This is the first report of the application of silver-ion impregnated high-performance liquid chromatography (Ag(+)-HPLC) to the separation of complex mixtures of conjugated linolenic acid (CLA) isomers present in commercial CLA sources and foods and in biological specimens. This method showed a clear separation of CLA isomers into three groups related to their trans,trans, cis,trans or trans,cis, and cic,cis configuration of the conjugated double-bond system. In addition, this method separated individual positional isomers of the conjugated diene system within each geometrical isomeric group. Following Ag(+)-HPLC isolation, gas chromatography (GC)-electron impact mass spectrometry, and GC-direct deposition-Fourier transformed infrared spectroscopy were used to confirm the identity of two major positional isomers in the cis/trans region, i.e., delta 8,10- and delta 11,13-octadecadienoic acids, which had not been chromatographically resolved previously. Furthermore, the potential of this method was demonstrated by showing different Ag(+)-HPLC profiles exhibiting patterns of isomeric distributions for biological specimens from animals fed a diet containing a commercial CLA preparation, as well as for a commercial cheese product.

Pigs were fed a commercial conjugated linoleic acid (CLA) mixture, prepared by alkali isomerization of sunflower oil, at 2% of the basal diet, from 61.5 to 106 kg live weight, and were compared to pigs fed the same basal diet with 2% added sunflower oil. The total lipids from liver, heart, inner back fat, and omental fat of pigs fed the CLA diet were analyzed for the incorporation of CLA isomers into all the tissue lipid classes. A total of 10 lipid classes were isolated by three-directional thin-layer chromatography and analyzed by gas chromatography (GC) on long capillary columns and by silver-ion high-performance liquid chromatography (Ag+-HPLC); cholesterol was determined spectrophotometrically. Only trace amounts (<0.1%; by GC) of the 9,11-18:2 cis/trans and trans,trans isomers were observed in pigs fed the control diet. Ten and twelve CLA isomers in the diet and in pig tissue lipids were separated by GC and Ag+- HPLC, respectively. The relative concentration of all the CLA isomers in the different lipid classes ranged from 1 to 6% of the total fatty acids. The four major cis/trans isomers (18.9% 11 cis,13 trans-18:2; 26.3% 10 trans,12 cis-18:2; 20.4% 9 cis,11 trans-18:2; and 16.1% 8 trans, 10 cis-18:2) constituted 82% of the total CLA isomers in the dietary CLA mixture, and smaller amounts of the corresponding cis,cis (7.4%) and trans,trans (10.1%) isomers were present. The distribution of CLA isomers in inner back fat and in omental fat of the pigs was similar to that found in the diet. The liver triacylglycerols (TAG), free fatty acids (FFA), and cholesteryl esters showed a similar pattern to that found in the diet. The major liver phospholipids showed a marked increase of 9 cis,11 trans-18:2, ranging from 36 to 54%, compared to that present in the diet. However, liver diphosphatidylglycerol (DPG) showed a high incorporation of the 11 cis,13 trans-18:2 isomer (43%). All heart lipid classes, except TAG, showed a high content of 11 cis,13 trans-18:2, which was in marked contrast to results in the liver. The relative proportion of 11 cis,13 trans-18:2 ranged from 30% in the FFA to 77% in DPG. The second major isomer in all heart lipids was 9 cis,11 trans-18:2. In both liver and heart lipids the relative proportions of both 10 trans,12 cis-18:2 and 8 trans, 10 cis-18:2 were significantly lower compared to that found in the diet. The FFA in liver and heart showed the highest content of trans,trans isomers (31 to 36%) among all the lipid classes. The preferential accumulation of the 11 cis,13 trans-18:2 into cardiac lipids, and in particular the major phospholipid in the inner mitochondrial membrane, DPG, in both heart and liver, appears unique and may be of concern. The levels of 11 cis,13 trans-18:2 naturally found in foods have not been established.

Commercial cheese products were analyzed for their composition and content of conjugated linoleic acid (CLA) isomers. The total lipids were extracted from cheese using petroleum ether/diethyl ether and methylated using NaOCH3. The fatty acid methyl esters (FAME) were separated by gas chromatography (GC), using a 100-m polar capillary column, into nine minor peaks besides that of the major rumenic acid, 9c,11t-octadecadienoic acid (18:2), and were attributed to 19 CLA isomers. By using silver ion-high performance liquid chromatography (Ag+ -HPLC), CLA isomers were resolved into seven trans,trans (5-9%), three cis/trans (10-13%), and five cis,cis (<1%) peaks, totaling 15, in addition to that of the 9c,11t-18:2 (78-84%). The FAME of total cheese lipids were fractionated by semipreparative Ag+ -HPLC and converted to their 4,4-dimethyloxazoline derivatives after hydrolysis to free fatty acids. The geometrical configuration of the CLA isomers was confirmed by GC-direct deposition-Fourier transform infrared, and their double bond positions were established by GC-electron ionization mass spectrometry. Reconstructed mass spectral ion profiles of the m + 2 allylic ion and the m + 3 ion (where m is the position of the second double bond in the parent conjugated fatty acid) were used to identify the minor CLA isomers in cheese. Cheese contained 7t,9c-18:2 and the previously unreported 11t,13c-18:2 and 12c,14t-18:2, and their trans,trans and cis,cis geometric isomers. Minor amounts of 8,10-, and 10,12-18:2 were also found. The predicted elution orders of the different CLA isomers on long polar capillary GC and Ag+ -HPLC columns are also presented.

Operating from one to six silver ion-high-performance liquid chromatography (Ag+-HPLC) columns in series progressively improved the resolution of the methyl esters of conjugated linoleic acid (CLA) isomeric mixtures from natural and commercial products. In natural products, the 8 trans, 10 cis-octadecadienoic (18:2) acid was resolved from the more abundant 7 trans, 9 cis-18:2, and the 10 trans, 12 cis-18:2 was separated from the major 9 cis, 11 trans-18:2 peak. In addition, both 11 trans, 13 cis-18:2 and 11 cis, 13 trans-18:2 isomers were found in natural products and were separated; the presence of the latter, 11 cis, 13 trans-18:2, was established in commercial CLA preparations. Three Ag+-HPLC columns in series appeared to be the best compromise to obtain satisfactory resolution of most CLA isomers found in natural products. A single Ag+-HPLC column in series with one of several normal-phase columns did not improve the resolution of CLA isomers as compared to that of the former alone. The 20:2 conjugated fatty acid isomers 11 cis, 13 trans-20:2 and 12 trans, 14 cis-20:2, which were synthesized by alkali isomerization from 11 cis, 14 cis-20:2, eluted in the same region of the Ag+-HPLC chromatogram just before the corresponding geometric CLA isomers. Therefore, CLA isomers will require isolation based on chain length prior to Ag+-HPLC separation. The positions of conjugated double bonds in 20:2 and 18:2 isomers were established by gas chromatography-electron ionization mass spectrometry as their 4,4-dimethyloxazoline derivatives. The double-bond geometry was determined by gas chromatography-direct deposition-Fourier transform infrared spectroscopy and by the Ag+-HPLC relative elution order.

The identity of a previously unrecognized conjugated linoleic acid (CLA) isomer, 7 trans, 9 cis-octadecadienoic acid (18:2) was confirmed in milk, cheese, beef, human milk, and human adipose tissue. The 7 trans, 9 cis-18:2 isomer was resolved chromatographically as the methyl ester by silver ion-high-performance liquid chromatography (Ag+-HPLC); it eluted after the major 9 cis, 11 trans-18:2 isomer (rumenic acid) in the natural products analyzed. In the biological matrices investigated by Ag+-HPLC, the 7 trans, 9 cis-18:2 peak was generally due to the most abundant minor CLA isomer, ranging in concentration from 3 to 16% of total CLA. By gas chromatography (GC) with long polar capillary columns, the methyl ester of 7 trans, 9 cis-18:2 was shown to elute near the leading edge of the major 9 cis, 11 trans-18:2 peak, while the 4,4-dimethyloxazoline (DMOX) derivative permitted partial resolution of these two CLA isomers. The DMOX derivative of this new CLA isomer was analyzed by gas chromatography-electron ionization mass spectrometry (GC-EIMS). The double bond positions were at delta7 and delta9 as indicated by the characteristic mass spectral fragment ions at m/z 168, 180, 194, and 206, and their allylic cleavages at m/z 154 and 234. The cis/trans double-bond configuration was established by GC-direct deposition-Fourier transform infrared as evidenced from the doublet at 988 and 949 cm(-1) and absorptions at 3020 and 3002 cm(-1). The 7 trans, 9 cis-18:2 configuration was established by GC-EIMS for the DMOX derivative of the natural products examined, and by comparison to a similar product obtained from treatment of a mixture of methyl 8-hydroxy- and 11-hydroxyoctadec-9 cis enoates with BF3 in methanol.

Conjugated linoleic acid (CLA) mixtures were isomerized with p-toluenesulfinic acid or I2 catalyst. The resultant mixtures of the eight cis/trans geometric isomers of 8,10-, 9,11-, 10,12-, and 11,13-octadecadienoic (18:2) acid methyl esters were separated by silver ion-high-performance liquid chromatography (Ag+-HPLC) and gas chromatography (GC). Ag+-HPLC allowed the separation of all positional CLA isomers and geometric cis/trans CLA isomers except 10,12-18:2. However, one of the 8,10 isomers (8cis, 10trans-18:2) coeluted with the 9trans,11cis-18:2 isomer. There were differences in the elution order of the pairs of geometric CLA isomers resolved by Ag+-HPLC. For the 8,10 and 9,11 CLA isomers, cis,trans eluted before trans,cis, whereas the opposite elution pattern was observed for the 11,13-18:2 geometric isomers (trans,cis before cis,trans). All eight cis/trans CLA isomers were separated by GC on long polar capillary columns only when their relative concentrations were about equal. Large differences in the relative concentration of the CLA isomers found in natural products obscured the resolution and identification of a number of minor CLA isomers. In such cases, GC-mass spectrometry of the dimethyloxazoline derivatives was used to identify and confirm coeluting CLA isomers. For the same positional isomer, the cis,trans consistently eluted before the trans,cis CLA isomers by GC. High resolution mass spectrometry (MS) selected ion recording (SIR) of the molecular ions of the 18:1, 18:2, and 18:3 fatty acid methyl esters served as an independent and highly sensitive method to confirm CLA methyl ester peak assignments in GC chromatograms obtained from food samples by flame-ionization detection. The high-resolution MS data were used to correct for the nonselectivity of the flame-ionization detector.

No abstract

A commercial mixture of conjugated linoleic acid (CLA) isomers, reportedly consisting of six components, was recently resolved into 12 peaks attributed to CLA isomers using silver‐ion high performance liquid chromatography (Ag+‐HPLC). In this study, the coupling of two analytical silver‐ion high performance liquid chromatography columns (tandem‐column Ag+‐HPLC) in series led to the enhanced resolution of CLA isomers. Many CLA isomers were baseline resolved and the pair 18 : 2 8,10 c/t and 18 : 2 7,9 c/t found in cheese products, was resolved for the first time. In this work, a similar commercial CLA mixture was separated into 16 peaks, while CLA isomers from cheese also gave rise to 16 peaks. As expected, the CLA isomers were separated into three geometric groups in the order trans,trans, cis/trans, and cis,cis. Semi‐preparative Ag+‐HPLC, followed by gas chromatography–mass spectroscopy of the dimethyloxazoline derivatives, was used to confirm the identity of the newly resolved positional CLA isomers. The double bond configuration of CLA isomers was established by gas chromatography–Fourier transform infrared spectroscopy. Two minor t,t CLA isomers found in cheese, presumably 18 : 2 t6t8 and 18 : 2 t13t15, were also separated. The CLA isomeric composition of 16 commercial cheese products was determined.