Usability of volatile organic compounds from exhaled breath compared with those from ruminal fluid, serum, urine, and milk to identify diet-specific metabolite profiles in lactating dairy cows

To investigate dietary influences on the volatilome, the volatile subcategory of the metabolome, we performed a comparative untargeted volatilome analysis of exhaled breath, ruminal fluid, serum, urine, and milk from lactating Holstein cows fed different diets. Thirty-two cows (83.3 ± 31.40 DIM, 30.6 ± 5.03 kg of milk/d) were assigned to 4 diets. The experiment lasted 16 wk. Throughout the experiment, half of the animals were fed a hay-based diet (HD; n = 16), and the other half were fed a silage-based diet (SIL; n = 16). In experimental wk 5 to 12, half of the animals in each group received the control concentrate (CON), and the other half was fed with the CON supplemented with a blend of essential oils (EXP). We hypothesized that the basal diet and the essential oils influence the volatile organic compound (VOC) profiles of the cows through potential changes in ruminal fermentation, digestion, and metabolism (hypothesis 1). Furthermore, we hypothesized that the potential effects of essential oils would have a delayed onset and a carryover effect (hypothesis 2). Every 4 experimental weeks (i.e., in wk 4, wk 8, wk 12, and wk 16), samples of exhaled breath, ruminal fluid, serum, urine, milk, and feed were collected for dynamic headspace extraction and gas chromatographic analysis of VOC in their gaseous phase. Milk yield, milk composition, BW, and feed intake were recorded regularly. Linear mixed models and multivariate and univariate data analyses were performed. The total DMI and basal diet intake was similar between cows fed HD and SIL diets. However, SIL cows consumed less of the concentrate, NDF, and water-soluble carbohydrates and more starch than HD cows. The SIL cows had a higher milk production than the HD cows. No effect was found regarding the concentrate type on feed intake or milk production. Irrespective of diet, 2,957 VOC were detected in the gaseous phase of serum; 2,771 in exhaled breath; 1,016 in urine; 1,001 in milk; and 921 in ruminal fluid. Across the experimental wk 4, 8, 12, and 16, the basal diet altered the VOC profiles of ruminal fluid, urine, and exhaled breath but not those of serum and milk. The concentrate type affected only the VOC profiles of the exhaled breath. Most diet-influenced VOC in the affected biological matrices were identified as dietary components. The experimental week influenced the VOC profiles of all matrices, especially those of exhaled breath. The VOC profile of exhaled breath strongly correlated with that of urine, followed by that of ruminal fluid, milk, and serum. This study provides the first description of diet- and time-specific VOC profiles from the biological matrices of dairy cows. The identified discriminatory VOC seem suitable as markers to discriminate between HD and SIL cows. Exhaled breath may be a promising, sensitive, and less invasive tool to follow diet- and time-related metabolic changes.