INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

In the 1930s, Booth et al. (3) noted that summer milk had a greater absorbance at 233 nm than milk produced in the winter, and later work indicated that rats fed summer milk grew better than those fed winter milk even if the fat content was similar (2). In 1963, Riel (28) noted that summer milk fat had more conjugated dienoic acid than winter milk fat. More recently, conjugated linoleic acid (CLA) has been shown to inhibit chemically induced tumors (1, 10, 17), prevent atherosclerosis (24), and improve the protein-to-fat ratio in experimental animals (8).

CLAs can be produced by alkaline isomerization, but there are as many as 16 isomers which are not fully characterized (26, 29). Ruminant nutritionists have attempted to increase the naturally occurring CLA content of cow's milk via diet changes and alterations of ruminal fermentation (9). Recent work indicated that polyunsaturated oil supplements could increase the CLA content of milk, but these diet-dependent increases were often small or transitory (9, 19).

Many ruminal bacteria are inhibited by long-chain fatty acids (25), and gram-positive bacteria are more sensitive than gram-negative species (12). Polyunsaturated fatty acids are particularly toxic (21), but some ruminal bacteria are able to saturate the double bonds via a process known as biohydrogenation (27). In the 1960s, Kepler et al. (22) studied the biohydrogenation of Butyrivibrio fibrisolvens and demonstrated that linoleic acid (LA) was first converted to cis-9, trans-11 CLA. The reductase steps were inhibited by oxygen, but the LA isomerase could continue to produce cis-9, trans-11 CLA even if oxygen was present (15).

Since B. fibrisolvens A38 has a greater CLA-producing capacity than other ruminal bacteria, it has often been used as a model of CLA production (16, 20, 27). Washed-cell suspensions of B. fibrisolvens produced CLA, but the CLA production of growing cultures was not examined (15, 20). The following question then arose: is CLA a normal end product or is it simply an artifact of cells that could not biohydrogenate? Recent work indicated that mammalian tissues could also produce cis-9, trans-11 CLA from trans-octadecenoic acid (trans-C_18:1), and the significance of ruminal CLA production has been questioned (9).

The experiments described here sought to define more precisely the effect of LA on the biohydrogenation and CLA production of B. fibrisolvens.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Bacterial growth. B. fibrisolvens A38 was grown anaerobically at 39°C in basal medium containing (per liter) 292 mg of K₂HPO₄, 292 mg of KH₂PO₄, 480 mg of (NH₄)₂SO₄, 480 mg of NaCl, 100 mg of MgSO₄ · 7H₂O, 64 mg of CaCl₂ · 2H₂O, 4,000 mg of Na₂CO₃, 600 mg of cysteine hydrochloride, 10 g of Trypticase (BBL Microbiology Systems, Cockeysville, Md.), 2.5 g of yeast extract, and branched-chain volatile fatty acids (1 mmol each of isobutyrate, isovalerate, and 2-methylbutyrate), plus hemin, vitamins, and trace minerals (6). Glucose (2 mg/ml, final concentration) was prepared as a separate solution and was added after autoclaving. Cultures were grown anaerobically under O₂-free CO₂ in 150- by 18-mm tubes that were capped with butyl rubber stoppers and aluminum seals. Growth rate was estimated from the increase in optical density (OD) at 600 nm (1-cm cuvettes). Cultures were sometimes grown in serum bottles (160 ml) that were prepared in a similar fashion. The relationship of OD and bacterial protein was typically 220 mg of protein/liter/OD unit.

Fatty acid preparation. Concentrated LA (Sigma Chemical Co., St. Louis, Mo.) and CLA (75% cis-9, trans-11 isomer; Matreya, Inc., Pleasant Gap, Pa.) solutions (0.1 g/ml of water with 20% bovine serum albumin) were sterilely filtered (pore size, 0.2 µm). The stability of the emulsion could successfully be maintained with 20% bovine serum albumin throughout the experimental periods. LA and CLA stock solutions were then serially diluted in sterile anaerobic water to decrease the concentration. The LA and CLA solutions were then added to cultures or washed-cell suspensions (10 µl/ml).

Washed-cell suspensions. Cultures (typically 10 ml) were harvested by centrifugation (3,500 × g, 5°C, 10 min), and cell pellets were washed twice with anaerobic medium lacking Trypticase, yeast extract, ammonia, and glucose and resuspended in K₂HPO₄ (50 mM, pH 7.5) that was prepared anaerobically or aerobically. Cell ODs were typically 1. All incubations were performed at 39°C. The pH of the potassium phosphate buffer was decreased by adding HCl. When the pH was higher than 7.5, Tris buffer (50 mM) was used, and pH was adjusted with NaOH.

Fatty acid analyses. Cultures or washed cell suspensions (typically 10 ml) were extracted with a mixture of organic solvents (2 ml; 1 part hexane to 3 parts isopropanol to 1 part acetone; 1 min using a Vortex mixer). The suspensions were then centrifuged (1,000 × g, 3 min, 20°C). The solvent layer (top) was removed and flushed with nitrogen until dry. The fatty acids were then dissolved in toluene (1 ml). The fatty acids were methylated as previously described by Kim and Liu (23). Fatty acid methyl esters were separated by a Supelcowax-10 fused silica capillary column (60 m by 0.53 mm, 0.5-µm film thickness; Supelco., Inc, Bellefonte, Pa.) using a Hewlett Packard model HP5890 gas chromatograph equipped with a flame ionization detector and model HP3392 integrator. The conditions were as follows: helium flow, 2.4 ml/min; injector, 200°C; detector, 250°C; column chamber temperature, initially 40°C (5 min) and then increased to 220°C at 20°C/min and held for 30 min. A sample (1 µl) containing 0.5 to 5 µg of LA or CLA was injected into the column in a splitless mode. Heptadecanoic acid (C_17:0) was used as an internal standard. cis-9, trans-11 octadecadienoic acid (>98% purity) was used as a CLA standard. The recovery of CLA was 83%, and that of C_17:0 was 80%. A known standard mixture of fatty acids was used to identify other fatty acids. This protocol was able to separate eight isomers of LA, but it could not differentiate cis, trans versus trans, cis configurations in the same position. B. fibrisolvens A38 produced only the cis-9, trans-11 isomer.

Statistical analyses and design. All incubations were performed three times. Mean values and standard deviations of the mean are shown.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

When B. fibrisolvens A38 was incubated in basal medium lacking fatty acids, the culture grew rapidly (0.46 h¹) and the maximal cell density was approximately 1.0 (Fig. 1a). Stationary-phase cells lysed, and the OD at 24 h was only 0.6. When low concentrations of LA (as little as 35 µM) were added to the growth medium at inoculation, growth was not observed. Similar concentrations of a CLA mixture (75% cis-9, trans-11 CLA) also inhibited growth.

View larger version (16K): [in this window] [in a new window]   FIG. 1.   The effect of LA concentration on the maximum and final ODs of B. fibrisolvens cultures (a). LA was added to actively growing cultures at an OD of 0.3, and the final OD was measured at 24 h. LA additions caused an increase in OD, but the OD of inoculated controls could be subtracted to determine the culture OD. (b) Effects of LA hydrogenated end products and CLA.

Actively growing cultures tolerated higher concentrations of LA (Fig. 1a), and virtually all of the LA was converted to hydrogenated products (primarily trans-C_18:1 and small amounts of stearic acid) (Fig. 1b). If the LA concentration was 350 µM or greater, growth was inhibited, hydrogenated products declined, CLA accumulated, and more of the cells lysed. When the LA concentration was 1,800 µM, growth was completely inhibited and LA was not converted to either hydrogenated products or CLA.

When washed stationary-phase cells were incubated anaerobically with 350 µM LA, most of the LA was converted to hydrogenated products (Fig. 2a). The CLA concentration was as high as 13 µM (Fig. 2b). However, if the cells were incubated for more than 2 min, the CLA concentration declined, and after only 30 min CLA could no longer be detected. Washed cells that were incubated aerobically produced little hydrogenated product (Fig. 2a), and the CLA accumulation was at least twofold greater (Fig. 2b). The CLA concentration eventually declined, but the concentration at 30 min was greater than 15 µM. Aerobic-cell suspensions had a pH optimum for CLA production of 7.5, but pH values from 5.5 to 8.5 did not have a marked impact on the CLA production (Fig. 3).

View larger version (16K): [in this window] [in a new window]   FIG. 2.   Conversion of LA to hydrogenated products (trans-C18:1 and C18:0) (a) or CLA (b) by washed B. fibrisolvens cells. The initial LA concentration was 350 µM, and the cell OD was 1. The incubations were performed in triplicate, and the values are the means ± the standard deviations.

View larger version (15K): [in this window] [in a new window]   FIG. 3.   Effect of pH on CLA production of washed B. fibrisolvens cells that were incubated aerobically. The initial LA concentration was 350 µM, and the cell OD was 1. The incubations were performed in triplicate, and the values are the means ± the standard deviations.

B. fibrisolvens A38 cultures that had been treated with more than 50 µM LA could not be transferred successively, but cultures that were gradually adapted to increasing amounts of LA (0 to 50 µM) grew repeatedly with 35 µM LA. LA-adapted cells that were washed and incubated aerobically produced less CLA and more hydrogenated products than unadapted cells (Fig. 4). When unadapted cells were given a 350 µM dose of LA, CLA increased rapidly, but a second dose of LA did not cause a further increase in CLA (Fig. 5a). Washed-cell suspensions that were given a larger dose of LA (700 µM) produced approximately the same amount of CLA as those given a single dose of 350 µM or two doses of 350 µM. When washed-cell suspensions were provided with LA at concentrations ranging from 0 to 350 µM, CLA concentrations increased but only at LA concentrations lower than 350 µM. Cells that were preincubated with 35 µM CLA produced less additional CLA than those that were not preincubated with CLA (Fig. 6). The CLA production was greater if more cells were added, and the relationship between CLA and cell density was linear (Fig. 7).

View larger version (20K): [in this window] [in a new window]   FIG. 4.   Conversion of LA to CLA (open bars) or hydrogenated products (dark bars) by washed B. fibrisolvens cells that were incubated aerobically. The initial LA concentration was 350 µM, and the cell OD was 1. The unadapted cells had not been grown with LA, but the adapted cells had been repeatedly transferred with 35 µM LA. The incubations were performed in triplicate, and the values are the means ± the standard deviations.

View larger version (14K): [in this window] [in a new window]   FIG. 5.   CLA production by washed B. fibrisolvens cells that were incubated aerobically. (a) Cells received two doses of LA (350 µM) or a single dose of LA (700 µM) as indicated by the arrows. (b) The LA concentration was increased from 0 to 2,800 µM. The cell OD was 1. The incubations were performed in triplicate, and the values are the means ± the standard deviations.

View larger version (19K): [in this window] [in a new window]   FIG. 6.   CLA production by washed B. fibrisolvens cells that were incubated aerobically with 350 µM LA. Control cells were not preincubated with CLA; preincubation was done with 35 µM CLA. The incubations were performed in triplicate, and the values are the means ± the standard deviations.

View larger version (15K): [in this window] [in a new window]   FIG. 7.   Effect of cell OD on the conversion of LA to CLA by washed cells of B. fibrisolvens that were incubated aerobically with 350 µM LA. Open symbols show a 2-min incubation, and closed symbols show a 5-min incubation. Dotted line represents the regression line of 2-min and 5-min incubations.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Henderson (12) noted that B. fibrisolvens was very sensitive to long-chain fatty acids, but he did not examine the effect of LA or CLA. Our results indicated that unadapted B. fibrisolvens A38 inocula could not initiate growth if the LA or CLA was included at a concentration of 15 µM. Other strains of B. fibrisolvens did not produce as much CLA as A38 did. When the LA concentration was increased in a stepwise fashion, B. fibrisolvens A38 tolerated more LA, but the maximal LA concentration that allowed growth was only 35 µM. Growing cultures (OD, 0.3) were, however, 10-fold more LA resistant, but these cultures could not be transferred successively.

Dawson and Kemp (7) noted that saturated fatty acids were less toxic to ruminal bacteria than polyunsaturated fatty acids and suggested that biohydrogenation was a detoxification mechanism. Growing B. fibrisolvens A38 cultures biohydrogenated LA, but this capacity could be overcome by a high LA concentration. If the LA concentration was greater than 350 µM, biohydrogenation was incomplete, CLA could be detected, and the cultures were no longer viable. These results indicated that CLA was not a normal end product of growing cultures, and CLA accumulated only if biohydrogenation was inhibited.

Washed-cell suspensions that were incubated anaerobically biohydrogenated LA, with little CLA accumulation, and CLA eventually declined to undetectable levels. When the washed cells were incubated aerobically, biohydrogenation was strongly inhibited, the CLA production was at least twofold greater, and CLA persisted. These results supported the idea that CLA accumulation was due to an inhibition of biohydrogenation.

Our washed-cell suspensions were typically prepared from cultures that had not been exposed to LA. However, experiments with adapted cultures indicated that the CLA production was even lower and even more of the LA was converted to hydrogenated products. These results indicated that CLA production was not being constrained by the absence of an "inducer." Kepler and Tove (20) noted that the LA isomerase had a broad pH range, and we also observed CLA production at pH values ranging 5.5 to 8.5.

Kepler and Tove (20) incubated their B. fibrisolvens extracts for less than 2 min, and we also noted a very rapid increase in CLA production. However, the prolonged incubation did not cause a further increase in CLA. The LA isomerase did not recycle like a normal enzyme to catalyze more substrate, and the CLA production was highly cell density dependent. The CLA production increased if more LA was added, but only if the LA concentration was low. Because CLA was as toxic as LA, there was no advantage in releasing large amounts of free CLA.

When B. fibrisolvens cells were sonicated, the membrane fraction (pelleted by centrifugation at 150,000 × g) produced at least 10-fold more CLA than the cytoplasmic extract, and the result indicated that the LA isomerase was a membrane-bound enzyme (data not shown). Hughes and Tove (13, 14) extracted B. fibrisolvens cells with solvent, and they were able to purify the reductase of biohydrogenation. The LA isomerase activity was also found in the membrane fraction, but the preparations always had large amounts of contaminating carbohydrate (20).

Metabolically active B. fibrisolvens A38 cells produced hydrogenated end products rather than CLA, but B. fibrisolvens is a bacterium that lyses once it reaches stationary phase. Because membrane fractions retained their ability to convert LA to CLA, it is conceivable that dead or lysed cells could produce CLA in the rumen. When Harfoot et al. (11) incubated particle-associated ruminal bacteria with large amounts of sucrose and LA, biohydrogenation was favored and CLA was not detected. However, if the sucrose was omitted, some of the LA was converted to CLA.

When cattle are fed diets rich in fiber, the CLA content of milk often increases, but the ratio of CLA to total milk fat is usually less than 20 mg/g (28). CLA production in cattle fed hay and grain can be enhanced by LA supplements (oils), and in these cases the ratio of CLA to total milk fat can be greater than 20 mg/g (19). Both of these effects are consistent with the properties of B. fibrisolvens. B. fibrisolvens is a hemicellulose-digesting bacterium that is found in high numbers when cattle are fed hay or grass (4), and high concentrations of LA stimulated CLA production by B. fibrisolvens A38.

The study of CLA production in ruminants is complicated by the fact that ruminal fermentation is not the only source of CLA. If biohydrogenation is incomplete, some trans-C_18:1 can pass from the rumen (18), and trans-C_18:1 can be converted to CLA by the -9 desaturase of the mammary and adipose tissues (30). When the mammalian -9 desaturase was inhibited by a postruminal infusion of sterculate, endogenous CLA synthesis decreased 40%, but postruminal trans-C_18:1 addition caused only a small increase in the level of CLA in milk (5 to 7 mg/g) (9).

Our results indicate that B. fibrisolvens A38 can produce significant amounts of CLA if the LA concentration is high enough to inhibit biohydrogenation, but further work will be needed to define the role of B. fibrisolvens and other ruminal bacteria in ruminant CLA production. Because the LA isomerase does not seem to release free CLA and appears to be feedback inhibited, traditional schemes of cloning and overexpression to increase CLA production could be ineffective.