If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
The effects of monensin on transition cow metabolism may be dependent on modulation of feeding behavior, rumen pH, and expression of key metabolic genes. Multiparous Holstein cows were used to determine the effects of monensin (400 mg/cow daily) on these variables. Cows were randomly assigned, based on calving date, to control or monensin treatments (n = 16 per treatment) 21 d before their expected calving date, and cows remained on treatments through 21 d postpartum. Feeding behavior and water intake data were collected daily. Liver biopsies were conducted after assessing BCS and BW on d −21, −7, 1, 7, and 21 relative to calving for analysis of triglyceride (TG) content as well as mRNA abundance of cytosolic phosphoenolpyruvate carboxykinase, carnitine palmitoyltransferase 1a, and apolipoprotein B. Blood samples were collected 21, 7, and 4 d before expected calving and 1 (day of calving), 4, 7, 14, and 21 d postpartum for nonesterified fatty acid, β-hydroxybutyrate, glucose, insulin, and haptoglobin analyses. Ruminal pH was collected every 5 min on d 1 through 6 postpartum via a wireless indwelling probe. On d 7 postpartum, a caffeine clearance test was performed to assess liver function. Data were analyzed using mixed models with repeated measures over time. Monensin decreased mean plasma β-hydroxybutyrate (734 vs. 616 ± 41 μM) and peak concentrations (1,076 vs. 777 ± 70 μM on d 4 postpartum). Monensin also decreased time between meals prepartum (143 vs. 126 ± 5.0 min) and postpartum (88.8 vs. 81.4 ± 2.9 min), which was likely related to a smaller ruminal pH standard deviation in the first day after cows changed to a lactation ration (0.31 vs. 0.26 ± 0.015). Monensin also increased liver mRNA abundance of carnitine palmitoyltransferase 1a (0.10 vs. 0.15 ± 0.002 arbitrary units), which corresponded to a slower rate of liver TG accumulation from d −7 to +7 (412 vs. 128 ± 83 mg of TG/g of protein over this time period). No significant effects of monensin supplementation were observed on milk production, liver cytosolic phosphoenolpyruvate carboxykinase, apolipoprotein B, plasma nonesterified fatty acid, glucose, insulin, or haptoglobin. No effects on disease incidence were detected, but sample size was small for detecting such effects. Overall, results confirm that the effects of monensin on transition cows extend beyond altered propionate flux.
The weeks surrounding parturition are a critical time in the life cycle of a high-producing dairy cow. During this period, cows make many metabolic adjustments to support the transition from pregnancy to lactation. Furthermore, dairy cows produce milk in excess of their ability to consume energy, resulting in a period of negative energy balance in early lactation (
). In recent years, monensin has been used to help mitigate the effects of negative energy balance, presumably by promoting ruminal production of glucogenic precursors (
Under normal physiological conditions, monensin alters ruminal digestion in a manner that augments propionate production rate and concentration in the rumen (
), which could improve the overall energetic balance of the transition cow. Thus, an increased supply of glucose is often assumed to be the primary benefit of monensin supplementation. Observations of
support the hypothesis that monensin decreases the acetate:propionate ratio in transition cows. In contrast, when propionate kinetics were measured during the periparturient period, monensin did not affect ruminal propionate production (
). This evidence suggests that the beneficial effects of monensin likely extend beyond gluconeogenic flux, and may even be independent of changes in gluconeogenesis.
Transition cow health is directly linked to DMI, partly because negative energy balance is not as severe in early lactation cows with higher intakes (
), but feeding behavior was not measured in either study. Research from feedlot cattle suggests dietary monensin could modulate intake patterns, thus decreasing dramatic changes in rumen pH while cattle are adapting to a high-energy diet (
). Furthermore, data obtained from midlactation cows subjected to SARA indicates that administering monensin in feed increases meal frequency during the challenge and recovery periods (
). The hypothesis that monensin affects transition cow rumen pH has not been extensively investigated. Monensin administered through a controlled-release capsule increased transition cow rumen pH (
), but these data must be interpreted with caution because ruminal samples were collected using an esophageal tube. Research conducted using indwelling probes to measure pH showed no difference between monensin supplemented cows and control cows (
used a controlled-release capsule, which may not modulate intake as much as dietary monensin. Furthermore, neither of these studies examined differences in variance of rumen pH.
The primary objectives of this study were to determine the effects of monensin on transition cow feeding behavior and metabolic parameters. The secondary objectives were to assess the effects of monensin on ruminal pH and productivity of transition cows.
Materials and methods
All experimental procedures were approved by the Institutional Animal Care and Use Committee at Kansas State University.
Design and Treatments
Thirty-two multiparous Holstein transition cows from the Kansas State University Dairy Cattle Teaching and Research Facility (Manhattan) were used in a randomized complete block design. Cows were blocked by expected calving date (16 blocks) and randomly assigned within block to 1 of 2 treatments (n = 16 cows per treatment) 21 d before their expected calving date. Cows remained on their respective treatments through 21 d postpartum. The treatment group received monensin (Rumensin; Elanco Animal Health, Greenfield, IN) as a top-dress at a rate of 400 mg/cow per day, and the control group received no monensin for the duration of the study. Cows were dried off an estimated 45 d before calving. Monensin was excluded from the far-off dry cow ration to help ensure that no cows entering the study were influenced by prior monensin exposure. Cows entered the study from January to June 2010.
Monensin was premixed into a ground corn carrier and 0.91 kg of the premix was offered daily to each cow in the treatment group. Monensin treatments were administered by top dressing and manually mixing the premix into the upper 33% of each TMR once per day. The ground corn carrier was top dressed to the control cows at the same rate in the same manner. The monensin dose was selected based on previous research (
requirements (Table 1). Samples of all dietary ingredients were collected weekly and stored at −20°C. Upon study completion, feed ingredients were composited into bimonthly samples for wet chemistry analysis of CP, ADF, NDF, ether extract, and ash by Dairy One Forage Laboratory (Ithaca, NY).
Table 1Ingredient and nutrient composition of diets
A portion of the corn grain (0.91kg/cow daily) served as the top-dress carrier for 400mg of monensin (Rumensin; Elanco Animal Health, Greenfield, IN) for supplemented cows. The same amount of corn alone was top-dressed for control cows.
Prepartum premix consisted of 42.6% vitamin E premix, 11.9% Se premix, 11.6% trace mineral salt, 10.8% limestone, 9.71% vitamin A premix, 6.47% 4-plex, 4.31% vitamin D premix, 2.17% magnesium oxide, and 0.53% ethylenediamine dihydroiodide.
2 A portion of the corn grain (0.91 kg/cow daily) served as the top-dress carrier for 400 mg of monensin (Rumensin; Elanco Animal Health, Greenfield, IN) for supplemented cows. The same amount of corn alone was top-dressed for control cows.
3 Prepartum premix consisted of 42.6% vitamin E premix, 11.9% Se premix, 11.6% trace mineral salt, 10.8% limestone, 9.71% vitamin A premix, 6.47% 4-plex, 4.31% vitamin D premix, 2.17% magnesium oxide, and 0.53% ethylenediamine dihydroiodide.
4 Postpartum premix consisted of 48.4% limestone, 27.3% sodium bicarbonate, 12.6% trace mineral salt, 6.04% magnesium oxide, 2.33% 4-plex, 1.51% Se premix, 1.16% vitamin E premix, 0.46% vitamin A premix, 0.21% vitamin D premix, and 0.03% ethylenediamine dihydroiodide.
Cows were dried off and moved into a freestall pen approximately 45 d prepartum where cows received a low-energy diet (≈1.35 Mcal/kg) containing no monensin. Dry cows were moved into the maternity barn approximately 1 wk before starting the study. Cows were allowed ad libitum access to the designated treatment rations by an electronic gating system (Roughage Intake System; Insentec B.V., Marknesse, the Netherlands), 1 cow assigned per gate. After parturition, cows were moved into a tie-stall facility where they remained through 21 d postpartum. Individual feed bunks in the tie-stall facility were suspended from load cells and bunk weight was monitored continuously by computer. Feed weights and times were stored before and immediately after any deviation in bunk weight. Dry cows were fed twice daily (0800 and 1530 h) to accommodate the capacity of the feeding system used prepartum, and lactating cows were fed once daily (1500 h) to minimize the time during which feeding behavior could not be recorded. All feeding activity, including meal length and size, were recorded electronically. As-fed feed intake of each cow was recorded on a daily basis. As-fed ration consumption was adjusted for DM content for determination of meal and daily DMI. Dry matter percentage was determined weekly for the corn silage and bimonthly for each concentrate ingredient; these values were used to determine ration DM for each week. Water was offered ad libitum, and individual water consumption was also recorded daily throughout the study. During summer months, the maternity barn and tie-stall facility were both cooled using evaporative pads.
Cows were milked 3 times daily in a milking parlor, and milk yields were recorded at each milking. Milk samples were collected from each milking beginning at 4 DIM and continuing through 21 DIM. Samples were analyzed for concentrations of fat, true protein, lactose (B-2000 Infrared Analyzer; Bentley Instruments Inc., Chaska, MN), urea nitrogen (MUN spectrophotometer; Bentley Instruments Inc.), and somatic cells (SCC 500, Bentley Instruments Inc.; Heart of America DHIA, Manhattan, KS). Data from individual milkings were averaged by cow-day, using a statistical model to account for the random effect of milking. Energy-corrected milk yield was calculated as: 0.327 × milk yield + 12.86 × fat yield + 7.65 × protein yield (
Body weight was measured 2 h before feeding on d −21 and −7 relative to expected calving, and on d 1, 7, and 21 postpartum. Immediately after BW was obtained, liver samples were collected via needle biopsy. For collection of liver tissue, an area spanning the tenth and eleventh ribs was shaved, aseptically prepared, and anesthetized with 3.5 mL of subcutaneous lidocaine hydrochloride (Agri Laboratories Ltd., St. Joseph, MO). After 5 min, a #10 blade (Feather Safety Razor Co. Ltd., Kita-Ku, Osaka, Japan) was used to make a stab incision into the body wall. A 14-gauge × 15 cm biopsy needle (SABD-1415–15-T; US Biopsy, Franklin, IN) was inserted through the incision and 200 mg of tissue was collected (a total of 10 biopsies). Liver samples were snap frozen in liquid nitrogen immediately after collection, and then stored at −80°C until subsequent analysis. Blood samples were collected from the coccygeal vessels after each biopsy and also 2 h before feeding on d −4 relative to expected calving date and on d 4 and 14 postpartum. Approximately 14 mL of blood was collected into 2 tubes, one containing potassium EDTA and the other containing potassium oxalate with sodium fluoride as a glycolytic inhibitor (Vacutainer; Becton, Dickinson and Co., Franklin Lakes, NJ). Blood samples were centrifuged at 2,000 × g for 10 min immediately after sample collection, and plasma was collected and frozen at −20°C until subsequent analysis of glucose, NEFA, BHBA, insulin, and haptoglobin concentrations. Body condition score was evaluated by 3 trained investigators on the same day (±1 d) as BW and liver sample collections.
Indwelling ruminal pH probes (Rumen Sensors; Kahne Ltd., Auckland, New Zealand) were delivered to the rumen as an oral bolus after liver biopsy on d 1 postpartum. These probes measured rumen pH every 5 min and used a radio frequency to transmit this data to a computer. Electronic data were collected in real time and were also stored on the probe. Stored data were downloaded during the biopsy on d 7 to ensure collection of all recorded data. Measurement of rumen pH was limited to the first 7 DIM because of concern about drift in pH measurements from probes remaining in situ for more than 7 d. Probes removed from 2 cows (cannulated for previous studies) on d 7 postpartum generated pH values of 6.94 and 6.96 in pH 7.0 buffer and 4.01 and 3.97 in pH 4.0 buffer, suggesting that data analyzed here were valid.
Liver function was assessed using an in vivo caffeine metabolism test on d 7 postpartum (
). To conduct this test, jugular catheters (#1411; Mila International, Erlanger, KY) were inserted and caffeine was administered as caffeine-sodium benzoate (C4144; Sigma-Aldrich Co., St. Louis, MO) in a sterile pyrogen-free normal saline solution (25 mg of caffeine/mL of solution). Caffeine was infused intravenously in a bolus dose at the rate of 1 mg of caffeine/kg of BW, and the d 7 postpartum BW was used to calculate the amount of caffeine to infuse. Caffeine infusions were initiated at feeding time. Blood samples were collected immediately before infusion and at 30-min intervals for 180 min following infusion using disposable 5-mL syringes. Blood was immediately transferred to a tube containing potassium EDTA (Vacutainer; Becton, Dickinson and Co.). Plasma was collected as described above and frozen at −20°C until analysis. Catheter patency was maintained by flushing with 6 mL of sterile 3.5% sodium citrate solution following each collection.
Facilities and equipment were observed daily for abnormalities. Postpartum cows (and prepartum cows with abnormalities) underwent a health inspection daily, including monitoring for urine ketones (ReliOn ketone test strips; Bayer Healthcare LLC., Mishawaka, IN) and rectal temperature. Health records were kept throughout the study to register the incidence of ketosis, left displaced abomasum, retained placenta, metritis, milk fever, lameness, and other abnormalities. Ketosis was defined as a urine ketone concentration >80 mg/dL, or urine ketone concentrations >40 mg/dL for 2 or more consecutive days. Mastitis was defined as an SCC greater than 200,000 at any milking after d 3 (
. If cows displayed signs of any disorder described they were treated according to on-site standard operating procedures. Cows were removed from the study if they were diagnosed with a displaced abomasum (n = 3) or severe lameness (n = 1). Data obtained from these cows before removal from the study were included in all analyses. Cows diagnosed with a displaced abomasum were removed on d 7 (control), 8 (monensin), and 13 (control) postpartum, respectively. The cow removed for lameness issues was removed on d 7 (control) postpartum.
Liver and Plasma Analyses
Approximately 20 mg of liver was placed in 500 μL of chilled phosphate-buffered saline (pH 7.4) and homogenized. The homogenate was centrifuged at 2,000 × g for 10 min at 4°C and 100 μL of the supernatant was then removed for free glycerol and total protein analysis. Triglyceride (TG) content was measured using a method adapted from
. The remaining liver homogenate was incubated with 100 μL of lipase (porcine pancreatic lipase, MP Biomedicals LLC, Solon, OH) for 16 h at 37°C, and glycerol content was then determined by an enzymatic glycerol phosphate oxidase method (#F6428; Sigma-Aldrich Co.). Triglyceride content was calculated based on the difference between glycerol concentrations before and after lipase digestion. Total protein content of the original homogenate was analyzed by a Coomassie blue (
) colorimetric method (kit #23236; Thermo Scientific, Pierce, Rockford, IL). To avoid potential bias introduced by differences in moisture content of liver samples, liver TG concentration was normalized by protein concentration, which is unaltered in fatty liver (
The mRNA abundance of cytosolic phosphoenolpyruvate carboxykinase (cPEPCK), carnitine palmitoyltransferase 1a (CPT1a), apolipoprotein B (Apo B), and ribosomal protein subunit 9 (RPS9) in liver tissue was determined by real-time PCR. Total RNA was isolated from liver samples using a commercial kit (RNeasy Lipid Tissue Mini Kit; Qiagen Inc., Valencia, CA) and spectroscopy was used to quantify RNA (NanoDrop 1000; NanoDrop Technologies Inc., Wilmington, DE). Coding DNA was then synthesized from 2 μg of total RNA using a commercial kit (High-Capacity cDNA Reverse Transcription kit; Applied Biosystems, Foster City, CA). Quantitative real-time PCR was performed in triplicate with one-twentieth of the cDNA product in the presence of 200 nM gene-specific forward and reverse primers (Table 2) using SYBR green fluorescent detection (ABI 7500 Fast; Applied Biosystems). Messenger RNA abundance was quantified using the delta delta cycle threshold (Ct) method, with RPS9 used to normalize values. Abundance of RPS9 mRNA within liver tissue did not differ in response to treatment or across days in the study (all P > 0.50), making it a valid reference gene.
Table 2Primers used for quantitative real-time PCR detection of transcripts in liver tissue
Amplicons span an exon-exon boundary, as predicted by aligning the specified sequence to the bovine genome using Splign (http://www.ncbi.nlm.nih.gov/sutils/splign).
CPT1a
DV820520.1
CTTCCCATTCCGCACTTTC
616–719
CCATGTCCTTGTAATGAGCCA
cPEPCK
NM_174737.2
CGAGAGCAAAGAGATACGGTGC
427–562
TGACATACATGGTGCGACCCT
Apo B
XM_583270.3
TCCTTGATTCCACATGCAGCT
8,610–8,720
GGTGTGCAAAGGATGCGTTAG
RPS9
DT860044.1
GAACAAACGTGAGGTCTGGAGG
233–344
ATTACCTTCGAACAGACGCCG
1 CPT1a = carnitine palmitoyl transferase 1a; cPEPCK = cytosolic phosphoenolpyruvate carboxykinase; Apo B = apolipoprotein B; RPS9 = ribosomal protein subunit 9.
3 Amplicons span an exon-exon boundary, as predicted by aligning the specified sequence to the bovine genome using Splign (http://www.ncbi.nlm.nih.gov/sutils/splign).
Plasma was analyzed for NEFA using an enzymatic colorimetric procedure (NEFA-HR; Wako Chemicals USA Inc., Richmond, VA), glucose by a colorimetric kit (kit #439–90901; Wako Chemicals USA Inc.), insulin by a bovine-specific sandwich ELISA (#10–1201–01; Mercodia AB, Uppsala, Sweden), haptoglobin by a bovine-specific ELISA (kit #2410–7 using 3,000-fold dilution; Life Diagnostics Inc., West Chester, PA), and BHBA using an enzymatic reaction (kit #H7587–58; Pointe Scientific Inc., Canton, MI). Absorbance was read on a spectrophotometer (PowerWave XS; BioTek Instruments Inc., Winooski, VT) and calculations were performed using Gen5 software (BioTek Instruments Inc.).
High-performance liquid chromatography was used to quantify plasma caffeine following the procedures of
. Briefly, 500 μL of plasma was mixed with 500 μL of 0.8 M perchloric acid. The mixture was centrifuged at 14,000 × g for 20 min at 21°C, and 400 μL of the clarified supernatant was transferred to an autosampler vial containing 20 μL of 4 M NaOH. Vials were capped and 50 μL was injected into a Discovery BIO Wide Pore C18 guard column (2 cm × 4 mm, 5-μm particle size; Supelco #568272-U; Sigma-Aldrich Co.) and a Discovery BIO Wide Pore C18 column (25 cm × 4.6 mm, 5-μm particle size; Supelco #568222-U, Sigma-Aldrich). The photochemical reaction was carried out in a mobile phase consisting of 20 mM phosphate buffer (pH 3.0)-acetonitrile, 85:15 (vol/vol), at a flow rate of 1 mL/min. Absorbance was read at 273 nm using an Acutect 500 UV/Vis detector (#06–653–5; Thermo Fisher Scientific, Waltham, MA).
Data and Statistical Analysis
Feeding behavior variables were calculated from logged data that included the start and end weights as well as start and end times of meals. To generate meaningful meal pattern data, feeding bouts are grouped into meals, but no broadly accepted definition exists of what constitutes a meal. Previous reports have used minimum thresholds for intermeal intervals ranging from 2 to more than 40 min (
). Selecting a threshold was complicated in this study because cows started in a pen setting and then moved into tie-stalls. Therefore, approaches to data analysis were examined from 4 studies conducted by different laboratories. In these studies, minimum intermeal intervals were defined as 7.5 (
). For our feeding systems and housing situations, a 12-min intermeal interval was determined to be appropriate, because it is within the range of previously used thresholds and because it generated meal frequencies similar to those reported in the studies above. Therefore, meals were combined if the intermeal interval was less than 12 min. After combining meals accordingly, any meal <0.2 kg of DM was excluded from behavior analysis. Ruminal pH data were analyzed to determine mean pH, standard deviation, amount of time spent below pH 5.8, and area (pH × min) under pH 5.8 for each cow each day. The threshold of 5.8 is representative of SARA according to previously established standards (
). Caffeine elimination half-lives were determined by plotting post-infusion caffeine concentrations over time for each animal, performing an exponential regression on these values (y = a × e−bx, where a = intercept and b = slope), and using the equation half-life = ln 2/slope (
). The distribution volume was determined as the dose of caffeine infused divided by the intercept.
Differences in caffeine elimination were determined using the REML procedure of JMP (version 8.0; SAS Institute Inc., Cary, NC) with the fixed effect of treatment and the random effect of cow. Other data were analyzed using mixed models with repeated measures over time (SAS 9.1; SAS Institute Inc.). Spatial power covariance structures were used to model repeated measures over time within cow for data with time points that were not equally spaced (BW, BCS, and plasma and liver variables). For data collected daily (DMI, pH, feeding behavior, and production variables), autoregressive [AR(1)] covariance structures were used. Fixed effects were treatment, day relative to parturition, and treatment × day. Individual cows were treated as a random effect. Values were deemed outliers and omitted from analysis when studentized residuals were >|3.0|. After initial outlier removal, the model was repeated and studentized residuals >|3.5| were deemed outliers. No more than 3% of data were removed from any single analysis. Plasma haptoglobin, insulin, liver TG, and CPT1a and Apo B mRNA data were log-transformed before analysis to achieve normal residual distributions, and the reported means and standard errors were back-transformed (
). Prepartum and postpartum measures were analyzed separately for DMI, water intake, and feeding behavior. Effects were declared significant at P < 0.05, and trends are discussed at P < 0.10.
Results and discussion
No significant difference occurred between treatments in deviation from expected calving date (−2.0 vs. −0.8 ± 1.0 d for the control and monensin, respectively; P = 0.40). Actual calving dates ranged from 10 d before expected calving to 7 d after expected calving.
Feed Intake, Rumen pH, BW, Body Condition, and Milk Production
Daily DMI is shown in Figure 1A. Intake of DM decreased before parturition and increased after parturition in both groups (P < 0.001), which resembles intake patterns observed in many transition cow studies (
). Dry matter intake was not affected by treatment pre-or postpartum. This was surprising because other transition cow research has shown monensin to have an effect.
reported 1.2 kg/d lower DMI of cows fed 30 g of monensin/ton of DM compared with cows receiving no monensin from 1 wk prepartum through 3 wk postpartum.
), although treatment means were not reported in these abstracts.
Figure 1Dry matter intake (A), intermeal interval (B), and meal size (C) during the experimental period. A. An effect of day pre-and postpartum was observed (P < 0.001); prepartum standard error of the means = 0.88, postpartum standard error of the means = 0.87. B. Monensin shortened intermeal interval prepartum (P < 0.03) and tended to shorten intermeal interval postpartum (P < 0.08). An effect of day postpartum was observed (P < 0.001); prepartum standard error of the means = 10.6, postpartum standard error of the means = 4.77 († denotes P < 0.10; * denotes P < 0.05). C. An effect of day pre-and postpartum on meal size was observed (P < 0.001), and a treatment × day interaction was detected postpartum (P < 0.02); prepartum standard error of the means = 0.18, postpartum standard error of the means = 0.09.
Although intakes were similar, monensin supplementation tended to decrease time between meals (Table 3 and Figure 1B; P < 0.08), which is consistent with reports showing that inclusion of monensin in feedlot diets results in more consistent feed intake patterns throughout the day (
). Even though the intermeal interval was smaller with monensin supplementation, the number of meals consumed per day and average meal duration did not differ between treatments. Although meal sizes were similar overall (Table 3), a treatment × day interaction was observed for postpartum meal size (effects on d 8, 15, 20, and 21; Figure 1C). The small increase in meal frequency, coupled with similar to larger meal sizes, resulted in small, nonsignificant increases in DMI for monensin-supplemented cows during the postpartum period.
observed similar meal pattern results during a SARA challenge with midlactation cows, and concluded that monensin affects feeding behavior during times when rumen pH is low. In their study, mean ruminal pH was greater than the 5.9 observed in our study (
). This implies that the cows in our study were experiencing SARA to a greater degree, and that decreased intermeal interval in fresh cows on monensin treatment may have been related to the low ruminal pH observed. However, monensin's effects in the current study were not dependent on the presence of SARA. Cows received a low-energy diet prepartum that should not have caused SARA, yet we still observed a shorter intermeal interval prepartum, suggesting that another mechanism must be involved in this response to monensin.
Table 3Feed and water intake and feeding behavior during the experimental period
As expected, DMI was noticeably different pre-and postpartum; however, the dramatic decrease in meal length (Table 3) for postpartum cows compared with prepartum cows likely reflects differences in feeding behavior of cows in tie-stall versus pen housing rather than a true stage of production effect.
indirectly examined this hypothesis and found that cows housed in a freestall barn consumed dramatically fewer and larger meals than cows housed in tie-stall facilities (
). These 3 studies all investigated early lactation cows, and it was suggested that social interactions result in less frequent access to feed in a freestall situation, and that when cows gain access to the feed bunk, they remain there even when not eating (
Daily water consumption did not differ between treatments throughout the experimental period (Table 3). Water intake of both groups was steady throughout the 21-d prepartum period. Postpartum water intake started around 80 L/d and, as expected, increased (P < 0.001) to approximately 115 L/d by d 21.
Mean ruminal pH and the total time per day that ruminal pH was below 5.8 were not affected by treatment or day (P > 0.28; Table 4) during the first 6 d postpartum. Some researchers suggest that a key role of monensin is to alter ruminal pH, but that lactate needs to exceed approximately 5 mM for monensin to have such an effect (
Effects of monensin on ruminal forage degradability and total tract diet digestibility in lactating dairy cows during grain-induced subacute ruminal acidosis.
). Thus, fresh cow diets likely do not contain enough rapidly fermentable carbohydrates to promote excessive levels of lactic acid. If this is true, it would explain why monensin typically increases pH in feedlot studies (
) but few effects have been observed in dairy cattle. The lack of an effect on mean ruminal pH and time below 5.8 coincides with results from both transition (
Effects of monensin on ruminal forage degradability and total tract diet digestibility in lactating dairy cows during grain-induced subacute ruminal acidosis.
); however, these studies did not examine differences in variance of ruminal pH. In our study, cows supplemented with monensin had a smaller standard deviation of ruminal pH (Figure 2; P < 0.02) on d 1 postpartum, but no differences were detected beyond d 1. The more stable ruminal pH in monensin-supplemented cows could have facilitated quicker adaptation of ruminal microflora to the lactation diet, and may be related to the shorter intermeal interval in this group. Notably, an effect of day was detected for area under pH 5.8 (P < 0.03), with a relatively steady increase from 113 min × pH units on d 1 to 169 ± 18 min × pH units on d 6 postpartum (data not shown).
Table 4Rumen pH parameters from d 1 through 7 and results of d 7 caffeine clearance test
Figure 2Standard deviation of rumen pH until d 7 postpartum. Monensin decreased the standard deviation of ruminal pH for the first day after calving (P < 0.02). Standard error of the means = 0.02 (* denotes P < 0.05).
Cows receiving monensin tended to have a lower BW on d 1 postpartum (P < 0.09), but no other differences were observed between treatments for BCS or BW (data not shown). On average, cows lost 0.6 BCS units (3.3 to 2.7) and 110 kg of BW (767 to 657 kg) during the experiment. Milk production (39.0 vs. 39.3 ± 1.7 kg/d for control and monensin, respectively; P = 0.92) and concentrations of fat, protein, lactose, and SNF did not differ (P > 0.18) between dietary treatments, but MUN was higher for monensin-supplemented cows (11.8 vs. 10.4 ± 0.42 mg/dL; P < 0.02). No clear explanation exists for the observed effect on MUN. It has been shown that impaired liver function associated with lipid accumulation results in decreased ureagenic capability (
), but neither liver TG content nor the caffeine clearance test demonstrated dramatic effects of monensin on liver health. The MUN response could also be a result of monensin's ruminal protein sparing effect, allowing more escape protein to reach the small intestine.
observed a rise in blood urea nitrogen in early postpartum cows given monensin controlled-release capsules and suggested that it could be related to gluconeogenesis from nonessential AA, because the prerequisite deamination results in increased urea production. Blood urea nitrogen distributes freely into body fluids, including milk, so this proposed mechanism could help explain why MUN was higher in cows fed monensin. This is something of a paradox, however, because if monensin increases propionate supply, it may decrease the need to use AA for glucose production, potentially limiting ureagenesis. A simpler explanation is to consider that monensin-supplemented cows consumed, on average, an additional 0.23 kg of CP/d, with no increase in milk protein yield. If metabolizable protein supply did not limit milk protein yield in either group, then the increase in MUN for the monensin treatment was an expected response to increased CP intake.
Metabolic and Endocrine Changes
Plasma NEFA, BHBA, glucose, and insulin concentrations are displayed in Figure 3. As expected, plasma NEFA concentrations increased dramatically from 222 ± 80 μM 21 d before expected calving to a peak of 878 ± 80 μM on d 1 postpartum (Figure 3A; P < 0.001). Monensin supplementation did not significantly alter NEFA concentrations throughout the study. This was somewhat unexpected, because a meta-analysis including 24 studies with plasma NEFA data demonstrated that monensin could decrease NEFA concentration (
). However, the small mean response to monensin (36.6 μEq/L) reported in the meta-analysis was similar to the numerical difference between treatments in this study, suggesting that the current study was simply underpowered to detect such an effect.
Figure 3Plasma concentrations of NEFA (A), BHBA (B), glucose (C), and insulin (D) during the experimental period. A. No treatment effects were detected, but a day effect was observed (P < 0.001); standard error of the means = 84.9. B. Effects of treatment (P < 0.05), day (P < 0.001), and treatment × day interaction (P < 0.01) were detected; standard error of the means = 73.6 (* denotes P < 0.05). C. No treatment effects were detected, but a day effect was observed (P < 0.001); standard error of the means = 2.38. D. Significant effects of day (P < 0.001) and day × treatment interaction (P < 0.05) were detected. Cows receiving monensin tended to have higher plasma insulin concentrations on d 7 postpartum. Standard errors of the means are shown on D († denotes P < 0.10).
Monensin treatment decreased plasma BHBA over the course of the entire study (734 vs. 616 ± 40.9 μM; P < 0.05) with a significant effect of day (P < 0.001) and a treatment × day interaction (P < 0.01; Figure 3B). Most notably, monensin significantly decreased plasma BHBA on d 4 postpartum (777 vs. 1,077 ± 71 μM; P < 0.01). The effect on BHBA is not surprising given that almost all relevant publications have reported similar decreases in BHBA (
). Lower BHBA concentrations are likely a result of more complete FA oxidation in the liver. Monensin can increase the supply of propionate to the liver (
Plasma glucose concentrations decreased after parturition in both groups (P < 0.001), but monensin did not affect plasma glucose concentrations pre-or postpartum. Meta-analysis indicated that monensin can increase plasma glucose concentration of transition cows, but increases were not consistently reported (
did not observe changes in blood glucose concentration, but observed an increase in distribution space and glucose pool size when feeding monensin to prepartum cows, suggesting increased uptake of glucose by peripheral tissues in response to monensin. If monensin does, in fact, alter clearance of plasma glucose, then plasma glucose concentration is a poor proxy for gluconeogenic flux and glucose turnover data are required to evaluate the effects of monensin on this pathway accurately. Data in this area are limited, but in one study, rate of appearance of glucose was 10% higher for monensin-supplemented cows on an equivalent DMI basis (
Overall, monensin treatment did not affect plasma insulin concentration. However, effects of day (P < 0.001) and treatment × day interaction (P < 0.05) were significant. The treatment × day interaction showed a tendency for higher plasma insulin concentration in monensin-fed cows on d 7 postpartum (P < 0.08). In the meta-analysis by
, data from 5 relevant reports in which plasma insulin was measured did not demonstrate an effect of monensin on insulin concentration in transition cows. Increased insulin concentration appears unlikely to be a primary mechanism by which monensin alters periparturient metabolism.
A day effect (P < 0.001) was detected for plasma haptoglobin concentration (Figure 4) as a result of an increase in the early postpartum period. Haptoglobin is an acute phase protein that increases in concentration during times of inflammation (
); therefore, we expected the effect of day during the transition period. Monensin did not significantly alter haptoglobin concentrations. The numerical difference between treatments on d 1 postpartum, however, seemed to correspond with the findings of
, who reported elevated haptoglobin in diseased transition heifers given monensin. We attempted to assess whether a similar differential response could be found in our data, but were unable to conduct a valid analysis because only 1 cow on the monensin treatment had an observed health disorder before d 4 postpartum.
Figure 4Plasma concentrations of haptoglobin during the experimental period. No treatment differences were detected; however, an effect of day was detected (P < 0.001). Standard errors of the means are shown on the graph.
Many metabolic fuels are oxidized or synthesized by the liver, making the health and function of this organ extremely important to early lactation dairy cows. In the present study, TG content, mRNA abundance of key genes, and caffeine clearance were used as measures of the liver's metabolic function.
Liver TG content throughout the experiment is shown in Figure 5A. All cows experienced an increase in liver TG following parturition (P < 0.001). A trend for a treatment × day interaction was detected (P < 0.09), driven primarily by a tendency for increased liver TG content in monensin-treated cows on d −7 (P < 0.09) and the numerical difference in the opposite direction observed on d 7 postpartum. Little data exists reporting transition cow liver TG content in response to monensin, but one group reported a tendency for animals administered monensin to have lower liver TG content (P = 0.12) and higher glycogen content (P = 0.02) 3 wk into lactation (
Figure 5Triglyceride (TG) content (A) and mRNA abundance of carnitine palmitoyl transferase 1a (CPT1a; B) in liver tissue during the experimental period. A. A tendency for a treatment × day effect was detected (P < 0.09). Liver TG content increased in both groups during the study period (P < 0.001); standard errors of the means are shown on the graph († denotes P < 0.10). B. Monensin significantly increased abundance of CPT1a relative to the control gene ribosomal protein subunit 9 (RPS9; P < 0.04); standard errors of the means are shown on the graph (* denotes P < 0.05).
A major source of metabolic fuel and substrate for liver TG synthesis in transition cows are NEFA. Carnitine palmitoyl transferase 1a is important for translocating FA from the cytosol into the mitochondria, making it a central component for determining oxidative flux of FA within the liver (
). Results from this study indicated that liver CPT1a mRNA was greater in cows fed monensin (0.10 vs. 0.15 ± 0.002 arbitrary units, P < 0.04; Figure 5B). The difference was driven mainly by effects on d −7 and 1 (both P < 0.05). These results suggest a novel mechanism underlying the role of monensin in improving the overall health of transition cows. This response led us to evaluate whether treatment altered the rate of liver TG accumulation during this time period. A contrast statement was used to determine whether the increase in liver TG from d −7 to d 7 differed by treatment. The increase in liver TG concentration was significantly greater for the control compared with monensin (412 vs. 128 ± 83 mg of TG/g of protein over this period, P = 0.03). This differential rate of TG accumulation coincides with the treatment effects on liver CPT1a mRNA abundance. Previous findings also have suggested that monensin increases the liver's capacity to export very low density lipoprotein (
). This stimulated us to measure expression of Apo B in liver tissue from d −7 to 7, but treatment had no effect on Apo B mRNA abundance during this period (P = 0.42). Expression of Apo B was upregulated over this time period (0.18, 0.35, and 1.13 ± 0.28 arbitrary units for d −7, 1, and 7, respectively). Expression of Apo B appeared to adapt to the increased TG synthesis in the liver of these transition cows, but monensin did not regulate lipoprotein secretion through Apo B transcription in this experiment.
The evidence that monensin lowers transition cow plasma ketone concentrations is substantial (
explain the effect on ketones through increased supply of propionate for hepatic gluconeogenesis. To our knowledge, our report is the first to consider the effects of monensin on abundance of CPT1a in dairy cows, and suggests that monensin has a positive effect on lipid metabolism in the liver. Although β-oxidation of FA can result in either complete oxidation of acetyl-CoA or ketone production, increased CPT1a abundance could benefit liver function because increased mitochondrial oxidation could 1) limit TG synthesis, 2) decrease reliance on peroxisomal oxidation with its subsequent production of reactive oxygen species, and 3) prevent accumulation of lipid metabolites that may impair metabolic function (FA-CoA, ceramides, and peroxides, among others). Nevertheless, these findings should not be over-interpreted because our results are limited to the transcript level, and we did not observe obvious corresponding decreases in liver TG content or hepatic inflammation (as measured by plasma haptoglobin concentration).
The liver is the primary site for gluconeogenesis, and cPEPCK is thought to be a rate-determining enzyme for hepatic gluconeogenesis from propionate, lactate, and AA (
). The relative abundance of liver cPEPCK mRNA differed across day of study (P < 0.02), but was not different between treatments. The slight change through the transition period (an 85% increase from d 1 to 21 postpartum) is consistent with previous reports (
), so it was surprising that no differences were detected in this study. Because cPEPCK is one of several rate-determining gluconeogenic enzymes, this is consistent with the lack of treatment effect on plasma glucose concentration, although mRNA abundance is not a measure of enzyme activity and plasma glucose concentration is not a measure of gluconeogenic flux.
Liver diseases, including fatty liver, decrease activity of the cytochrome P450 (CYP450) enzyme complex (
), and activity of CYP450 can serve as an index of normal liver function. We assessed metabolism of caffeine because it is metabolized by CYP450 and has few side effects when administered at low doses (
). Results of the caffeine challenge tests are shown in Table 4. Caffeine elimination half-life was 226 and 208 ± 13.5 min for control and monensin cows, respectively, and no differences were detected (P = 0.34). The volume of distribution of caffeine also was calculated, and similar results across treatments suggest no bias in administering the caffeine dose; thus, our results do not indicate that CYP450 activity was enhanced or impaired by monensin treatment. Reports of caffeine clearance in dairy cows are limited.
examined hepatic function of later-lactation dairy cows and reported an average elimination half-life of 228 min (range was 156 to 414 min), which was remarkably similar to our results. In the current study, caffeine elimination half-life also was correlated with log-transformed plasma haptoglobin concentration (P = 0.03, r = 0.41), similar to results from a separate study conducted in our laboratory (
). These positive correlations between markers of impaired liver function and liver inflammation are consistent with their utility as gauges of metabolic health in transition cows.
Management and Health
The incidences of health disorders are shown in Table 5. No differences were found between treatment groups. Because only 32 experimental units were used for this study, detecting differences in disease incidence would have been difficult. A more powerful assessment of the effects of monensin on cow health is the meta-analysis conducted by
. These researchers combined data from 16 papers; overall, monensin significantly decreased the relative risk of ketosis, displaced abomasum, and mastitis. The BHBA response observed in our study is consistent with the means reported in the meta-analysis (
). Therefore, monensin likely lowers the risk of diseases such as ketosis and displaced abomasa because these diseases are related to energy status, although our study lacked statistical power to detect these differences.
Table 5Incidence of health disorders during the experimental period
In this first report of monensin's effects on feeding behavior combined with ruminal pH dynamics in transition cows, monensin increased meal frequency and minimized ruminal pH variance in the first day after cows changed to a lactation ration. Monensin supplementation also significantly increased liver mRNA abundance of CPT1a, a key mediator of liver mitochondrial FA oxidation, and decreased the rate of liver TG accumulation in the 2 wk around parturition. Consistent with previous results, monensin significantly decreased peak plasma BHBA concentrations in postpartum cows, but did not alter concentrations of plasma NEFA, glucose, or insulin in the postpartum period. Despite the observed beneficial effects on metabolism, no significant effects on milk production or disease incidence were detected.
Acknowledgments
This paper is contribution no. 11-386-J from the Kansas Agricultural Experiment Station (Manhattan). The authors thank Elanco Animal Health (Greenfield, IN) for partially funding this research and especially for providing financial assistance to build feeding behavior systems at the Kansas State University Dairy Unit (Manhattan). We also thank contributors at Kansas State University: Michael Scheffel, Cheryl Armendariz, and all of the undergraduate and graduate students that helped with feeding and data collection, particularly Jaymelynn Farney, Claire Legallet, and Michelle Sullivan. We are grateful to Marina von Keyserlingk and her laboratory group at the University of British Columbia (Vancouver, BC, Canada) for technical assistance with feeding behavior data processing.
References
Aiello R.J.
Armentano L.E.
Effects of volatile fatty acids on propionate metabolism and gluconeogenesis in caprine hepatocytes.
Effects of monensin on ruminal forage degradability and total tract diet digestibility in lactating dairy cows during grain-induced subacute ruminal acidosis.
00089-6&id=gr1.jpg){kind=link}
00089-6&id=gr2.jpg){kind=link}
00089-6&id=gr3.jpg){kind=link}
00089-6&id=gr4.jpg){kind=link}
00089-6&id=gr5.jpg){kind=link}