Effect of the timing of sodium acetate infusion on the daily rhythms of milk synthesis and plasma metabolites and hormones in Holstein cows

Dairy cows have a daily pattern of feed intake which influences ruminal fermentation and nutrient absorption. Milk synthesis also exhibits a daily rhythm and is altered by the timing of feed availability. Nutrients can regulate physiological rhythms, but it is unclear which specific nutrients affect the rhythms of milk synthesis in the cow. The objective of this study was to determine the effect of the timing of acetate infusion on the daily rhythms of feed intake, milk synthesis, milk fatty acids, plasma insulin and metabolites, and core body temperature. Ten lactating ruminally cannulated Holstein cows (127 ± 24.6 d in milk; mean ± standard deviation) were arranged in a 3 × 3 Latin square de-sign. Treatments were ruminal infusions of 600 g/d of acetate either continuously throughout the day (CON) or over 8 h/d during the day (day treatment, DT; 0900 to 1700 h) or the night (night treatment, NT; 2100 to 0500 h). Experimental periods were 14 d with a 7-d washout between periods. Cows were milked every 6 h during the final 7 d of each experimental period to determine the daily pattern of milk synthesis. Blood samples were taken to represent every 4 h across the day and plasma glucose, insulin, β-hydroxybutyrate, urea nitrogen, and acetate concentration were measured. An intravaginal temperature logger was used to measure core body temperature. Data were analyzed with cosinor-based rhythmometry to test the fit of a cosine function with a period of 24 h and to determine the acrophase (time at peak) and amplitude (peak to mean) of each rhythm. Milk yield fit a daily rhythm for all treatments and DT and NT phase-delayed the rhythm and DT increased the robustness of the rhythm. Milk protein concentration fit a daily rhythm for all treatments and DT increased robustness, whereas NT phase-delayed the rhythm. Plasma acetate concentration also fit a daily rhythm in all treatments. Plasma acetate peaked at ~1600 h in CON and DT and at 0053 h in NT, reflecting the timing of treatment infusions. There was a daily rhythm in plasma β-hydroxybutyrate that reflected the plasma acetate rhythm. Core body temperature fit a rhythm for all treatments, but the amplitude of the rhythm was smaller than previously observed. In conclusion, the timing of acetate infusion influences peripheral rhythms of milk synthesis and plasma metabolites.


INTRODUCTION
Dairy cows have a daily pattern of feed intake, with higher rates of intake occurring after feed delivery and during the afternoon and early evening.The differences in feed intake causes variation in the amounts of fermentable nutrients entering the rumen across the day and drives changes in the daily rhythms of ruminal fermentation and nutrient absorption (Salfer et al., 2018).Milk production also exhibits a daily pattern, with milk yield being highest in the morning and milk fat and protein concentrations peaking in the evening (Rottman et al., 2014).The timing of feed intake also modifies the daily pattern of milk synthesis (Salfer and Harvatine, 2020).Nutrients can act as circadian regulators (Oosterman et al., 2015;Pickel and Sung, 2020), but it is not clear if there are specific nutrients driving the daily patterns of milk production.
Circadian rhythms are biological rhythms that repeat approximately every 24 h and allow organisms to physiologically anticipate recurring changes in their environment.The master circadian clock is located in the suprachiasmatic nucleus of the hypothalamus and is most potently regulated by light-dark cycles.Biological clocks in peripheral tissues are regulated by the master circadian clock, but can also be entrained by other signals, including nutrient intake (Patke et al., 2020).Tahara et al. (2018) provided evidence of in vivo entrainment of circadian rhythms in the liver, kidney, and submandibular gland by acetate, butyrate, propionate, and lactate or by a high fiber diet in mice.

Effect of the timing of sodium acetate infusion on the daily rhythms of milk synthesis and plasma metabolites and hormones in Holstein cows
C. Matamoros, 1 I. J. Salfer, 1,2 P. A. Bartell, 1 and K. J. Harvatine 1 * Acetate is the major short chain fatty acid (FA) produced in the rumen and is the main substrate for milk fat synthesis (Palmquist et al., 1969;Bauman et al., 1970).Recent reports have demonstrated that increasing acetate increases milk fat yield mostly through an increase in mammary de novo lipogenesis (Urrutia et al., 2019;Matamoros et al., 2021).
It is not known if the timing of acetate availability influences the pattern of milk and milk fat synthesis.The objective of the current study was to determine the effect of the time of acetate infusion on the daily rhythm of milk production, plasma metabolites and hormones, and core body temperature.Our hypothesis was that the timing of acetate infusion modifies the daily rhythms of milk synthesis because acetate is a major energy source for the mammary gland and substrate for milk fat synthesis and its supplementation increases milk fat synthesis in the cow.Short chain FA, including acetate, have been reported to entrain rhythms in the mouse (Tahara et al., 2018).This investigation is one of a set of 3 experiments investigating the timing of nutrient infusion with the others investigating the timing of infusion of long-chain FA and protein (Salfer, 2019).

Animals and Experimental Design
All experimental procedures were approved by the Pennsylvania State University Institutional Animal Care and Use Committee (PRAMS #201446144).Ten lactating, ruminally cannulated, multiparous Holstein cows [127 ± 24.6 DIM (mean ± SD); minimum = 96 and maximum = 169 at the start of the experiment] were randomly assigned to one of 6 treatment sequences in a 3 × 3 Latin square design.Experimental periods were 14 d with 7-d washout periods.Cows were housed in a tiestall barn at the Pennsylvania State University Dairy Production Research and Teaching Center.The experiment was conducted from September 24 to November 27, 2018, and the animals were kept in a light regiment of 18 h light and 6 h dark (lights off from 2300 to 0500 h each day).The stalls were equipped with a feed intake monitoring system comprised of a feed tub hanging from a load monitor.Feeding behavior, including the number and characteristics of meals and the rate of feed intake, was analyzed as described by Rottman et al. (2015) and Niu et al. (2017).

Treatments and Diet
All treatments provided 600 g/d of acetate as anhydrous sodium acetate (Niacet Corp.) dissolved in 5 L of distilled water, as this dose has been shown to elicit the maximum response in milk fat synthesis (Urrutia and Harvatine, 2017b).Treatments were ruminally infused either as a continuous infusion throughout the day (control, CON) or for 8 h/d during the day (0900 to 1700 h; day treatment, DT) or the night (2100 to 0500 h; night treatment, NT).Acetate was infused through acid-resistant tubing (Masterflex L/S Precision Norprene 06404-14, Cole-Parmer) with adjustable peristaltic pumps (Masterflex L/S drive 7520-35 or 7521-50, Cole-Parmer) with standard pump heads (Masterflex L/S Standard Pump Head 7014-21, Cole-Parmer).All cows were not infused while they were at the parlor during milkings (times described below).All cows were fed once a day the same TMR at 0600 h at 110% of intake (Table 1).Feed samples were taken at the beginning of sampling for each of the experimental periods and were analyzed for DM, CP, starch, and NDF by Cumberland Valley Analytical Services Inc. (Waynesboro, PA).Feed ash content was determined by complete combustion of the OM of dried feed samples in a furnace at 500°C for 6 h.Fatty acid content of feeds was determined after extraction and methylation in a one-step procedure using methanolic HCL and were quantitated using GC-FID as described by Rico et al. (2014).

Milk Sample and Analysis
Cows were milked 4 times per day at 0500, 1100, 1700, and 2300 h during the last 7 d of each treatment periods to evaluate the daily rhythm of milk and milk component synthesis.Milk yield for each time point was determined by an integrated milk meter (AfiMilk: SAE Afikim) and corrected using a stall deviation, which was calculated using data from the entire milking herd of the Dairy Research and Teaching Center (>200 cows) as described by Andreen et al. (2020).Milk was sampled during the last 2 d of the treatment period at all milkings.One sample was stored at 4°C with preservative (Bronolab-WII, Advanced Instruments Inc.) until analyzed for fat, protein, and MUN by Fourier transform infrared spectroscopy (Fossomatic Milko-Scan FT+ and FC; Foss Electric; Dairy One).A second milk sample was composited by cow within time point for each period according to milk yield and a fat cake was extracted by centrifugation at 1,300 × g for 20 min at 4°C for milk FA analysis.Milk FA profile was determined after FA were extracted with hexane: isopropanol (Hara and Radin, 1978), transmethylated with sodium methoxide (Chouinard et al., 1999), and quantified using GC-FID as described by Urrutia and Harvatine (2017a).

Plasma Samples and Analysis
Blood samples were collected from a tail vessel using vacuum tubes containing potassium EDTA (Greiner Bio-One North America Inc.) 6 times over the last 3 d of each experimental period to represent every 4 h of a day (0030, 0430, 0830, 1230, 1630, 2030 h).Blood samples were placed on ice and immediately transported to the on-farm laboratory.Plasma was collected after centrifugation for 1,300 × g at 4°C for 15 min.Multiple aliquots were stored at −20°C until further analysis.Plasma acetate was determined after derivatization into a propyl ester and quantification with a GC-MS as described in Cai et al. (2017).Plasma glucose (PGO enzyme P 7119, Sigma-Aldrich; Raabo and Terkildsen, 1960), BHB (β-Hydroxybutyrate LiquiColor, Stanbio Chemistry), nonesterified FA (Wako HR Series NEFA-HR kit, Wako Chemicals USA; Ballou et al., 2009), and insulin (PI-12K Radioimmunoassay, Millipore Corp.; 90% cross-reactive with bovine insulin) were determined with commercially available assay kits.

Core Body Temperature
Core body temperature was determined with the use of a temperature logger (DS1922L or DS1921G-F5, iButtonLink) attached to a hormone-free intravaginal implant placed centrally in the vagina.Temperature was recorded at 10 min intervals at a resolution of 0.5°C over the last 6 d of each treatment period.Raw data were binned by averaging measurements over 2 h intervals before analysis.

Statistical Analysis
Milk yield at each milking over the last 2 d of the experimental period were summed and averaged across each day to test the overall effect of treatment on milk production and analyzed with a model that included the fixed effect of treatment and the random effect of period and cow using the Fit Model module of JMP Pro 13.2.1 (SAS Institute Inc.).Data for temporal analysis of daily rhythms of milk synthesis was averaged at each time point across the days observed.Time course data for milk production, plasma metabolites and hormones, and body temperature were fit to the linear form of a cosine function with a 24-h period using random regression in SAS 9.4 (SAS Institute Inc.).The model included the fixed effect of treatment, cosine terms, and their interaction and the random effects of cow, period, and day (period).A cosine function with a 12-h period was also tested for a significant fit to the daily patterns of plasma metabolites and core body temperature to test for harmonic rhythms, and this function was kept if it improved model fit, according to the estimated corrected Akaike and Bayesian information criterions (Schwarz, 1978;Sugiura, 1978).The fit of the 24-h cosine function was determined using a zero-amplitude test and the amplitude, acrophase, and significance were determined as previously described (Niu et al., 2014).Least squares means at each time point were also analyzed with a model that included the fixed effects of treatment, time of day, and their interaction and the random effects of cow, period, and day (period).In all analyses, data points with studentized residuals outside of ± 3.5 were declared outliers and removed.Differences were declared at P ≤ 0.05, and tendencies were acknowledged at 0.05 < P ≤ 0.10.High resolution images were created with an Excel add-in (Microsoft Corp.;Kraus, 2014).

DMI and Feeding Behavior
There was no effect of treatment on DMI (P = 0.99) or meal frequency, length, and size (P = 0.63, 0.58, and 0.97, respectively; Table 2).There was also no effect of treatment or treatment by time interaction on the rate of feed intake, calculated as kilograms per hour or percentage of daily intake per hour (P = 0.88 and 0.96 for the main effect of treatment and P = 0.64 and 0.64 for the interaction, respectively; Figure 1).

Milk and Milk Component Synthesis
There was no effect of treatment on total milk, milk fat, or milk protein synthesis (Table 2).A daily rhythm of milk yield was detected for all treatments (Figure 2A; P = 0.02, 0.01, and 0.01 for CON, DT, and NT, respectively).When compared with CON, DT increased the amplitude of the daily rhythm of milk synthesis by 69% and phase-delayed the rhythm by 2.1 h, whereas NT did not change the amplitude but phased-delayed the rhythm by 1.0 h.Milk fat yield did not fit a daily rhythm in any treatment (Figure 2B; P = 0.79, 0.99, and 0.24 for CON, DT, and NT, respectively).Milk fat concentration tended to fit a rhythm in CON and NT (P = 0.07 and 0.09, respectively) and DT induced a rhythm in fat concentration (Figure 2C; P < 0.01).
The daily peak in milk fat concentration was 159% and 91% greater in DT compared with CON and NT, respectively, and was phase-delayed by 3.0 and 4.9 h.
Milk protein yield did not fit a daily rhythm for CON, but it tended to fit a rhythm for DT and fit a rhythm for NT (Figure 2D; P = 0.40, 0.06, and <   within each panel shows the P-value for the main effect of treatment and time and their 2-way interaction from a mixed model that also includes the random effects of cow and period.The waveform properties, as determined by cosinor analysis, are shown below each panel and include the amplitude (Amp; difference between peak and mean), acrophase (Acro; time at peak of rhythm, h), and the P-value of the zero-amplitude test.Superscripts within each column denote differences among estimates (P ≤ 0.05).0.01, respectively).Milk protein concentration fit a rhythm for all treatments (Figure 2E; P < 0.01 for all).When compared with CON, DT increased the amplitude of the rhythm of milk protein concentration by 175% but did not change the phase of the rhythm and NT did not change the amplitude of the rhythm but phase-delayed the rhythm by 7.6 h.Similarly, MUN concentration fit a daily rhythm in all treat- A table within each panel shows the P-value for the main effect of treatment and time and their 2-way interaction from a mixed model that also includes the random effects of cow and period.The waveform properties, as determined by cosinor analysis, are shown below each panel and include the amplitude (Amp; difference between peak and mean), acrophase (Acro; time at peak of rhythm, h), and the P-value of the zero-amplitude test.Superscripts within each column denote differences among estimates (P ≤ 0.05).ments (Figure 2F; P < 0.01 for all).When compared with CON, DT increased the amplitude of the rhythm by 64% and NT decreased the amplitude by 34% and phase-delayed the rhythm by 3.5 h.

Milk FA Profile
The yield of de novo FA did not fit a daily rhythm for CON and DT (P = 0.70 and 0.82, respectively), but NT induced a daily rhythm (Figure 3A; P < 0.01).Similarly, the yield of mixed sourced FA did not fit a rhythm for CON and DT, but NT induced a daily rhythm (Figure 3B; P < 0.01).The yield of preformed FA and odd and branched-chain FA did not fit a daily rhythm for any treatments, except for NT which induced a daily rhythm in yield of odd and branchedchain FA (Figure 3C and D; P = 0.02).

Plasma Metabolites
Plasma acetate concentration fit a daily rhythm for all treatments (Figure 4A; P = 0.04, < 0.01, and 0.04 for CON, DT, and NT, respectively).When compared with CON, DT increased the amplitude of the daily rhythm of plasma acetate by 130% and phase-delayed the rhythm by 52 min, whereas NT decreased the amplitude by 28% and advanced the phase of the rhythm by 15.2 h.Plasma BHB fit a daily rhythm in CON and DT (Figure 4B; P < 0.01, for both), but not in NT (P = 0.2).When compared with CON, DT increased the amplitude of the daily rhythm of plasma BHB by 85%, but did not change the phase of the rhythm, and NT decreased the amplitude by 44% and phase-advanced the rhythm by 1.9 h.
There was no daily rhythm for plasma glucose concentration in control and NT (P = 0.13 and 0.28 for CON and NT, respectively), but DT tended to fit a rhythm (Figure 4C; P = 0.06).Plasma insulin concentration fit a daily rhythm for CON and DT (P < 0.01 for both), but not for NT (Figure 4D; P = 0.13).When compared with CON, DT did not change the amplitude or phase of the insulin rhythm.There was no daily rhythm for plasma nonesterified FA concentration in DT and NT (P = 0.92 and 0.26 for DT and NT, respectively), but CON tended to fit a rhythm (Figure 4E; P = 0.07).Blood urea nitrogen concentration fit a daily rhythm in all treatments (Figure 4F; P < 0.01 for both CON and DT and P = 0.03 and NT).When compared with CON, DT increased the amplitude of the rhythm by 98% and phase-delayed the rhythm by 37 min, and NT decreased the amplitude of the rhythm by 35% and phase-delayed the rhythm by 2.5 h.The advance in the phase of the rhythm of BUN concentration of NT was higher than that of DT.

Body Temperature
Core body temperature fit a daily rhythm in all treatments (Figure 5; P < 0.01 for all).When compared with CON, DT decreased the amplitude by 22% but did not change the phase of the rhythm and NT did not change the amplitude but phase-advanced the rhythm by 2.8 h.

DISCUSSION
Dairy cows have a clear daily pattern of feed intake, with highest eating activity during the day, whether grazing or eating a total mixed ration (Albright, 1993;Haley et al., 2000).Feeding time, frequency, or restricting feed availability to part of the day affects the daily pattern of intake, but not overall DMI (DeVries and Von Keyserlingk, 2005;DeVries et al., 2005;Salfer and Harvatine, 2020).There were no effects of treatment on overall DMI or feeding behavior in our current study, similar to what was found in the companion studies assessing the timing of long-chain FA and protein infusion (Salfer, 2019).It does not appear that the timing of infusion of these nutrients at the level investigated is a major determinant of overall DMI or the daily pattern of intake.
Recent work with ruminal infusions and feeding sodium acetate have reported no effects on milk yield (Matamoros et al., 2021;Urrutia et al., 2019), but the timing of acetate infusion appears to have an effect on the daily rhythm of milk synthesis.In the current experiment, peak milk yield occurred during the early morning in all treatments, similar to what was found in our previous work (Rottman et al., 2014;Salfer, 2019).The companion studies investigating long-chain FA and protein infusions similarly observed that day infusions phase-advanced and increased the amplitude of the daily rhythm of milk synthesis, indicating that the effects are not nutrient specific (Salfer, 2019).Infusions at night had inconsistent effects on the daily rhythm of milk synthesis, with acetate and protein phase-advancing the rhythm, whereas long-chain FA had no effect.
Ruminal infusion and dietary supplementation of sodium acetate both consistently increase milk fat yield relative to a no supplement control (Urrutia and Harvatine, 2017b;Urrutia et al., 2019;Matamoros et al., 2021).A daily rhythm of milk fat yield has been reported, with higher fat yield observed in the afternoon (e.g., Rottman et al., 2014).In the current experiment, however, there was no detectable rhythm for milk fat yield after any treatments, an outcome similar to the companion studies where no detectable rhythm of milk fat synthesis was observed during infusion of protein or long-chain FA during the night or day (Salfer, 2019).
Milk protein percent and yield also exhibited a daily rhythm, with milk protein synthesis higher in the after- noon (Rottman et al., 2014).Restricting feed intake to a portion of the day changed the daily rhythm of milk protein synthesis, indicating regulation by the timing of nutrient absorption (Salfer and Harvatine, 2020).Changing the timing of protein and long-chain FA infu-sion also had inconsistent results on the rhythm of milk protein synthesis.Acetate and protein infusions during the day increased the amplitude of the daily rhythm of protein concentration (Salfer, 2019).In contrast, long-chain FA infusions during the day dampened the Effects of time of acetate infusion on the daily rhythms of plasma metabolites and hormones.Treatments were 600 g/d of acetate as anhydrous sodium acetate dissolved in 5 L of distilled water and were ruminally infused continuously throughout the day (CON), for 8 h/d from 0900 to 1700 h (DT), or for 8 h/d from 2100 to 0500 h (NT).Panels show the effects of timing of acetate infusion on the daily rhythm of (A) plasma acetate concentration (mmol/mL), (B) plasma BHB (µmol/L), (C) plasma glucose (mg/dL), (D) plasma insulin concentration (µIU/L), (E) plasma nonesterified fatty acids concentration (NEFA; µEq/L), and (F) BUN (mg/dL).Data are presented as LSM with SEM bars, and the best fit cosine function is shown.A table within each panel shows the P-values for the main effects of treatment and time and their 2-way interaction from a mixed model that also includes the random effects of cow and period.Waveform parameters, as determined by cosinor analysis, are shown below each panel and include the amplitude (Acro; time at peak of rhythm, h), acrophase (Acro; time at peak of rhythm, h), and the P-value of the zero-amplitude test.Superscripts within each column denote differences among estimates (P ≤ 0.05).rhythm of milk protein concentration (Salfer, 2019).Milk urea N is highly related to BUN and is a common metric for efficiency of nitrogen use (Nousiainen et al., 2004).The acrophase of the rhythm of MUN tracked the acrophase of milk protein concentration.In the current experiment, peak MUN concentration lagged that of milk protein concentration by ~6 h for CON and DT, but for NT it was 10 h later.The magnitude of change in the daily rhythms of MUN and milk protein concentration were similar in DT but not in NT, as the amplitude of the rhythm of milk protein concentration did not change, but that of MUN decreased when compared with CON.
Fasting cows for 7 h during the day or the night also modified the daily rhythms of de novo, mixed, and preformed FA (Salfer and Harvatine, 2020).In the current study, there was no detectable rhythm in milk FA yield for control and DT, but NT induced a rhythm of milk de novo, mixed, and odd and branched-chain FA, all of which at least partially originate from mammary de novo lipogenesis (Vlaeminck et al., 2015;Palmquist and Harvatine, 2020).In previous work, acetate supplementation increased mammary de novo lipogenesis (e.g., Matamoros et al., 2021) and also stimulated lipogenic enzymes in primary bovine mammary epithelial cells (Song et al., 2020;Zhao et al., 2020).Matamoros et al. (2021) also did not observe a difference in responses at the morning and evening milkings in cows supplemented with acetate and milked twice per day, but these effects may have been blurred because of the timing of milking relative to the rhythms.
Dietary acetate supplementation has been shown to increase plasma acetate in the period of the day with the highest rate of intake, likely because acetate is more available for absorption (Matamoros et al., 2021).The acrophase of the rhythms of plasma acetate for CON and DT were similar to that reported in ad libitum fed cows, with or without acetate supplementation (Allen et al., 2005;Matamoros et al., 2021).These findings are likely the result of endogenous acetate synthesis being highest during the peak of fermentation following periods of high intake, such as after fresh feed delivery and in the afternoon.Consequently, night infusions of acetate would shift the acrophase of the rhythm, as we observed.The rate of absorption of acetate from the rumen is expected to be affected by several factors, including rumen pH and rumen and plasma acetate concentrations (Masson and Phillipson, 1951).Night infusions of acetate likely decreases the amplitude of the daily rhythm of plasma acetate by providing additional acetate during the period of the day when less fermentation is normally occurring because of lower rates of intake at night.The acrophase of plasma acetate preceded the acrophase of milk de novo FA by ~2.9 h in NT.Purdie et al. (2008) suggested that there is a time-lag between plasma acetate and milk fat output, but they did not quantify the lag.The time-lag could be attributed to either mammary acetate uptake, de novo lipogenesis, or intracellular trafficking and excretion of milk fat globules into the lumen.
Acetate can be converted into butyrate in the rumen and, subsequently, into BHB in the rumen epithelium  within the panel shows the P-values for the main effects of treatment and time and their 2-way interaction from a mixed model that also includes the random effects of cow and period.Waveform parameters, as determined by cosinor analysis, are shown below the panel and include the amplitude (Amp; difference between peak and mean), acrophase (Acro; time at peak of rhythm, h), and the P-value of the zero-amplitude test.Superscripts within each column denote differences among estimates (P ≤ 0.05).(Sutton et al., 2003).Dietary supplementation with acetate increases plasma BHB concentrations with a peak concentration that is concurrent with peak plasma acetate (Urrutia et al., 2019;Matamoros et al., 2021).In DT and CON, the acrophase of the rhythm of acetate preceded the acrophase of the rhythm of BHB by ~2 h and the changes in the amplitude of plasma BHB was proportional to the changes in the amplitude of plasma acetate.Interestingly, NT eliminated the rhythm of plasma BHB, even though there was a rhythm of plasma acetate.It is not clear, however, if the rhythms of plasma acetate and BHB are driven by a rhythm of ruminal absorption and metabolism.
There is a daily rhythm of plasma glucose concentration in lactating dairy cows that can be modulated by fasting cows during either the day or night (Salfer and Harvatine, 2020).There was no detectable rhythm in plasma glucose concentration after any of the current treatments, except for a tendency in DT.The lack of a daily rhythm in glucose is similar to what was observed after infusion of long-chain FA during the day or night; or after protein infusion during the day (Salfer, 2019).Hepatic gluconeogenesis is the main source of plasma glucose in dairy cows; thus, it is important to consider the effect of timing of nutrient absorption on the liver circadian clock (Aschenbach et al., 2010).Because acetate supplementation increases de novo FA synthesis, an increase in peripheral glucose utilization is likely necessary to provide the additional NADPH needed to support the increase in lipogenesis (Purdie et al., 2008;Matamoros et al., 2021;Danes et al., 2020).More research is needed, however, to understand the effects of the time of absorption of acetate and long-chain FA on the circadian rhythms of hepatic gluconeogenesis and peripheral glucose utilization.
Plasma insulin concentration has a daily rhythm in cows and the properties of this rhythm can be modified by fasting cows for 7 h during the day or the night (Lefcourt et al., 1999;Salfer and Harvatine, 2020).The acrophase of the rhythm of plasma insulin in CON and DT was similar to previous reports for ad libitum fed animals (Lefcourt et al., 1999;Rottman et al., 2014;Salfer and Harvatine, 2020).The amplitude of the rhythm was also similar to that of cows that had been fasted for 7 h (Salfer and Harvatine, 2020).Considering that there was no detectable rhythm for plasma insulin in NT, acetate infusions may have dampened the rhythm of plasma insulin by increasing energy availability during the night, a time of day when nutrient absorption is expected to be the lowest.It is important to note that circadian and ultradian rhythms of insulin secretion and insulin sensitivity are well described in cows and other species (Lefcourt et al., 1999;Rakshit et al., 2015;Stenvers et al., 2019).Blood urea N is the major end product of N metabolism in the ruminant and its concentration is related to efficiency of N use (Nousiainen et al., 2004).The concentrations of urea N in milk and blood are positively correlated, but little is known about the relationship of their daily rhythms.Furthermore, BUN has a daily rhythm in dairy cows that is affected by restriction of feed or specific nutrients to a defined portion of the day (Lefcourt et al., 1999;Salfer, 2019;Salfer and Harvatine, 2020).When compared with experiments that restricted the infusion of long-chain FA or protein to the day, acetate infused during the day increased the amplitude of the rhythm of BUN similar to what has been reported for infusion with long-chain FA, whereas protein infusions completely dampened the rhythm when compared with CON.Infusions of acetate or longchain FA decreased or did not change, respectively, the amplitude of the rhythm of BUN compared with CON when infused during the night, whereas protein infusions increased the amplitude of the rhythm (Salfer et al., 2019).The acrophase of BUN for all treatments agrees with previous reports (Gustafsson and Palmquist, 1993;Lefcourt et al., 1999;Ikuta et al., 2005).It is important to note that the magnitude of changes in the acrophase and the amplitude of the rhythms of BUN and MUN remained constant for all treatment, suggesting that they are at least partially responsive to each other.
The daily rhythm of core body temperature is at least partially regulated by the light-responsive oscillator in the suprachiasmatic nucleus, and information on body temperature provides insight into the regulation and phasing of the master circadian clock in the hypothalamus.Fasting cows during either the night or day shifted the acrophase of core body temperature rhythms such that the peak of the rhythm occurs near the middle of the feeding period in lactating dairy cows (Salfer and Harvatine, 2020).Importantly, core body temperature is related to rumen temperature, and rumen temperature appears to exhibit a daily rhythm as well.This rhythm in rumen temperature is likely driven by the circadian pattern of feed intake (Piccione et al., 2014).Changes in the phasing of the core body temperature rhythm was greatest when feed availability was restricted to a portion of the day, when compared with the effects of infusing acetate or long-chained FA during a portion of the day (Salfer et al., 2019;Salfer and Harvatine, 2020).

CONCLUSIONS
Changing the timing of sodium acetate infusions modified the robustness and phase entrained daily rhythms in the lactating dairy cow.Despite robust rhythms in plasma acetate and BHB, 2 metabolic fuels of mammary de novo lipogenesis, only acetate infusion during the day induced a rhythm in milk fat percentage.However, infusion during the night elicited a robust rhythm in FA that originate at least partially from de novo lipogenesis, indicating that increasing mammary acetate supply may either stimulate mammary de novo lipogenic capacity or that acetate supply is limiting during during this period.Overall, this demonstrates that daily dynamics in acetate supply have the potential to entrain the daily rhythms of important physiological processes in the mammary gland and likely also within other metabolically important tissues.

Figure 1 .Figure 2 .
Figure 1.The effects of timing of ruminal infusion of sodium acetate on rate of feed intake across the day, as observed by an automated system over 2-h intervals.Treatments provided 10 mol/d of acetate as sodium acetate anhydrous dissolved in 5 L/d of distilled water either in a continuous ruminal infusion throughout the day (CON), for 8 h/d from 0900 to 1700 h (DT), or for 8 h/d from 2100 to 0500 h (NT).Data shown as eating rate in (A) kilograms per hour and (B) percent of daily intake per hour.A table within each panel shows the P-value for the main effect of treatment and time and their 2-way interaction from a mixed model that also includes the random effects of cow and period.Data are presented as LSM with SEM bars.

Figure 2 (
Figure 2 (Continued).Effects of time of acetate infusion on the daily rhythms of milk and milk component synthesis.Treatments were 600 g/d of acetate as anhydrous sodium acetate dissolved in 5 L of distilled water and were ruminally infused continuously throughout the day (CON), for 8 h/d from 0900 to 1700 h (DT), or for 8 h/d from 2100 to 0500 h (NT).Panels show the effects of timing of acetate infusion on the daily rhythms of (A) milk yield (kg/milking), (B) milk fat yield (kg/milking), (C) milk fat concentration (%), (D) milk protein yield (kg/ milking), (E) milk protein concentration (%), and (F) MUN (mg/dL).Data are presented as LSM with SEM bars, and the best fit cosine function is shown.A table within each panel shows the P-value for the main effect of treatment and time and their 2-way interaction from a mixed model that also includes the random effects of cow and period.The waveform properties, as determined by cosinor analysis, are shown below each panel and include the amplitude (Amp; difference between peak and mean), acrophase (Acro; time at peak of rhythm, h), and the P-value of the zero-amplitude test.Superscripts within each column denote differences among estimates (P ≤ 0.05).

Figure 3 .
Figure 3. Effects of time of acetate infusion on the daily rhythms of milk fatty acid (FA) yield.Treatments were 600 g/d of acetate as anhydrous sodium acetate dissolved in 5 L of distilled water and were ruminally infused continuously throughout the day (CON), for 8 h/d from 0900 to 1700 h (DT), or for 8 h/d from 2100 to 0500 h (NT).Panels show the effects of timing of acetate infusion on the daily rhythms of (A) de novo synthesized FA (Σ < 16 C FA), (B) mixed source FA (Σ 16 C FA), (C) preformed FA (Σ > 16 C FA), and (D) total sum of odd-and branched-chain FA (OBCFA).Data are shown as grams per milking and are presented as LSM with SEM bars, and the best fit cosine function is shown.A table within each panel shows the P-value for the main effects of treatment and time and their 2-way interaction from a mixed model that also includes the random effects of cow and period.The waveform properties, as determined by cosinor analysis, are shown below each panel and include the amplitude (Amp; difference between peak and mean), acrophase (Acro; time at peak of rhythm, h), and the P-value of the zero-amplitude test.Superscripts within each column denote differences among estimates (P ≤ 0.05).

Figure 4 .
Figure 4. Effects of time of acetate infusion on the daily rhythms of plasma metabolites and hormones.Treatments were 600 g/d of acetate as anhydrous sodium acetate dissolved in 5 L of distilled water and were ruminally infused continuously throughout the day (CON), for 8 h/d from 0900 to 1700 h (DT), or for 8 h/d from 2100 to 0500 h (NT).Panels show the effects of timing of acetate infusion on the daily rhythm of (A) plasma acetate concentration (mmol/mL), (B) plasma BHB (µmol/L), (C) plasma glucose (mg/dL), (D) plasma insulin concentration (µIU/L), (E) plasma nonesterified fatty acids concentration (NEFA; µEq/L), and (F) BUN (mg/dL).Data are presented as LSM with SEM bars, and the best fit cosine function is shown.A table within each panel shows the P-values for the main effects of treatment and time and their 2-way interaction from a mixed model that also includes the random effects of cow and period.Waveform parameters, as determined by cosinor analysis, are shown below each panel and include the amplitude (Acro; time at peak of rhythm, h), acrophase (Acro; time at peak of rhythm, h), and the P-value of the zero-amplitude test.Superscripts within each column denote differences among estimates (P ≤ 0.05).

Figure 4 (
Figure 4 (Continued).Effects of time of acetate infusion on the daily rhythms of plasma metabolites and hormones.Treatments were 600 g/d of acetate as anhydrous sodium acetate dissolved in 5 L of distilled water and were ruminally infused continuously throughout the day (CON), for 8 h/d from 0900 to 1700 h (DT), or for 8 h/d from 2100 to 0500 h (NT).Panels show the effects of timing of acetate infusion on the daily rhythm of (A) plasma acetate concentration (mmol/mL), (B) plasma BHB (µmol/L), (C) plasma glucose (mg/dL), (D) plasma insulin concentration (µIU/L), (E) plasma nonesterified fatty acids concentration (NEFA; µEq/L), and (F) BUN (mg/dL).Data are presented as LSM with SEM bars, and the best fit cosine function is shown.A table within each panel shows the P-values for the main effects of treatment and time and their 2-way interaction from a mixed model that also includes the random effects of cow and period.Waveform parameters, as determined by cosinor analysis, are shown below each panel and include the amplitude (Acro; time at peak of rhythm, h), acrophase (Acro; time at peak of rhythm, h), and the P-value of the zero-amplitude test.Superscripts within each column denote differences among estimates (P ≤ 0.05).

Figure 5 .
Figure 5. Effects of time of acetate infusion on the daily rhythm of core body temperature.Treatments were 600 g/d of acetate as anhydrous sodium acetate dissolved in 5 L of distilled water and were ruminally infused continuously throughout the day (CON), for 8 h/d from 0900 to 1700 h (DT), or for 8 h/d from 2100 to 0500 h (NT).Data are presented as LSM with SEM bars, and the best fit cosine function is shown.A table within the panel shows the P-values for the main effects of treatment and time and their 2-way interaction from a mixed model that also includes the random effects of cow and period.Waveform parameters, as determined by cosinor analysis, are shown below the panel and include the amplitude (Amp; difference between peak and mean), acrophase (Acro; time at peak of rhythm, h), and the P-value of the zero-amplitude test.Superscripts within each column denote differences among estimates (P ≤ 0.05).

Table 1 .
Matamoros et al.: ACETATE AND CIRCADIAN RHYTHMS Ingredient and nutrient composition (all % of DM) of experimental diet

Table 2 .
Matamoros et al.: ACETATE AND CIRCADIAN RHYTHMSThe effect of timing of ruminal infusion of sodium acetate on DMI, feeding behavior, and milk yield and composition Matamoros et al.: ACETATE AND CIRCADIAN RHYTHMS