (a temperature above the boiling point) as the starting point of the oven program because the resolution obtained was adequate with a shorter analysis time (
10 min/chromatogram).
The spectrometer was tuned manually to get the relative abundances of m/z 69, 131, 264, and 502 of perfluorotributylamine and the maximum resolution from their respective [M + 1] isotopic masses.
To assess both the isotopic enrichment and the individual fatty acid concentrations (both unlabeled and 1-13C-labeled) in the same GC/MS analysis, we designed a dual acquisition program in single-ion monitoring (SIM) mode. With this program two signals (M and [M + 1]) were recorded for each fatty acid, taking into account the loss of a methyl group (-15) from the molecular ion, typical of TMS derivatives. The following m/z ratios were acquired: 313.25 and 314.25 for TMS-palmitate, 327.27 and 328.27 for TMS-heptadecanoate, and 339.27 and 340.27 for TMS-oleate (Table 1
). Fig. 1
in the online data supplement [available at Clinical Chemistry Online (http://www.clinchem.org/content/vol48/issue6/) provides examples of the full mass spectra of the TMS esters of palmitic, heptadecanoic, oleic, [1-13C]palmitic, and[1-13C]oleic acids.
|
|
calculations
Calibration mixtures containing labeled and unlabeled fatty acids (either palmitic or oleic) were prepared gravimetrically to obtain 0%, 1.06%, 2.10%, 5.00%, 9.70%, 17.79%, and 30.06% molar percent excesses (MPE) of labeled over unlabeled compounds. Aliquots of these calibration mixtures were treated and derivatized as described above.
The basal ratio (Rb) of Am + 1/Am was obtained from the calibration mixture corresponding to 0% MPE (MPE0) by the SIM program (where A is the peak area for each fatty acid derivative at a given retention time). The remaining calibration mixtures were also analyzed, and their MS-derived MPE (MPEi) was calculated according to the equation:
 | (1) |
Calibration curves for isotopic enrichment were constructed by plotting the MPEi values against the real MPE obtained gravimetrically (13). In addition, five different calibration curves were constructed for each assayed unlabeled fatty acid (palmitic and oleic). Each curve was obtained by diluting the preceding calibration mixture by 50% [1 mL of mixture + 1 mL of chloroform; see Table 1 in the online data supplement (http://www.clinchem.org/content/vol48/issue6/)]. This allowed us to select the best calibration curve for deriving the fatty acid concentration in plasma samples, in terms of the internal standard peak area that better fit that of the sample. The concentration of each 1-13C-labeled fatty acid was finally calculated by multiplying its MPE by the concentration of its corresponding unlabeled compound.
ethical issues
The study was performed in accordance with the latest revision of the Helsinki Declaration, and informed consent was obtained from all participants.
|
Results |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsDiscussionReferences |
|---|
4 mL, this assessment was performed with 1.5 mL of plasma to evaluate the precision and accuracy of the method when there was a shortage of sample (Table 2
).
|
gc separation of fatty acids
The use of TMS esters allowed good GC separation of the individual fatty acids we aimed to identify (namely, palmitic and oleic acids) in all lipoprotein samples from every individual we studied (n = 15). Particularly, we were able to separate the TMS ester of oleic acid (cis-C18:1 n-9) from that of its stereoisomer, elaidic acid (trans-C18:1 n-9). An example is shown in Fig. 1
.
ms analysis
Typical chromatograms of the [M + 1] (top chromatograms) and M (bottom chromatograms) masses of TMS-palmitate and TMS-oleate, as acquired by the SIM program, are shown in Fig. 2, A and B
, respectively.
|
linearity
A major issue was to determine whether the Rb remains unchanged with increasing amounts of fatty acids in the sample. Accordingly, we obtained the CV based on 10 injections from solutions containing 0.2 and 2 nmol of either the TMS-palmitate or TMS-oleate calibrators. The CV was 0.25% for 0.2 nmol of palmitate, 0.40% for 0.2 nmol of oleate, 0.99% for 2 nmol of palmitate, and 1.4% for 2 nmol of oleate. The variation in MPE ranged from ±0.06 to ±0.34, meaning that, at the worst, the uncertainty of the method is <0.34%.
On the other hand, we checked the influence of the voltage of the repeller plate on the isotopic ratio. Processing the same sample with repeller voltages of 1, 5, 9, and 12 V yielded ratios of 0.2390, 0.2441, 0.2439, and 0.2469 for TMS-palmitate and 0.2606, 0.2667, 02744, and 0.2732 for TMS-oleate, respectively.
To verify that the slopes of the calibration curves for MPE were independent of the volume of sample injected, we calculated these curves for each TMS ester at different injection volumes (0.5, 1, and 2 µL). As shown in Fig. 3
, the calibration curves were almost identical regardless of the amount of analyte.
|
precision of ms measurements
The precision of the MS measurements was assessed in plasma samples. Table 3A
shows the within-day variation for repeated MPE measurements of TMS esters in ChM-associated triglycerides from baseline and 360-min samples from a single individual. For both TMS-palmitate and TMS-oleate in either sample, the CV was always <1%, producing a within-day variability in MPE measurements <0.25%.
|
The between-day variation in [M + 1]/M measurements is shown in Table 3B
with the baseline plasma samples of free fatty acids and ChM-associated triglycerides from 15 individuals. Again, the CV was <1%, and MPE varied <0.26%.
detection limit
The detection limit for the naturally occurring oleic and palmitic acids was obtained from repeated analyses of a solution containing a known amount of internal standard (2.3 µmol/L C17:0) and no palmitic and oleic acid added. The concentrations of fatty acids were derived from the areas of m/z 313.25 and m/z 339.27, as well as for m/z 327.27 for the internal standard, at the specific retention times by means of the corresponding calibration curve. The detection limit was then calculated as the mean plus 3 SD of these repeated measures. The detection limit was 0.1157 and 0.1613 µmol/L for palmitic and oleic acid, respectively.
Similarly, the detection limit for the 1-13C-labeled fatty acids was calculated from repeated MPE analyses of a solution with known amounts of the naturally occurring fatty acid but devoid of the 1-13C-labeled compound. The detection limit for MPE was calculated as the mean plus 3 SD of these repeated measures. The detection limits were 0.50 MPE for [1-13C]palmitate and 0.77 MPE for [1-13C]oleate.
recovery
The recoveries of naturally occurring and 1-13C-labeled fatty acids were assessed in separate experiments. To evaluate the recovery of naturally occurring fatty acids, we added a known amount of exogenous triglycerides to a pool of plasma, and 50 µL of glyceryl triheptadecanoate and increasing amounts (5, 10, 15, 20, and 25 µL) of both tripalmitine and triolein were added to aliquots of plasma (1 mL). The recoveries in the presence of 47–235 µg of added triglycerides were 91–132% (median, 100%) for palmitic and 72–83% (median, 80%) for oleic acid (n = 5 for each).
To assess the recovery of 1-13C-labeled fatty acids, we added increasing amounts (5–50 µL) of two calibration solutions containing [1-13C]palmitic and [1-13C]oleic acids to 500-µL aliquots of pooled plasma. For 5–50 µg of labeled fatty acid, the recoveries were 83–100% (median, 92%) for palmitic acid and 71–111% (median, 99%) for oleic acid (n = 7 for each).
|
Discussion |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsDiscussionReferences |
|---|
Silylation is another potential method for derivatization. Although experience with long-chain fatty acids is limited (14), tert-butyldimethylsilyl (TBDMS) esters have been used for chromatographic analysis of short-chain fatty acids (15)(16). Furthermore, this agent has been useful in GC/MS analysis of amino acids (17). On the basis of these data, we initially used TBDMS-trifluoroacetamide as a derivatizing agent. In early experiments with fatty acid calibration mixtures, the TBDMS esters obtained were very stable and suitable for assessing isotopic enrichment of both TBDMS-palmitate (M = 313; [M + 1] = 314) and TBDMS-oleate (M = 339; [M + 1] = 340). However, in plasma samples, the retention time of the TBDMS ester of oleic acid was very similar to that of its trans isomer, elaidic acid. This would interfere with tracer measurements because we administer the [1-13C]-labeled cis isomer (oleic acid). To obtain better separation of chromatographic peaks, changes in the chromatographic method, such as isothermal elution and increasing the number of column plates, were made, but they failed to yield complete resolution between oleic and elaidic TBDMS esters. This was achieved only by use of TMS-trifluoroacetamide instead of TBDMS-trifluoroacetamide as derivatizing agent (Fig. 1
). Although TBDMS esters are more apolar and thereby would provide better resolution in a DB-1 column than would TMS esters, they are also heavier and have higher boiling points. This implies that TBDMS esters require a higher elution temperature. Unfortunately, this counteracted the effect of their apolarity on resolution, and oleic and elaidic acids were not completely resolved when we used TBDMS as derivatizing agent. Moreover, TMS esters maintained constant isotopic ratios with increasing amounts of the analyte, either natural or isotopically enriched (Fig. 3
).
In contrast to previous reports (8)(9), the isotopic ratio that we measured was independent of the amount of the analyte in the sample within a 10-fold concentration range. In fact, the CV of the isotopic ratio was <1.4%, meaning that the uncertainty of the method was at worst <0.34% in terms of variation of MPE. Some methodologic peculiarities might account for these differences. As discussed earlier, we used TMS esters as the derivatized compounds because they yield more intense peaks than do methyl esters, thus nonspecifically contributing to a decrease in the variability. On the other hand, the chemical structure of methyl esters favors the concentration dependence of the isotopic ratio. In contrast to earlier reports (8), recent reports have shown that this is mostly because, during ionization, these compounds produce a McLafferty rearrangement ion at m/z 74. This ion is a strong proton donor that induces a self-chemical ionization reaction of M0, thereby spuriously increasing [M + 1]+ (18)(19). This process should not take place when TMS esters are used.
Another major feature of the present protocol is that MPEs were derived from calibration curves constructed from MPEs obtained gravimetrically. This procedure obviates some problems inherent to the use of crude MPEs derived from ratios of spectrometric areas. These problems include changes in these ratios depending on the concentration of analyte, the nonlinear increase in MS response with growing amounts of tracer, and variations of the GC/MS conditions with time (e.g., loss of column performance with use or tuning conditions).
The design of a dual acquisition program is an additional advantage of the method that deserves mention. With a single assay, it allows acquisition of m/z peaks to estimate both isotopic enrichment and the concentration of the naturally occurring fatty acid. In addition to minimizing the probabilities of error, it saves time and improves the output of the instrument.
All these methodologic actions have ultimately rendered a quite precise GC/MS method for tracing 1-13C-labeled palmitic and oleic acids in biological samples, with within- and between-day CVs <1% and a high recovery rate. The method appears to be particularly precise for 1-13C-labeled palmitic acid. Obviously, the throughput of the method is far from being optimal because it requires a very long column (60 m), which increases the run time (>37 min). It must be noted that the method is not useful for resolving the isomers of polyunsaturated fatty acids (C18 or longer), as can be inferred from Fig. 1
.
In conclusion, the proposed method provides a suitable tool for tracing 1-13C-labeled palmitic and oleic acids in biological samples. This may be particularly useful for kinetic studies on the absorption, metabolism, and clearance of these compounds in a wide range of disease conditions, including chronic liver disease, pancreatic insufficiency, inflammatory bowel disease, diabetes mellitus, and hyperlipidemia. Gaining further insight on the metabolic fate of different dietary fatty acids in these conditions in both the clinical and the experimental setting would contribute to improving the design of nutritional strategies (i.e., enteral and parenteral feeding) and dietary recommendations for these patients.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsDiscussionReferences |
|---|
-hydroxylase activity in experimental animals. Eur J Biochem 1983;136:167-173.[Medline]
[Order article via Infotrieve]
-ketoisocaproic acid and leucine and evaluation of -1-13C-ketoisocaproic acid and 1-13C-leucine enrichment by gas chromatography-mass spectrometry. Anal Biochem 1994;220:308-314.[Medline]
[Order article via Infotrieve]
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  E. Cabre, J. M Hernandez-Perez, L. Fluvia, C. Pastor, A. Corominas, and M. A Gassull Absorption and transport of dietary long-chain fatty acids in cirrhosis: a stable-isotope-tracing study Am. J. Clinical Nutrition, March 1, 2005; 81(3): 692 - 701. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
; y = 0.614x; R2 = 0.9959), and 2 µL (•; y = 0.9809x; R2 = 0.9981).