Assessing Rumen Biohydrogenation and Its Manipulation in Vivo, In Vitro, In Situ

Veerle Fieveza

Bruno Vlaemincka

Tom Jenkinsb

Francis Enjalbertc

Michel Doreaud

a Laboratory of Animal Nutritionand Animal Product Quality,Ghent University,Melle, Belgium

b Department of Animal andVeterinary Sciences,Clemson University,Clemson, USA

c INRA, UMR1289 TissusAnimaux, Nutrition, Digestion,Ecosystème, Métabolisme,Castanet-Tolosan, France;INP-ENSAT, Castanet-Tolosan,France;ENVT, Toulouse, France

d INRA, URH, Theix,Saint-Genès Champanelle,France

Assessing rumen biohydrogenation and itsmanipulation in vivo, in vitro and in situ

Reference techniques to study rumen biohydrogenation (BH) rely on the comparison ofintake and duodenal or (ab)omasal flows of polyunsaturated fatty acids (PUFA),whereas the net BH of PUFA to their saturated end-products gives a quantitativemeasure of accumulating BH intermediates. The current review paper aims at evalu-ating alternative in vivo, in vitro and in sacco techniques to simulate reference in vivoresults of unprotected PUFA sources, as well as strategies for overcoming or manip-ulating BH. In vivo rumen sampling approaches show potential but require furtherinvestigation, whereas in sacco results are inappropriate. In vitro 24-h batch incuba-tions and continuous cultures approach in vivo BH of unprotected C18 PUFA sourcesand are useful to assess the pH value at which dissociation of calcium salts of fattyacids occurs, but overestimate the degree of rumen inertness of formaldehyde-treatedoil(seeds) and marine products. Batch or continuous cultures provide an accurateestimate (0.6) of the proportion of hydrogenated C18 PUFA that are converted into theirsaturated end-product, except for incubations with high levels of fermentable sub-strate (.1.0 g/100 mL) in combination with high amounts of C18 PUFA (.0.5 mg/mL) orin incubations with EPA and DHA.

Keywords: Rumen biohydrogenation, in vivo, in vitro, in sacco, lipolysis.

740 DOI 10.1002/ejlt.200700033 Eur. J. Lipid Sci. Technol. 109 (2007) 740–756

1 Introduction

The process of biohydrogenation (BH) reduces the rumenoutflow of polyunsaturated fatty acids (PUFA) and con-tributes to accumulation of cis and trans isomers in rumi-nant products, including conjugated linoleic acid (CLA)and trans monoenes. Hence, the extent and type of therumen BH process will determine both the amounts andstructures of fatty acids leaving the rumen. As the fattyacid structure determines its physiological features,interest has grown in the process of rumen BH. Conse-quently, information is needed on (i) the disappearance ofPUFA, (ii) the production of the saturated end-products ofthe BH process, and (iii) the nature and the amount of theaccumulating BH intermediates. Measures and experi-mental techniques discussed in the current paper arelimited to the former two. Experimental techniques dis-cussed include in vivo reference techniques as well as insacco and in vitro approaches.

An increasing number of novel technologies and strate-gies to control and/or prevent rumen BH are currentlybeing investigated. These are aimed at by-passing

rumen PUFA metabolism or ensuring the accumulationof desired intermediate compounds. However, the re-sponse to protection technology often is variable andover-processing might reduce digestibility. Hence, rou-tine methods are required to assess the effectiveness ofthese technologies or strategies, both in terms of abso-lute data extrapolation to in vivo values as well as theirranking. In terms of protection technology, three mainprocess types can be considered: (1) chemical protec-tion, e.g. through encapsulation in a protein matrix fol-lowed by aldehyde treatment, direct formaldehyde treat-ment of oilseeds or formation of a whey gel complex;(2) formation of calcium salts and amides of fatty acids;and (3) technological treatments of oilseeds, such asextrusion, roasting, cracking, etc. Strategies to manip-ulate the extent of rumen BH and the accumulation ofBH intermediates mostly deal with antimicrobial addi-tives, including fish oil. The reliability of in vitro tech-niques to assess the effectiveness of strategies forovercoming or manipulating BH by ruminal microorgan-isms is evaluated.

Due to data availability, BH of linoleic (18:2n-6) andlinolenic (18:3n-3) acid is the main focus of thispaper, although some aspects of BH of eicosa-pentaenoic (EPA) and docosahexaenoic (DHA) acidare addressed.

Correspondence: Veerle Fievez, Laboratory for Animal Nutritionand Animal Product Quality, Ghent University, Proefhoeves-traat 10, 9090 Melle, Belgium. Phone: 132 9 2649002, Fax: 1329 2649099, e-mail: [email protected]

© 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.ejlst.com

Rev

iew

Art

icle

Eur. J. Lipid Sci. Technol. 109 (2007) 740–756 Rumen in vivo, in vitro and in situ biohydrogenation 741

2 Terminology and calculations

2.1 Rumen lipolysis

Lipolysis is the hydrolysis (release) of fatty acids from theglycerol backbone of triacylglycerols, galactolipids orphospholipids. Consequently, separation of esterified(EFA) and non-esterified fatty acids (NEFA) is a pre-requisite for the distinct evaluation of lipolysis and BH.

The extent of rumen lipolysis is calculated from the max-imal loss of EFA over a specific time period. This methodof expression can be used to calculate the mean lipolysisof all fatty acids as well as the extent to which an individ-ual fatty acid (FAi) is released from its EFAi fraction. Cal-culations based on in vivo data, effluents of continuousfermentors and batch in vitro results are similar, but theamounts of EFAi in the inoculum should be taken intoaccount in the latter case.

Lipolysis EFAi ¼ 100� EFAi0h � EFAith

EFAi0h

with EFAi 0h = EFA intake or in vitro supply of EFAi at thestart of the incubation, and EFAi th = duodenal or (ab)omasal flow of EFAi, amount of EFA in continuous fer-mentor effluents or amount of EFA recovered in batch invitro incubation flasks after t hours of incubation.

2.2 Rumen BH of PUFA

Assessment of rumen BH in the forestomach of rumi-nants is analogous to the determination of apparentrumen carbohydrate or protein digestibility [1]. Hence,rumen BH is calculated as the disappearance of PUFAi

relative to their intake. It should be noted that this repre-sents a combined measure of two different processes,i.e. the release of fatty acids from the esterified lipid andthe first step of PUFA BH, i.e. isomerization for 18:2n-6and 18:3n-3. This calculation can be applied both to invivo flow data and effluent data of continuous fermentorsas well as to batch in vitro measurements, but inoculumfatty acids have to be included in the latter calculations.Alternatively, fatty acids that do not originate from thetest product can be excluded by subtracting theamounts of fatty acids in blank (without the test product)incubation flasks (e.g. [2]):

Biohydrogenation PUFAi ¼ 100� PUFAi0h � PUFAith

PUFAi0h

with PUFAi 0h = PUFA intake or in vitro PUFA supply (in g)at the start of the incubation and PUFAi th = duodenal or(ab)omasal flow of PUFA (g), PUFA (g) in continuous fer-mentor effluents or PUFA (g) recovered in batch in vitroincubation flasks after t hours of incubation.

2.3 Production of BH intermediates andsaturated end-products

Calculation of rumen BH as shown above might be ap-propriate to assess protection of PUFA, but it does notprovide information on the accumulation of BH inter-mediates. The net duodenal flow or in vitro production ofthe saturated BH end-product (e.g. 18:0, 20:0 or 22:0)proportional to the amount of PUFA precursors lost duringrumen BH (i.e. 18:2n-6 1 18:3n-3, EPA and DHA,respectively) gives a quantitative measure of the overallPUFA BH efficiency. This approach is most appropriatefor C20 and C22 fatty acids, as dietary C20 and C22 PUFAare almost exclusively EPA and DHA, respectively, while itis hampered for C18 PUFA due to the production of 18:0from dietary oleic acid. The net amount of 18:0 resultingfrom BH of C18 PUFA can be calculated as the differenceof the duodenal or in vitro gain of 18:0 and the amount ofoleic acid lost. This requires the assumption that theduodenal flow or in vitro production of trans mono-unsaturated fatty acids originating from isomerization ofoleic acid is of minor importance [3].

Nevertheless, this approach does not provide informationon the nature of the accumulating intermediates. Indeed,rumen fatty acid metabolism is a multistep process,requiring measures to distinctively assess the conversionefficiency of each step. Detailed knowledge on the pre-dominant BH pathways allows assessing the inhibition orstimulation of specific intermediate BH steps. Calcula-tions are complicated due to the movement of fatty acidsamong pools, which is illustrated by e.g. Boeckaert et al.[4], Ribeiro et al. [5] and Harvatine and Allen [6]. However,limited knowledge on possible (secondary) BH pathwaysas well as an insufficient number of data in most experi-ments limits the current development of multi-compart-ment models including all intermediates [5]. As a con-sequence, these approaches exclude the identification ofpossible shifts of BH routes, which is further consideredout of the scope of the current paper.

3 Statistical analysis

Where sufficient literature information is available, wehave attempted to formulate quantitative models. Twomajor principles were taken into account in this statisticalapproach: (1) Observations within a given study havemore in common than observations across studies, and(2) inference is to future, unknown studies. This meansthat the factor ‘study’ was considered random as thestudies included in this review only represent a randomsample of a larger set of potential studies and the interestis not in evaluating the effect of a specific study but rather


742 V. Fievez et al. Eur. J. Lipid Sci. Technol. 109 (2007) 740–756

to derive models that are applicable to all potential stud-ies within a population. Hence, relations and regressionequations presented in the current paper were developedusing a mixed-model regression analysis as described bySt-Pierre [7]. Data were analyzed according to the follow-ing model: Yij = B0 1 B1Xij 1 si* 1 bi*Xij 1 eij; with i = the ith

study, j = the jth observation within the ith study, B0 1 B1Xij

= the fixed effect part of the model, si* 1 bi*Xij 1 eij = therandom effect part of the model and with the means of si*and bi* equaling 0.

To illustrate effects, it was further attempted to graphicallypresent regression results. However, results from amixed-model regression cannot simply be presented as aY vs. X graph as the observations come from a (j 1 2)-dimensional space, which should first be collapsed into atwo-dimensional space. To account for this loss ofdimensions, adjusted observations should be calculated.The latter is done by adding the residual from each indi-vidual observation to the predicted value of the studyregression [7]. These adjusted observations for studyeffects were used to calculate determination coefficients.

4 Reference in vivo techniques: Duodenal,omasal and abomasal flows

As with digestibility studies, determination of rumen PUFABH is typically assessed through the use of digestibilitymarkers and the collection of digesta via sampling fromthe omasal canal or from abomasal or duodenal cannulas.Almost all published experiments with data on fatty acidruminal metabolism used duodenal cannulas. Omasalcanal measurements were used by Shingfield et al. [8] andLundy et al. [9], whereas an abomasal cannula was usedin five experiments in Australia and Canada between1972 and 1987 [10–12]. To correct for non-representativedigesta sampling, a double-marker method is suggestedthat allows recombination of the fluid and particle phase[13]. Apparently, the choice of appropriate markers,representative of rumen fatty acid outflow, and theiraccurate analysis are essential. However, the number andnature of markers and the flow calculations largely varyamong experiments. In experiments with measurementson digestive lipids, Ru, Yb, Dy, La (radioactive or as chlo-ride and acetate), Cr (as Cr2O3, mordanted, or 51Cr), lignin,C31 and/or C33 alkanes and acid-insoluble ash havebeen used as particle phase marker; Cr-EDTA, Co-EDTA(radioactive or not) and PEG have been used as liquidphase marker. Out of 81 published studies and 7 unpub-lished experiments from INRA-Theix (either partially pub-lished or submitted for publication) estimating in vivo fattyacid flows, there were 49 that used two markers. Whenone marker was used, it was mostly chromic oxide in any

form. This dataset is too limited to allow comparisonsbetween markers and flow calculations. However, anoverview of the use of flow markers has been made byOwens and Hanson [14] who did not draw conclusions onthe superiority of any marker.

4.1 Rumen lipolysis

Although lipolysis is a prerequisite of rumen BH, in vivostudies reporting dietary and digesta EFA and NEFAfractions are scarce [15–17]. Omasal or abomasal sam-pling offers the greatest opportunity for the correctassessment of rumen lipolysis, without the risk of intest-inal hydrolysis. A duodenal cannula should be insertedcranial to the common bile and pancreatic duct to provideevidence that the production of the unesterified fattyacids did not take place in the small intestine, as specifiedby Atkinson et al. [15].

4.2 Rumen BH of PUFA

Jenkins [18] estimated the average in vivo BH of unpro-tected PUFA sources from slopes of the regression lines,estimating ruminal loss of dietary PUFA per unit of PUFAconsumed, after adjustment for the random effect ofstudy. Slopes were 828 6 17 and 875 6 22 g/kg for18:2n-6 and 18:3n-3, respectively. However, 18:2n-6 and18:3n-3 BH data in recent studies are calculated usingdata obtained with improved analytical methods allowingthe chromatographic separation of numerous 18:2 and18:1 isomers. It appears now that 18:2n-6 only representa part of the 18:2 fatty acids. Different conjugatedisomers (CLA) and non-conjugated isomers, especially18:2t11c15, account for a large proportion of the total18:2, depending on the diet. Loor et al. [19] separated15 isomers of 18:2 and showed that the percentage of18:2n-6 in total 18:2 in duodenal chyme varied between77% for a forage-based diet and 34% for a concentrate-rich diet supplemented with linseed oil. Obviously, thismight result in a bias in the calculated BH when isomersare not separated. Indeed, in all determinations of fattyacid flows published before 2001, one chromatographicpeak was evidenced in the 18:2 area, which was con-sidered as 18:2n-6 whereas it represented in fact a sum ofco-eluted isomers. This explains why in recent studies the18:2n-6 BH is often higher than in earlier studies (800 and650 g/kg on average for diets containing less or morethan 60% of concentrate, respectively; [20]). We haveevaluated this bias using 60 dietary treatments from16 experiments published from 2001 and two unpub-lished experiments in which the separation between 18:2isomers was made. The disappearance of 18:2n-6 was857 g/kg, which is slightly higher than the formerly



reported value obtained by Jenkins [18]. However, the BHcalculated based on the disappearance of total 18:2isomers was 805 g/kg. A relationship between BH oftotal 18:2 and 18:2n-6 was established: BH18:2 n-6 =0.450(SE = 0.0503; p ,0.001) 1 0.506(SE = 0.06151; p ,0.001)6BH18:2

(R2 = 0.748; n = 60). As the proportion of 18:3n-3 in total18:3 duodenal fatty acids is higher than 0.90, it can beexpected that the bias on BH is of lesser importance forthis fatty acid.

4.3 Production of BH intermediates andsaturated end-products

Calculation of net 18:0 production from hydrogenated18:2n-6 and 18:3n-3 is hampered when based on inade-quate separation of 18:1 isomers. As dietary 18:1 almostexclusively represents oleic acid, some attempts to esti-mate the net 18:0 production could be made assuming aconstant ruminal loss of oleic acid per unit of oleic acidconsumed. Hence, duodenal flow of 18:0 has been inte-grated as dependent variable in a multiple linear model(Fig. 1) with the amount of hydrogenated C18 PUFA andoleic acid intake as independent variables. The propor-tional conversion of hydrogenated 18:2n-6 and 18:3n-3 to18:0 might then be estimated from its respective coeffi-cient in this multiple linear regression, which suggeststhat on average 0.599 of the hydrogenated C18 PUFA werecompletely hydrogenated to 18:0. The data consideredhere include all published experiments from 1987 untilnow that quantified intake and duodenal flow of C18 fattyacids, but excluding dietary treatments where protectedfats, fish oil or fermentation modifiers (e.g. monensin)were used. Variation within the dataset, such as forage/concentrate ratio (wt/wt, min–max, 500 : 500–820 : 180 indairy cattle diets), experimental animal (e.g. dairy cattle,finishing beef steers, sheep), dry matter intake (DMI) (min–max, 1.1–27.0 kg/day), dietary PUFA content (min–max,8.3–30.5 g/kg DM) or variation in BH of oleic acid (notquantified), might be the reason for some deviation of theobserved values from the trend line in Fig. 1. Two obser-vations by Loor et al. [19], with dairy cattle fed high-con-centrate diets (350 : 650 forage/concentrate, wt/wt) sup-plemented with linseed or soybean oil, were excludedfrom the multiple linear regression. Nevertheless, thesedata are presented in Fig. 1 and illustrate a possible in-hibitory effect of increased concentrate levels on the pro-duction of the saturated BH end-product.

Fish oil is another well-known inhibitor of BH to 18:0 asend-product [21]. Effects of fish oil on rumen metabolismwere suggested to be attributed to the action of non-esterified EPA and DHA [22], but to our knowledge,attempts to quantitatively assess their overall effect onduodenal flow of 18:0 have not yet been made [21].

Fig. 1. Net duodenal gain of 18:0 (18:0 – 0.643618:1intake) as estimated from the multiple linear regressionmodel [18:0 (g/kg DMI) = 3.22(SE = 1.37; p = 0.023) 1

0.599(SE = 0.061; p ,0.001)6C18 PUFAlost (g/kg DMI) 1

0.643(SE = 0.061; p ,0.001)618:1 intake (g/kg DMI), R2 = 0.890,n = 48] in relation to the amount of 18:2n-6 and 18:3n-3 lostby rumen BH (C18 PUFA lost) [6, 8, 9, 24, 25, 42, 44, 45, 57,75–86]. Observations represented by grey squares [43]were not included in the multiple linear model developmentand illustrate a possible inhibition of complete 18:2n-6 and18:3n-3 BH at higher dietary concentrate levels.

Introduction of an interaction term (amount of C18 PUFAlost * EPA 1 DHA intake) in the former regression (Fig. 1)allowed assessment of the effect of EPA and DHA (g/kgDMI) on the conversion efficiency of unesterified C18 PUFAto 18:0. As the BH of dietary 18:1, and hence the coefficientof 18:1 in the multiple linear regression, might be lower forMegalac® or formaldehyde-treated linseed, fish oil sup-plementation data using these sources were not includedin the database, e.g. [23] and some treatments of [24] and[25]. The negative coefficient of the interaction term con-firms the inhibitory effect of EPA and DHA on the produc-tion of 18:0 from C18 PUFA and suggested the conversionof C18 PUFA to 18:0 to be reduced by 20% for each unit ofEPA 1 DHA intake (g/kg DMI). The latter value assumesnegligible effects of EPA and DHA on the BH of oleic acidand accumulation of its isomers. A negligible effect of EPAand DHA on the BH of oleic acid seems justified [21],whereas the effect on the accumulation of its isomers hasnot yet been described. Moreover, this approach assumesEPA and DHA to exert an equally inhibitory effect, which isunlikely (e.g. [26]). However, the limited number of fish oil-supplemented duodenal flow data (n = 7) does not allow amore detailed evaluation.

5 Alternative in vivo approaches

5.1 Calculations based on concentrationsinstead of flows

The need for quantitative flow measures was suggestedto be redundant in rumen fatty acid studies, as rumenfatty acid metabolism is essentially limited to hydrolysis



and BH. Hence, total fatty acid intake can be assumed toequal ruminal fatty acid outflow, which would allow toapply the former equation (see Section 2.2.) to calculateBH, but expressing fatty acid intake and flow relative tothe total fat or to the total C18 fatty acids [27]. However,Doreau and Ferlay [28] showed evidence of some rumenloss of fatty acids through oxidation and/or absorption,whereas bacterial de novo synthesis of fatty acids alsooccurs [29]. In addition, quantification of total C18 fattyacids has been tedious due to extensive rumen fatty acidmetabolism resulting in the formation of a wide range ofidentified as well as unidentified C18 fatty acids. This sug-gests that estimations of BH based on fatty acid con-centrations or total 18-carbon fatty acids might bebiased. From a database of experiments in which fattyacid duodenal flows were determined (n = 218), weassessed the importance of this bias through the com-parison of BH calculated from quantitative measurementsof intake and duodenal flow with calculations based on18:2n-6 and 18:3n-3 concentrations in total fatty acidsand proportional to the sum of C18 fatty acids. As the ratiobetween fatty acid duodenal flow and fatty acid intakedepends on the fatty acid content of the diet (see forexample [28]), dietary fatty acid content was included inthe regression equation. From the equations presented inTab. 1, BH can be predicted from data obtained usingconcentrations in total fatty acids or in C18 fatty acids,with good accuracy (R2 between 0.85 and 0.97).

5.2 BH estimates through rumen sampling

From an animal welfare point of view, use of experimentalanimals with duodenal cannulas should be minimized.

(Ab)omasal sampling through the rumen cannula mightoffer an alternative, but direct rumen sampling is the leastdemanding technique. However, continuous lipid intakehampers calculation of rumen BH, as variations in fattyacid concentrations are insufficient [30]. Alternatively,Doreau and Gachon [31] ruminally infused a pulse dose of300 g of a linseed product into cows and then followedthe fatty acid kinetics in consecutive samples of rumenjuice and total contents. This method was used for lin-seed oil, rolled linseeds and extruded linseeds togetherwith duodenal flow measurements [21]. The mean pro-portions of the various C18 fatty acids in the rumen couldbe estimated from the areas under the curves of thesefatty acids during 24 h of sampling and was shown to besimilar to values obtained from sampling the duodenalchime effluent. Differences between treatments were alsosimilar to duodenal flow measurements and ruminalkinetics. This shows the potential of this technique, butestimation of PUFA BH requires a dynamic model, toadjust for factors such as (i) time of lipid infusion into therumen, (ii) lipid hydrolysis, and (iii) the liquid outflow rateof the rumen. Alternatively, PUFA BH can be estimatedfrom rumen fatty acid pools determined during a fastingperiod (16 h) as described by Lourenco et al. [32]. BH wasestimated from 18:3n-3 intake and its hydrogenation ratebased on rumen pool sizes of 18:3n-3 and acid detergentlignin clearance rates. In this approach, acid detergentlignin clearance rates were assumed to equal rumenpassage rates of fatty acids, which is justified as fattyacids are adsorbed onto particles. Despite the low num-ber of rumen evacuations (n = 3), this approach appar-ently generated reliable BH measures, but no referenceduodenal flow measurements were available to unequi-vocally assess this technique. However, both rumen

Tab. 1. Prediction of the 18:2n-6 and 18:3n-3 BH calculated based on total intake and duodenal flows (BH18:2 flow andBH18:3 flow) from calculations based on (i) 18:2n-6 [BH18:2 (%FA)] and 18:3n-3 [BH18:3 (%FA)] concentrations in total fatand (ii) proportional to the sum of 18-carbon fatty acids [BH18:2 (%C18) and BH18:3 (%C18)] (n = 218).

Intercept Independent variable R2

Estimate SE§ p$ Variable Estimate SE§ p$

BH18:2flow NS BH18:2 (%FA) 0.930 0.0092 ,0.001 0.896dietary fat content 0.009 0.0013 ,0.001

BH18:2flow NS BH18:2 (%C18) 0.928 0.0105 ,0.001 0.857dietary fat content 0.010 0.0015 ,0.001

BH18:3flow 20.101 0.0202 ,0.001 BH18:3 (%FA) 1.075 0.0223 ,0.001 0.961dietary fat content 0.005 0.0009 ,0.001

BH18:3flow 20.127 0.0239 ,0.001 BH18:3 (%C18) 1.102 0.0263 ,0.001 0.970dietary fat content 0.006 0.0010 ,0.001

§ SE, standard error of the regression coefficient.$ p, p-value of the regression coefficient.# NS, the intercept is not significantly different from zero.



sampling techniques might have some limitations due tomassive perturbation of the substrate pool size duringfasting and/or effects of unsaturated fatty acid overloadon rumen BH enzymes or pathways when supplying apulse dose of PUFA.

6 Batch in vitro techniques

Rumen metabolism of PUFA has been studied in vitro forseveral decades at different laboratories using slightlydifferent methods. Most often, 24-h incubations are car-ried out. However, as effects on rates of lipolysis and BHmight be masked after 24 h of incubation, someresearchers perform time series of in vitro incubations toallow estimations of kinetic parameters such as lag timeand rate of lipolysis or BH. An overview of the in vitroincubation characteristics of the studies used in thisreview is given in Tab. 2.

6.1 In vitro conditions

Because culture conditions affect microbial metabolismin vitro, factors characterizing batch in vitro incubationsand their effect on the extent and kinetics of rumen lipo-lysis and BH are discussed below.

6.1.1 Rumen inoculum from adapted ornon-adapted donors

Experiments evaluating the effect of adaptation of thedonor animal to dietary PUFA on in vitro BH are availablefor 18:2n-6 sources only [33]. From this study, it wasconcluded that dietary supplementation of soybean oil tothe donor cows did not affect in vitro BH of 18:2n-6, whichis consistent with the overall conclusions from experi-ments summarized in Tab. 2. Adaptation of the donoranimal might be of minor importance for 18:2n-6- and18:3n-3-rich sources since rumen microbes of donor ani-mals are continuously exposed to lower amounts of thesefatty acids through ingestion of forages or grains. How-ever, this basal diet generally does not contain any of thelonger PUFA such as EPA and DHA. Based on milk CLAconcentration after fish meal supplementation, Abu Gha-zaleh et al. [34] and Roy et al. [35] suggested that therumen microbial population required a gradual adaptationof several weeks.

6.1.2 Particulate material

In a few in vitro studies, undegradable particles have beenadded to create optimal conditions for solid-adherent

bacteria and rumen microbial BH (e.g. [2]), as earlierstudies showed that BH took place on the surfaces offeed particles [36]. In most incubations, particles are pro-vided either by the inoculum or by the fermentation sub-strate, which is required as a carbon source becauseanaerobes do not generate energy (ATP) from PUFA. Inthis respect, the entire removal of inoculum particlesthrough centrifugation should not be considered andmight explain the reduced BH activity in incubations byChoi et al. [37].

6.1.3 Introduction of PUFA in incubationmedium

Several methods have been described for the introduc-tion of water-insoluble PUFA into the incubation medium.Fellner et al. [38] proposed the preparation of 18:2n-6emulsions through sonication. Martin and Jenkins [39]described this technique as unsuccessful for unesterified18:2n-6, but emulsions of soybean oil triglycerides werehomogeneous and evenly distributed. In the laboratory atGhent University, fatty acids in oil or fat are added as asolvent-oil/fat solution, with the solvent being evaporatedby N2 flushing. Deswysen et al. [40] ensured fatty aciddispersion in an aqueous environment through the non-ionic surfactant Tween-80. The latter group illustrated apossible effect of the mode of PUFA introduction on the invitro rumen fatty acid metabolism.

6.1.4 Gas pressure

Among laboratories, a multitude of devices are used forbatch in vitro simulation of the rumen environment, withsome allowing gases produced during fermentation toaccumulate in the incubators, while others permit gasescape to maintain gas pressure at low levels. Supple-mentation of non-esterified 18:2n-6 or 18:3n-3 in incuba-tors where gases were continuously collected in syringesand accumulated for 24 h resulted in enhanced produc-tion of 18:0 and lower amounts of BH intermediatesaccumulated (J. P. Jouany, B. Lassalas, T. T. Chow, D.Demeyer, V. Fievez, unpublished results). This is in accord-ance with the increased 18:0 concentrations in incubationswith sunflower oil under a 50 : 50 H2/CO2 atmospherecompared to a 100% CO2 atmosphere [4]. This stimulatingeffect of gas pressure accumulation has also been sug-gested for other hydrogen-consuming reactions, such asmethane production [41].

Obviously, in vitro conditions might affect estimates ofrumen BH and their kinetics, which underlines the needfor a detailed description of the incubation characteristicsand the inclusion of a control treatment under the same



Tab. 2. Overview of characteristics of 24-h batch in vitro incubations used in the current review and indication whether thestudy resulted in appropriate (3) simulation or not (–) of in vivo 18:2n-6 or 18:3n-3 BH and production of the BH end-product(18:0). The 18:0 cell is empty when the reported results did not allow evaluation of 18:0 production. In the final row, somecharacteristics are summarized that impair appropriate in vivo simulation of both 18:2n-6 and 18:3n-3 BH as well as 18:0production. ‘IND’ (independent) is mentioned when the whole range of the above-mentioned features ensures appropriatesimulation.

Incubated PUFA Substrate Gas§ Inoculum pH Totalvolume[mL]

BH 18:0

main source concen-tration§§

[mg/mL]

intro$ type¤ concen-tration#

[mg/mL]

ani-mal##

adap-tation{

time{{

[h]pre-treatment{ dilution{{

start/24¥

[2] 18:218:3

seed 0.240.50

0.5 mm C1U 10 – C – �12 0.8-mm sieve 1 : 1 6.9/UN 120 3 3

[3] 18:218:3

oil 0.3–1.90.04

weight F 18.8 – C – 1 1.6-mm sieve 1 : 1 7.0/6.4 160 3

3

–

[4] 18:218:3

oil 0.2–0.60.1–0.5

hexane evapor. F 16 3 SH – 15 1-mm sieve 1 : 4 7.0/5.8 25 3

3

–

[26] 18:218:3

corn 0.30.01

finely ground TMR 10 – C cheese-cloth (2) 1 : 4¥¥ UN/UN 52 3

3

3

[27] 18:218:3

oil 0.030.06

weight F 10 3 ST – �12 cheese-cloth (2) 1 : 4 UN/UN 50 3

3

3

[33] 18:2 oil orNEFA

0.1–0.5 weight F 9.6 – C – 12 cheese-cloth (2) 1 : 3¥¥ UN/UN 208 3

[37] 18:2 NEFA 0.2 albumin none – 3 C – 5 particle free –} 6.8/UN 10 –}} –}}

[40] 18:2 NEFA 0.04–0.2 tween 80 F 8 3 C – 1 0.5-mm sieve 1 : 4 7.1/6.4 50 –}} –}}

[40] 18:2 NEFA 0.04–0.2 hexane evapor. F 8 3 C – 1 0.5-mm sieve 1 : 4 7.1/6.4 50 3}}

3}}

[47] 18:2 oil 0.3–2.7 hexane evapor. F 16 3 SH – 15 1-mm sieve 1 : 4 7.0/5.7 25 3

[51] 18:218:3

oil orseed

0.21.0

weightgrinder

F 15 semi SH – 15 cheese-cloth (4) 1 : 9 6.7/UN 200 3

3

–

[60] 18:218:3

oil 0.2–0.60.3

hexane evapor. C 10 3 SH – 15 1-mm sieve 1 : 4 7.0/6.0 50 3

3

3

[87] 18:2 oil 0.3–0.5 weight FC 9.6 – C – 2 cheese-cloth (2) 1 : 4¥¥ UN/UN 52 3 3

[88] 18:2 NEFA 0.1–0.3 ethanol TMR 9.6 3 C – 3 cheese-cloth (2) 18:2 UN/UN 52 3 3

[89] 18:2 oil 0.7 UN SM 10 3 ST 3 �12 cheese-cloth (2) 1 : 4 UN/UN 50 3

Impaired simulation �0.5 tween 80 –albumin?

none �15 IND UN IND IND IND nobuffer

,6.0 IND

§ In vitro gas accumulation (3) or not (–); semi indicates that gas is allowed to accumulate for a certain period (3, 6 or 9 h)after which it is released.§§ Initial concentration in incubation flask of predominant PUFA.$ Mode of introduction of the PUFA source into incubation flask in case of oils or NEFA, with ‘evapor.’ indicating that sol-vent has been evaporated prior to incubation, ‘albumin’ indicating the use of bovine serum albumin to ensure that fattyacids remain in suspension and ‘tween 80’ indicating the use of the non-ionic surfactant Tween-80 to disperse fatty acids inthe aqueous buffer solution. All seeds were weighed into the incubation flasks and the mode of grinding (coffee grinder orthrough 0.5 mm mesh width) has been indicated as ‘grinder’ and 0.5 mm, respectively.$$ Type of fermentation substrate added (forage, F; concentrate, C; total mixed ration similar to donor cow’s diet, TMR;soybean meal, SM; undegradable straw or hay particles, U).# Concentration of fermentable substrate; without undegradable straw when added.## Type of donor animal: sheep (SH), steer (ST) or dairy cow (C).{ Adaptation of animal (3) or not (–) to the incubated PUFA source.{{ Time of inoculum sampling after last meal or concentrate feeding.{ Numbers between brackets indicate layers of cheesecloth. Particle-free inoculum obtained through centrifugation at50006g.{{ Rumen incolum/buffer (vol/vol).¥ pH at start (ST) of the incubation and minimum pH after 24 h of incubation; UN indicates values were not reported.¥¥ Reducing solution was added together with buffer and inoculum (0.04 of final volume).} Washed cell suspensions were added to sterile rumen contents in ratio 3 : 10, no buffer solution added.}} 6-h [37] and 9-h [40] instead of 24-h in vitro incubation.



conditions, to allow correct interpretation of data on theaccumulation of BH intermediates.

6.2 In vitro simulation of BH of unprotectedPUFA sources

The amount (mg/mL) of 18:2n-6 or 18:3n-3 lost during24-h incubations from NEFA, oils or oilseeds that have notbeen subjected to lipid protection technology is propor-tional to the amount (mg/mL) incubated (Fig. 2). Theoverall linear relationship includes both NEFA and EFA,indicating that 24-h in vitro BH is not limited by lipolysis orimpaired by the presence of EFA in the oleosomes of oil-seeds. The linear relationship further illustrates 24-h invitro BH to be independent of the amount of fat added tothe incubation flask, although the rate of lipolysis and BHhas been suggested to decrease with increasing amountsof soybean oil or PUFA in the culture substrate [3, 33]. Theslopes of these overall linear equations reveal that the in

Fig. 2. Disappearance of 18:2n-6 and 18:3n-3 during 24-h batch in vitro incubations with pure unesterified C18:2n-6, oils or oilseeds [2–4, 26, 27, 33, 47, 51, 60, 61, 87–89].Fermentation substrates were mainly hay or lyophilizedgrass or alfalfa, but total mixed rations (TMR) have beenadded in some experiments n check TMRn. [18:2 lost(mg/mL) = 0.816(SE = 0.0111; p ,0.001)618:2 initial (mg/mL),R2 = 0.986, n = 43; 18:3 lost (mg/mL) =0.882(SE = 0.0054; p ,0.001)618:3 initial (mg/mL), R2 = 0.999,n = 19].

vitro BH of 18:3n-3 (882 g/kg) exceeds the BH of 18:2n-6(816 g/kg), in agreement with in vivo observations.Moreover, 24-h in vitro and in vivo (825 and 875 g/kg for18:2n-6 and 18:3n-3, respectively; [18]) slopes weresimilar.

Average in vitro BH of EPA (367 6 295 g/kg, n = 28) andDHA (225 6 242 g/kg, n = 52) are not consistent with invivo results (781 6 123 g/kg, n = 17 and 763 6 134 g/kg,n = 17 for EPA and DHA, respectively) following fish oiland microalgae supplementation [23–25, 42–45]. Largedifferences between in vitro and in vivo BH data of EPAand DHA might be attributed to the absence of these fattyacids in the diet of the inoculum donor animals. Moreover,an exponential decline in the 24-h BH of EPA and DHAhas been suggested by Gulati et al. [46] from their dose-response study. Although inter-experimental variationhampered the establishment of a global dose-responserelation, the exponential decline holds within most in vitroexperiments (Fig. 3). Variable composition of marinesources as well as a lack of information on EPA and DHA

Fig. 3. Biohydrogenation of EPA and DHA during 24-hbatch in vitro incubations relative to the initial amount ofEPA or DHA added through supplementation of pureunesterified EPA or DHA, fish oil or microalgae [4, 26, 47,51, 60, 61, 87, 90]. Data from dose-response studies areconnected with lines.



lipolysis might contribute to inter-experimental differ-ences. Indeed, accumulation of non-esterified EPA [4, 47]and to a larger extent DHA [26] has been shown to exert anegative feedback on the BH of EPA and/or DHA.

6.3 In vitro simulation of production of BHintermediates and the saturated BHend-product from unprotected PUFA sources

The adequate identification of oleic acid in the batch invitro incubations under consideration allowed the calcu-lation of the net gain of 18:0 [3] which has been related tothe amounts of C18 PUFA lost. Oleic acid concentrationsin 24-h incubations of Akraim et al. [2] were not reportedand in this case an average BH value of oleic acid (491 g/kg) has been assumed to calculate the oleic acid remain-ing after 24 h. The linear relationship revealed that a con-stant proportion (0.666; Fig. 4) of hydrogenated C18PUFA was transformed into the saturated end-product,18:0 (Fig. 4), which was not different from the in vivo con-version of hydrogenated C18 PUFA to 18:0 (0.599). Thesimilarity of the coefficients refutes the opinion that batchin vitro conditions generally stimulate the accumulation ofBH intermediates. Nevertheless, accumulation of BHintermediates is stimulated under in vitro conditions withhigh substrate levels (.1.0 g/100 mL) in combination withrelatively high (.0.5 mg/mL) C18 PUFA supplementationas NEFA or oils (Fig. 4). This might have been provoked bya pH effect, which was on average 5.8 at the end of the24-h incubations of Boeckaert [4]. Indeed some compre-

Fig. 4. Net gain of 18:0 in relation to 18:2n-6 and 18:3n-3lost by rumen BH (C18 PUFA lost) during 24-h batch invitro incubations with pure unesterified 18:2n-6, oils oroilseeds and with substrate concentrations and C18PUFA from oils below (diamonds) [net 18:0 produced (g/kg DMI) = 0.666(SE = 0.054; p ,0.001)6C18 PUFA lost (g/kgDMI), R2 = 0.968, n = 15] and beyond (triangles) 1 g/100 mL and 0.5 mg/mL, respectively [2–4, 26, 27, 51, 60,87, 88]. Observations represented by grey squares [3]were not included in the linear model development andillustrate a possible inhibitory effect of increasingamounts of C18 PUFA as oils at higher substrate con-centrations (1.6 g/100 mL) on the production of the satu-rated BH end-product from hydrogenated 18:2n-6 and18:3n-3.

hensive studies by Van Nevel and Demeyer [48] andTroegeler-Meynadier et al. [3] showed that a lower rumenpH was associated with increased accumulation of BHintermediates. However, incubations with relatively highPUFA supplementation in combination with higher sub-strate levels all have been performed with sheep inocula(Tab. 2). Hence, a higher sensitivity of microbes fromsheep compared to cow or steer inocula in terms ofaccumulation of BH intermediates cannot be excluded.

6.4 In vitro assessment of protectiontechnology and additives

Formaldehyde-treated seeds, calcium salts and fish oilsupplementation are used as prototypes to assess thereliability of batch in vitro incubations to estimate rumenBH and production of the saturated end-products of pro-tected PUFA sources or when supplementing anti-microbial BH modifiers. The choice for these prototypeswas based on the number of available in vitro and in vivodata. To our knowledge, the effect of heat treatment, i.e.extrusion, on PUFA BH has been assessed using both invitro and in vivo approaches for one seed only [2, 49].Comparison of other in vitro and in vivo data is limited asdifferent technological processes and different oilseedswere used, which illustrates the need for further in vitro-invivo comparisons of processed oilseeds.

6.4.1 Assessment of effectiveness of protectiontechnology: Formaldehyde treatment

A procedure to protect polyunsaturated oil droplets fromrumen BH through their encapsulation in formaldehyde-treated casein has been described decades ago [50].Published reports on both in vitro and in vivo BH of thesame oil sources are limited ([51] vs. [45] and [52] vs. [53];Tab. 3), but indicate that in vitro procedures tend to over-estimate the degree of rumen inertness. From this, Gulatiet al. [54] suggested that effective in vivo protection onlycan be guaranteed for products showing in vitro BHvalues of 250 g/kg or lower. A further comparison of invitro and in vivo studies (Tab. 3), although not with thesame products, suggests oilseed pretreatment to beessential for the formation of the inert formaldehyde-pro-tein matrix, as extensive PUFA BH has been observedboth in vitro and in vivo when using formaldehyde-treatedoilseeds without pretreatment. In conclusion, batch invitro screenings might provide useful information on theextent of protection of chemically treated fat supple-ments, but should include both a negative (unprotected)as well as a positive control of which in vivo data areavailable, to enable ranking of products for efficacy andsome speculation on the expected in vivo results.



Tab. 3. Overview of extent of in vivo or 24-h in vitro BH of the predominant C18 PUFA of formaldehyde-treated (FT) oilsources and their experimental control treatment.

Reference Oil source Pretreatment§ Control PUFA (donor)animal

BH [g/kg]

Con FT

In vivo[23] linseed no – 18:3 sheep – 927[45] linseed HCOOH linseed oil 18:3 sheep 920 852[95] canola seed alkaline crushed canola seed 18:2 steer 750 648[53] canola/soybean emulsification yellow grease 18:2 steer 714 570[50] linseed oil emulsification linseed oil 18:3 sheep 981$ 425$

In vitro[52] canola/soybean no – 18:2 sheep – 800[51] linseed no ground linseed 18:3 sheep 850 721[51] linseed HCOOH ground linseed 18:3 sheep 850 482[51] linseed NaOH ground linseed 18:3 sheep 850 333[52] canola{ unknown – 18:2 sheep – 232[52] canola/soybean{ emulsification – 18:2 sheep – 200[52] canola/soybean emulsification – 18:2 sheep – 160[96] cottonseed oil emulsification sunflower oil 18:2 sheep 960 150

§ Procedure prior to formaldehyde treatment of protein. ‘Emulsification’ has been mentioned when no information hasbeen provided on the chemical compounds used to prepare the emulsions.$ Indicative data, as dietary 18:3 was not reported and has been assumed as 550 g/kg fatty acids, based on average 18:3concentrations of linseed oil [6] and alfalfa hay [97].# Source used in vivo by Sinclair et al. [45].{ Source used in vivo by Tymchuk et al. [98], without in vivo BH information.{ Source used in vivo by Zinn et al. [53].

6.4.2 Assessment of effectiveness of protectiontechnology: Formation of calcium salts

The level of protection from rumen metabolism for cal-cium salts has been investigated extensively, with invivo reports varying from no protection to about 40% ofthe PUFA being protected against rumen BH. Variationcould be related to dissociation of Ca salts, which isdetermined both by the rumen pH as well as the degreeof unsaturation of the PUFA, as pKa values generallyseem related to unsaturation of the soaps [55]. Indeed,in the same manner as lipolysis from glycerides is nec-essary, dissociation of calcium salts is needed beforeBH. Moreover, the extent of dissociation of calciumsalts is usually much lower than the extent of lipolysis,so that dissociation is the limiting factor of BH of PUFAsalts. In their comprehensive in vitro study, Van Neveland Demeyer [56] illustrated that the formation of cal-cium salts to protect PUFA against rumen BH is inef-fective at rumen pH of 6.2 or lower. This threshold pHvalue seems to hold in vivo. Indeed, significant reduc-tions in BH of 18:2 and 18:3 were observed in pH ran-ges above 6.2 only, both under in vitro and in vivo con-ditions (Fig. 5). Data are presented relative to their con-trol due to relatively high experiment-to-experiment

variation in BH of the controls (ranging from 750 to990 g/kg). The study by Ferlay et al. [57] has beenexcluded because of aberrantly low BH estimates,whereas a more recent study of Lundy et al. [9] was notincluded due to the lack of rumen pH information.Based on dietary concentrate (500 g/kg) and maizesilage (450 g/kg) proportions in the latter study, anaverage rumen pH below 6.2 is expected, which couldexplain minor effects on BH when feeding soybean fattyacids as calcium salts or as soybean oil.

As specific (conjugated) fatty acids exert a number ofphysiological effects, efforts are intensified to formulaterumen-protected lipid supplements enriched in thesefatty acids (e.g. [58]). With calcium salts being one of themajor commercially applied rumen protection formula-tions, attempts are made to apply this technology todevelop supplements containing these specific (con-jugated) fatty acids (e.g. [59]). However, as no informationis available on pKa values of calcium soaps containing (amixture of) these specific positional and geometric iso-mers, in vitro screenings of the effectiveness of thesesupplements at resisting rumen BH should include incu-bations at different pH values as well as the NEFA sup-plement as a negative control.



Fig. 5. Effect of formation of Ca salts on BH of18:2n-6 (triangles) and 18:3n-3 (diamonds),expressed relative to the control treatmentwithin the experiment. Open symbols indicatethat BH of Ca salts does not differ from con-trol; full symbols indicate significant differ-ences between Ca salts and control. Greysymbols are in vitro data from Van Nevel andDemeyer [56]; black symbols represent in vivodata from Fotouhi and Jenkins [91], Enjalbertet al. [86, 92], Aldrich et al. [93], Klusmeyerand Clarck [94], and Wu et al. [76].

6.4.3 Fish oil as a BH modifier

Net 18:0 production from C18 PUFA was negligible during24-h batch in vitro incubations (Fig. 6), irrespective of theamount of EPA 1 DHA incubated, although some 18:0has been produced in incubations with 9.2 g EPA/kg DMI[26]. This suggests batch in vitro incubations to be inap-propriate to simulate the dose-dependent inhibition of BHby EPA 1 DHA as observed in vivo (see Section 4.3), witha complete inhibition of the 18:0 production from C18

PUFA for EPA 1 DHA intakes above 5 g/kg DMI. How-ever, it should be noted that EPA 1 DHA intakes werebelow the in vivo threshold of 5 g/kg DMI in the study ofChow et al. [60] only and the in vitro EPA 1 DHA supplyranged from 3 to 37 g/kg DMI. Accordingly, more in vitrostudies with lower amounts of EPA and DHA are neededto appropriately evaluate the in vitro potential to simulatein vivo fish oil supplementation. Results reported byWasowska et al. [61] have been omitted from this study,as BH of 18:2n-6 and 18:3n-3 and the accumulation ofintermediates were aberrant, most probably due to theincubation of rumen inoculum alone, without buffer.

7 Continuous-culture in vitro techniques

To our knowledge, only eight papers reported the evalua-tion of rumen BH using the continuous-culture technique,sampling either the daily effluent [62–66] or the con-tinuous-culture vessel after the installation of steady-statePUFA concentrations [38, 39, 67] or 2 h after a pulse feed-ing [68] (Tab. 4).

7.1 Continuous-culture simulation of BH ofunprotected PUFA sources

BH calculations based on daily effluent data are similar tothose based on in vivo duodenal flows, whereas BH has

Fig. 6. Effect of fish oil supplementation (triangles) on netgain of 18:0 from C18:2n-6 and C18:3n-3 lost by rumenBH (C18 PUFA lost) during 24-h batch in vitro incubationswith substrate concentrations below 1 g/100 mL [26, 60,87]. Net 18:0 gain under circumstances without fish oilsupplementation (diamonds) has been added as a com-parison.

been calculated from differences in PUFA concentrationsof the input and the steady-state rumen fluid for studiesreported by Fellner et al. [38, 67] and Martin and Jenkins[39]. Aberrantly low 18:2n-6 BH values (0–650 g/kg) havebeen calculated from the latter study upon supplementa-tion of soybean oil, although pH (5.0 and 6.7), substrateconcentrations (0.05 and 0.1 g/100 mL) and dilution rates(0.05 and 0.1 h–1) were in line with the other studiesincluded in the dataset. However, both rumen inoculum(particle-free fluid obtained after centrifugation of thesqueezed rumen contents instead of squeezed rumencontents), fermentation substrate (mixed soluble carbo-hydrates instead of complete feeds) and PUFA-to-fer-mentable substrate ratio considerably differed. Particu-larly the absence of particulate material might haveimpaired normal rumen BH. Hence, this study was furtherexcluded from the dataset. Observations by Jenkins et al.[68] were not included in this dataset as samples weretaken from the rumen vessel 2 h after the morning feed-ing, which impaired the calculation of the BH.



Tab. 4. Overview of characteristics of continuous-culture fermentors used in the current review and indication whether thestudy resulted in appropriate (3) simulation or not (–) of in vivo 18:2n-6 or 18:3n-3 BH and production of the BH end-product(18:0). In the final row, some characteristics are summarized that impair appropriate in vivo simulation of both 18:2n-6 and18:3n-3 BH as well as 18:0 production. ‘IND’ (independent) is mentioned when the whole range of the above-mentionedfeatures ensures appropriate simulation.

Incubated PUFA Substrate Inoculum Dilution rate§ Adap-tation$

pH Totalvolume

BH 18:0

main source concen-tration

intro# C/F{ amount ani-mal{

pre-treatment¥ solid liquid

[g/kg DMI] [g DM/day] [h–1] [h–1] [days] [mL]

[38] 18:2 NEFA 29 emul 50 : 50 24 C cheese-cloth (2) 0.068 UN 6.8 700 3 3

[39] 18:2 oil 53–106 emul Soluble carboh. 0.6–2.4 ST particle free 0.05–0.10 UN 6.5 500 – –[63] 18:2

18:3grass 8–15

5–7– 0 : 100 14 C cheese-cloth (2) 0.046 0.063 5 5.9–6.3 700 3

3

3

[64] 18:218:3

orchardgrass

3–96–9

– 0 : 100–32 : 68 50 C cheese-cloth (1) 0.07 0.18 UN 6.8–7.1 1200 3

3

3

[64] 18:218:3

red clover 2–82–6

– 0 : 100–32 : 68 50 C cheese-cloth (1) 0.07 0.18 UN 6.8–7.1 1200 3

3

–

[65] 18:218:3

alfalfa 2–42.5–11

– 0 : 100–8 : 92 100 C cheese-cloth (2) 0.055 0.10 7 6.2 1750 3

3

3

[66] 18:218:3

NEFAalfalfa

15–352.3

emul 62 : 38 120 C cheese-cloth (2) 0.04–0.08 0.12 7 6.5 1780 3

3

3

[66] 18:218:3

NEFAalfalfa

152.3

emul 62 : 38 120 C cheese-cloth (2) 0.04 0.12 7 5.8 1780 – –

[67] 18:2 NEFA 15–30 emul 50 : 50 24 C cheese-cloth (2) 0.068 UN 6.8 700 3 3

[68] 18:2 oil 30 feed 70 : 30 28 C cheese-cloth (2) 0.068 4 6.8 700 3 3

Impairedsimulation

red clover �50 IND exclusivelysoluble carboh.

IND UN particle free IND IND �5.8 IND

§ One value for solid and liquid dilution rates indicates single-flow system.$ Number of adaptation days of continuous system prior to experimental sampling.# Mode of introduction to ensure dispersion of fatty acids in case of oils or NEFA; ‘emul’ refers to emulsions that are pre-pared through sonication.{ Concentrate/forage ratio (C/F); in study [39], soluble carbohydrates (mixture of cellobiose, glucose, maltose and xylose)were used; in study [65], the concentrate was sucrose.{ Type of donor animal: steer (ST) or dairy cow (C).¥ Particle-free inoculum obtained through consecutive filtering through four layers of cheesecloth and centrifugation at1506g.

Overall, a highly significant linear relation between 18:2n-6 or 18:3n-3 inputs and their ruminal loss is observed(Fig. 7). This suggests a negligible effect on the BH of18:2n-6 and 18:3n-3 of varying continuous-culture con-ditions, such as the type of the continuous-culture system(singlevs. dual flow), samplingmode (from rumen vessel vs.effluent), solid (0.04–0.08 h–1) and liquid (0.063–0.18 h–1)dilution rates, pH (6.0–7.25), 18:2n-6 (2–35 mg/g DMI) or18:3n-3 (2–11 mg/g DMI) concentrations, forage/con-centrate ratio of the substrate (100 : 0–30 : 70, wt/wt) andsubstrate concentrations (0.1–0.24 g/100 mL). However,continuous-culture simulations seem to overestimate invivo BH, as suggested from the slope of the linearregressions (Fig. 7) (948 and 955 g/kg for the 18:2n-6 and18:3n-3 regression, respectively). Nevertheless, thismight be due to the low number of continuous-culturestudies in combination with the high amounts of unester-

ified fatty acids in the majority of these experiments.Indeed, the 27 observations used in the linear regressionof 18:2n-6 included six infusions of unesterified 18:2n-6[38, 66, 67] and 16 NEFA-rich forages, either as silage(n = 4) [63] or frozen fresh grass (n = 12) [64]. Both ensilingas well as freezing followed by thawing of forage has beenshown to increase the amount of forage NEFA [69,70].The linear regression of 18:3n-3 included 25 observa-tions, of which 4 stem from ensiled forages [63], 12 fromthawed grass after frozen storage [64] and 9 from hay [65,66]. Oilseed sources were completely absent in the con-tinuous-culture database.

Significantly less 18:2n-6 has been lost in continuouscultures at pH 5.8 [66], which is in accordance with theprotective effect of low pH as suggested from the meta-analysis of duodenal fatty acid flows by Glasser et al. [21].



Fig. 7. Disappearance of 18:2n-6 and 18:3n-3 in single-(n = 3) and dual-flow (n = 25) continuous-culture fermen-tors with pure unesterified 18:2n-6, fresh, ensiled or driedforages [38, 63–67]. Substrate concentrations varied be-tween 0.1 and 0.24 g/100 mL. Grey diamond indicatesobservation at pH 5.8 [66], which has not been includedin the linear regression. [18:2 lost (g/kg DMI) =0.948(SE = 0.0115; p ,0.001)618:2 initial (g/kg DMI), R2 = 0.999,n = 27; 18:3 lost (g/kg DMI) = 0.955(SE = 0.0182; p ,0.001)618:3initial (g/kg DMI), R2 = 0.995, n = 25].

However, the deviation from the linear curve of this low-pH observation is greater than suggested from the in vivometa-analysis (reduction of BH of 109 g/kg for each pHunit decrease).

7.2 Continuous-culture simulation ofproduction of BH intermediates and thesaturated BH end-product from unprotectedPUFA sources

The production of 18:0 is a linear function of the amount ofC18 PUFA lost, with its slope (0.533; Fig. 8) being similar tothe slope observed under in vivo conditions (Fig. 1).However, in the experiment by Qiu et al. [66], this linearrelation was not observed at a rumen pH of 5.8 (Fig. 8).Similarly, Fig. 8 also shows that a smaller proportion ofhydrogenated 18:2n-6 and 18:3n-3 from red clover wascompletely saturated to 18:0 [64]. Hence, these observa-

Fig. 8. Formation of 18:0 in single- (n = 3) and dual-flow(n = 25) continuous-culture fermentors with pure unester-ified 18:2n-6, oil, fresh, ensiled or dried forage [38, 63–67]in relation to 18:2n-6 and 18:3n-3 lost by rumen BH (C18

PUFA lost). Substrate concentrations varied between 0.1and 0.24 g/100 mL. Grey diamonds indicate observationsat pH 5.8 [66] and cultures with red clover [64] and havenot been included in the linear regression. [18:0 produced(g/kg DMI) = 0.533(SE = 0.054; p ,0.001)6C18 PUFA lost (g/kgDMI) 1 0.860(SE = 0.071; p ,0.001)618:1 intake (g/kg DMI),R2 = 0.809, n = 18].

tions were not taken into account for the linear regressiondevelopment. No 18:1 was detected in gamagrass stud-ied by Eun et al. [63], which resulted in aberrantly low 18:1intake values when no or low levels of corn were added.Accordingly, observations of these treatments wereexcluded. From the limited number of observations, con-tinuous-culture conditions could be suggested to simu-late in vivo conversion of hydrogenated C18 PUFA to 18:0relatively well. Moreover, this conversion was independ-ent of the continuous-culture conditions, which seems animprovement compared to the batch in vitro systems.However, supplementation of fatty acids in continuous-culture studies were less than the amounts provokingaccumulation of BH intermediates during in vitro batchincubations (at least 1.5 g/100 mL substrate in combina-tion with at least 0.5 mg/mL C18 PUFA as NEFA or oils). Onthe other hand, fish oil supplementation (2 g EPA 1 DHA/kg DMI) to continuous cultures with solid dilution rates of0.03 and 0.06 h–1 completely inhibited 18:0 formationfrom hydrogenated C18 PUFA of soybean and linseed oil[62]. Indeed, under these continuous-culture conditions,the net 18:0 formed (18:0 formed minus oleic acid lost) didnot significantly differ from 0. This would indicate thatcontinuous cultures are no improvement over batch invitro cultures to quantitatively assess the in vivo dose re-sponse of fish oil supplementation. However, as theamount of currently available experimental data fromcontinuous cultures is limited, further studies are neededto evaluate the full potential of this methodology.

8 In sacco technique

The in sacco methodology is considered a convenientcompromise between the expensive in vivo approach



using duodenal cannulas and the in vitro proceduresusing a non-physiological environment. Traditionally, insacco procedures were applied to assess rumen proteinand fiber degradation, and only a limited number ofstudies used this technique to assess rumen BH [2, 71–74]. This might not be surprising as the in situ dis-appearance of PUFA might originate both from physicallosses of (esterified) PUFA from the bag as well as fromBH, which would result in an overestimation of PUFA BH.In an attempt to correct for these physical losses ofPUFA from the nylon bags, Enjalbert et al. [73] suggestedthat calculations should rely on the decrease in PUFAproportions in the bag residue relative to proportions inthe original substrate. This requires the assumption of anequal BH of PUFA that have left the bag as thoseremaining. Equal losses of 18:3n-3, 18:2n-6 and oleicacid have been demonstrated with oilseeds, based on invitro incubations of nylon bags in sterilized rumen fluidwith added pancreatin [73]. This experiment proved thatall dietary fatty acids were released to the same extent,but does not demonstrate equal BH of fatty acidsremaining in the bags versus those that have left the bag.This is unlikely as EFA in intact protective structures suchas oleosomes probably will be predominant in the lipidresidues in the bags, whereas these structures will havebeen destroyed in the lipids that have left the bags. Thiscould explain both the low BH of PUFA and the low 18:0proportions in 24-h incubated bags, when compared toin vitro [73] or in vitro and in vivo [2, 49] measurements.However, it cannot explain the higher 18:1 trans propor-tions in the bag residues compared to in vitro or in vivo.These higher amounts of trans fatty acids eventuallymight be related to temporarily high concentrations ofnon-esterified PUFA within the nylon bag, creating amicroenvironment which might be particularly toxic tothe more sensitive hydrogenating microbes responsiblefor the final BH step (conversion of 18:1 trans to 18:0).Mixing straw with the incubated oilseed might havedecreased the hydrophobicity and diluted the con-centrations of non-esterified PUFA in the bags, whichresulted in increased BH rates and production of thesaturated BH end-product [74]. However, in situ BHremained considerably lower compared to in vivo obser-vations, even when the incubated oilseeds were mixedwith straw or hay and when cows adapted to a PUFA-rich diet were used. This illustrates the limited applica-bility of this method to simulate in vivo BH.

9 General overview on in vitro rumen fattyacid metabolism and future prospects

Overall, both batch and continuous-culture in vitro sys-tems show the potential to appropriately simulate BH of

18:3n-3, 18:2n-6 and production of 18:0 from unpro-tected fatty acid sources, provided that some considera-tions are taken into account, which are summarized in thefinal rows of Tabs. 2 and 4: (1) The amount of PUFA aswell as the fermentation substrate should be limited to50 g/kg substrate DM and 10 mg/mL incubation fluid,respectively. (2) The inoculum should not be treated toremove all particles, but supplementation of additionalundegradable particulate material is not needed.(3) Adaptation of the donor animal did not affect in vitroC18 BH results, but less decisive conclusions could bemade for the type of donor animal, mainly because of thelow number of studies with sheep and steer inoculum.(4) Weighing, solvent solutions with or without solventevaporation, and emulsification through sonication allseem appropriate techniques to introduce fatty acidsources into the incubator, and both grinding through astandard sieve mesh as well as the use of a coffee grinderare appropriate oil seed pretreatments. (5) Buffer solu-tions should avoid pH shifts below 6.0, but the inoculum/buffer ratio is of less importance.

In vitro testing at different pH values is needed to assessthe effectiveness of calcium salts at resisting rumen BH.However, overestimation of rumen inertness impairsdirect extrapolation of in vitro BH results of technologi-cally protected PUFA sources. Hence, incubation seriesof these and other protected sources should include anegative (unprotected) as well as a positive control ofwhich in vivo data are available, to enable some specula-tion on the expected in vivo results.

In vitro BH of EPA and DHA seems dose dependent andgenerally underestimates in vivo BH. Moreover, in vitrosupplementation of EPA and DHA most often completelyinhibits 18:0 production, unlike in vivo circumstanceswhere a dose-dependent inhibition of the trans 18:1 to18:0 reduction has been suggested. Hence, optimizationof the in vitro methodology to improve simulation of EPAand DHA rumen metabolism remains challenging. Adap-tation of the inoculum donor animal to EPA or DHA sup-plements should be a first attempt to improve in vitrosimulations.

Acknowledgments

The authors are grateful to the OECD for providing a grantto V.F. to participate in the biohydrogenation workshop(an official pre-conference event of the 4th Euro Fed LipidCongress). B.V. is a Postdoctoral Fellow of the Fund forScientific Research-Flanders (Belgium).



References

[1] G. J. Faichney: Marker methods for measuring digesta flow.Br J Nutr. 1993, 70, 663–664.

[2] F. Akraim, M. C. Nicot, P. Weill, F. Enjalbert: Effects of pre-conditioning and extrusion of linseed on the ruminal biohy-drogenation of fatty acids. 2. In vitro and in situ studies.Anim Res. 2006, 55, 261–271.

[3] A. Troegeler-Meynadier, M. C. Nicot, C. Bayourthe, R. Mon-coulon, F. Enjalbert: Effects of pH and concentrations oflinoleic and linolenic acids on extent and intermediates ofruminal biohydrogenation in vitro. J Dairy Sci. 2003, 86,4054–4063.

[4] C. Boeckaert, B. Vlaeminck, J. Mestdagh, V. Fievez: In vitroexamination of DHA-edible micro algae: 1. Effect on rumenlipolysis and biohydrogenation of linoleic and linolenic acids.Anim Feed Sci Technol. 2007, 136, 63–79.

[5] C. V. D. M. Ribeiro, M. L. Eastridge, J. L. Firkins, N. R. St-Pierre, D. L. Palmquist: Kinetics of fatty acid biohydrogena-tion in vitro. J Dairy Sci. 2007, 90, 1405–1416.

[6] K. J. Harvatine, M. S. Allen: Fat supplements affect fractionalrates of ruminal fatty acid biohydrogenation and passage indairy cows. J Nutr. 2006, 136, 677–685.

[7] N. R. St-Pierre: Invited Review: Integrating quantitativefindings from multiple studies using mixed model method-ology. J Dairy Sci. 2001, 84, 741–755.

[8] K. J. Shingfield, S. Ahvenjarvi, V. Toivonen, A. Arola, K. V. V.Nurmela, P. Huhtanen, J. M. Griinari: Effect of dietary fish oilon biohydrogenation of fatty acids and milk fatty acid con-tent in cows. Anim Sci. 2003, 77, 165–179.

[9] F. P. Lundy, E. Block, W. C. Bridges, J. A. Bertrand, T. C.Jenkins: Ruminal biohydrogenation in Holstein cows fedsoybean fatty acids as amides or calcium salts. J Dairy Sci.2004, 87, 1038–1046.

[10] G. J. Faichney, G. A. White: Formaldehyde treatment ofconcentrate diets for sheep. 1. Partition of digestion oforganic-matter and nitrogen between stomach and intes-tines. Aust J Agric Res. 1977, 28, 1055–1067.

[11] J. P. Hogan: Intestinal digestion of subterranean clover bysheep. Aust J Agric Res. 1973, 24, 587–598.

[12] B. G. White, J. R. Ingalls, H. R. Sharma, J. A. Mckirdy: Theeffect of whole sunflower seeds on the flow of fat and fattyacids through the gastrointestinal tract of cannulated Hol-stein steers. Can J Anim Sci. 1987, 67, 447–459.

[13] G. J. Faichney: Measurement in sheep of the quantity andcomposition of rumen digesta and of the fractional outflowrates of digesta constituents. Aust J Agric Res. 1980, 31,1129–1137.

[14] F. N. Owens, C. F. Hanson: External and internal markers forappraising site and extent of digestion in ruminants. J DairySci. 1992, 75, 2605–2617.

[15] R. L. Atkinson, E. J. Scholljegerdes, S. L. Lake, V. Nayigi-hugu, B. W. Hess, D. C. Rule: Site and extent of digestion,duodenal flow, and intestinal disappearance of total andesterified fatty acids in sheep fed a high-concentrate dietsupplemented with high-linoleate safflower oil. J Anim Sci.2006, 84, 387–396.

[16] D. Bauchart, F. Legay-Carmier, M. Doreau: Ruminal hydro-lysis of dietary triglycerides in dairy cows fed lipid-supple-mented diets. Reprod Nutr Dev. 1990, Suppl 2, 187S.

[17] R. Bickerstaffe, E. F. Annison, D. E. Noakes: Quantitativeaspects of fatty acid biohydrogenation, absorption andtransfer into milk fat in lactating goat, with special referenceto cis-isomers and trans-isomers of octadecenoate andlinoleate. Biochem J. 1972, 130, 607–617.

[18] T. C. Jenkins: Protection of fatty acids against ruminal bio-hydrogenation. Eur J Lipid Sci Technol. 2007, submitted.

[19] J. J. Loor, K. Ueda, A. Ferlay, Y. Chilliard, M. Doreau: Biohy-drogenation, duodenal flow, and intestinal digestibility oftrans fatty acids and conjugated linoleic acids in response todietary forage:concentrate ratio and linseed oil in dairycows. J Dairy Sci. 2004, 87, 2472–2485.

[20] M. Doreau, Y. Chilliard: Digestion and metabolism of dietaryfat in farm animals. Br J Nutr. 1997, 78, S15–S35.

[21] F. Glasser, P. Schmidely, D. Sauvant, M. Doreau: Digestion offatty acids in ruminants: A meta-analysis of flows and varia-tion factors. 2. C18 fatty acids. Animal. 2007, submitted.

[22] V. Fievez, F. Dohme, M. Danneels, K. Raes, D. Demeyer: Fishoils as potent rumen methane inhibitors and associatedeffects on rumen fermentation in vitro and in vivo. Anim FeedSci Technol. 2003, 104, 41–58.

[23] S. Chikunya, G. Demirel, M. Enser, J. D. Wood, R. G. Wilk-inson, L. A. Sinclair: Biohydrogenation of dietary n-3 PUFAand stability of ingested vitamin E in the rumen, and theireffects on microbial activity in sheep. Br J Nutr. 2004, 91,539–550.

[24] A. M. Wachira, L. A. Sinclair, R. G. Wilkinson, K. Hallett, M.Enser, J. D. Wood: Rumen biohydrogenation of n-3 poly-unsaturated fatty acids and their effects on microbial effi-ciency and nutrient digestibility in sheep. J Agric Sci. 2000,135, 419–428.

[25] N. D. Scollan, M. S. Dhanoa, N. J. Choi, W. J. Maeng, M.Enser, J. D. Wood: Biohydrogenation and digestion of longchain fatty acids in steers fed on different sources of lipid. JAgric Sci. 2001, 136, 345–355.

[26] A. A. AbuGhazaleh, T. C. Jenkins: Disappearance of doc-osahexaenoic and eicosapentaenoic acids from cultures ofmixed ruminal microorganisms. J Dairy Sci. 2004, 87, 645–651.

[27] Z. Wu, D. L. Palmquist: Synthesis and biohydrogenation offatty acids by ruminal microorganisms in vitro. J Dairy Sci.1991, 74, 3035–3046.

[28] M. Doreau, A. Ferlay: Digestion and utilization of fatty acidsby ruminants. Anim Feed Sci Technol. 1994, 45, 379–396.

[29] B. Vlaeminck, V. Fievez, D. Demeyer, R. J. Dewhurst: Effectof forage:concentrate ratio on fatty acid composition ofrumen bacteria isolated from ruminal and duodenal digesta.J Dairy Sci. 2006, 89, 2668–2678.

[30] J. J. Loor, K. Ueda, A. Ferlay, Y. Chilliard, M. Doreau: ShortCommunication: Diurnal profiles of conjugated linoleic acidsand trans fatty acids in ruminal fluid from cows fed a highconcentrate diet supplemented with fish oil, linseed oil, orsunflower oil. J Dairy Sci. 2004, 87, 2468–2471.

[31] M. Doreau, S. Gachon: Kinetics of ruminal fatty acid con-centration as a tool to evaluate the rate of production of fattyacid from microbial hydrogenation. Reprod Nutr Dev. 2004,44 Suppl. 1, 61.

[32] M. Lourenco, B. Vlaeminck, M. Bruinenberg, D. Demeyer, V.Fievez: Milk fatty acid composition and associated rumenlipolysis and fatty acid hydrogenation when feeding foragesfrom intensively managed or semi-natural grasslands. AnimRes. 2005, 54, 471–484.

[33] T. M. Beam, T. C. Jenkins, P. J. Moate, R. A. Kohn, D. L.Palmquist: Effects of amount and source of fat on the ratesof lipolysis and biohydrogenation of fatty acids in ruminalcontents. J Dairy Sci. 2000, 83, 2564–2573.

[34] A. A. AbuGhazaleh, D. J. Schingoethe, A. R. Hippen, K. F.Kalscheur: Conjugated linoleic acid increases in milk whencows fed fish meal and extruded soybeans for an extendedperiod of time. J Dairy Sci. 2004, 87, 1758–1766.



[35] A. Roy, A. Ferlay, K. J. Shingfield, Y. Chilliard: Examination ofthe persistency of milk fatty acid composition responses toplant oils in cows given different basal diets, with particularemphasis on trans-C-18:1 fatty acids and isomers of con-jugated linoleic acid. Anim Sci. 2006, 82, 479–492.

[36] C. G. Harfoot, R. C. Noble, J. H. Moore: Food particles as asite for biohydrogenation of unsaturated fatty acids inrumen. Biochem J. 1973, 132, 829–832.

[37] N. J. Choi, J. Y. Imm, S. Oh, B. C. Kim, H. J. Hwang, Y. J.Kim: Effect of pH and oxygen on conjugated linoleic acid(CLA) production by mixed rumen bacteria from cows fedhigh concentrate and high forage diets. Anim Feed SciTechnol. 2005, 124, 643–653.

[38] V. Fellner, F. D. Sauer, J. K. G. Kramer: Steady-state rates oflinoleic-acid biohydrogenation by ruminal bacteria in con-tinuous-culture. J Dairy Sci. 1995, 78, 1815–1823.

[39] S. A. Martin, T. C. Jenkins: Factors affecting conjugatedlinoleic acid and trans C18:1 fatty acid production by mixedruminal bacteria. J Anim Sci. 2002, 80, 3347–3352.

[40] D. Deswysen, C. Van Dang, M. Focant, Y. Larondelle: Effectof linoleic acid concentrations on in vitro ruminal biohy-drogenation patterns. 4th Euro Fed Lipid Congress – Fats,Oils and Lipids for a Healthier Future, Workshop RumenBiohydrogenation, Madrid (Spain) 2006.

[41] J. P. Jouany, B. Lassalas: Gas pressure inside a rumen invitro system stimulates the use of hydrogen. Reprod NutrDev. 2002, 42 Suppl. 1, S64.

[42] M. R. F. Lee, J. K. S. Tweed, A. P. Moloney, N. D. Scollan: Theeffects of fish oil supplementation on rumen metabolism andthe biohydrogenation of unsaturated fatty acids in beefsteers given diets containing sunflower oil. Anim Sci. 2005,80, 361–367.

[43] J. J. Loor, K. Ueda, A. Ferlay, Y. Chilliard, M. Doreau: Intest-inal flow and digestibility of trans fatty acids and conjugatedlinoleic acids (CLA) in dairy cows fed a high-concentrate dietsupplemented with fish oil, linseed oil, or sunflower oil. AnimFeed Sci Technol. 2005, 119, 203–225.

[44] X. Qiu, M. L. Eastridge, J. L. Firkins: Effects of dry matterintake, addition of buffer, and source of fat on duodenal flowand concentration of conjugated linoleic acid and trans-11C-18:1 in milk. J Dairy Sci. 2004, 87, 4278–4286.

[45] L. A. Sinclair, S. L. Cooper, S. Chikunya, R. G. Wilkinson, K.G. Hallett, M. Enser, J. D. Wood: Biohydrogenation of n-3polyunsaturated fatty acids in the rumen and their effects onmicrobial metabolism and plasma fatty acid concentrationsin sheep. Anim Sci. 2005, 81, 239–248.

[46] S. K. Gulati, J. R. Ashes, T. W. Scott: Hydrogenation ofeicosapentaenoic and docosahexaenoic acids and theirincorporation into milk fat. Anim Feed Sci Technol. 1999, 79,57–64.

[47] F. Dohme, V. Fievez, K. Raes, D. I. Demeyer: Increasinglevels of two different fish oils lower ruminal biohydrogena-tion of eicosapentaenoic and docosahexaenoic acid in vitro.Anim Res. 2003, 52, 309–320.

[48] C. J. Van Nevel, D. I. Demeyer: Influence of pH on lipolysisand biohydrogenation of soybean oil by rumen contents invitro. Reprod Nutr Dev. 1996, 36, 53–63.

[49] F. Akraim, M. C. Nicot, P. Weill, F. Enjalbert: Effects of pre-conditioning and extrusion of linseed on the ruminal biohy-drogenation of fatty acids. 1. In vivo studies. Anim Res.2006, 55, 83–91.

[50] T. W. Scott, L. J. Cook, S. C. Mills: Protection of dietarypolyunsaturated fatty acids against microbial hydrogenationin ruminants. J Am Oil Chem Soc. 1971, 48, 358–364.

[51] L. A. Sinclair, S. L. Cooper, J. A. Huntington, R. G. Wilkinson,K. G. Hallett, M. Enser, J. D. Wood: In vitro biohydrogenationof n-3 polyunsaturated fatty acids protected against ruminal

microbial metabolism. Anim Feed Sci Technol. 2005, 124,579–596.

[52] S. K. Gulati, M. R. Garg, T. W. Scott: Rumen protected pro-tein and fat produced from oilseeds and/or meals by form-aldehyde treatment; their role in ruminant production andproduct quality: A review. Aust J Exp Agric. 2005, 45, 1189–1203.

[53] R. A. Zinn, S. K. Gulati, A. Plascencia, J. Salinas: Influence ofruminal biohydrogenation on the feeding value of fat in fin-ishing diets for feedlot cattle. J Anim Sci. 2000, 78, 1738–1746.

[54] J. R. Ashes, S. K. Gulati, L. J. Cook, T. W. Scott, J. B. Don-nelly: Assessing the biological effectiveness of protectedlipid supplements for ruminants. J Am Oil Chem Soc. 1979,56, 522–527.

[55] P. S. Sukhija, D. L. Palmquist: Dissociation of calcium soapsof long-chain fatty acids in rumen fluid. J Dairy Sci. 1990, 73,1784–1787.

[56] C. J. Van Nevel, D. I. Demeyer: Effect of pH on biohy-drogenation of polyunsaturated fatty acids and their Ca-salts by rumen microorganisms in vitro. Arch Anim Nutr.1996, 49, 151–157.

[57] A. Ferlay, J. Chabrot, Y. Elmeddah, M. Doreau: Ruminal lipidbalance and intestinal digestion by dairy cows fed calciumsalts of rapeseed oil fatty acids or rapeseed oil. J Anim Sci.1993, 71, 2237–2245.

[58] S. K. Gulati, S. M. Kitessa, J. R. Ashes, E. Fleck, E. B. Byers,Y. G. Byers, T. W. Scott: Protection of conjugated linoleicacids from ruminal hydrogenation and their incorporationinto milk fat. Anim Feed Sci Technol. 2000, 86, 139–148.

[59] M. J. de Veth, S. K. Gulati, N. D. Luchini, D. E. Bauman:Comparison of calcium salts and formaldehyde-protectedconjugated linoleic acid in inducing milk fat depression. JDairy Sci. 2005, 88, 1685–1693.

[60] T. T. Chow, V. Fievez, A. P. Moloney, K. Raes, D. Demeyer, S.De Smet: Effect of fish oil on in vitro rumen lipolysis, appar-ent biohydrogenation of linoleic and linolenic acid andaccumulation of biohydrogenation intermediates. AnimFeed Sci Technol. 2004, 117, 1–12.

[61] I. Wasowska, M. R. G. Maia, K. M. Niedzwiedzka, M. Czau-derna, J. M. C. Ramalho Ribeiro, E. Devillard, K. J. Shing-field, R. J. Wallace: Influence of fish oil on ruminal biohy-drogenation of C18 unsaturated fatty acids. Br J Nutr. 2006,95, 1199–1211.

[62] A. A. AbuGhazaleh, B. N. Jacobson: The effect of pH andpolyunsaturated C18 fatty acid source on the production ofvaccenic acid and conjugated linoleic acids in ruminal cul-tures incubated with docosahexaenoic acid. Anim Feed SciTechnol. 2007, 136, 11–22.

[63] J. S. Eun, V. Fellner, J. C. Burns, M. L. Gumpertz: Fermen-tation of eastern gamagrass (Tripsacum dactyloides [L.] L.)by mixed cultures of ruminal microorganisms with or withoutsupplemental corn. J Anim Sci. 2004, 82, 170–178.

[64] J. J. Loor, W. H. Hoover, T. K. Miller-Webster, J. H. Herbein,E. Polan: Biohydrogenation of unsaturated fatty acids incontinuous culture fermenters during digestion of orchard-grass or red clover with three levels of ground corn supple-mentation. J Anim Sci. 2003, 81, 1611–1627.

[65] C. V. D. M. Ribeiro, S. K. R. Karnati, M. L. Eastridge: Biohy-drogenation of fatty acids and digestibility of fresh alfalfa oralfalfa hay plus sucrose in continuous culture. J Dairy Sci.2005, 88, 4007–4017.

[66] X. Qiu, M. L. Eastridge, K. E. Griswold, J. L. Firkins: Effectsof substrate, passage rate, and pH in continuous culture onflows of conjugated linoleic acid and trans C-18:1. J DairySci. 2004, 87, 3473–3479.



[67] V. Fellner, F. D. Sauer, J. K. G. Kramer: Effect of nigericin,monensin, and tetronasin on biohydrogenation in con-tinuous flow-through ruminal fermenters. J Dairy Sci. 1997,80, 921–928.

[68] T. C. Jenkins, V. Fellner, R. K. McGuffey: Monensin by fatinteractions on trans fatty acids in cultures of mixed ruminalmicroorganisms grown in continuous fermentors fed corn orbarley. J Dairy Sci. 2003, 86, 324–330.

[69] V. Fievez, M. Ensberg, T. T. Chow, D. Demeyer: Effects offreezing and drying grass products prior to fatty acidextraction on grass fatty acid and lipid class composition – atechnical note. Commun Agric Appl Biol Sci. 2004, 69, 93–102.

[70] M. Lourenco, G. Van Ranst, V. Fievez: Differences in extentof lipolysis in red or white clover and ryegrass silages inrelation to polyphenol oxidase activity. Commun Agric ApplBiol Sci. 2005, 70, 169–172.

[71] R. Perrier, B. Michalet-Doreau, D. Bauchart, M. Doreau:Assessment of an in situ technique to estimate the degra-dation of lipids in the rumen. J Sci Food Agric. 1992, 59,449–455.

[72] P. Y. Chouinard, J. Levesque, V. Girard, G. J. Brisson: Dietarysoybeans extruded at different temperatures: Milk compo-sition and in situ fatty acid reactions. J Dairy Sci. 1997, 80,2913–2924.

[73] F. Enjalbert, P. Eynard, M. C. Nicot, A. Troegeler-Meynadier,C. Bayourthe, R. Moncoulon: In vitro versus in situ ruminalbiohydrogenation of unsaturated fatty acids from a raw orextruded mixture of ground canola seed/canola meal. JDairy Sci. 2003, 86, 351–359.

[74] A. Agazzi, C. Bayourthe, M. C. Nicot, A. Troegeler-Meyna-dier, R. Moncoulon, F. Enjalbert: In situ ruminal biohy-drogenation of fatty acids from extruded soybeans: Effectsof dietary adaptation and of mixing, with lecithin or wheatstraw. Anim Feed Sci Technol. 2004, 117, 165–175.

[75] M. R. Weisbjerg, C. F. Borsting, T. Hvelplund: Fatty acidmetabolism in the digestive tract of lactating cows fed tallowin increasing amounts at 2 feeding levels. Acta Agric ScandSect A Anim Sci. 1992, 42, 106–114.

[76] Z. Wu, O. A. Ohajuruka, D. L. Palmquist: Ruminal synthesis,biohydrogenation, and digestibility of fatty acids by dairycows. J Dairy Sci. 1991, 74, 3025–3034.

[77] B. J. Wonsil, J. H. Herbein, B. A. Watkins: Dietary and rum-inally derived trans-18:1 fatty acids alter bovine milk lipids. JNutr. 1994, 124, 556–565.

[78] S. K. Duckett, J. G. Andrae, F. N. Owens: Effect of high-oilcorn or added corn oil on ruminal biohydrogenation of fattyacids and conjugated linoleic acid formation in beef steersfed finishing diets. J Anim Sci. 2002, 80, 3353–3360.

[79] C. D. Avila, E. J. DePeters, H. Perez-Monti, S. J. Taylor, R. A.Zinn: Influences of saturation ratio of supplemental dietaryfat on digestion and milk yield in dairy cows. J Dairy Sci.2000, 83, 1505–1519.

[80] M. Murphy, P. Uden, D. L. Palmquist, H. Wiktorsson: Rumenand total diet digestibilities in lactating cows fed diets con-taining full-fat rapeseed. J Dairy Sci. 1987, 70, 1572–1582.

[81] J. Pantoja, J. L. Firkins, M. L. Eastridge, B. L. Hull: Fatty aciddigestion in lactating dairy cows fed fats varying in degree ofsaturation and different fiber sources. J Dairy Sci. 1996, 79,575–584.

[82] K. F. Kalscheur, B. B. Teter, L. S. Piperova, R. A. Erdman:Effect of fat source on duodenal flow of trans-C18:1 fattyacids and milk fat production in dairy cows. J Dairy Sci.1997, 80, 2115–2126.

[83] J. J. Loor, J. H. Herbein, T. C. Jenkins: Nutrient digestion,biohydrogenation, and fatty acid profiles in blood plasmaand milk fat from lactating Holstein cows fed canola oil orcanolamide. Anim Feed Sci Technol. 2002, 97, 65–82.

[84] B. S. Oldick, J. L. Firkins: Effects of degree of fat saturationon fiber digestion and microbial protein synthesis when dietsare fed twelve times daily. J Anim Sci. 2000, 78, 2412–2420.

[85] E. J. Scholljegerdes, B. W. Hess, G. E. Moss, D. L. Hixon, D.C. Rule: Influence of supplemental cracked high-linoleate orhigh-oleate safflower seeds on site and extent of digestion inbeef cattle. J Anim Sci. 2004, 82, 3577–3588.

[86] F. Enjalbert, M. C. Nicot, C. Bayourthe, M. Vernay, R. Mon-coulon: Effects of dietary calcium soaps of unsaturated fattyacids on digestion, milk composition and physical proper-ties of butter. J Dairy Res. 1997, 64, 181–195.

[87] A. A. AbuGhazaleh, T. C. Jenkins: Short Communication:Docosahexaenoic acid promotes vaccenic acid accumula-tion in mixed ruminal cultures when incubated with linoleicacid. J Dairy Sci. 2004, 87, 1047–1050.

[88] A. A. AbuGhazaleh, L. D. Holmes, B. N. Jacobson, K. F.Kalscheur: Short Communication: Eicosatrienoic acid anddocosatrienoic acid do not promote vaccenic acid accu-mulation in mixed ruminal cultures. J Dairy Sci. 2006, 89,4336–4339.

[89] P. V. Reddy, J. L. Morrill, T. G. Nagaraja: Release of free fattyacids from raw or processed soybeans and subsequenteffects on fiber digestibilities. J Dairy Sci. 1994, 77, 3410–3416.

[90] V. I. Fievez, C. J. van Nevel, D. I. Demeyer: Lipolysis andbiohydrogenation of PUFA’s from fish oil during in vitroincubations with rumen contents. Proc Nutr Soc. 2000, 59,193A–193A.

[91] N. Fotouhi, T. C. Jenkins: Ruminal biohydrogenation of lino-leoyl methionine and calcium linoleate in sheep. J Anim Sci.1992, 70, 3607–3614.

[92] F. Enjalbert, D. Griess, M. C. Nicot: Effect of different formsof unsaturated fatty acids in sheep on digestibility, and esti-mation of ruminal biohydrogenation. Rev Med Vet. 1994,145, 477–483.

[93] C. G. Aldrich, N. R. Merchen, J. K. Drackley: The effect ofroasting temperature applied to whole soybeans on site ofdigestion by steers. 1. Organic-matter, energy, fiber, andfatty-acid digestion. J Anim Sci. 1995, 73, 2120–2130.

[94] T. H. Klusmeyer, J. H. Clark: Effects of dietary fat and proteinon fatty acid flow to the duodenum and in milk produced bydairy cows. J Dairy Sci. 1991, 74, 3055–3067.

[95] H. S. Hussein, N. R. Merchen, G. C. Fahey: Effects ofchemical treatment of whole canola seed on digestion oflong-chain fatty acids by steers fed high or low forage diets.J Dairy Sci. 1996, 79, 87–97.

[96] S. K. Gulati, T. W. Scott, J. R. Ashes: In-vitro assessment offat supplements for ruminants. Anim Feed Sci Technol.1997, 64, 127–132.

[97] R. J. Dewhurst, W. J. Fisher, J. K. S. Tweed, R. J. Wilkins:Comparison of grass and legume silages for milk produc-tion. 1. Production responses with different levels of con-centrate. J Dairy Sci. 2003, 86, 2598–2611.

[98] S. M. Tymchuk, G. R. Khorasani, J. J. Kennelly: Effect offeeding formaldehyde- and heat-treated oil seed on milkyield and milk composition. Can J Anim Sci. 1998, 78, 693–700.

[Received: February 6, 2007; accepted: June 1, 2007]


Assessing Rumen Biohydrogenation and Its Manipulation in Vivo, In Vitro, In Situ

Documents

Transcript of Assessing Rumen Biohydrogenation and Its Manipulation in Vivo, In Vitro, In Situ