INVITED REVIEW: Research on ruminal biohydrogenation: Achievements, gaps in knowledge, and future approaches from the perspective of dairy science.

Scientific knowledge about ruminal biohydrogenation (BH) has improved greatly since this metabolic process was empirically confirmed in 1951. For years, BH had mostly been perceived as a process to be avoided to increase the post-ruminal flow of UFA from the diet. Two milestones changed this perception and stimulated great interest in BH intermediates themselves: In 1987, the in vitro anticarcinogenic properties of CLA were described, and in 2000, the inhibition of milk fat synthesis by trans-10 cis-12 CLA was confirmed. Since then, numerous BH metabolites have been described in small and large ruminants, and the major deviation from the common BH pathway (i.e., the trans-10 shift) has been reasonably well established. However, there are some less well-characterized alterations, and the comprehensive description of new BH intermediates (e.g., using isotopic tracers) has not been coupled with research on their biological effects. In this regard, the low quality of some published fatty acid profiles may also be limiting the advance of knowledge in BH. Furthermore, although BH seems to no longer be considered a metabolic niche inhabited by a few bacterial species with a highly specific metabolic capability, researchers have failed to elucidate which specific microbial groups are involved in the process and the basis for alterations in BH pathways (i.e., changes in microbial populations or their activity). Unraveling both issues may be beneficial for the description of new microbial enzymes involved in ruminal lipid metabolism that have industrial interest. From the perspective of diary science, other knowledge gaps that require additional research in the coming years are evaluation of the relationship between BH and feed efficiency and enteric methane emissions, as well as improving our understanding of how alterations in BH are involved in milk fat depression. Addressing these issues will have relevant practical implications in dairy science.

Assessment of milk metabolites as biomarkers for predicting feed efficiency in dairy sheep.

Estimating feed efficiency (FE) in dairy sheep is challenging due to the high cost of systems that measure individual feed intake. Identifying proxies that can serve as effective predictors of FE could make it possible to introduce FE into breeding programs. Here, 39 Assaf ewes in first lactation were evaluated regarding their FE by 2 metrics, residual feed intake (RFI) and feed conversion ratio (FCR). The ewes were classified into high, medium and low groups for each metric. Milk samples of the 39 ewes were subjected to untargeted metabolomics analysis. The complete milk metabolomic signature was used to discriminate the FE groups using partial least squares discriminant analysis. A total of 41 and 26 features were selected as the most relevant features for the discrimination of RFI and FCR groups, respectively. The predictive ability when utilizing the complete milk metabolomic signature and the reduced data sets were investigated using 4 machine learning (ML) algorithms and a multivariate regression method. The orthogonal partial least squares algorithm outperformed other ML algorithms for FCR prediction in the scenarios using the complete milk metabolite signature (R2 = 0.62 ± 0.06) and the 26 selected features (R2 = 0.62 ± 0.15). Regarding RFI predictions, the scenarios using the 41 selected features outperformed the scenario with the complete milk metabolite signature, where the multilayer feedforward artificial neural network (R2 = 0.18 ± 0.14) and extreme gradient boosting (R2 = 0.17 ± 0.15) outperformed other algorithms. The functionality of the selected metabolites implied that the metabolism of glucose, galactose, fructose, sphingolipids, amino acids, insulin, and thyroid hormones was at play. Compared with the use of traditional methods, practical applications of these biomarkers might simplify and reduce costs in selecting feed-efficient ewes.

Relationship between feed efficiency and resilience in dairy ewes subjected to acute underfeeding.

Selection of dairy sheep based on production levels has caused a loss of rusticity, which might compromise their future resilience to nutritional challenges. Although refocusing breeding programs toward improved feed efficiency (FE) is expected, more-efficient ewes also seem to be more productive. As a first step to examine the relationship between FE and resilience in dairy sheep, in this study we explored the variation in the response to and the recovery from an acute nutritional challenge in high-yielding Assaf ewes phenotypically divergent for FE. First, feed intake, milk yield and composition, and body weight changes were recorded individually over a 3-wk period in a total of 40 sheep fed a total mixed ration (TMR) ad libitum. Data were used to calculate their FE index (FEI, defined as the difference between the actual and predicted intake estimated through net energy requirements for maintenance, production, and weight change). The highest and lowest FE ewes (H-FE and L-FE groups, respectively; 10 animals/group) were selected and then subjected to the nutritional challenge (i.e., withdrawing the TMR and limiting their diet only to the consumption of straw for 3 d). Afterward, sheep were fed again the TMR ad libitum. Temporal patterns of variation in performance traits, and ruminal fermentation and blood parameters were examined. A good consistency between FEI, residual feed intake, and feed conversion ratio was observed. Results supported that H-FE were more productive than L-FE sheep at similar intake level. Average time trends of milk yield generated by a piecewise model suggest that temporal patterns of variation in this trait would be related to prechallenge production level (i.e., H-FE presented quicker response and recovery than L-FE). Considering all studied traits, the overall response to and recovery from underfeeding was apparently similar or even better in H-FE than in L-FE. This would refute the initial hypothesis of a poorer resilience of more-efficient sheep to an acute underfeeding. However, the question remains whether a longer term feed restriction might impair the ability of H-FE ewes to maintain or revert to a high-production status, which would require further research.

Effect of lipid supplementation on the endogenous synthesis of milk cis-9,trans-11 conjugated linoleic acid in dairy sheep and goats: A tracer assay with 13C-vaccenic acid.

A major proportion of milk rumenic acid (RA; cis-9,trans-11 CLA) is synthesized through mammary Δ9-desaturation of vaccenic acid (VA; trans-11 18:1). Diet composition may determine the relative contribution of this endogenous synthesis to milk RA content, with effects that might differ between ruminant species. However, this hypothesis is mostly based on estimated values, proxies of stearoyl-CoA desaturase (SCD) activity, and indirect comparisons between publications in the literature. With the aim of providing new insights into this issue, in vivo Δ9-desaturation of 13C-labeled VA (measured via milk 13C-VA and -RA secretion) was directly compared in sheep and goats fed a diet without lipid supplementation or including 2% of linseed oil. Four Assaf sheep and 4 Murciano-Granadina goats were used in a replicated 2 × 2 crossover design to test the effects of the 2 dietary treatments during 2 consecutive 25-d periods. On d 22 of each period, 500 mg of 13C-VA were i.v. injected to each animal. Dairy performance, milk fatty acid profile, including isotope analysis, and mammary mRNA abundance of genes coding for SCD were examined on d 21 to 25 of each period. Supplementation with linseed oil improved milk fat concentration and increased the content of milk VA and RA. However, the isotopic tracer assay suggested no variation in the relative proportion of VA desaturated to milk RA, and the percentage of this CLA isomer deriving from SCD activity would remain constant regardless of dietary treatment. These results put into question a major effect of lipid supplementation on the endogenous synthesis of milk RA and support that mammary Δ9-desaturation capacity would not represent a limiting factor when designing feeding strategies to increase milk RA content. The lack of diet-induced effects was common to caprines and ovines, but inherent interspecies differences in mammary lipogenesis were found. Thus, the higher proportions of VA desaturation and endogenous synthesis of milk RA in sheep supported a greater SCD activity compared with goats, a finding that was not associated with the similar mRNA abundance of SCD1 in the 2 species. On the other hand, transfer efficiency of the isotopic tracer to milk was 37% higher in caprine than in ovine, suggesting a greater efficiency in mammary fatty acid uptake from plasma in caprine.

Rumen biohydrogenation and milk fatty acid profile in dairy ewes divergent for feed efficiency.

A sustainable increase in livestock production would require selection for improved feed efficiency, but the mechanisms underlying this trait and explaining its large individual variation in dairy ruminants remain unclear. This study was conducted in lactating ewes to test the hypothesis that rumen biohydrogenation (BH) would differ between high- and low-efficiency animals, and these differences would be reflected in rumen fatty acid (FA) profile and affect milk FA composition. A second aim was to identify differences in FA that may serve as biomarkers of feed efficiency. Data of daily feed intake and milk yield and composition, as well as body weight, were collected individually over a 3-wk period in 40 ewes. The difference between the mean actual and predicted feed intake (estimated through metabolizable energy requirements for maintenance, production, and body weight change) over the period was used as the feed efficiency index (FEI) to select 8 of the highest feed efficiency (H-FE) and 8 of the lowest feed efficiency (L-FE) animals. In addition, residual feed intake (RFI) was estimated as the residual term from the regression of feed intake on various energy sinks. Rumen and milk FA composition were characterized by using gas chromatography, and results were analyzed using a statistical model that included the fixed effect of the group (H-FE vs. L-FE). The FEI averaged -0.29 ± 0.046 and 0.81 ± 0.084 in H-FE and L-FE, respectively, whereas RFI averaged -0.16 ± 0.084 and 0.18 ± 0.082, respectively. The correlation coefficient between both metrics was 0.69. Feed intake was similar in both groups, but H-FE showed greater milk yield, with increases in lactose content and yield, and in milk protein and fat production. Results from rumen FA profiles included a lower proportion of 18:2n-6, cis-9 18:1, and of several of their BH metabolites, and a greater concentration of 18:0, which may indicate that the apparent BH would be more complete in more efficient sheep. Milk FA analysis suggested that the greater fat yield in the H-FE group was mostly explained by increased de novo FA synthesis, whereas their milk would have lower proportions of cis-9 18:1 and C20 to 22n-6 polyunsaturated FA than L-FE. Stepwise multiple linear regression suggested that milk C20 to 22n-6 PUFA might be convenient biomarkers to discriminate more efficient dairy sheep. Further research is needed to validate these findings (e.g., under different dietary conditions).

Endogenous synthesis of milk cis-9,trans-11 conjugated linoleic acid in dairy ewes: Quantification using 13C-labeled vaccenic acid and comparison with estimates based on cobalt administration.

Isotopic tracers are used to directly quantify the effect of mammary Δ9-desaturation on milk fatty acid (FA) composition, but very few studies have applied this method to measuring the endogenous synthesis of rumenic acid (RA; cis-9,trans-11 conjugated linoleic acid) in cows and goats, and no publications exist in ewes. In sheep, knowledge about the contribution of stearoyl-coenzyme A desaturase (SCD) to milk FA secretion is derived mostly from indirect estimates based on inhibition of the enzyme by oral administration of cobalt, a cost-effective method that has not been validated to date. To fill both gaps, we conducted an isotopic tracer assay in sheep to quantify the proportion of endogenous RA in milk for the first time in this species. We then compared the results with estimates derived from a Co administration assay performed on the same animals. First, 5 lactating ewes received an intravenous injection of 200 mg of [1-13C]trans-11 18:1 (vaccenic acid, VA), the precursor for RA production by SCD activity. At -24, -15, 0, 9, 24, 33, 48, 57, 72, 81, and 96 h post-injection, we recorded milk yield and collected milk samples to examine fat concentration and FA profile. We conducted compound-specific isotope analysis of VA and RA by gas chromatography-combustion isotope ratio mass spectrometry. Afterward, in the Co administration assay, ewes received a daily dose of 7 mg of Co/kg of body weight for 5 d. We analyzed milk samples for composition before and on the last days of cobalt dosing. On average, 17% of the injected amount of [1-13C]VA was transferred to milk within 96 h post-injection, and up to 29% of the VA taken up by the mammary gland was desaturated to milk RA. Under our conditions, the mean proportion of this conjugated linoleic acid isomer deriving from Δ9-desaturation represented 82 to 90% of the amount secreted in milk. However, the proportion estimated in the Co assay with calculations based on individual FA concentrations was lower (on average, 46%). When we calculated the same estimates based on changes in Δ9-desaturation ratios after Co dosing, the higher values of endogenous RA (75%) did not differ from the results of the isotopic tracer assay. Nevertheless, correlation analysis between the results obtained through [1-13C]VA or Co administration revealed no significant relationship, which would prevent acceptance of the latter as a reliable alternative to isotopic labeling to examine mammary Δ9-desaturation in dairy ewes.

Effect of dietary lipids and other nutrients on milk odd- and branched-chain fatty acid composition in dairy ewes.

Milk odd- and branched-chain fatty acids (OBCFA) are largely derived from bacteria leaving the rumen, which has encouraged research on their use as biomarkers of rumen function. Targeted research has examined relationships between these fatty acids (FA) and dietary components, but interactions between the effects of lipids and other nutrients on milk OBCFA are not well characterized yet. Furthermore, factors controlling milk OBCFA in sheep are largely unknown. Thus, the present meta-analysis examined relationships between diet composition and milk OBCFA using a database compiled with lot observations from 14 trials in dairy ewes fed lipid supplements. A total of 47 lots received lipid supplements, whereas their respective controls (27 lots) were fed the same basal diets without lipid supplementation. Relationships between milk OBCFA and dietary components were first assessed through a principal component analysis (PCA) and a correlation analysis. Then, responses of milk OBCFA to variations in specific dietary components (selected on the basis of the PCA) were examined in more detail by regression analysis. According to the loading plot, dietary unsaturated C18 FA loaded opposite to major milk OBCFA (e.g., 15:0, 15:0 anteiso, and 17:0) and were strongly correlated with principal component 1, which described 46% of variability. Overall, regression equations supported this negative, and generally linear, relationship between unsaturated C18 FA levels and milk OBCFA. However, the influence of C20-22 n-3 polyunsaturated FA and saturated FA was more limited. The PCA also suggested that dietary crude protein is not a determinant of milk OBCFA profile in dairy ewes, but significant relationships were observed between some OBCFA and dietary fiber or starch, consistent with a potential role of these FA as biomarkers of rumen cellulolytic and amylolytic bacteria. In this regard, regression equations indicated that iso FA would show opposite responses to increasing levels of acid detergent fiber (positive linear coefficients) and starch (negative linear coefficients). Lipid supplementation would not largely affect these associations, supporting the potential of OBCFA as noninvasive markers of rumen function under different feeding conditions (i.e., with or without lipid supplementation). Because consumption of these FA may have nutritional benefits for humans, the use of high-fiber/low-starch rations might be recommended to maintain the highest possible content of milk OBCFA in dairy sheep.

In vitro biohydrogenation of 13C-labeled α-linolenic acid in response to ruminal alterations associated with diet-induced milk fat depression in ewes.

The basis for marine lipid-induced milk fat depression (MFD) has not been established yet, but recent reports suggest the putative contribution of shifts in the ruminal metabolism of α-linolenic acid (ALA). To test this hypothesis, an isotopic tracer approach was used in batch cultures of rumen microorganisms with inoculum collected from cannulated ewes fed either a total mixed ration without lipid supplementation (control inoculum) or the same diet supplemented with 2% of fish oil, which is known to cause MFD in lactating sheep (FO-MFD inoculum). The [1-13C]ALA was added at a dose of 1% of incubated dry matter and the proportions of 13C-labeled fatty acids (FA) were examined after 24 h of incubation, using complementary gas chromatography and gas chromatography-combustion isotope ratio mass spectrometry (GC-C-IRMS) analyses. Expected differences in FA profiles were confirmed between control and FO-MFD inocula (e.g., large decreases in 18:0 and increases in most 18:1 and 18:2 intermediates, particularly trans isomers, to fish oil supply). The biohydrogenation of 13ALA was extensive and yielded multiple metabolites, with a total of 48 chromatographic peaks showing 13C enrichment, regardless of the inoculum type. However, although ALA was biohydrogenated through common pathways under standard or MFD conditions, large changes in the accumulation of 13C-labeled FA suggest important differences in the relative contribution of each specific route. First, increased accumulation of trans-11-containing FA in FO-MFD incubations was accompanied by a general repression of the trans-13/14 pathway (supported by lower trans-13+14 18:1 or trans-11,trans-13 18:2 proportions), together with a lower production of cis FA (e.g., cis-9, -12, and -15 18:1 and some cis,cis 18:2). Results also downplayed the relevance of the trans-11 to trans-10 shift as an effective marker of diet-induced MFD in sheep, and challenged the involvement of some trans-10 intermediates (e.g., trans-10 18:1 and trans-10,cis-15 18:2) in the low-fat milk syndrome in this species. Conversely, increased abundance of most 18:3 intermediates (including some unidentified isomers) might be representative of ruminal alterations related to fish oil supplementation in ewes. Further research is necessary to examine the potential association between these findings and MFD in lactating animals.

Effects of shearing 2 breeds of dairy ewes during lactation under mild winter conditions.

The lactational effects of shearing (CO, control unshorn; SH, shorn) were investigated in 48 dairy ewes of 2 breeds (Lacaune, LC, n = 24; Manchega, MN, n = 24) having a similar stage of lactation (120 ± 6 d in milk) and body frame (65.1 ± 1.5 kg of body weight and 2.4 ± 0.1 body condition score), but differing in fleece and milk production. Ewes were penned indoors, adapted to the diet (alfalfa hay ad libitum and fixed amount of concentrate), and allocated for 30 d in 8 balanced groups to which the experimental treatments were applied. All ewes were sheared on the same day. Feed intake by pen and milk yield by ewe were recorded daily. Individual samples of milk (d -3, 3, 5, 7, and 15) and blood (d -7, 3, 7, and 15) were collected, as well as body weight and body condition score measured (d -15, 0, and 15), related to shearing. Pooled milk samples per pen were also collected before and after shearing for milk fatty acid analysis (d -3 and 15). Average temperatures in the barn before (12.6 ± 0.7°C) and after (13.7 ± 0.4°C) shearing were mild. Fleece was heavier in MN than in LC (1.04 ± 0.10 vs. 0.75 ± 0.09 kg/ewe) and tended to cover more body surface in MN than in LC ewes. Responses to shearing varied according to breed, the rectal temperature after shearing only decreasing significantly in the MN (-0.36 ± 0.09°C). Feed intake increased in the LC-SH (5%), when compared with LC-CO, but did not vary in the MN ewes. Ingestibility of the alfalfa hay, expressed as filling units for sheep and monitored in 2 groups of 6 dry and unshorn ewes of each breed (73.0 ± 2.5 kg of body weight and 3.1 ± 0.2 body condition score), was constant throughout the experiment (0.99 ± 0.03 filling units for sheep/kg of dry matter). Regarding milk production, LC-SH ewes yielded 10% more milk (1.38 ± 0.06 vs. 1.52 ± 0.05 kg/d) than LC-CO ewes, but no differences were detected in MN ewes (0.74 ± 0.03 kg/d, on average). No differences in the concentration of major milk components by effect of the shearing treatment were detected in either breed, but LC-SH ewes yielded 9% more milk protein than did LC-CO ewes. No relevant effects of shearing were detected on milk fatty acid profiles, although MN ewes showed lower C4:0, C6:0, C14:0, trans-11 and trans-12 C18:1 contents than did LC ewes. Moreover, no changes by effect of shearing were detected in plasma glucose, nonesterified fatty acids, cortisol, or insulin values in either breed, or in body weight or body condition score. In conclusion, shearing dairy ewes during lactation under mild winter conditions is a suitable management option that may increase feed intake and milk production, without deleterious effects on milk composition.

Milk fat depression in dairy ewes fed fish oil: Might differences in rumen biohydrogenation, fermentation, or bacterial community explain the individual variation?

Dairy ewes show large individual variation in the extent of diet-induced milk fat depression (MFD) but reasons behind this variability remain uncertain. Previous results offered no convincing support for these differences being related to relevant changes in the milk fatty acid (FA) profile, including potentially antilipogenic FA, or in the transcript abundance of candidate genes involved in mammary lipogenesis. Therefore, we hypothesized that alterations in the processes of rumen biohydrogenation and fermentation, as well as in the bacterial community structure, might account for individual variation in fish oil-induced MFD severity. To test this explanation, 15 ewes received a total mixed ration without lipid supplementation (control; n = 5) or supplemented with 20 g of fish oil/kg of dry matter [10 animals divided into those showing a strong (RESPON+; -25.4%; n = 5) or a mild (RESPON-; -7.7%; n = 5) decrease in milk fat concentration] for 5 wk. Rumen fermentation parameters, biohydrogenation metabolites, and bacterial structure and diversity were analyzed in rumen samples collected before and after treatments. Although the fish oil supplementation increased the concentration of demonstrated or putative antilipogenic FA (e.g., cis-9 16:1, cis-11 18:1, or trans-10,cis-12 CLA), surprisingly, none of them differed significantly in relation to the extent of MFD (i.e., between RESPON- and RESPON+), and this was the case only for a few minor FA (e.g., cis-6+7 16:1 or 17:0 anteiso). Changes in total volatile FA, acetate, and propionate concentrations were associated with MFD severity, with higher decreases in more susceptible animals. Individual responses were not related to shifts in rumen bacterial structure but some terminal restriction fragments compatible with Clostridiales, Ruminococcaceae, Lachnospiraceae, and Succiniclasticum showed greater abundances in RESPON-, whereas some others that may correspond to Prevotella, Mogibacterium, and Quinella-related spp. were more abundant in RESPON+. Overall, the results suggest that individual variation in MFD severity in dairy ewes fed fish oil cannot be fully explained by differences in the processes of rumen biohydrogenation and fermentation or in the bacterial community, and further research would be necessary to elucidate the large variability in the responsiveness to MFD-inducing marine lipids.

In vitro ruminal biohydrogenation of eicosapentaenoic (EPA), docosapentaenoic (DPA), and docosahexaenoic acid (DHA) in cows and ewes: Intermediate metabolites and pathways.

A great deal of uncertainty still exists about intermediate metabolites and pathways explaining the biohydrogenation (BH) of 20- and 22-carbon polyunsaturated fatty acids (PUFA). Therefore, this study was conducted to provide further insight into the ruminal metabolism of 20:5 n-3 (EPA), 22:5 n-3 (DPA), and 22:6 n-3 (DHA), the main n-3 PUFA present in the marine lipids used in dairy ruminant feeding, and to examine potential differences between bovine and ovine. To meet this aim, we investigated the 20- and 22-carbon metabolites accumulated during in vitro incubation of EPA, DPA, and DHA with rumen inocula from cows and ewes. The PUFA were added at a dose of 2% incubated dry matter and digesta samples were analyzed after 24 h of incubation using complementary gas-liquid chromatography of fatty acid methyl esters and gas chromatography-mass spectrometry of 4,4-dimethyloxazoline derivatives. Results suggested that the main BH pathway of EPA and DPA would proceed via the reduction of the double bond closest to the carboxyl group (cis-5 in EPA and cis-7 in DPA); curiously, this mechanism seemed of much lower importance for DHA. Thus, DPA would not be a major intermediate product of DHA and their BH might actually follow separate pathways, with the accumulation of numerous unique metabolites in each case. A principal component analysis supported this hypothesis, with a clear separation between PUFA treatments in the score and loading plots. Within EPA and DPA groups, cow and ewe samples loaded separately from each other but not distant. No conjugated 20:5, 22:5, or 22:6 isomer compatible with the initial product of EPA, DPA, or DHA metabolism, respectively, was identified in the ruminal digesta, although this would not unequivocally exclude their transient formation. In this regard, results from DPA incubations provided the first indication that the metabolism of this very long chain PUFA may involve the formation of conjugated double bond structures. The BH of EPA, DPA, and DHA resulted in the appearance of several tentative trans-10-containing metabolites, showing a general trend to be more abundant in the digesta of ewes than in that of cows. This finding was speculated to have some relationship with the susceptibility of dairy sheep to marine lipid-induced milk fat depression. Differences in the relative proportion of intermediate products would also suggest an influence of ruminant species on BH kinetics, with a process that would likely be slower and less complete in cows than in ewes.

Individual variation of the extent of milk fat depression in dairy ewes fed fish oil: Milk fatty acid profile and mRNA abundance of candidate genes involved in mammary lipogenesis.

Dairy ewes are less prone than cows to milk fat depression (MFD) but suffer from this syndrome when marine lipids are added to their diet to modulate milk fatty acid (FA) profile. However, there are large individual differences in MFD extent, and the reasons behind this variability are uncertain. On this basis, a study was conducted in lactating sheep to test the hypotheses that individual susceptibility to the low-fat milk condition may be explained by differences in (1) the milk concentration of some FA, particularly antilipogenic FA, or (2) the transcriptional regulation of mammary lipogenesis. For 5 wk, 15 ewes received a total mixed ration supplemented with 0 (control; n = 5) or 20 g of fish oil/kg of dry matter [10 animals selected out of 22 and divided into those showing marked (RESPON+; n = 5) or mild (RESPON-; n = 5) MFD]. Milk production and composition, including a comprehensive FA profile, were examined on 3 consecutive days before and after treatments. Candidate gene expression was also analyzed before the start of the trial and at its end using RNA isolated from milk somatic cells. According to the experimental design, the fish-oil-induced decrease in milk fat concentration was much stronger in RESPON+ (-25.4%) than in RESPON- (-7.7%). Milk from all ewes fed the supplemented diet showed rather uniform changes in the proportion of potentially healthy FA (such as cis-9,trans-11 18:2, trans-11 18:1, or 20:5n-3) and of those with confirmed or putative antilipogenic effects (e.g., cis-9 16:1, trans-10 and cis-11 18:1, trans-9,cis-11 18:2, and 10-oxo-18:0), without significant variation between RESPON+ and RESPON-. It was not possible to relate the very few exceptions to this general trend (e.g., in cis-7 16:1 and 22:6n-3) to responsiveness. Major mechanisms involved in mammary lipogenesis, specifically the uptake and de novo synthesis of FA, appeared to be unequally inhibited in ewes displaying different degrees of MFD, with molar yields of >16C FA being unaffected in RESPON-. However, this was not reflected in candidate gene expression. Supplementation with fish oil showed a tendency to lower the mRNA abundance of lipogenic genes such as ACSS2, FASN, LPIN1, FADS2, and INSIG1, but only SCD and GPAT4 tended to differ between RESPON- and RESPON+. Overall, these results offer no convincing support for the initial hypotheses, so further research must be pursued to explain the individual variation in MFD severity.

Endogenous synthesis of milk oleic acid in dairy ewes: In vivo measurement using 13C-labeled stearic acid.

The use of stable isotopes is a reliable and risk-free alternative to radioactive tracers for directly examining in vivo fatty acid (FA) metabolism. However, very limited information is available in ruminants, and none is available in sheep. Therefore, we conducted an experiment in dairy ewes to determine, for the first time in this species, the uptake, Δ9-desaturation, and secretion of 13C-labeled stearic acid (SA) into milk with the aim of measuring in vivo endogenous synthesis of milk oleic acid (OA) and stearoyl-CoA desaturase activity. Six lactating Assaf ewes fed a total mixed ration (forage:concentrate ratio = 30:70) received an intravenous injection of 2 g of 13C-labeled SA. At -24, -15, 0, 4, 8, 12, 16, 20, 24, 36, 48, 60, and 72 h postinjection (p.i.), milk yield was recorded and milk samples were collected to examine fat concentration and FA composition, including compound-specific isotope analysis of SA and OA by gas chromatography-combustion isotope ratio mass spectrometry. Over the p.i. period, the SA proportion ranged from 7.6 to 8.3% of total FA, with a maximum 13C enrichment of 1.9%, whereas OA was more abundant (14.3-15.4% of total FA) and had lower 13C enrichments (up to 0.69%). On average, 15% of the isotopic tracer was transferred to milk within 72 h p.i., and 47 to 50% of the SA taken up by the mammary gland would have been desaturated to OA. The proportion of oleic acid being synthesized endogenously was estimated to represent between 48 and 57% of the amount secreted in milk. Further research under different dietary conditions is recommended.

mRNA abundance of genes involved in mammary lipogenesis during fish oil- or trans-10,cis-12 CLA-induced milk fat depression in dairy ewes.

Milk fat depression (MFD) caused by trans-10,cis-12 18:2 is known to be mediated in cows and ewes by downregulation of mammary lipogenic genes. However, transcriptional mechanisms underlying marine lipid-induced MFD have not been well defined yet and the few available studies in ovine are not consistent. This trial was conducted to directly compare changes in animal performance, milk fatty acid composition, and particularly mammary mRNA abundance of candidate lipogenic genes and transcription factors in response to the inclusion of fish oil or trans-10,cis-12 18:2 in the dairy sheep diet. To meet this objective, 12 lactating Assaf ewes (on average, 64 days in milk, producing 1.72 kg of milk/d with 5.17% of fat) were divided into 3 groups and offered a total mixed ration without supplementation (control) or supplemented with 2.4% dry matter of fish oil (FO treatment) or 1% dry matter of a commercial product rich in trans-10,cis-12 18:2 (CLA treatment) for 39 d. Measurements and samplings were conducted before starting the treatments and at the end of the trial. Milk samples were used for RNA extraction from somatic cells. Feed intake was not affected by lipid supplements, and as designed, reductions in milk fat concentration (-31%) were similar in the 2 treatments, although the unpredicted increase in milk production with FO counteracted the anticipated reduction in milk fat yield. Nevertheless, this did not preclude the detection of FO-induced decreases in the mRNA abundance of candidate lipogenic genes [e.g., acyl-CoA synthetase short-chain family member 2 (ACSS2), fatty acid synthase (FASN), and lipin 1 (LPIN1)], thus supporting the hypothesis that transcriptional regulation would be a relevant component of this type of MFD in sheep. Expected CLA-induced downregulation of some genes, such as FASN or sterol regulatory element binding transcription factor 1 (SREBF1), could not be detected in our samples, which might be related, at least in part, to high inter-individual variation and relatively advanced lactation stage (on average 102-103 d in milk on d 38 and 39). Overall, direct comparison of the effects of dietary FO and CLA on transcript abundance of candidate lipogenic genes and transcription factors suggest that there might be relevant differences in the transcriptional control mechanisms underlying the MFD induced by each kind of supplement (i.e., fish oil or CLA).

In vitro response to EPA, DPA, and DHA: Comparison of effects on ruminal fermentation and biohydrogenation of 18-carbon fatty acids in cows and ewes.

The modulation of milk fat nutritional quality through fish oil supplementation seems to be largely explained by the action of n-3 very long chain polyunsaturated fatty acids (PUFA) on ruminal biohydrogenation (BH) of C18 fatty acids (FA). However, relationships among this action, disappearance of those PUFA in the rumen, and potential detrimental consequences on ruminal fermentation remain uncertain. This study compared the effect of 20:5n-3 (eicosapentaenoic acid; EPA), 22:5n-3 (docosapentaenoic acid; DPA), and 22:6n-3 (docosahexaenoic acid; DHA) on rumen fermentation and BH of C18 FA and was conducted simultaneously in cows and sheep to provide novel insights into interspecies differences. The trial was performed in vitro using batch cultures of rumen microorganisms with inocula collected from cannulated cows and ewes. The PUFA were added at a dose of 2% incubated dry matter, and treatment effects on ruminal C18 FA concentrations, PUFA disappearances, and fermentation parameters (gas production, ammonia and volatile FA concentrations, and dry matter and neutral detergent fiber disappearances) were examined after 24 h of incubation. A principal component analysis suggested that responses to PUFA treatments explained most of the variability; those of ruminant species were of lower relevance. Overall, EPA and DHA were equally effective for inhibiting the saturation of trans-11 18:1 to 18:0 and had a similar influence on ruminal fermentation in cows and sheep (e.g., reductions in gas production and acetate:propionate ratio). Nevertheless, DHA further promoted alternative BH pathways that lead to trans-10 18:1 accumulation, and EPA seemed to have specific effects on 18:3n-3 metabolism. Only minor variations attributable to DPA were observed in the studied parameters, suggesting a low contribution of this FA to the action of marine lipids. Although most changes due to the added PUFA were comparable in bovine and ovine, there were also relevant specificities, such as a stronger inhibition of 18:0 formation in cows and a greater increase in 18:3n-3 metabolites in sheep. No direct relationship between in vitro disappearance of the incubated PUFA and effect on BH (in particular, inhibition of the last step) was found in either cows or ewes, calling into question a putative link between extent of disappearance and toxicity for microbiota. Conversely, an association between the influence of these PUFA on ruminal lipid metabolism and fermentation may exist in both species. In vivo verification of these findings would be advisable.

Comparison of feed intake, digestion and rumen function among domestic ruminant species grazing in upland vegetation communities.

This study aimed to compare feed intake, digestion, rumen fermentation parameters and bacterial community of 5 beef cows, 12 crossed ewes and 12 goats grazing together in spring-early summer on heather-gorse vegetation communities with an adjacent area of improved pasture. Organic matter intake (OMI) and digestibility (OMD) were estimated using alkane markers. Ruminal fluid samples were collected for measuring fermentation parameters, and studying the bacterial community using terminal restriction fragment length polymorphism (T-RFLP). Spot samples of urine were taken to determine purine derivative (PD) and creatinine concentrations to estimate microbial protein synthesis in the rumen. Herbaceous species were the main dietary component in all animal species. Cattle had higher (p < 0.05) daily OMI (g/kg LW0.75 ) and OMD, whereas sheep and goats showed similar values. The highest ammonia concentration was observed in sheep. Total VFA, acetate and butyrate concentrations were not influenced by animal species, while propionate concentrations in goats were 1.8 times lower (p < 0.05) than in sheep. Acetate:propionate ratio was greater (p < 0.05) in goats, whereas cattle excreted more allantoin (p < 0.05). Estimated supply of microbial N was higher in cows (p < 0.01), whereas the efficiency of microbial protein synthesis was lower (p < 0.01) in this animal species. Hierarchical clustering analysis indicated a clear effect of animal species on rumen bacterial structure. Differences among animal species were also observed in the relative frequency of several T-RFs. Certain T-RFs compatible with Lachnospiraceae, Proteobacteria and Clostridiales species were not found in goats, while these animals showed high relative frequencies of some fragments compatible with the Ruminococcaceae family that were not detected in sheep and cattle. Results suggest a close relationship between animals' grazing behaviour and rumen bacterial structure and its function. Goats seem to show a greater specialization of their microbial populations to deal with the greater fibrous and tannin content of their diet.

Does supplemental 18:0 alleviate fish oil-induced milk fat depression in dairy ewes?

Supplementation of dairy ewe diet with marine lipids may be an effective strategy for modulating milk fatty acid composition but induces milk fat depression (MFD). This syndrome has been associated with a shortage of 18:0 for uptake and Δ(9)-desaturation that may impair the capacity of the mammary gland to achieve an adequate fluidity for milk fat secretion. On this basis, it was suggested that supplemental 18:0 may contribute to alleviate marine lipid-induced MFD in sheep. To test this hypothesis, 12 lactating ewes were allocated to 1 of 3 lots and used in a 3×3 Latin square design with 3 periods of 28 d each and 3 experimental treatments: a total mixed ration without lipid supplementation (control) or supplemented with 20 g/kg of DM of fish oil alone (FO) or in combination with 20 g/kg of DM of 18:0 (FOSA). Diets were offered ad libitum, and animal performance and rumen and milk fatty acid composition were studied at the end of each period. After completing the Latin square trial and following a change-over design, the in vivo digestibility of supplemental 18:0 was estimated using 6 lactating sheep. As expected, diet supplementation with fish oil increased the milk content of some potentially health-promoting fatty acids (e.g., cis-9,trans-11 18:2, trans-11 18:1, 20:5n-3, 22:5n-3, and 22:6n-3), but reduced milk fat concentration and yield (-20% in both FO and FOSA treatments). Thus, although reductions in milk 18:0 and cis-9 18:1 output caused by FO (-81 and -51%, respectively) were partially reversed with FOSA diet (-49 and -27%, respectively), the addition of 18:0 to the diet did not prove useful to alleviate MFD. This response, which could not be fully accounted for by the low digestibility coefficient of supplemental 18:0, may challenge the theory of a shortage of this fatty acid as a mechanism to explain fish oil-induced MFD in sheep. Effects of FO and FOSA on rumen and milk fatty acid composition would support that increases in the concentration of some candidate milk fat inhibitors (e.g., cis-9 16:1 or 10-oxo-18:0) might play a relevant role in this type of MFD.

Fish oil-induced milk fat depression and associated downregulation of mammary lipogenic genes in dairy ewes.

Several studies in dairy cows have shown a relationship between milk fat depression (MFD) and alterations caused in lipogenic gene expression by dietary nutrients. However, information on small ruminants is not only scarce but also inconsistent. Therefore, this experiment was conducted in dairy ewes to study the effect of a diet known to induce MFD on milk fatty acid (FA) composition and mRNA abundance of key candidate genes involved in mammary lipogenesis. Twelve lactating Assaf ewes (on average 63d in milk) were randomly assigned to 2 treatments consisting of a total mixed ration based on alfalfa hay and concentrates (50:50), supplemented with 0 (control) or 17g of fish oil/kg of diet dry matter (FO). Profiles of milk FA and mRNA abundance of candidate genes in biopsied mammary tissue were examined before starting the treatments and after 1 and 4.5wk on the diets. As expected, FO induced MFD and modified milk FA composition. Compared with the control, reductions in milk fat concentration and yield were not detected on d 7, but reached up to 25 and 22%, respectively, on d 30. However, increases in confirmed or putative antilipogenic FA (trans-10,cis-12 and trans-9,cis-11 18:2, cis-9 16:1, cis-11 18:1, and oxo-FA) were already established on the early stage of the treatment and lasted until the end of the feeding period. These changes were accompanied by decreases in the mRNA abundance of genes encoding lipogenic enzymes. The coordinated nature of the downregulation, which tended to affect most studied metabolic pathways, including FA activation (ACSS1), de novo synthesis (ACACA and FASN), uptake and transport (LPL and FABP3), desaturation (SCD1), and esterification (AGAPT6), supports the involvement of a central regulator of milk fat synthesis. In this regard, without ruling out the potential contribution of PPARG, our results suggest that SREBF1 would have a relevant role in the MFD syndrome in sheep fed FO. Among the other studied transcription factors, the tendency to a downregulation of INSIG1 was associated with that of SREBF1, whereas no variation was detected for SCAP or THRSP. Fish oil had no significant effects on the transcript abundance of CD36, GPAM, DGAT1, LPIN1, and XDH. Overall, changes in potential antilipogenic FA and mRNA abundance of candidate lipogenic genes support a relationship between them and suggest that FO-induced MFD in dairy ewes would be mediated by transcriptional mechanisms.

Isolation of RNA from milk somatic cells as an alternative to biopsies of mammary tissue for nutrigenomic studies in dairy ewes.

Nutrigenomic studies of mammary lipogenesis in ruminants often rely on the use of mammary tissue (MT) collected either by biopsy or at slaughter. However, isolating RNA from milk would be a useful and cost-effective technique that may avoid distress to the animal and facilitate the collection of samples in time series experiments. This assay was therefore conducted to test the hypothesis that RNA extracted from milk somatic cells (MSC) in dairy sheep would be a feasible alternative to the performance of MT biopsies for nutrigenomic analyses. To meet this objective, 8 lactating Assaf ewes were divided in 2 groups and offered a total mixed ration without supplementation (control) or supplemented with 2.4% dry matter of fish oil, which was known not only to elicit milk fat depression but also to downregulate the expression of some candidate genes involved in mammary lipogenesis. Total RNA was extracted from MSC and biopsied MT to examine whether the potential changes in the abundance of transcripts was similarly detected with both RNA sources. Milk fatty acid profile was also analyzed by gas chromatography, and variations in mRNA abundance were determined by reverse transcription quantitative PCR. Values of RNA integrity number were always ≥7.7. The expected and designed decrease of milk fat concentration with fish oil (-29%), was associated with a lower transcript abundance of genes coding for enzymes involved in fatty acid activation (ACSS1), de novo synthesis (ACACA and FASN), uptake from plasma lipids (LPL), and esterification of fatty acids to glycerol (LPIN1), as well as of a transcription factor that may regulate their expression (INSIG1). Stable mRNA levels were showed in other candidate genes, such as FABP3, GPAT4, or SCD. Changes due to the dietary treatment were similarly detected with both RNA sources (MSC and MT biopsies), which supports the initial hypothesis and would validate the use of milk as an alternative RNA source for nutrigenomic analyses in dairy sheep.

Comparison of ruminal lipid metabolism in dairy cows and goats fed diets supplemented with starch, plant oil, or fish oil.

Direct comparison of cow and goat performance and milk fatty acid responses to diets known to induce milk fat depression (MFD) in the bovine reveals relevant species-by-diet interactions in ruminal lipid metabolism. Thus, this study was conducted to infer potential mechanisms responsible for differences in the rumen microbial biohydrogenation (BH) due to diet and ruminant species. To meet this objective, 12 cows and 15 goats were fed a basal diet (control), a similar diet supplemented with 2.2% fish oil (FO), or a diet containing 5.3% sunflower oil and additional starch (+38%; SOS) according to a 3 × 3 Latin square design with 25-d experimental periods. On the last day of each period, fatty acid composition (by gas chromatography) and bacterial community (by terminal-RFLP), as well as fermentation characteristics, were measured in rumen fluid samples. Results showed significant differences in the response of cows and goats to dietary treatments, although variations in some fermentation parameters (e.g., decreases in the acetate-to-propionate ratio due to FO or SOS) were similar in both species. Main alterations in ruminal BH pathways potentially responsible for MFD on the SOS diet (i.e., the shift from trans-11 to trans-10 18:1 and related increases in trans-10,cis-12 18:2) tended to be more pronounced in cows, which is consistent with an associated MFD only in this species. However, changes linked to FO-induced MFD (e.g., decreases in 18:0 and increases in total trans-18:1) were stronger in caprine rumen fluid, which may explain their unexpected susceptibility (although less marked than in bovine) to the negative effect of FO on milk fat content. Altogether, these results suggest that distinct ruminal mechanisms lead to each type of diet-induced MFD and confirm a pronounced interaction with species. With regard to microbiota, differences between cows and goats in the composition of the rumen bacterial community might be behind the disparity in the microorganisms affected by the experimental diets (e.g., Ruminococcaceae, Lachnospiraceae, and Succinivibrionaceae in the bovine, and Pseudobutryrivibrio, Clostridium cluster IV, Prevotella, and Veillonellaceae in the caprine), which hindered the assignation of bacterial populations to particular BH steps or pathways. Furthermore, most relevant variations in microbial groups corresponded to as yet uncultured bacteria and suggest that these microorganisms may play a predominant role in the ruminal lipid metabolism in both cows and goats.