An HPLC method for determining desirable or undesirable conjugated linoleic acid

A binary gradient elution system with photodiode array detection at 234.5 nm was used for analysis of underivatized conjugated linoleic acid (CLA). The CLA isomers were substantially re­ tained on the columns and were distinct from saturated, unsaturated non-conjugated fatty acids, background fluctuations and endogenous substances in milk and intestinal digesta samples. Al l CLA isomers appeared as a pair of large peaks at 14.1 ±0.1 and 14.6±0.1 min of elution, respectively. Trace amounts of other unidentified conjugated dienes in these biological samples were detected after 18 min of elution. A l l endogenous substances in milk, meat and intestinal digesta samples are trans­ parent in the applied UV monitoring range (234±2 nm). Elimination of pre-column derivatization yields a less expensive, more specific and simpler analytical tool for routine determination of conju­ gated dienes in milk, meat and intestinal digesta.


INTRODUCTION
Feed ingested by ruminants is subjected to microbial attack in the rumen. Com plex plant lipids are hydrolyzed rapidly and polyunsaturated fatty acids, consisting mainly o f linolenate, undergo extensive hydrogenation and double bond migration to yield a mixture o f geometrical (cis, trans) and positional (9)(10)(11)(9)(10)(11)(12)(13)) isomers o f linoleic acid (C18:2 n-6; i.e., C L A ) (Chouinard et al., 1999;Ostrowska et al., 1999;M i r et al., 2000). Cis-9 trans-11 linoleic acid is the major isomer o f the C L A mixture (Jahreis et al., 1997;M i r et al., 2000). The amount and type o f feed can influence the microbial population o f the rumen and this can influence hydro genation and formation o f trans isomers o f milk fat ( M F ) . In western countries the contribution o f milk products to total trans fatty acid (/-FAs) intake is esti mated as 6-25% (Henninger et al., 1994). The /-FA content in food is also an indication o f the degradation o f naturally occurring unsaturated fatty acids in plants during commercial hydrogenation (Heinig et al., 1998). Due to the func tion o f fatty acids (FAs) as carriers o f fat-soluble vitamins, components o f nerve cells, membranes, hormones, and biliary acids they are o f great importance for living organisms (Heinig et al., 1998). In fact, recent studies have shown that /-FAs have, in general, a negative impact on lipoprotein metabolism (Heinninger et al., 1994). Unexpectedly, certain /-FAs as C L A (mainly cis-9 trans-11 linoleic acid) are recognized as having antioxidative and anticarcinogenic properties (Ostrowska et al., 1999;M i r et al., 2000). Moreover, C L A has been shown to stimulate immune responses and protect against arteriosclerosis (Mir et al., 1999). Therefore, studies on methods o f accurately determining the C L A content in milk, meat and digesta and differentiating C L A from saturated and unsaturated FAs with emphasis on /-Fas, were undertaken.

Reagents
A l l chemicals were o f analytical grade; acetonitrile, water and methanol were of HPLC grade. Triethylamine and 2,4-dibromoacetophenone were from Merck. Palmitic, stearic acids, C L A and all other cis and/or trans fatty acids were pur chased from Sigma. A n internal standard (nonanoic acid) and other saturated fatty acids were from Fluka. A l l other chemical reagents were from POCH (Gliwice, Poland). The mobile phases were filtered through a 0.45 \xm membrane ( M i l l i pore) and then degassed for 2-3 min in vacuum with ultrasonication prior to use.

Chromatographic equipment
The instrument used consisted o f a Water 625LC system, which included a controller for gradient elution, two Waters 501 pumps and a Water 515 pump. The apparatus is coupled to a Waters 717plus WISP autosampler and a Waters 996 photodiode array detector. Data acquisition was performed on an Optimus Pen tium 5P60 computer with Millennium 2001 software. Separations were performed C Z A U D E R N A M . ET A L . on two Nova Pak C 1 8 columns (4 |Lim, 250 x 4.6 m m I.D., Waters) in conjunction with a Waters guard C 1 8 column o f 10 x 6 m m I.D. Two HPLC grade solvents were used; solvent A was acetonitrile, while solvent B was water. A binary gradient elution system with U V detection at 234.5 nm (Table 1) was used for analysis o f underivatized C L A in milk, meat and intestinal digesta samples. Preparation and hydrolysis of biological samples M i l k , meat and duodenal digesta samples were collected from sheep. A l l sam ples were frozen, lyophilized, and the obtained residues were stored in sealed tubes at -20°C until analyzed. M i l k (-50 mg) and duodenal digesta samples (-105 mg) were hydrolyzed with 4 m l o f 2 M NaOH at ~85°C for 35 min and then the hydro lyzed samples were acidified with 4 M HC1 to p H ~2. The free fatty acids were extracted four times with 3 ml o f dichloromethane. The organic layer was removed under a gentle stream o f argon and the residue dissolved in 1 m l o f acetonitrile. The obtained C L A solutions were injected onto HPLC columns. Undecanoic acid (an internal standard) and other saturated and unsaturated non-conjugated fatty acids were assayed after pre-column derivatization (Czauderna and Kowalczyk, 2001).

RESULTS A N D DISCUSSION
Larger quantities o f CLA are found in animal sources than in vegetable sources, particularly in products o f ruminant origin such as dairy products. Therefore, the subject o f the presented work was to study the C L A content in milk, meat and intestinal digesta. In fact, ruminal and duodenal fatty acid compositions signifi cantly influence the essential fatty acid profile o f milk, red meat, etc. C L A repre sents a mixture o f geometric and positional isomers o f CI8:2 n-6. Presently, we know o f seventeen C L A isomers (Ha et al., 1989), however, cis, trans /trans, cis 8,10-; -9,11 ;-10,12-; and -77,73-octadecadienoic acid accounts for only the main isomers. Considering the above, it is essential to use improved methods for routine quantification o f conjugated dienes with emphasis on C L A in milk and intestinal digesta samples. As shown in Figure 1, in our HPLC system (Table 1) the underi vatized C L A isomers were substantially retained on two C 1 8 columns and were completely separated from background interference and endogenous compounds in milk and duodenal digesta samples. The underivatized C L A samples eluted as pair o f large peaks clearly distinct from critical di-unsaturated fatty acids (particu larly linoleic and linolelaidic acids) and from all unidentified endogenous species in milk and intestinal digesta by using unique U V monitoring at 234.5 nm. Indeed, the underivatized conjugated dienes showed a very high band in the spectral range from 210 to 250 nm due to the presence o f conjugated double bonds. As can be seen from chromatographic runs o f milk, duodenal digesta and a standard o f the C L A isomers, the procedure resulted in reproducible separation of two C L A peaks, which eluted at 14.1±0.1 and 14.6±0.1 min, respectively. Moreover, we found that trace quantities o f other unidentified conjugated dienes in milk (-1.5 \xg/g D M ) and duodenal digesta (-0.05 (Lig/g D M ) samples were eluted after 18 min o f the HPLC run. It is possible that these dienes contained more than eighteen carbon atoms since the retention times o f these species significantly increased with de creasing polarity, i.e., increasing alkyl chain length. As expected, all conjugated diene peaks were absent from the blank when this gradient elution system was used.
C L A , quite likely the cis/trans 10,12 isomer, has been shown to decrease milk fat secretion (Mire et a l , 2000) and body fat deposition (Dudgeon et a l , 1997;Dunce et al., 1998). Thus manipulation o f the diet o f dairy cattle may be a means of increasing the C L A content of duodenal digesta and milk of dairy cows (Table 2). Moreover, appropriate dietary supplementation can reduce the levels o f saturated fatty acids o f milk and meat (Best et al., 2000). Therefore, the current HPLC me- Retention time, min Figure 1. The part of a typical HPLC chromatogram for the underivatized CLA isomers. The analyses were performed using the binary gradient elution system with UV detection at 234.5 nm. Peaks: 1 and 2 -the CLA isomers. Chromatogram A: line I -a milk sample (CLA content: ~ 11.4 ug/g DM); line II-a duodenal digesta sample (CLA content: ~ 0.55 ug/g DM). Chromatogram B: a standard of the CLA isomers thod without pre-column derivatization provides an accurate and specific analyti cal tool for routine quantification o f conjugated dienes in biological samples. The low detection (0.30 pg) and quantification (0.98 pg) limits, universality as well as a very simple procedure for preparation o f free fatty acids point to satisfactory reproducibility o f the presented method. The concentration o f underivatized C L A in the assayed biological samples was calculated using C L A standards and unde canoic acid (an internal standard) as a measure o f extraction yield. Simultaneous quantification o f an internal standard and other saturated, and unsaturated non conjugated fatty acids was achieved using the HPLC method with pre-column derivatization (Czauderna and Kowalczyk., 2001).

CONCLUSIONS
The presented HPLC method provides a simple, universal and reproducible analytical tool for routine quantification o f conjugated dines in various types o f biological samples.