Advertisement

Hepatic betaine-homocysteine methyltransferase and methionine synthase activity and intermediates of the methionine cycle are altered by choline supply during negative energy balance in Holstein cows

Open AccessPublished:July 10, 2019DOI:https://doi.org/10.3168/jds.2018-16204

      ABSTRACT

      Although choline requirements are unknown, enhanced postruminal supply may decrease liver triacylglycerol (TAG) storage and increase flux through the methionine cycle, helping cows during a negative energy balance (NEB). The objective was to investigate effects of postruminal choline supply during NEB on hepatic activity of betaine-homocysteine methyltransferase (BHMT), methionine synthase (MTR), methionine adenosyltransferase, transcription of enzymes, and metabolite concentrations in the methionine cycle. Ten primiparous rumen-cannulated Holstein cows (158 ± 24 d postpartum) were used in a replicated 5 × 5 Latin square design with 4-d treatment periods and 10 d of recovery (14 d/period). Treatments were unrestricted intake with abomasal infusion of water (A0), restricted intake (R; 60% of net energy for lactation requirements to induce NEB) with abomasal infusion of water (R0) or R plus abomasal infusion of 6.25, 12.5, or 25 g/d of choline ion. Liver tissue was collected on d 5 after the infusions ended, blood on d 1 to 5, and milk on d 1 to 4. Statistical contrasts were A0 versus R0 (CONT1) and tests of linear (L), quadratic (Q), and cubic (C) effects of choline dose. Plasma choline increased with R (CONT1) and choline (L). Although R decreased milk yield (CONT1), choline increased milk yield and liver phosphatidylcholine (PC), but decreased TAG (L). No differences were observed in plasma PC or very-low-density lipoprotein concentrations with R or choline. Activity and mRNA abundance of BHMT were greater with R (CONT1) and increased with choline (L). Although activity of MTR was lower with R (CONT1), it tended to increase with choline (L). No effect of R was detected for activity of methionine adenosyltransferase, but it changed cubically across dose of choline. Those responses were associated with linear increases in the concentrations of liver tissue (+13%) and plasma methionine concentrations. The mRNA abundance of CPT1A, SLC22A5, APOA5, and APOB, genes associated with fatty acid oxidation and lipoprotein metabolism, was upregulated by choline (Q). Overall, enhanced supply of choline during NEB increases hepatic activity of BHMT and MTR to regenerate methionine and PC, partly to help clear TAG. The relevance of these effects during the periparturient period merits further research.

      Key words

      INTRODUCTION

      Dairy cows enter a period of negative energy balance (NEB) when they transition from pregnancy to lactation (i.e., the periparturient period). During this transition, lipid mobilization increases, resulting in greater nonesterified fatty acid (NEFA) uptake by the liver and increased triacylglycerol (TAG) accumulation (
      • Loor J.J.
      • Bionaz M.
      • Drackley J.K.
      Systems physiology in dairy cattle: Nutritional genomics and beyond..
      ). This accumulation of TAG may be linked to increased NEFA uptake activating sterol regulatory element-binding protein 1c, which recent data indicate could promote transcription of lipogenic genes (
      • Zhu Y.
      • Liu G.
      • Du X.
      • Shi Z.
      • Jin M.
      • Sha X.
      • Li X.
      • Wang Z.
      • Li X.
      Expression patterns of hepatic genes involved in lipid metabolism in cows with subclinical or clinical ketosis..
      ), and the inherently low rate of very-low-density lipoprotein (VLDL) secretion in ruminants (
      • Bauchart D.
      • Gruffat D.
      • Durand D.
      Lipid absorption and hepatic metabolism in ruminants..
      ). A chronic state of TAG accumulation in the liver can lead to mitochondrial dysfunction and inflammation (
      • Li X.
      • Huang W.
      • Gu J.
      • Du X.
      • Lei L.
      • Yuan X.
      • Sun G.
      • Wang Z.
      • Li X.
      • Liu G.
      SREBP-1c overactivates ROS-mediated hepatic NF-κB inflammatory pathway in dairy cows with fatty liver..
      ;
      • Du X.
      • Shen T.
      • Wang H.
      • Qin X.
      • Xing D.
      • Ye Q.
      • Shi Z.
      • Fang Z.
      • Zhu Y.
      • Yang Y.
      • Peng Z.
      • Zhao C.
      • Lv B.
      • Li X.
      • Liu G.
      • Li X.
      Adaptations of hepatic lipid metabolism and mitochondria in dairy cows with mild fatty liver..
      ). To release TAG from the liver it is packaged in VLDL; thus, a way to reduce TAG accumulation may be to stimulate VLDL synthesis and export.
      Choline may stimulate VLDL synthesis in the liver through synthesis of phosphatidylcholine (PC), the main phospholipid component of VLDL (
      • Vance J.E.
      Assembly and secretion of lipoproteins.
      ). In nonruminants, PC synthesis and VLDL secretion is reduced during choline deficiency, leading to the development of fatty liver (
      • Yao Z.M.
      • Vance D.E.
      Reduction in VLDL, but not HDL, in plasma of rats deficient in choline..
      ;
      • Fast D.G.
      • Vance D.E.
      Nascent VLDL phospholipid composition is altered when phosphatidylcholine biosynthesis is inhibited: Evidence for a novel mechanism that regulates VLDL secretion..
      ). Thus, supplementation of choline during the transition period may help improve VLDL synthesis and reduce liver TAG accumulation. However, studies supplementing rumen-protected choline (RPC) to dairy cows have yielded mixed results, with reductions in hepatic TAG reported in some experiments (
      • Cooke R.F.
      • Silva Del Rio N.
      • Caraviello D.Z.
      • Bertics S.J.
      • Ramos M.H.
      • Grummer R.R.
      Supplemental choline for prevention and alleviation of fatty liver in dairy cattle..
      ;
      • Zom R.L.G.
      • van Baal J.
      • Goselink R.M.A.
      • Bakker J.A.
      • de Veth M.J.
      • van Vuuren A.M.
      Effect of rumen-protected choline on performance, blood metabolites, and hepatic triacylglycerols of periparturient dairy cattle..
      ;
      • Elek P.
      • Gaal T.
      • Husveth F.
      Influence of rumen-protected choline on liver composition and blood variables indicating energy balance in periparturient dairy cows..
      ), but not in others (
      • Hartwell J.R.
      • Cecava M.J.
      • Donkin S.S.
      Impact of dietary rumen undegradable protein and rumen-protected choline on intake, peripartum liver triacylglyceride, plasma metabolites and milk production in transition dairy cows..
      ;
      • Piepenbrink M.S.
      • Overton T.R.
      Liver metabolism and production of cows fed increasing amounts of rumen-protected choline during the periparturient period..
      ;
      • Zhou Z.
      • Vailati-Riboni M.
      • Trevisi E.
      • Drackley J.K.
      • Luchini D.N.
      • Loor J.J.
      Better postpartal performance in dairy cows supplemented with rumen-protected methionine compared with choline during the peripartal period..
      ).
      During its metabolism, choline can be oxidized to betaine (Bet), which is used as a methyl donor in the synthesis of Met from homocysteine via the enzyme betaine homocysteine methyltransferase (BHMT;
      • Li Z.
      • Vance D.E.
      Phosphatidylcholine and choline homeostasis..
      ). Methionine can also be synthesized from homocysteine via 5-methyltetrahydrofolate-homocysteine methyltransferase (MTR). Compared with rats, hepatic BHMT activity was lower and MTR greater in sheep, suggesting a crucial role of the latter in de novo synthesis of Met (
      • Xue G.P.
      • Snoswell A.M.
      Comparative studies on the methionine synthesis in sheep and rat tissues..
      ). Despite those general differences, enhanced supply of Bet can increase Met synthesis in liver of sheep, underscoring flexibility of BHMT to substrate availability (
      • Xue G.P.
      • Snoswell A.M.
      Comparative studies on the methionine synthesis in sheep and rat tissues..
      ). Because choline may be synthesized endogenously via methyl group donation from Met to S-adenosyl-methionine (SAM) via activity of methionine adenosyltransferase (MAT;
      • Vance D.E.
      • Walkey C.J.
      • Cui Z.
      Phosphatidylethanolamine N-methyltransferase from liver..
      ), enhanced supply of choline may help increase flux through the Met cycle. This could spare Met so that it can be used for its key roles in protein synthesis, antioxidant production, and methyl group donation (
      • Finkelstein J.D.
      Methionine metabolism in mammals..
      ), rather than making choline. Whether choline supply enhances Met synthesis in dairy cows during NEB, and the mechanisms involved, has not been determined.
      Dairy cattle have no known requirement for choline (
      NRC
      Nutrient Requirements of Dairy Cattle.
      ). Previous studies with periparturient cows have used approximately 12.5 g/d of choline provided as RPC (
      • Zahra L.C.
      • Duffield T.F.
      • Leslie K.E.
      • Overton T.R.
      • Putnam D.
      • LeBlanc S.J.
      Effects of rumen-protected choline and monensin on milk production and metabolism of periparturient dairy cows..
      ;
      • Zom R.L.G.
      • van Baal J.
      • Goselink R.M.A.
      • Bakker J.A.
      • de Veth M.J.
      • van Vuuren A.M.
      Effect of rumen-protected choline on performance, blood metabolites, and hepatic triacylglycerols of periparturient dairy cattle..
      ;
      • Zhou Z.
      • Vailati-Riboni M.
      • Trevisi E.
      • Drackley J.K.
      • Luchini D.N.
      • Loor J.J.
      Better postpartal performance in dairy cows supplemented with rumen-protected methionine compared with choline during the peripartal period..
      ), which has been suggested as the optimal amount. Recent work by
      • de Veth M.J.
      • Artegoitia V.M.
      • Campagna S.R.
      • Lapierre H.
      • Harte F.
      • Girard C.L.
      Choline absorption and evaluation of bioavailability markers when supplementing choline to lactating dairy cows..
      with lactating cows in positive nutrient balance reported that abomasal infusions of choline chloride for 5 d at 12.5 and 25 g/d increased the appearance of choline and its metabolites Bet and phosphocholine in arterial plasma and increased portal flux of choline. In pregnant, nonlactating feed restricted cows (restricted to 31% of their net energy requirements), increasing choline ion supply from 0 to 25 g/d with RPC reduced hepatic TAG concentration linearly. Thus, in the context of physiologic adaptations during periods of NEB, there clearly would be a benefit in determining not only optimal amounts of postruminally delivered choline but also mechanisms in the one-carbon metabolism pathway responsive to choline supply.
      The hypothesis for the present study was that incremental postruminal supply of choline during a feed restriction-induced NEB would decrease liver TAG accumulation, increase activity of BHMT, and alter concentrations of intermediates in the Met cycle. Our specific objective was to investigate the effects of postruminal choline supply during a feed restriction-induced NEB on hepatic activity of BHMT, MTR, MAT, transcription of enzymes, and metabolite concentrations in the methionine cycle, and their relationship with liver TAG.

      MATERIALS AND METHODS

      Experimental Design and Dietary Treatments

      All procedures involving animals received approval from the Institutional Animal Care and Use Committee at the University of Illinois (protocol no. 16176). Ten primiparous rumen-cannulated Holstein cows (158 ± 24 postpartum) were used in a 5 × 5 replicated Latin square design. Periods were 14 d in length, with 4-d treatment periods (d 1–4) and 10 d of recovery (d 5–14). Treatment periods of 4 d were chosen based on preliminary observations that suggested concentrations on NEFA and BHB (indicators of NEB) stabilized and began to decrease after 4 d of feed restriction at 60% of NEL requirements (unpublished data). The 10-d recovery periods were also chosen based on preliminary observations that suggested that milk yield and concentrations of NEFA and BHB had returned to baseline levels 10 d after feed restriction at 60% of NEL requirements (unpublished data). Additionally, a restriction at 60% of NEL requirements was chosen based on previous work using this restriction model to mimic the periparturient period (
      • Moyes K.M.
      • Drackley J.K.
      • Morin D.E.
      • Loor J.J.
      Greater expression of TLR2, TLR4, and IL6 due to negative energy balance is associated with lower expression of HLA-DRA and HLA-A in bovine blood neutrophils after intramammary mastitis challenge with Streptococcus uberis..
      ). Treatments in the present study were ad libitum intake with abomasal infusion of water (A0), restricted intake (R; 60% of NEL requirements) with abomasal infusion of water (R0) or R plus abomasal infusion of 6.25 (R6.25), 12.5 (R12.5), or 25 g/d of choline ion (R25). The diet was provided as a TMR and formulated to meet
      NRC
      Nutrient Requirements of Dairy Cattle.
      requirements (Table 1). Diets were fed once daily at 0800 h. All cows returned to full feed intake during the 10-d recovery period.
      Table 1Ingredient composition of the diet
      ItemContent
      Ingredient, % of DM
       Alfalfa hay9.26
       Ajipro-L-G3
      Ajinomoto Heartland Inc. (Chicago, IL).
      0.11
       Canola meal15.15
       Corn gluten meal4.56
       Corn silage41.33
       Ground shelled corn23.24
       ProVAAL2 AADvantage
      Perdue AgriBusiness (Salisbury, MD).
      0.96
       Smartamine M
      Adisseo (Alpharetta, GA).
      0.07
       Soy hulls1.69
       Vitamin-mineral mix
      Contained a minimum of 12.5% Ca, 10.4% Na, 2.2% Mg, 8.0% K, 0.1% S, 7.1% Se, 244.5 kIU of vitamin A/kg, 48.9 kIU of vitamin D3/kg, and 0.922 kIU of vitamin E/kg.
      3.57
       Zinpro Availa-Dairy
      Zinpro Corporation (Eden Prairie, MN).
      0.06
      Chemical analysis, % of DM
       CP16.00
       ADF20.00
       NDF18.19
       Ether extract3.05
      NEL, Mcal/kg1.71
      1 Ajinomoto Heartland Inc. (Chicago, IL).
      2 Perdue AgriBusiness (Salisbury, MD).
      3 Adisseo (Alpharetta, GA).
      4 Contained a minimum of 12.5% Ca, 10.4% Na, 2.2% Mg, 8.0% K, 0.1% S, 7.1% Se, 244.5 kIU of vitamin A/kg, 48.9 kIU of vitamin D3/kg, and 0.922 kIU of vitamin E/kg.
      5 Zinpro Corporation (Eden Prairie, MN).
      The choline was provided as aqueous choline chloride (CC, 70%, Balchem Corporation, New Hampton, NY). The doses of 6.25, 12.5, and 25 g/d represent the amount of choline ion that was supplied postruminally, and were achieved by providing 12.0, 23.9, and 47.9 g/d of CC (CC 70%, Balchem Corporation). Doses above and below the suggested requirement of 12.5 g/d of choline allow for the determination of linear and quadratic responses of postruminal choline ion supplementation. Abomasal infusions were accomplished by passing an infusion line through the rumino-omasal orifice into the abomasum by way of the rumen cannula. An infusion line (Tygon tubing, Cole-Parmer, Vernon Hills, IL) fitted with a 60-mL plastic bottle with the bottom removed and with an approximately 8-cm rubber flange was inserted into the abomasum. The tubing was in place for the duration of the study and was checked daily for proper placement. The infusion apparatus was connected to an external infusion line via a hole in the rumen cannula plug. The internal and external lines were connected by way of quick-disconnect fitting, to allow disconnection for milking. The treatment solutions were provided by mixing the 70% CC with deionized tap water and were infused continuously using rotary peristaltic pumps with their corresponding 1,200-mL bag (Sentinel Enteral Feeding Pump, Alcor Scientific Inc., Smithfield, RI). A daily volume of 3,600 mL/cow (three 1,200-mL bags/d) was infused during the 4-d treatment period. Infusions were initiated after cows returned from the 0400 h milking on d 1, and were ended on the morning of d 5 at the 0400 h milking. Since the cows were milked 3 times a day (0400, 1200, and 1930 h), a new infusion bag was hooked up after each milking.

      Feed and Production Parameter Measurements

      Dry matter of individual feed ingredients was determined weekly and the ration was adjusted accordingly to maintain DM ratios of the ingredients. Weekly samples of ingredients and TMR were frozen at −20°C and composited monthly for analysis by standard wet chemistry techniques at a commercial laboratory (Dairy One, Ithaca, NY).
      Body weight was recorded on d 1 after the morning milking, and on d 5 after the morning milking. Milk production was recorded daily, and samples were collected at each milking on d 1 to 4, and on d 14. Composite samples were prepared each day in proportion to milk yield at each milking and were preserved (800 Broad Spectrum Microtabs II; D & F Control Systems Inc., San Ramon, CA), and analyzed for contents of fat, protein, SNF, MUN, and SCC in a commercial laboratory (Dairy One). Based on milk sample analysis, FCM was calculated daily as follows: FCM = (0.4324 × kg of milk yield) + (16.216 × kg of milk fat) (
      • Hutjens M.F.
      Benchmarking your feed efficiency, feed costs and income over feed cost.
      ). The concentration of free choline in milk was measured using a commercial kit (no. KA1662, Abonova, Taipei City, Taiwan) where free choline is oxidized by choline oxidase to Bet and hydrogen peroxide, which reacts with a specific dye to form a pink-colored product. Energy balance (EB) was calculated on d 1 and 5 using NRC equations, as described previously (
      • Moyes K.M.
      • Drackley J.K.
      • Salak-Johnson J.L.
      • Morin D.E.
      • Hope J.C.
      • Loor J.J.
      Dietary-induced negative energy balance has minimal effects on innate immunity during a Streptococcus uberis mastitis challenge in dairy cows during midlactation..
      ).

      Blood Collection and Analyses

      Blood was sampled from the coccygeal vein after the morning milking on d 1 to 5, approximately 3 h before feeding. Samples were collected into vacutainer tubes containing lithium heparin (BD Vacutainer, BD and Co., Franklin Lakes, NJ) and placed on ice. Plasma was obtained by centrifugation at 2,000 × g for 15 min at 4°C and aliquots stored at −80°C until further analysis. Plasma concentrations of BHB were analyzed using a commercial kit (no. 700190, Cayman Chemical, Ann Arbor, MI) and plasma NEFA as described by
      • Johnson M.M.
      • Peters J.
      Technical note: An improved method to quantify nonesterified fatty acids in bovine plasma..
      . Concentrations of free choline, VLDL (no. K615–100 and KK613–100, respectively; Biovision Inc., Milpitas, CA), and PC (No. 10009926; Cayman Chemical) in plasma were quantified using commercial kits. Plasma concentrations of AA and AA derivatives were measured on d 5 using an ultra-performance liquid chromatography–MS (Waters, Milford, MA) with the derivatization method (AccQ-Tag Derivatization) provided by Waters.

      Liver Biopsy and Tissue Analysis

      Liver biopsies were conducted on the morning of d 5 after infusions were ended using a similar technique described previously (
      • Khan M.J.
      • Jacometo C.B.
      • Graugnard D.E.
      • Corrêa M.N.
      • Schmitt E.
      • Cardoso F.
      • Loor J.J.
      Overfeeding dairy cattle during late-pregnancy alters hepatic PPARα-regulated pathways including hepatokines: Impact on metabolism and peripheral insulin sensitivity..
      ;
      • Vailati Riboni M.
      • Meier S.
      • Priest N.V.
      • Burke C.R.
      • Kay J.K.
      • McDougall S.
      • Mitchell M.D.
      • Walker C.G.
      • Crookenden M.
      • Heiser A.
      • Roche J.R.
      • Loor J.J.
      Adipose and liver gene expression profiles in response to treatment with a nonsteroidal antiinflammatory drug after calving in grazing dairy cows..
      ). Briefly, the skin was shaved and disinfected, and the skin and body wall were anesthetized with 7 mL of 2% lidocaine HCl (VetOne, Boise, ID). A stab incision was made through the skin in the right 11th intercostal space, through which an 18-gauge by 10.2-cm bone marrow probe (Monoject-8881–247087, Medtronic, Minneapolis, MN) was inserted into the liver and used to obtain approximately 2 g (wet weight) of liver. No more than 3 separate penetrations with the biopsy probe were performed. Samples were snap frozen in liquid nitrogen and subsequently stored at −80°C.
      Liver TAG was measured as described by
      • Zhou Z.
      • Vailati-Riboni M.
      • Trevisi E.
      • Drackley J.K.
      • Luchini D.N.
      • Loor J.J.
      Better postpartal performance in dairy cows supplemented with rumen-protected methionine compared with choline during the peripartal period..
      and
      • Batistel F.
      • Arroyo J.M.
      • Bellingeri A.
      • Wang L.
      • Saremi B.
      • Parys C.
      • Trevisi E.
      • Cardoso F.C.
      • Loor J.J.
      Ethyl-cellulose rumen-protected methionine enhances performance during the periparturient period and early lactation in Holstein dairy cows..
      . Total RNA was extracted from liver tissue using the miRNeasy mini kit (no. 217004, Qiagen, Hilden, Germany). The Nano-Drop ND-1000 spectrophotometer (Nano-Drop Technologies, Wilmington, DE) was used to measure the concentration of RNA, whereas the quality of RNA was evaluated using the Agilent Bioanalyzer system (Agilent 2100 Bioanalyzer, Agilent Technologies, Santa Clara, CA). All samples had an RNA integrity number greater than 7. The quantitative PCR was performed as described previously (
      • Osorio J.S.
      • Ji P.
      • Drackley J.K.
      • Luchini D.
      • Loor J.J.
      Smartamine M and MetaSmart supplementation during the peripartal period alter hepatic expression of gene networks in 1-carbon metabolism, inflammation, oxidative stress, and the growth hormone–insulin-like growth factor 1 axis pathways..
      ), and full details of the primer design and evaluation and the quantitative PCR performance and primer information are included in the Supplemental Materials and Supplemental Tables S1 and S2 (https://doi.org/10.3168/jds.2018-16204). Activities of BHMT and MTR were measured as described previously (
      • Zhou Z.
      • Garrow T.A.
      • Dong X.
      • Luchini D.N.
      • Loor J.J.
      Hepatic activity and transcription of betaine-homocysteine methyltransferase, methionine synthase, and cystathionine synthase in periparturient dairy cows are altered to different extents by supply of methionine and choline..
      ;
      • Batistel F.
      • Alharthi A.S.
      • Yambao R.R.C.
      • Elolimy A.A.
      • Pan Y.X.
      • Parys C.
      • Loor J.J.
      Methionine supply during late-gestation triggers offspring sex-specific divergent changes in metabolic and epigenetic signatures in bovine placenta..
      ). Activity of MAT was measured using a commercial kit (no. IK00401, Arthus Biosystems, LLC, Richmond, CA). Protein of enzyme assay samples was measured via the Bradford assay (no. 500–0205, Bio-Rad Laboratories Inc., Hercules, CA). Liver PC content was analyzed using a commercial kit (no. ab83377, Abcam, Cambridge, United Kingdom). This kit uses an enzyme-coupled reaction to hydrolyze PC and release choline, which subsequently oxidizes an OxiRed probe to generate absorbance.
      Approximately 150 mg of frozen tissue was extracted for metabolomics analysis using the 2-step protocol described by
      • Wu H.
      • Southam A.D.
      • Hines A.
      • Viant M.R.
      High-throughput tissue extraction protocol for NMR- and MS-based metabolomics..
      . Targeted metabolomics [liquid chromatography (LC)-MS] was performed to quantify 4 metabolites related to one-carbon metabolism: homocysteine, SAM, S-5′-adenosyl-homocysteine (SAH), and Met. Samples were analyzed with the 5500 QTRAP LC/MS/MS system (Sciex, Framingham, MA) in the Metabolomics Laboratory of the Roy J. Carver Biotechnology Center, University of Illinois at Urbana-Champaign. Details of the system and analysis are provided in the Supplemental File, and the multiple reaction monitoring information used for quantification is in Supplemental Table S3 (https://doi.org/10.3168/jds.2018-16204).

      Statistical Analyses

      Data were analyzed as a replicated 5 × 5 Latin square with repeated measures of day, when needed, using the MIXED procedure of SAS 9.4 (SAS Institute Inc., Cary, NC).
      The model tested the random effects of square and cow within square, and the fixed effects period, treatment, day, and the interaction of treatment × day. Covariance structures compared were compound symmetry, autoregressive order 1, autoregressive heterogeneous order 1, unstructured, and Toeplitz. The compound symmetry structure was used based on the corrected Akaike information criterion. The Kenward-Roger degrees of freedom approximation was used to determine the denominator degrees of freedom for tests of fixed effects. The CONTRAST statement of SAS was used to perform 4 orthogonal contrasts: CONT1 = A0 compared with R0, and the linear, quadratic, and cubic effects of R0, R6.25, R12.5, and R25. Contrast coefficients were adjusted using the IML procedure to reflect uneven spacing of treatments. Significance was set at P ≤ 0.05, and tendencies at P ≤ 0.10. When a polynomial contrast was significant or tended to be significant, equations for the trends were calculated. Pearson correlation coefficients between dependent variables in R cows were calculated using PROC CORR of SAS. The correlations were used to assess the relationship between liver TAG and key metabolites and components of Met and choline metabolism during R. The variables used were liver TAG, Met and PC, plasma free choline, PC, VLDL, and Met, milk free choline, and hepatic activity of BHMT, MTR, and MAT.
      Variables were assessed for normality of distribution using the Shapiro-Wilk test. The mRNA abundance data were log2-scale transformed to fit normal distribution of residuals. Tables and graphs contain the log2 back-transformed means that resulted from the statistical analysis. Least squares means and standard errors were determined using the LSMEANS statement of SAS and least squares means separation between time points was performed using the PDIFF statement. When treatment × day interactions were significant, the SLICE option was used to separate means.

      RESULTS

      Production Responses

      Production results are presented in Table 2. Restriction reduced DMI (P < 0.001; CONT1), BW (P < 0.001), and EB (P < 0.001). Restriction reduced milk yield (P < 0.001; CONT1), whereas choline dose had a positive linear effect (P = 0.05; Y = 0.07783X + 23.6325). Fat-corrected milk was reduced by R (P = 0.001; CONT1). A positive quadratic effect was also observed FCM (P = 0.05; Y = −0.00221X2 + 0.06404X + 21.72), with the R12.5 treatment having the greatest yield. Protein percentage and yield (P = 0.003 and P < 0.001, respectively; CONT1), and lactose percentage (P < 0.001) decreased with R, whereas fat percentage (P = 0.001), solids percentage (P = 0.01), and MUN (P < 0.001) increased with R (CONT1). A negative linear effect was observed for milk fat percentage (P = 0.04; Y = −0.01431X + 3.0340), whereas a tendency was observed for a negative linear effect for protein percentage (P = 0.10; Y = −0.00178X + 3.2753). Lactose yield also tended to increase linearly (P = 0.07; Y = 0.003749X + 1.1240). Treatment × day interactions (Supplemental Figures S1 and S2; https://doi.org/10.3168/jds.2018-16204) were observed for BW (P < 0.001), DMI (P = 0.01) milk yield (P = 0.002), protein yield (P = 0.003), MUN (P = 0.02), and EB (P = 0.003). The interactions for DMI and EB were due to A0 cows having greater intakes and EB compared with cows that were feed-restricted on all days (P ≤ 0.05). Interactions for MY and protein yield were observed on d 2 to 4, with cows in the R treatments having lower yields compared with A0 cows (P ≤ 0.05). The interaction for BW was due to the R treatments inducing lower BW compared with A0 cows on the morning of d 5 when infusions ended (P ≤ 0.05).
      Table 2Least squares means and associated pooled SEM for BW and production responses in Holstein cows (n = 10/treatment) fed 5 different treatments over a period of 4 d: ad libitum intake with abomasal water infusion, or restricted intake at 60% of NEL requirements with abomasal infusions of 0, 6.25, 12.5, or 25 g/d of choline
      Choline was provided via 70% aqueous solution of choline chloride (Balchem Corporation, New Hampton, NY) mixed with water.
      ItemTreatment
      A0 = ad libitum intake, abomasal infusion of water; R0 = restricted intake, abomasal infusion of water; R6.25 = restricted intake, abomasal infusion of 6.25 g/d of choline; R12.5 = restricted intake, abomasal infusion of 12.5 g/d of choline; R25 = restricted intake, abomasal infusion of 25 g/d of choline.
      SEMContrast, P-value
      Contrasts: 1 = A0 compared with R0. Linear, quadratic, and cubic effects of treatments R0, R6.25, R12.5, and R25.
      1Treatment (Trt)
      A0R0R6.25R12.5R25LinearQuadraticCubic
      DMI,
      DMI Trt × day = 0.01; BW Trt × day <0.001; milk yield Trt × day = 0.002; protein yield Trt × day = 0.003; MUN Trt × day = 0.02; energy balance Trt × day = 0.04.
      kg/d
      19.849.879.899.919.920.42<0.0010.890.990.64
      BW,
      DMI Trt × day = 0.01; BW Trt × day <0.001; milk yield Trt × day = 0.002; protein yield Trt × day = 0.003; MUN Trt × day = 0.02; energy balance Trt × day = 0.04.
      kg
      588.9567.9566.9564.4564.621.5<0.0010.480.720.78
      Milk yield,
      DMI Trt × day = 0.01; BW Trt × day <0.001; milk yield Trt × day = 0.002; protein yield Trt × day = 0.003; MUN Trt × day = 0.02; energy balance Trt × day = 0.04.
      kg/d
      30.4823.2424.2525.2025.250.74<0.0010.050.270.82
      Milk composition
       Fat, %2.483.122.872.802.720.160.0010.040.330.75
       Fat, kg/d0.770.720.690.700.680.050.330.480.770.65
       Protein, %3.443.313.243.233.230.040.0030.100.120.72
       Protein,
      DMI Trt × day = 0.01; BW Trt × day <0.001; milk yield Trt × day = 0.002; protein yield Trt × day = 0.003; MUN Trt × day = 0.02; energy balance Trt × day = 0.04.
      kg/d
      1.040.760.790.810.820.03<0.0010.070.410.76
       Lactose, %4.864.764.754.754.770.03<0.0010.750.320.98
       Lactose, kg/d1.471.111.151.201.200.05<0.0010.070.340.81
       Solids, %11.6512.1011.7711.7011.620.160.010.050.140.53
       Solids, kg/d3.532.802.842.942.930.120.010.250.530.64
       MUN,
      DMI Trt × day = 0.01; BW Trt × day <0.001; milk yield Trt × day = 0.002; protein yield Trt × day = 0.003; MUN Trt × day = 0.02; energy balance Trt × day = 0.04.
      mg/dL
      12.6214.8915.0815.2715.140.51<0.0010.670.610.90
       SCC134.10125.67153.30257.55191.5085.750.920.360.260.42
       FCM
      FCM = (0.4324 × kg of milk yield) + (16.216 × kg of milk fat).
      25.6221.7221.9122.1821.931.110.0010.830.050.89
      Energy balance,
      DMI Trt × day = 0.01; BW Trt × day <0.001; milk yield Trt × day = 0.002; protein yield Trt × day = 0.003; MUN Trt × day = 0.02; energy balance Trt × day = 0.04.
      Mcal/d
      4.32−8.63−8.62−8.99−8.960.83<0.0010.720.900.82
      1 Choline was provided via 70% aqueous solution of choline chloride (Balchem Corporation, New Hampton, NY) mixed with water.
      2 A0 = ad libitum intake, abomasal infusion of water; R0 = restricted intake, abomasal infusion of water; R6.25 = restricted intake, abomasal infusion of 6.25 g/d of choline; R12.5 = restricted intake, abomasal infusion of 12.5 g/d of choline; R25 = restricted intake, abomasal infusion of 25 g/d of choline.
      3 Contrasts: 1 = A0 compared with R0. Linear, quadratic, and cubic effects of treatments R0, R6.25, R12.5, and R25.
      4 DMI Trt × day = 0.01; BW Trt × day <0.001; milk yield Trt × day = 0.002; protein yield Trt × day = 0.003; MUN Trt × day = 0.02; energy balance Trt × day = 0.04.
      5 FCM = (0.4324 × kg of milk yield) + (16.216 × kg of milk fat).

      Plasma and Milk Metabolites

      The responses of plasma metabolites are presented in Table 3. Concentrations of BHB and NEFA were both increased with R (P = 0.02, P < 0.001, respectively; CONT1), indicating a NEB was achieved with our feed-restriction model. Additionally, plasma BHB concentrations tended to decrease linearly with choline dose (P = 0.09; Y = −0.00208X + 0.5008). Plasma free choline concentrations increased with R (P < 0.001; CONT1) and increased linearly with increasing choline (P = 0.001; Y = 0.2058X + 3.8221). No differences were observed in plasma VLDL or PC concentrations (P > 0.10). In milk, concentrations of free choline also increased linearly with dose of choline (P = 0.02; Y = 0.4982X + 105.61).
      Table 3Least squares means and pooled SEM for liver, plasma, and milk biomarker response in Holstein cows (n = 10/treatment) fed 5 different treatments over a period of 4 d: ad libitum intake with abomasal water infusion, or restricted intake at 60% of NEL requirements with abomasal infusions of 0, 6.25, 12.5, or 25 g/d of choline
      Choline was provided via 70% aqueous solution of choline chloride (Balchem Corporation, New Hampton, NY) mixed with water.
      Item
      NEFA = nonesterified fatty acids; PC = phosphatidylcholine; TAG = triacylglycerol; VLDL = very-low-density lipoprotein.
      Treatment
      A0 = ad libitum intake, abomasal infusion of water; R0 = restricted intake, abomasal infusion of water; R6.25 = restricted intake, abomasal infusion of 6.25 g/d of choline; R12.5 = restricted intake, abomasal infusion of 12.5 g/d of choline; R25 = restricted intake, abomasal infusion of 25 g/d of choline.
      SEMContrast, P-value
      Contrasts: 1 = A0 compared with R0. Linear, quadratic, and cubic effects of treatments R0, R6.25, R12.5, and R25.
      1Treatment (Trt)
      A0R0R6.25R12.5R25LinearQuadraticCubic
      Liver
       Liver TAG, mg·g of tissue−11.583.482.702.172.340.430.010.050.140.91
       Liver PC, mg·g of tissue−10.660.560.770.620.700.030.040.030.220.003
      Plasma
       BHB, mM0.420.500.480.480.450.030.020.090.990.82
       NEFA,
      NEFA Trt × day <0.001; plasma free choline Trt × day <0.001; milk free choline Trt × day <0.001.
      μEq·L−1
      120.37365.16328.81332.00340.0022.33<0.0010.520.190.40
       Plasma free choline,
      NEFA Trt × day <0.001; plasma free choline Trt × day <0.001; milk free choline Trt × day <0.001.
      μM
      2.673.895.016.268.740.46<0.0010.0010.880.93
       Plasma PC, μM1,6211,6861,6231,6821,6901310.600.820.780.60
       Plasma VLDL, μg·mL−10.260.250.310.280.250.030.840.520.150.29
      Milk
       Milk free choline,
      NEFA Trt × day <0.001; plasma free choline Trt × day <0.001; milk free choline Trt × day <0.001.
      μM
      97.71105.06109.84112.70117.873.810.140.020.710.89
      1 Choline was provided via 70% aqueous solution of choline chloride (Balchem Corporation, New Hampton, NY) mixed with water.
      2 NEFA = nonesterified fatty acids; PC = phosphatidylcholine; TAG = triacylglycerol; VLDL = very-low-density lipoprotein.
      3 A0 = ad libitum intake, abomasal infusion of water; R0 = restricted intake, abomasal infusion of water; R6.25 = restricted intake, abomasal infusion of 6.25 g/d of choline; R12.5 = restricted intake, abomasal infusion of 12.5 g/d of choline; R25 = restricted intake, abomasal infusion of 25 g/d of choline.
      4 Contrasts: 1 = A0 compared with R0. Linear, quadratic, and cubic effects of treatments R0, R6.25, R12.5, and R25.
      5 NEFA Trt × day <0.001; plasma free choline Trt × day <0.001; milk free choline Trt × day <0.001.
      Treatment × day effects (Supplemental Figure S3; https://doi.org/10.3168/jds.2018-16204) were observed for plasma NEFA (P < 0.001) and free choline (P < 0.001), and milk free choline (P < 0.001). The interaction for NEFA was due to R cows having greater concentrations on d 2 to 5 compared with A0 cows (P ≤ 0.05). In plasma, R25 cows had greater concentrations of free choline on d 2 to 5 compared with other treatments (P ≤ 0.05). Compared with R6.25, R12.5 cows had greater plasma free choline on d 2 (P = 0.02) and tended to have greater concentrations (P = 0.10) on d 4. The R12.5 cows also had greater free choline in plasma on d 2 to 5 when compared with R0 cows (P ≤ 0.05). The R6.25 treatment led to greater concentrations than R0 on d 3 and 5 (P ≤ 0.05). The treatment × day interaction for free choline in milk was mainly due to R25 cows having greater concentrations compared with other treatments on d 2 (P ≤ 0.05), whereas, on d 4, R25 and R12.5 cows had greater concentrations than R6.25 and R0 cows (P ≤ 0.05). Plasma concentrations of PC and VLDL were correlated positively on d 5 in R cows (r = 0.38; P = 0.02; Table 4). The concentrations of free choline in milk and plasma tended to be positively correlated (r = 0.31; P = 0.06). Plasma free choline was also correlated positively with plasma PC concentrations in R cows (r = 0.32; P = 0.04).
      Table 4Pearson correlation coefficients among hepatic concentrations of triacylglycerol (TAG), phosphatidylcholine (PC), and Met, plasma free choline, very-low-density lipoprotein (VLDL), Met and PC concentrations, milk free choline, and hepatic activity of betaine homocysteine methyltransferase (BHMT), 5-methyltetrahydrofolaare-homocysteine methyltransferase (MTR), and methionine adenosyltransferase (MAT) in Holstein cows fed to 60% of their NEL requirements with abomasal infusions of 0, 6.25, 12.5, or 25 g/d of choline for 4 d
      Choline was provided via 70% aqueous solution of choline chloride (Balchem Corporation, New Hampton, NY) mixed with water.
      VariableLiver TAGPlasma free cholineMilk free cholinePlasma PCLiver PCPlasma VLDLBHMT activityMTR activityMAT activityLiver Met
      Plasma free choline−0.16
      Milk free choline−0.230.31†
      Plasma PC−0.050.32
      P ≤ 0.05; †P ≤ 0.10.
      0.02
      Liver PC−0.010.48
      P ≤ 0.05; †P ≤ 0.10.
      0.180.45
      P ≤ 0.05; †P ≤ 0.10.
      Plasma VLDL−0.02−0.17−0.070.38
      P ≤ 0.05; †P ≤ 0.10.
      0.07
      BHMT activity−0.37
      P ≤ 0.05; †P ≤ 0.10.
      0.42
      P ≤ 0.05; †P ≤ 0.10.
      0.250.110.20−0.25
      MTR activity−0.42
      P ≤ 0.05; †P ≤ 0.10.
      −0.020.03−0.27†−0.18−0.200.24
      MAT activity0.190.15−0.080.130.070.190.03−0.47
      P ≤ 0.05; †P ≤ 0.10.
      Liver Met−0.190.17−0.160.31†0.120.06−0.05−0.04−0.08
      Plasma Met−0.36
      P ≤ 0.05; †P ≤ 0.10.
      0.18−0.030.05−0.010.0040.180.26−0.110.36
      P ≤ 0.05; †P ≤ 0.10.
      1 Choline was provided via 70% aqueous solution of choline chloride (Balchem Corporation, New Hampton, NY) mixed with water.
      * P ≤ 0.05; †P ≤ 0.10.

      Plasma AA

      Amino acid and derivative concentrations are presented in Table 5. The doubling of concentration of 3-methyl histidine (P < 0.001; CONT1) in response to R0 indicated enhanced muscle protein breakdown. Among the EAA, except for concentrations of arginine, phenylalanine, isoleucine, and leucine, all other AA were decreased (P < 0.05; CONT1) by R0. Similarly, among the NEAA, except for alanine, glutamate, glycine, and serine, all other AA were decreased (P < 0.05; CONT1) by R0. Methionine (P = 0.01; Y = 0.0155X + 2.9786) and threonine (P = 0.03; Y = 0.05720X + 8.6326) increased linearly across choline dose. Plasma serine also changed cubically (P = 0.04; Y = −0.00170X3 + 0.06582X2 – 0.6029X + 10.9680). Among AA derivatives, ornithine decreased with R0 (P < 0.001; CONT1). Additionally, plasma Met was negatively correlated (Table 4) with hepatic TAG in R cows (r = −0.36; P = 0.02).
      Table 5Least squares means and pooled SEM for plasma AA concentrations in Holstein cows (n = 10/treatment) fed 5 different treatments over a period of 4 d: ad libitum intake with abomasal water infusion, or restricted intake at 60% of NEL requirements with abomasal infusions of 0, 6.25, 12.5, or 25 g/d of choline
      Choline was provided via 70% aqueous solution of choline chloride (Balchem Corporation, New Hampton, NY) mixed with water.
      Item, μmoL/dLTreatment
      A0 = ad libitum intake, abomasal infusion of water; R0 = restricted intake, abomasal infusion of water; R6.25 = restricted intake, abomasal infusion of 6.25 g/d of choline; R12.5 = restricted intake, abomasal infusion of 12.5 g/d of choline; R25 = restricted intake, abomasal infusion of 25 g/d of choline.
      SEMContrast, P-value
      Contrasts: 1 = A0 compared with R0. Linear, quadratic, and cubic effects of treatments R0, R6.25, R12.5, and R25.
      1Treatment
      A0R0R6.25R12.5R25LinearQuadraticCubic
      EAA
       Arginine7.557.257.297.627.800.460.620.290.940.77
       Histidine6.004.193.904.104.670.75<0.0010.070.120.52
       Isoleucine9.357.828.778.428.390.620.070.250.370.52
       Leucine13.1212.4712.3612.7712.530.770.540.890.860.74
       Lysine7.706.126.266.626.600.410.010.350.660.73
       Methionine4.183.043.013.163.390.19<0.0010.010.540.57
       Phenylalanine4.314.173.764.033.890.150.460.250.620.29
       Threonine12.548.788.949.1310.190.59<0.0010.030.540.90
       Tryptophan4.433.153.473.473.270.19<0.0010.320.280.16
       Valine24.7917.4018.7518.4518.761.33<0.0010.520.640.65
      NEAA
       Alanine28.0225.5023.3423.9294.371.110.100.670.230.40
       Asparagine4.993.673.483.573.660.24<0.0010.800.360.50
       Aspartate0.900.650.610.660.670.04<0.0010.530.660.39
       Glutamate5.074.784.564.734.910.290.270.420.410.49
       Glutamine28.8425.7826.4426.9326.451.270.030.640.460.92
       Glycine49.8056.7148.0652.5952.273.250.140.630.270.16
       Proline9.787.997.277.727.730.520.0010.910.330.18
       Serine12.2410.979.3610.4010.480.530.070.980.120.04
       Tyrosine AA derivatives5.433.323.433.473.600.20<0.0010.250.900.89
       3-Methyl histidine0.370.630.670.620.650.03<0.0010.750.630.29
       Citrulline7.847.108.268.598.740.750.450.130.340.78
       Ornithine4.433.213.083.333.220.22<0.0010.790.870.36
      Sulfur-containing compounds
       Cystine1.331.351.331.361.390.080.890.840.810.64
       Taurine7.027.508.077.697.530.610.420.810.490.44
      1 Choline was provided via 70% aqueous solution of choline chloride (Balchem Corporation, New Hampton, NY) mixed with water.
      2 A0 = ad libitum intake, abomasal infusion of water; R0 = restricted intake, abomasal infusion of water; R6.25 = restricted intake, abomasal infusion of 6.25 g/d of choline; R12.5 = restricted intake, abomasal infusion of 12.5 g/d of choline; R25 = restricted intake, abomasal infusion of 25 g/d of choline.
      3 Contrasts: 1 = A0 compared with R0. Linear, quadratic, and cubic effects of treatments R0, R6.25, R12.5, and R25.

      Liver Composition and Metabolites

      Liver TAG content (Table 3) was greater with R (P = 0.01; CONT1). There was linear reduction in TAG content across choline doses (P = 0.05; Y = −0.4197X + 3.1316), with the lowest content observed with R12.5. Liver PC content decreased with R (P = 0.04; CONT1) and there was a significant cubic effect (P = 0.003; Y = 0.000198X3 – 0.00737X2 + 0.06653X – 0.5564). Of the 4 metabolites measured in liver tissue via LC-MS, homocysteine, and SAM were undetectable (Table 6). Concentrations of hepatic Met tended (P = 0.08; CONT1) to decrease in liver tissue with R0 versus A0, but increased linearly across choline dose (P = 0.05; Y = 1.4399X + 315.13). Hepatic SAH also changed cubically (P = 0.01; Y = 0.0786X3 – 2.6855X2 + 20.518X + 294.26); of the treatments the concentration of SAH was lowest for R12.5. The hepatic concentration of Met was also positively correlated (Table 4) with plasma Met concentrations (r = 0.36; P = 0.03) and tended to be positively correlated with plasma PC in R cows (r = 0.31; P = 0.07). Hepatic PC concentrations were correlated positively with plasma free choline (r = 0.48; P = 0.002) and plasma PC (r = 0.45; P = 0.01).
      Table 6Least squares means and pooled SEM for betaine homocysteine methyltransferase (BHMT), methionine adenosyltransferase (MAT), and 5-methyltetrahydrofolate-homocysteine methyltransferase (MTR) activity and liver concentrations on metabolites (measured via liquid chromatography-MS) in the methionine cycle in Holstein cows (n = 10/treatment) fed 5 different treatments over a period of 4 d: ad libitum intake with abomasal water infusion, or restricted intake at 60% of NEL requirements with abomasal infusions of 0, 6.25, 12.5, or 25 g/d of choline
      Choline was provided via 70% aqueous solution of choline chloride (Balchem Corporation, New Hampton, NY) mixed with water.
      ItemTreatment
      A0 = ad libitum intake, abomasal infusion of water; R0 = restricted intake, abomasal infusion of water; R6.25 = restricted intake, abomasal infusion of 6.25 g/d of choline; R12.5 = restricted intake, abomasal infusion of 12.5 g/d of choline; R25 = restricted intake, abomasal infusion of 25 g/d of choline.
      SEMContrast, P-value
      Contrasts: 1 = A0 compared with R0. Linear, quadratic, and cubic effects of treatments R0, R6.25, R12.5, and R25.
      1Treatment
      A0R0R6.25R12.5R25LinearQuadraticCubic
      Enzyme activity, nmol of product·h−1·mg of protein−1
       BHMT5.287.408.519.099.120.640.020.050.210.94
       MAT106.2109.6387.68107.3682.698.420.440.080.800.04
       MTR74.9559.2961.8563.2369.694.790.020.070.840.88
      Metabolite, ng·mg of protein−1
       Methionine386.61327.55324.87330.93354.6424.230.080.050.730.95
       HomocysteineND
      ND = not detectable.
      NDNDNDND
       S-5′-adenosyl-homocysteine304.26294.26334.69271.69318.4220.080.670.710.500.01
       S-5′-adenosyl-methionineNDNDNDNDND
      1 Choline was provided via 70% aqueous solution of choline chloride (Balchem Corporation, New Hampton, NY) mixed with water.
      2 A0 = ad libitum intake, abomasal infusion of water; R0 = restricted intake, abomasal infusion of water; R6.25 = restricted intake, abomasal infusion of 6.25 g/d of choline; R12.5 = restricted intake, abomasal infusion of 12.5 g/d of choline; R25 = restricted intake, abomasal infusion of 25 g/d of choline.
      3 Contrasts: 1 = A0 compared with R0. Linear, quadratic, and cubic effects of treatments R0, R6.25, R12.5, and R25.
      4 ND = not detectable.

      Methionine and Choline Metabolism Pathway

      Activity of BHMT, MTR, and MAT are in Table 6 and hepatic mRNA abundance for genes in the Met and choline metabolism pathways are presented in Table 7. Activity of BHMT was greater with R (P = 0.02 CONT1) and increased linearly with choline (P = 0.05; Y = 0.06312X + 7.8368). Abundance of BHMT was greater with R0 versus A0 (P = 0.0001; CONT1). A positive quadratic effect was also observed for BHMT abundance (P = 0.002; Y = −0.00021X2 + 0.01066X + 1.0256). Abundance of the isoform BHMT2 was greater with R (P = 0.003; CONT1). Activity of MTR decreased with R (P = 0.02; CONT1) but tended to increase linearly with choline supply (P = 0.07; Y = 0.4141X + 58.9313). No difference was observed in the mRNA abundance of MTR (P > 0.10). The activity of MAT changed cubically across choline dose (P = 0.04; Y = −0.02925X3 + 1.0164X2 – 8.1334X + 107.34), decreasing from R0 to R6.25, then increasing with R12.5 and then decreasing again to R25. The abundance of methionionine adenosyltransferase 1A (MAT1A) was lower with R (P = 0.02; CONT1). In contrast, methionine adenosyltransferase 2A (MAT2A) was greater with R (P < 0.001; CONT1). The mRNA abundance of phosphatidylethanolamine methyltransferase (PEMT) tended to be decreased by R relative to A0 (P = 0.09).
      Table 7Least squares means and pooled SEM for hepatic mRNA abundance in Holstein cows (n = 10/treatment) fed 5 different treatments over a period of 4 d: ad libitum intake with abomasal water infusion, or restricted intake at 60% of NEL requirements with abomasal infusions of 0, 6.25, 12.5, or 25 g/d of choline
      Choline was provided via 70% aqueous solution of choline chloride (Balchem Corporation, New Hampton, NY) mixed with water.
      Item
      ACADVL= acyl-CoA dehydrogenase very long chain; ACOX1 = acyl-CoA oxidase 1; APOA5 = apolipoprotein A5; APOB = apolipoprotein B; BHMT = betaine homocysteine methyltransferase; BHMT2 = betaine homocysteine methyltransferase 2; CD36 = CD36 molecule; CDO1 = cysteine dioxygenase; CEPT1 = choline/ethanolamine phosphotransferase 1; CHKA = choline kinase A; CHKB = choline kinase B; CPT1A = carnitine palmitoyltransferase 1A; CREB1 = cAMP responsive element binding protein 1; FABP1 = fatty acid binding protein 1; MAT1A = methionine adenosyltransferase 1A; MAT2A = methionine adenosyltransferase 2A; MTHFR = methylenetetrahydrofolate reductase; MTR = 5-methyltetrahydrofolate-homocysteine methyltransferase; MTTP = microsomal triglyceride transfer protein; PCYT1A = phosphate cytidylyltransferase 1, choline, a; PCYT1B = phosphate cytidylyltransferase 1, choline, b; PEMT = phosphatidylethanolamine methyltransferase; PPARA = peroxisome proliferator activated receptor α; SAHH = S-adenosylhomocysteine hydrolase; SLC22A5 = solute carrier family 22 member 5.
      Treatment
      A0 = ad libitum intake, abomasal infusion of water; R0 = restricted intake, abomasal infusion of water; R6.25 = restricted intake, abomasal infusion of 6.25 g/d of choline; R12.5 = restricted intake, abomasal infusion of 12.5 g/d of choline; R25 = restricted intake, abomasal infusion of 25 g/d of choline.
      SEMContrast, P-value
      Contrasts: 1 = A0 compared with R0; Treatment = R0 compared with the average of the 3 treatments (6.25, 12.5, and 25 g/d of choline). Linear and quadratic effects of treatments R0, R6.25, R12.5, and R25.
      1Treatment
      A0R0R6.25R12.5R25LinearQuadraticCubic
      Methionine and choline metabolism, log-2 backtransformed
      BHMT0.621.021.081.091.130.050.0010.0030.0020.91
      BHMT20.330.500.550.520.530.150.0030.640.700.56
      MAT1A1.331.071.151.131.050.090.020.870.330.49
      MAT2A0.710.991.131.081.080.06<0.0010.430.190.21
      MTHFR1.171.181.061.041.130.100.920.740.130.82
      MTR1.051.131.141.091.120.060.190.800.640.46
      PEMT0.820.720.760.780.780.060.090.340.430.87
      SAHH1.141.221.301.351.260.040.330.780.130.93
      CDP
      CDP = cytidine diphosphate.
      -choline pathway, log-2 backtransformed
      CEPT11.251.101.051.051.020.050.010.140.620.70
      CHKA0.870.980.850.870.890.090.160.440.140.42
      CHKB0.540.600.570.570.580.020.020.450.170.64
      PCYT1A0.870.971.090.960.980.050.150.610.530.07
      PCYT1B1.071.261.041.161.060.100.090.150.460.09
      Lipid metabolism, log-2 backtransformed
      ACADVL0.951.051.121.061.120.060.100.360.480.95
      ACOX11.171.041.051.051.070.080.110.640.910.83
      APOA50.710.971.151.181.190.080.0010.020.090.51
      APOB1.080.961.041.041.000.040.020.600.050.51
      CD360.651.230.971.051.170.09<0.0010.880.010.11
      CPT1A0.780.910.991.000.960.040.010.430.090.76
      CREB10.891.050.960.961.020.040.0040.750.040.55
      FABP11.390.810.900.880.880.07<0.0010.520.510.56
      MTTP1.211.111.111.001.070.050.120.270.150.16
      PPARA1.021.061.101.101.100.060.910.550.580.69
      SLC22A50.530.871.031.101.010.09<0.0010.130.010.62
      1 Choline was provided via 70% aqueous solution of choline chloride (Balchem Corporation, New Hampton, NY) mixed with water.
      2 ACADVL= acyl-CoA dehydrogenase very long chain; ACOX1 = acyl-CoA oxidase 1; APOA5 = apolipoprotein A5; APOB = apolipoprotein B; BHMT = betaine homocysteine methyltransferase; BHMT2 = betaine homocysteine methyltransferase 2; CD36 = CD36 molecule; CDO1 = cysteine dioxygenase; CEPT1 = choline/ethanolamine phosphotransferase 1; CHKA = choline kinase A; CHKB = choline kinase B; CPT1A = carnitine palmitoyltransferase 1A; CREB1 = cAMP responsive element binding protein 1; FABP1 = fatty acid binding protein 1; MAT1A = methionine adenosyltransferase 1A; MAT2A = methionine adenosyltransferase 2A; MTHFR = methylenetetrahydrofolate reductase; MTR = 5-methyltetrahydrofolate-homocysteine methyltransferase; MTTP = microsomal triglyceride transfer protein; PCYT1A = phosphate cytidylyltransferase 1, choline, a; PCYT1B = phosphate cytidylyltransferase 1, choline, b; PEMT = phosphatidylethanolamine methyltransferase; PPARA = peroxisome proliferator activated receptor α; SAHH = S-adenosylhomocysteine hydrolase; SLC22A5 = solute carrier family 22 member 5.
      3 A0 = ad libitum intake, abomasal infusion of water; R0 = restricted intake, abomasal infusion of water; R6.25 = restricted intake, abomasal infusion of 6.25 g/d of choline; R12.5 = restricted intake, abomasal infusion of 12.5 g/d of choline; R25 = restricted intake, abomasal infusion of 25 g/d of choline.
      4 Contrasts: 1 = A0 compared with R0; Treatment = R0 compared with the average of the 3 treatments (6.25, 12.5, and 25 g/d of choline). Linear and quadratic effects of treatments R0, R6.25, R12.5, and R25.
      5 CDP = cytidine diphosphate.
      Hepatic TAG concentration was correlated negatively (Table 4) with activity of BHMT (r = −0.37; P = 0.02) and MTR (r = −0.42; P = 0.01). Activity of BHMT was also positively correlated with plasma free choline concentration (r = 0.42; P = 0.01). The activities of MTR and MAT were negatively correlated with each other (r = 0.47; P = 0.003).

      Cytidine Diphosphate-Choline Pathway

      The mRNA abundance results for the cytidine diphosphate (CDP)-choline pathway are reported in Table 7. The abundance of choline/ethanolamine phosphotransferase 1 (CEPT1) was lower with R (P = 0.01; CONT1). A negative quadratic effect (P = 0.001) was also observed for CEPT1. Abundance of choline kinase B (CHKB) increased with R (P = 0.02; CONT1). The abundance of phosphate cytidyltransferase 1, choline, b (PCYT1B) tended to be increased with R (P = 0.09; CONT1).

      Lipid Metabolism

      The effects of R and choline supplementation on abundance of genes related to lipid metabolism are presented in Table 7. Restriction increased the abundance (CONT1) of apolipoprotein A5 (APOA5; P = 0.001), CD36 molecule (CD36; P < 0.001), cAMP responsive element binding protein 1 (CREB1; P = 0.004), solute carrier family 22 member 5 (SLC22A5, choline transporter; P < 0.001), and tended to increase abundance of carnitine palmitoyltransferase 1A (CPT1A; P = 0.08). In contrast, R decreased abundance of apolipoprotein B (APOB; P = 0.002) and fatty acid binding protein 1 (FABP1; P < 0.001). A positive linear effect was observed for the abundance of APOA5 (P = 0.02; Y = 0.01006X + 1.0381). Increasing dose of choline induced a positive quadratic effect (P = 0.05; Y = −0.00067X2 + 0.01856X + 0.9615) in APOB, whereas there was a quadratic tendency for CPT1A to increase across dose (P = 0.09; Y = −0.00067X2 + 0.01968X + 0.9171). A quadratic effect was observed for CD36 (P = 0.01; Y = 0.001660X2 – 0.04225X + 1.1947); abundance decreased from R0 to R6.25, but then increased with R12.5 and R25. A quadratic effect was observed for CREB1 (P = 0.04; Y = 0.000823X2 – 0.02223X + 1.0469) where abundance decreased from R0 to R6.25 but increased with greater doses. Last, a quadratic effect (P = 0.01; Y = −0.00202X2 + 0.05910X + 0.8712) was observed for SLC22A5 and was greatest in R12.5 cows.

      DISCUSSION

      Energy Balance Biomarkers and Milk Production

      In the present study, the drop in BW, EB, and milk yield coupled with greater plasma concentrations of NEFA and BHB with R indicated that our feed-restriction model successfully induced a NEB. While choline supply increased milk yield in the present study, other studies with abomasal infusions of choline have been inconsistent;
      • Sharma B.K.
      • Erdman R.A.
      Effects of dietary and abomasally infused choline on milk production responses of lactating dairy cows..
      observed increases in milk yield when 40 and 50 g/d of choline was infused for 3-week periods, whereas infusions of 12.5 and 25 g/d for 5 d by
      • de Veth M.J.
      • Artegoitia V.M.
      • Campagna S.R.
      • Lapierre H.
      • Harte F.
      • Girard C.L.
      Choline absorption and evaluation of bioavailability markers when supplementing choline to lactating dairy cows..
      did not alter milk yield. Inconsistencies in results among studies might be related to differences in length of infusion, as well as the use of feed restriction in the present study. However, perhaps the most relevant data in the context of experimental design is that of
      • de Veth M.J.
      • Artegoitia V.M.
      • Campagna S.R.
      • Lapierre H.
      • Harte F.
      • Girard C.L.
      Choline absorption and evaluation of bioavailability markers when supplementing choline to lactating dairy cows..
      , where no differences in milk yield or composition were detected when delivering 12.5 or 25 g/d of choline postruminally to cows in late lactation. In the present study, DMI is eliminated as the driver of increasing milk yield, due to the lack of differences in DMI between the restricted diets.
      One explanation for the increase in milk yield with increasing choline could be Bet production. Betaine is produced when choline is oxidized in the mitochondria and can act as an osmolyte (
      • Eklund M.
      • Bauer E.
      • Wamatu J.
      • Mosenthin R.
      Potential nutritional and physiological functions of betaine in livestock..
      ). Several studies with dairy cattle have observed increases in milk yield with rumen-protected Bet supplementation (
      • Wang C.
      • Liu Q.
      • Yang W.Z.
      • Wu J.
      • Zhang W.W.
      • Zhang P.
      • Dong K.H.
      • Huang Y.X.
      Effects of betaine supplementation on rumen fermentation, lactation performance, feed digestibilities and plasma characteristics in dairy cows..
      ;
      • Peterson S.E.
      • Rezamand P.
      • Williams J.E.
      • Price W.
      • Chahine M.
      • McGuire M.A.
      Effects of dietary betaine on milk yield and milk composition of mid-lactation Holstein dairy cows..
      ). Postruminal delivery of choline was previously observed to increase the concentration of Bet in milk (
      • de Veth M.J.
      • Artegoitia V.M.
      • Campagna S.R.
      • Lapierre H.
      • Harte F.
      • Girard C.L.
      Choline absorption and evaluation of bioavailability markers when supplementing choline to lactating dairy cows..
      ) suggesting that Bet concentrations in milk likely increased in the present study as well. The fact that we detected a linear increase in hepatic Bet concentration with increasing choline supply (reported in a companion paper;
      • Coleman D.N.
      • Alharthi A.
      • Lopreiato V.
      • Trevisi E.
      • Miura M.
      • Pan Y.X.
      • Loor J.J.
      Choline supply during negative energy balance alters hepatic cystathionine β-synthase activity, intermediates of the methionine cycle and transsulfuration pathway, and liver function in Holstein cows..
      ) supports the idea that Bet production was increased with enhanced postruminal choline supply. Hence, an increase in Bet may have led to an increase in milk yield through its function as an osmolyte, pulling water to the mammary gland. A second potential explanation for the increase in milk yield could be the tendency for greater lactose yield with increasing choline dose. Lactose serves as an osmoregulator that controls milk volume (
      • Kuhn N.J.
      • Carrick D.T.
      • Wilde C.J.
      Lactose synthesis: The possibilities of regulation..
      ). Thus, an increase in lactose synthesis might have promoted increased milk yield. Further work is needed to understand the mechanisms by which choline (or its metabolites) might alter lactose synthesis.
      The tendency for increasing milk protein yield with increasing choline dose was due to the increase in milk production with choline. The fact that R12.5 had the highest FCM yield compared with R6.25 and R25 was also likely due to the greater milk fat yield with R12.5 compared with the other 2 choline treatments. Additionally, previous work by
      • Sharma B.K.
      • Erdman R.A.
      Effects of dietary and abomasally infused choline on milk production responses of lactating dairy cows..
      indicated that increases in milk fat with abomasal infusions of choline may be associated with increased NEFA mobilization and utilization for milk fat. Although plasma NEFA was not different between choline doses, it is possible that more NEFA were going toward milk fat with R12.5, driving the increase in milk fat and FCM yields.

      Methionine and Choline Metabolism

      The significant increase in plasma and milk free choline concentrations across doses of choline, and the positive relationship between them, indicated that our abomasal infusions were successful at delivering choline postruminally. The data agree with
      • de Veth M.J.
      • Artegoitia V.M.
      • Campagna S.R.
      • Lapierre H.
      • Harte F.
      • Girard C.L.
      Choline absorption and evaluation of bioavailability markers when supplementing choline to lactating dairy cows..
      when they infused 12.5 and 25 g/d to cows in late lactation. The decrease in liver TAG content with choline is in accordance with previous work where RPC was fed for longer periods of time (
      • Cooke R.F.
      • Silva Del Rio N.
      • Caraviello D.Z.
      • Bertics S.J.
      • Ramos M.H.
      • Grummer R.R.
      Supplemental choline for prevention and alleviation of fatty liver in dairy cattle..
      ;
      • Zom R.L.G.
      • van Baal J.
      • Goselink R.M.A.
      • Bakker J.A.
      • de Veth M.J.
      • van Vuuren A.M.
      Effect of rumen-protected choline on performance, blood metabolites, and hepatic triacylglycerols of periparturient dairy cattle..
      ;
      • Elek P.
      • Gaal T.
      • Husveth F.
      Influence of rumen-protected choline on liver composition and blood variables indicating energy balance in periparturient dairy cows..
      ). While liver TAG content decreased with choline in the present study, the mechanism behind this reduction is unclear. Choline is used to synthesize PC in the liver, which can be incorporated into VLDL and exported (
      • Vance J.E.
      Assembly and secretion of lipoproteins.
      ), potentially helping decrease TAG storage in the liver. While choline increased liver PC, the greatest increase was with R6.25 and not R12.5. However, this agrees with previous work in periparturient cows where liver PC increased with 12.5 g/d of choline relative to controls (
      • Zhou Z.
      • Garrow T.A.
      • Dong X.
      • Luchini D.N.
      • Loor J.J.
      Hepatic activity and transcription of betaine-homocysteine methyltransferase, methionine synthase, and cystathionine synthase in periparturient dairy cows are altered to different extents by supply of methionine and choline..
      ). In nonruminants, the changes in liver PC are associated with changes in the export of VLDL (
      • Yao Z.M.
      • Vance D.E.
      Reduction in VLDL, but not HDL, in plasma of rats deficient in choline..
      ); however, concentrations of plasma VLDL were not altered by choline in the present study. Because 12.5 g/d resulted in the greatest reduction in liver TAG, these changes along with the lack of differences in plasma PC suggest that enhanced secretion of VLDL was not the main mechanism by which choline decreased liver TAG. This idea is supported by the lack of difference in the abundance of hepatic PEMT, the enzyme that converts phosphatidylethanolamine to PC (
      • Zeisel S.H.
      Choline: An important nutrient in brain development, liver function and carcinogenesis..
      ), and is further indication that this was not the avenue by which liver PC increased.
      The CDP-choline pathway can also generate PC. Choline kinases convert choline to phosphocholine to begin the pathway (
      • Fagone P.
      • Jackowski S.
      Phosphatidylcholine and the CDP-choline cycle..
      ). In the present study, the lack of difference in the abundance of CHKA and CHKB with choline supply suggested that choline did not enhance the production of phosphocholine. This is in contrast to the study by
      • de Veth M.J.
      • Artegoitia V.M.
      • Campagna S.R.
      • Lapierre H.
      • Harte F.
      • Girard C.L.
      Choline absorption and evaluation of bioavailability markers when supplementing choline to lactating dairy cows..
      , where abomasal infusions of choline increased plasma phosphocholine concentrations; however, those cows were not in NEB. The next step in the pathway is the conversion of phosphocholine to CDP-choline by phosphate cytidylyltransferase enzymes (
      • Fagone P.
      • Jackowski S.
      Phosphatidylcholine and the CDP-choline cycle..
      ). The CDP-choline is then used to synthesize PC via CEPT1. Changes in PCYT1A and PCYT1B were inconsistent; abundance of PCYT1B tended to decrease with R, whereas PCYT1A did not change. The mRNA abundance of these 2 enzymes also did not change significantly in the same pattern with increasing choline dose, and abundance of PCYT1A was similar in R0, R12.5, and R25 cows. Thus, this suggests that CDP-choline production might have been enhanced with R via PCYT1B, but the effects of choline on CDP-choline production are not clear.
      The decrease in CEPT1 abundance with R suggests that PC synthesis from phosphocholine may not have been increased with R. However, these changes do not agree with the increase in liver PC observed with increased choline supply, which suggests that the changes in transcription may not have resulted in corresponding changes in translation or activity. These results also contrast with previous work in periparturient cows where hepatic expression of CHKA, PCYT1A, and PCYT1B increased with RPC supplementation at 12.5 g/d of choline (
      • Zhou Z.
      • Garrow T.A.
      • Dong X.
      • Luchini D.N.
      • Loor J.J.
      Hepatic activity and transcription of betaine-homocysteine methyltransferase, methionine synthase, and cystathionine synthase in periparturient dairy cows are altered to different extents by supply of methionine and choline..
      ). A recent in vitro study revealed that supplementing bovine hepatocytes with choline increased mRNA abundance of CEPT1, CHKA, CHKB, and PCYT1A (
      • Zhou Y.F.
      • Zhou Z.
      • Batistel F.
      • Martinez-Cortés I.
      • Pate R.T.
      • Luchini D.L.
      • Loor J.J.
      Methionine and choline supply alter transmethylation, transsulfuration, and cytidine 5′-diphosphocholine pathways to different extents in isolated primary liver cells from dairy cows..
      ). The differences in the mRNA changes between the present and aforementioned studies could be related to the longer supplementation period by
      • Zhou Z.
      • Garrow T.A.
      • Dong X.
      • Luchini D.N.
      • Loor J.J.
      Hepatic activity and transcription of betaine-homocysteine methyltransferase, methionine synthase, and cystathionine synthase in periparturient dairy cows are altered to different extents by supply of methionine and choline..
      ; −21 d through 30 d around parturition) or the use of cells in the in vitro study.
      The role of Met in dairy cows during NEB, not only for milk protein synthesis, but also for antioxidant production and as a methyl donor, was well established in nonruminants decades ago (
      • Finkelstein J.D.
      Methionine metabolism in mammals..
      ). Methionine can be used to synthesize choline via methyl group donation from Met to SAM (
      • Vance D.E.
      • Walkey C.J.
      • Cui Z.
      Phosphatidylethanolamine N-methyltransferase from liver..
      ). Compared with nonruminants, the amount of choline synthesized from Met is relatively high in ruminants because degradation of choline in the rumen limits its bioavailability (
      • Snoswell A.M.
      • Xue G.-P.
      Methyl group metabolism in sheep..
      ). In turn, choline can provide methyl groups for the synthesis of Met from homocysteine via BHMT (
      • Li Z.
      • Vance D.E.
      Phosphatidylcholine and choline homeostasis..
      ). However, in ruminants the activity of BHMT is lower than that of MTR (
      • Xue G.P.
      • Snoswell A.M.
      Comparative studies on the methionine synthesis in sheep and rat tissues..
      ). In the present study, the increase in mRNA abundance and activity of BHMT along with liver and plasma Met content underscored an important role for BHMT and choline for de novo synthesis of Met during periods of NEB. As such, cows that are unable to consume enough Met from the diet during periods of NEB could in part alleviate a shortfall of this EAA via the BHMT reaction. The tendency for greater activity of MTR in response to increasing supply of choline agrees with in vitro mRNA abundance of MTR generated in hepatocytes (
      • Chandler T.L.
      • White H.M.
      Choline and methionine differentially alter methyl carbon metabolism in bovine neonatal hepatocytes..
      ;
      • Zhou Y.F.
      • Zhou Z.
      • Batistel F.
      • Martinez-Cortés I.
      • Pate R.T.
      • Luchini D.L.
      • Loor J.J.
      Methionine and choline supply alter transmethylation, transsulfuration, and cytidine 5′-diphosphocholine pathways to different extents in isolated primary liver cells from dairy cows..
      ). Together, those data indicate a role of choline in increasing Met synthesis via MTR. Additionally, the increases in liver and plasma Met with increasing choline supply provide further evidence that Met synthesis was increased by choline during a feed restriction-induced NEB. Overall, our data seem to underscore the importance of Met synthesis during times of NEB and suggest a greater reliance on BHMT during those times.
      Methionine adenosyltransferases convert Met to SAM (
      • Corrales F.J.
      • Pérez-Mato I.
      • Sánchez del Pino M.M.
      • Ruiz F.L.
      • Castro C.
      • García-Trevijano E.R.
      • Latasa U.
      • Martínez-Chantar M.L.
      • Martínez-Cruz A.
      • Avila M.A.
      • Mato J.M.
      Regulation of mammalian liver methionine adenosyltransferase..
      ). Although the activity of MAT in the present study was greater than the activities of BHMT and MTR, it is in line with previous work in sheep by
      • Radcliffe B.C.
      • Egan A.
      The effect of diet and of methionine loading on activity of enzymes in the transulfuration pathway in sheep..
      . The fact that R0 had the greatest MAT activity is in accordance with previous work where fasting was observed to increase the activity of MAT in the liver of mice (
      • Sakata S.F.
      • Okumura S.
      • Matsuda K.
      • Horikawa Y.
      • Maeda M.
      • Kawasaki K.
      • Chou J.Y.
      • Tamaki N.
      Effect of fasting on methionine adenosyltransferase expression and the methionine cycle in the mouse liver..
      ) and sheep (
      • Radcliffe B.C.
      • Egan A.
      The effect of diet and of methionine loading on activity of enzymes in the transulfuration pathway in sheep..
      ). In nonruminant liver, increased Met concentrations lead to greater SAM production via MAT (
      • Corrales F.J.
      • Pérez-Mato I.
      • Sánchez del Pino M.M.
      • Ruiz F.L.
      • Castro C.
      • García-Trevijano E.R.
      • Latasa U.
      • Martínez-Chantar M.L.
      • Martínez-Cruz A.
      • Avila M.A.
      • Mato J.M.
      Regulation of mammalian liver methionine adenosyltransferase..
      ). However, the linear increase in hepatic Met with increasing choline dose was associated with a cubic, rather than linear, change in MAT activity with activity being greatest in R12.5 cows, rather than R25. In the study by
      • Radcliffe B.C.
      • Egan A.
      The effect of diet and of methionine loading on activity of enzymes in the transulfuration pathway in sheep..
      , abomasal infusions of Met decreased MAT activity, suggesting that beyond a certain threshold increasing availability of cellular Met depresses MAT activity in ruminants. Thus, the fact that R25 cows had the lowest MAT activity could be related to them having the greatest hepatic Met concentrations. Additionally, homocysteine can be synthesized from SAH via SAHH. Although liver homocysteine could not be detected in the present study, suggesting rapid utilization, liver SAH concentration was lowest in R12.5 cows, suggesting a greater production of homocysteine.

      Lipid Metabolism

      When NEFA are taken up by the liver via CD36, they are preferentially stored in hepatocytes (
      • Loor J.J.
      • Bertoni G.
      • Hosseini A.
      • Roche J.R.
      • Trevisi E.
      Functional welfare—Using biochemical and molecular technologies to understand better the welfare state of peripartal dairy cattle..
      ). Thus, increases in the abundance of CD36 with R provide a link between NEFA uptake and TAG accumulation in the present study. Although PPARA was not altered, the changes in APOA5, SLC22A5, and CPT1A suggest greater PPARα activity in response to choline supply during R. These genes have been established as PPARα targets in nonruminant tissues, and evidence indicates that they are also targets in ruminants (
      • Bionaz M.
      • Chen S.
      • Khan M.J.
      • Loor J.J.
      Functional role of PPARs in ruminants: Potential targets for fine-tuning metabolism during growth and lactation..
      ). Because the activity of CPT1A is partly regulated at a transcriptional level in nonruminants (
      • Nakamura M.T.
      • Yudell B.E.
      • Loor J.J.
      Regulation of energy metabolism by long-chain fatty acids..
      ), the greater abundance of CPT1A (a gene involved in fatty acid oxidation) suggested an increase in mitochondrial β-oxidation. The increase in abundance of SLC22A5 mRNA with choline is in agreement with previous work in periparturient cows (
      • Goselink R.M.A.
      • van Baal J.
      • Widjaja H.C.A.
      • Dekker R.A.
      • Zom R.L.G.
      • de Veth M.J.
      • van Vuuren A.M.
      Effect of rumen-protected choline supplementation on liver and adipose gene expression during the transition period in dairy cattle..
      ). This transporter is associated with carnitine transport into cells, which has been observed to reduce liver TAG in dairy cattle through stimulation of FA oxidation (
      • Carlson D.B.
      • McFadden J.W.
      • D’Angelo A.
      • Woodworth J.C.
      • Drackley J.K.
      Dietary L-carnitine affects periparturient nutrient metabolism and lactation in multiparous cows..
      ).
      The mechanism behind these changes may be related to increased availability of methyl donors in the liver with enhanced choline supply. For instance, in the liver of mice increased flux through BHMT has been associated with altered DNA methylation of the promoter region of PPARA, resulting in increased abundance of CPT1A (
      • Wang L.
      • Chen L.
      • Tan Y.
      • Wei J.
      • Chang Y.
      • Jin T.
      • Zhu H.
      Betaine supplement alleviates hepatic triglyceride accumulation of apolipoprotein E deficient mice via reducing methylation of peroxisomal proliferator-activated receptor alpha promoter..
      ). In human hepatocytes, supplementation of choline induced a similar effect, upregulating CPT1A and downregulating fatty acid synthase (
      • Zhu J.
      • Wu Y.
      • Tang Q.
      • Leng Y.
      • Cai W.
      The effects of choline on hepatic lipid metabolism, mitochondrial function and antioxidative status in human hepatic C3A cells exposed to excessive energy substrates..
      ). Thus, even though PPARA abundance was not altered, increased synthesis of Met could have led to changes in DNA methylation, which may have mediated the increase in CPT1A. This potential mechanism provides a possible link for the negative relationship between BHMT and MTR activity and liver TAG. Further work is needed to fully elucidate the mechanisms behind the potential effects of choline on fatty acid metabolism in ruminants.
      The protein APOA5 is associated with VLDL and high-density lipoproteins, and (at least in nonruminants) is involved in lowering hepatic TAG content (
      • Ress C.
      • Moschen A.R.
      • Sausgruber N.
      • Tschoner A.
      • Graziadei I.
      • Weiss H.
      • Schgoer W.
      • Ebenbichler C.F.
      • Konrad R.J.
      • Patsch J.R.
      • Tilg H.
      • Kaser S.
      The role of apolipoprotein A5 in non-alcoholic fatty liver disease..
      ). While apolipoproteins are not closely regulated at the transcriptional level, the linear increase in APOA5 in response to increasing choline supply during R may be associated with the reduction in liver TAG. In addition, the quadratic increase in APOB (also related to VLDL production) with choline supply suggests an increase in VLDL assembly. Similar changes have been reported with long-term RPC supplementation (
      • Goselink R.M.A.
      • van Baal J.
      • Widjaja H.C.A.
      • Dekker R.A.
      • Zom R.L.G.
      • de Veth M.J.
      • van Vuuren A.M.
      Effect of rumen-protected choline supplementation on liver and adipose gene expression during the transition period in dairy cattle..
      ). Despite these molecular changes, choline supply did not alter plasma concentrations of VLDL and PC, or abundance of microsomal triglyceride transfer protein (MTTP). Therefore, it is unclear whether VLDL export was increased with enhance choline supply during our feed restriction-induced NEB. Previous work with bovine neonatal hepatocytes incubated with NEFA revealed an increase in VLDL with increasing choline dose, but this was not associated with a change in MTTP (
      • Chandler T.L.
      • White H.M.
      Choline and methionine differentially alter methyl carbon metabolism in bovine neonatal hepatocytes..
      ). Thus, despite a lack of effect on MTTP abundance, this suggests that VLDL export might have increased. Additionally, it should be noted that the lack of changes in plasma PC and VLDL in response to choline supply could have been due utilization by other tissues (e.g., the mammary gland); PC is one of the main choline metabolites in bovine milk (
      • Artegoitia V.M.
      • Middleton J.L.
      • Harte F.M.
      • Campagna S.R.
      • de Veth M.J.
      Choline and choline metabolite patterns and associations in blood and milk during lactation in dairy cows..
      ) and long-chain fatty acids in milk can be obtained from VLDL (
      • Palmquist D.L.
      • Mattos W.
      Turnover of lipoproteins and transfer to milk fat of dietary (1-carbon-14) linoleic acid in lactating cows..
      ;
      • Loor J.J.
      • Bandara A.B.
      • Herbein J.H.
      Characterization of 18:1 and 18:2 isomers produced during microbial biohydrogenation of unsaturated fatty acids from canola and soya bean oil in the rumen of lactating cows..
      ).

      Limitations

      While the use of a feed restriction-induced NEB provides the ability to mimic the DMI-induced NEB of the periparturient period, such a model does not fully mimic the NEB experienced by peripartal cows. This is because the DMI-induced NEB associated with the peripartal period lasts longer than 4 d and is also associated with the physical and hormonal stress of calving. Compared with cows in early lactation, use of cows in late lactation in the present study may also be a limitation due to the differences in milk production and metabolism. Last, although this study provides critical information on how different doses of choline alter milk production and hepatic metabolism, it was not designed to collect data on portal circulation and flux. Thus, to fully understand how different doses of choline alter metabolism, collection of such data should be considered in future studies.

      CONCLUSIONS

      Postruminal choline supply beyond the suggested 12.5 g/d during a feed restriction-induced NEB benefits milk production and liver fatty acid metabolism. While liver PC content and hepatic apoprotein mRNA abundance increased, no statistical changes were observed in plasma VLDL or PC. Transcriptional changes also suggest a role of choline in increasing fatty acid oxidation to further reduce liver TAG accumulation during a NEB. Beyond lipid metabolism, adaptations in the activity of BHMT and MTR are important to maintain Met synthesis during a NEB, especially when the supply of choline increases. This observation underscores the importance of choline supply when dairy cows experience periods of NEB.

      ACKNOWLEDGMENTS

      Ahmed A. Elolimy is a recipient of a PhD fellowship from Missions Sector, Higher Education Ministry, Egypt, to perform his PhD studies at the University of Illinois (Urbana).

      Supplementary Material

      REFERENCES

        • Artegoitia V.M.
        • Middleton J.L.
        • Harte F.M.
        • Campagna S.R.
        • de Veth M.J.
        Choline and choline metabolite patterns and associations in blood and milk during lactation in dairy cows..
        PLoS One. 2014; 9 (25157578)e103412
        • Batistel F.
        • Alharthi A.S.
        • Yambao R.R.C.
        • Elolimy A.A.
        • Pan Y.X.
        • Parys C.
        • Loor J.J.
        Methionine supply during late-gestation triggers offspring sex-specific divergent changes in metabolic and epigenetic signatures in bovine placenta..
        J. Nutr. 2019; 149: 6-17https://doi.org/10.1093/jn/nxy240
        • Batistel F.
        • Arroyo J.M.
        • Bellingeri A.
        • Wang L.
        • Saremi B.
        • Parys C.
        • Trevisi E.
        • Cardoso F.C.
        • Loor J.J.
        Ethyl-cellulose rumen-protected methionine enhances performance during the periparturient period and early lactation in Holstein dairy cows..
        J. Dairy Sci. 2017; 100: 7455-7467https://doi.org/10.3168/jds.2017-12689
        • Bauchart D.
        • Gruffat D.
        • Durand D.
        Lipid absorption and hepatic metabolism in ruminants..
        Proc. Nutr. Soc. 1996; 55 (8832781): 39-47
        • Bionaz M.
        • Chen S.
        • Khan M.J.
        • Loor J.J.
        Functional role of PPARs in ruminants: Potential targets for fine-tuning metabolism during growth and lactation..
        PPAR Res. 2013; 2013 (23737762)684159
        • Carlson D.B.
        • McFadden J.W.
        • D’Angelo A.
        • Woodworth J.C.
        • Drackley J.K.
        Dietary L-carnitine affects periparturient nutrient metabolism and lactation in multiparous cows..
        J. Dairy Sci. 2007; 90 (17582127): 3422-3441
        • Chandler T.L.
        • White H.M.
        Choline and methionine differentially alter methyl carbon metabolism in bovine neonatal hepatocytes..
        PLoS One. 2017; 12 (28152052)e0171080
        • Coleman D.N.
        • Alharthi A.
        • Lopreiato V.
        • Trevisi E.
        • Miura M.
        • Pan Y.X.
        • Loor J.J.
        Choline supply during negative energy balance alters hepatic cystathionine β-synthase activity, intermediates of the methionine cycle and transsulfuration pathway, and liver function in Holstein cows..
        J. Dairy Sci. 2019; 102: 8319-8331https://doi.org/10.3168/jds.2019-16406
        • Cooke R.F.
        • Silva Del Rio N.
        • Caraviello D.Z.
        • Bertics S.J.
        • Ramos M.H.
        • Grummer R.R.
        Supplemental choline for prevention and alleviation of fatty liver in dairy cattle..
        J. Dairy Sci. 2007; 90 (17430945): 2413-2418
        • Corrales F.J.
        • Pérez-Mato I.
        • Sánchez del Pino M.M.
        • Ruiz F.L.
        • Castro C.
        • García-Trevijano E.R.
        • Latasa U.
        • Martínez-Chantar M.L.
        • Martínez-Cruz A.
        • Avila M.A.
        • Mato J.M.
        Regulation of mammalian liver methionine adenosyltransferase..
        J. Nutr. 2002; 132 (12163696): 2377S-2381S
        • de Veth M.J.
        • Artegoitia V.M.
        • Campagna S.R.
        • Lapierre H.
        • Harte F.
        • Girard C.L.
        Choline absorption and evaluation of bioavailability markers when supplementing choline to lactating dairy cows..
        J. Dairy Sci. 2016; 99 (27771079): 9732-9744
        • Du X.
        • Shen T.
        • Wang H.
        • Qin X.
        • Xing D.
        • Ye Q.
        • Shi Z.
        • Fang Z.
        • Zhu Y.
        • Yang Y.
        • Peng Z.
        • Zhao C.
        • Lv B.
        • Li X.
        • Liu G.
        • Li X.
        Adaptations of hepatic lipid metabolism and mitochondria in dairy cows with mild fatty liver..
        J. Dairy Sci. 2018; 101 (30100495): 9544-9558
        • Eklund M.
        • Bauer E.
        • Wamatu J.
        • Mosenthin R.
        Potential nutritional and physiological functions of betaine in livestock..
        Nutr. Res. Rev. 2005; 18 (19079893): 31-48
        • Elek P.
        • Gaal T.
        • Husveth F.
        Influence of rumen-protected choline on liver composition and blood variables indicating energy balance in periparturient dairy cows..
        Acta Vet. Hung. 2013; 61 (23439292): 59-70
        • Fagone P.
        • Jackowski S.
        Phosphatidylcholine and the CDP-choline cycle..
        Biochim. Biophys. Acta. 2013; 1831: 523-532
        • Fast D.G.
        • Vance D.E.
        Nascent VLDL phospholipid composition is altered when phosphatidylcholine biosynthesis is inhibited: Evidence for a novel mechanism that regulates VLDL secretion..
        Biochim. Biophys. Acta. 1995; 1258 (7548179): 159-168
        • Finkelstein J.D.
        Methionine metabolism in mammals..
        J. Nutr. Biochem. 1990; 1 (15539209): 228-237
        • Goselink R.M.A.
        • van Baal J.
        • Widjaja H.C.A.
        • Dekker R.A.
        • Zom R.L.G.
        • de Veth M.J.
        • van Vuuren A.M.
        Effect of rumen-protected choline supplementation on liver and adipose gene expression during the transition period in dairy cattle..
        J. Dairy Sci. 2013; 96 (23200476): 1102-1116
        • Hartwell J.R.
        • Cecava M.J.
        • Donkin S.S.
        Impact of dietary rumen undegradable protein and rumen-protected choline on intake, peripartum liver triacylglyceride, plasma metabolites and milk production in transition dairy cows..
        J. Dairy Sci. 2000; 83 (11132863): 2907-2917
        • Hutjens M.F.
        Benchmarking your feed efficiency, feed costs and income over feed cost.
        Proc. Western Canadian Dairy Seminar. Red Deer, Alberta Canada. University of Alberta, Edmonton, Alberta, Canada2010 (Pages 3–10 in)
        • Johnson M.M.
        • Peters J.
        Technical note: An improved method to quantify nonesterified fatty acids in bovine plasma..
        J. Anim. Sci. 1993; 71 (8463162): 753-756
        • Khan M.J.
        • Jacometo C.B.
        • Graugnard D.E.
        • Corrêa M.N.
        • Schmitt E.
        • Cardoso F.
        • Loor J.J.
        Overfeeding dairy cattle during late-pregnancy alters hepatic PPARα-regulated pathways including hepatokines: Impact on metabolism and peripheral insulin sensitivity..
        Gene Regul. Syst. Bio. 2014; 8: 97-111https://doi.org/10.4137/GRSB.S14116
        • Kuhn N.J.
        • Carrick D.T.
        • Wilde C.J.
        Lactose synthesis: The possibilities of regulation..
        J. Dairy Sci. 1980; 63 (6766957): 328-336
        • Li X.
        • Huang W.
        • Gu J.
        • Du X.
        • Lei L.
        • Yuan X.
        • Sun G.
        • Wang Z.
        • Li X.
        • Liu G.
        SREBP-1c overactivates ROS-mediated hepatic NF-κB inflammatory pathway in dairy cows with fatty liver..
        Cell. Signal. 2015; 27 (26189441): 2099-2109
        • Li Z.
        • Vance D.E.
        Phosphatidylcholine and choline homeostasis..
        J. Lipid Res. 2008; 49 (18204095): 1187-1194
        • Loor J.J.
        • Bandara A.B.
        • Herbein J.H.
        Characterization of 18:1 and 18:2 isomers produced during microbial biohydrogenation of unsaturated fatty acids from canola and soya bean oil in the rumen of lactating cows..
        J. Anim. Physiol. Anim. Nutr. (Berl.). 2002; 86: 422-432
        • Loor J.J.
        • Bertoni G.
        • Hosseini A.
        • Roche J.R.
        • Trevisi E.
        Functional welfare—Using biochemical and molecular technologies to understand better the welfare state of peripartal dairy cattle..
        Anim. Prod. Sci. 2013; 53 (a): 931-953https://doi.org/10.1071/AN12344
        • Loor J.J.
        • Bionaz M.
        • Drackley J.K.
        Systems physiology in dairy cattle: Nutritional genomics and beyond..
        Annu. Rev. Anim. Biosci. 2013; 1 (b): 365-392https://doi.org/10.1146/annurev-animal-031412-103728
        • Moyes K.M.
        • Drackley J.K.
        • Morin D.E.
        • Loor J.J.
        Greater expression of TLR2, TLR4, and IL6 due to negative energy balance is associated with lower expression of HLA-DRA and HLA-A in bovine blood neutrophils after intramammary mastitis challenge with Streptococcus uberis..
        Funct. Integr. Genomics. 2010; 10: 53-61https://doi.org/10.1007/s10142-009-0154-7
        • Moyes K.M.
        • Drackley J.K.
        • Salak-Johnson J.L.
        • Morin D.E.
        • Hope J.C.
        • Loor J.J.
        Dietary-induced negative energy balance has minimal effects on innate immunity during a Streptococcus uberis mastitis challenge in dairy cows during midlactation..
        J. Dairy Sci. 2009; 92 (19700690): 4301-4316
        • Nakamura M.T.
        • Yudell B.E.
        • Loor J.J.
        Regulation of energy metabolism by long-chain fatty acids..
        Prog. Lipid Res. 2014; 53 (24362249): 124-144
        • NRC
        Nutrient Requirements of Dairy Cattle.
        7th rev. The National Academies Press, Washington, DC2001 (ed)
        • Osorio J.S.
        • Ji P.
        • Drackley J.K.
        • Luchini D.
        • Loor J.J.
        Smartamine M and MetaSmart supplementation during the peripartal period alter hepatic expression of gene networks in 1-carbon metabolism, inflammation, oxidative stress, and the growth hormone–insulin-like growth factor 1 axis pathways..
        J. Dairy Sci. 2014; 97 (25282416): 7451-7464
        • Palmquist D.L.
        • Mattos W.
        Turnover of lipoproteins and transfer to milk fat of dietary (1-carbon-14) linoleic acid in lactating cows..
        J. Dairy Sci. 1978; 61: 561-565
        • Peterson S.E.
        • Rezamand P.
        • Williams J.E.
        • Price W.
        • Chahine M.
        • McGuire M.A.
        Effects of dietary betaine on milk yield and milk composition of mid-lactation Holstein dairy cows..
        J. Dairy Sci. 2012; 95 (22981566): 6557-6562
        • Piepenbrink M.S.
        • Overton T.R.
        Liver metabolism and production of cows fed increasing amounts of rumen-protected choline during the periparturient period..
        J. Dairy Sci. 2003; 86 (12778583): 1722-1733
        • Radcliffe B.C.
        • Egan A.
        The effect of diet and of methionine loading on activity of enzymes in the transulfuration pathway in sheep..
        Aust. J. Biol. Sci. 1978; 31 (678217): 105-114
        • Ress C.
        • Moschen A.R.
        • Sausgruber N.
        • Tschoner A.
        • Graziadei I.
        • Weiss H.
        • Schgoer W.
        • Ebenbichler C.F.
        • Konrad R.J.
        • Patsch J.R.
        • Tilg H.
        • Kaser S.
        The role of apolipoprotein A5 in non-alcoholic fatty liver disease..
        Gut. 2011; 60 (21339203): 985-991
        • Sakata S.F.
        • Okumura S.
        • Matsuda K.
        • Horikawa Y.
        • Maeda M.
        • Kawasaki K.
        • Chou J.Y.
        • Tamaki N.
        Effect of fasting on methionine adenosyltransferase expression and the methionine cycle in the mouse liver..
        J. Nutr. Sci. Vitaminol. (Tokyo). 2005; 51 (16022199): 118-123
        • Sharma B.K.
        • Erdman R.A.
        Effects of dietary and abomasally infused choline on milk production responses of lactating dairy cows..
        J. Nutr. 1989; 119 (2918397): 248-254
        • Snoswell A.M.
        • Xue G.-P.
        Methyl group metabolism in sheep..
        Comp. Biochem. Physiol. B. 1987; 88 (3322658): 383-394
        • Vailati Riboni M.
        • Meier S.
        • Priest N.V.
        • Burke C.R.
        • Kay J.K.
        • McDougall S.
        • Mitchell M.D.
        • Walker C.G.
        • Crookenden M.
        • Heiser A.
        • Roche J.R.
        • Loor J.J.
        Adipose and liver gene expression profiles in response to treatment with a nonsteroidal antiinflammatory drug after calving in grazing dairy cows..
        J. Dairy Sci. 2015; 98 (25771063): 3079-3085
        • Vance D.E.
        • Walkey C.J.
        • Cui Z.
        Phosphatidylethanolamine N-methyltransferase from liver..
        Biochim. Biophys. Acta. 1997; 1348 (9370326): 142-150
        • Vance J.E.
        Assembly and secretion of lipoproteins.
        New Compr. Biochem.36. Elsevier, 2002 (Pages 505–526 in Vol)
        • Wang C.
        • Liu Q.
        • Yang W.Z.
        • Wu J.
        • Zhang W.W.
        • Zhang P.
        • Dong K.H.
        • Huang Y.X.
        Effects of betaine supplementation on rumen fermentation, lactation performance, feed digestibilities and plasma characteristics in dairy cows..
        J. Agric. Sci. 2010; 148: 487-495
        • Wang L.
        • Chen L.
        • Tan Y.
        • Wei J.
        • Chang Y.
        • Jin T.
        • Zhu H.
        Betaine supplement alleviates hepatic triglyceride accumulation of apolipoprotein E deficient mice via reducing methylation of peroxisomal proliferator-activated receptor alpha promoter..
        Lipids Health Dis. 2013; 12 (23497035): 34
        • Wu H.
        • Southam A.D.
        • Hines A.
        • Viant M.R.
        High-throughput tissue extraction protocol for NMR- and MS-based metabolomics..
        Anal. Biochem. 2008; 372 (17963684): 204-212
        • Xue G.P.
        • Snoswell A.M.
        Comparative studies on the methionine synthesis in sheep and rat tissues..
        Comp. Biochem. Physiol. B. 1985; 80 (4006442): 489-494
        • Yao Z.M.
        • Vance D.E.
        Reduction in VLDL, but not HDL, in plasma of rats deficient in choline..
        Biochem. Cell Biol. 1990; 68 (2344402): 552-558
        • Zahra L.C.
        • Duffield T.F.
        • Leslie K.E.
        • Overton T.R.
        • Putnam D.
        • LeBlanc S.J.
        Effects of rumen-protected choline and monensin on milk production and metabolism of periparturient dairy cows..
        J. Dairy Sci. 2006; 89: 4808-4818
        • Zeisel S.H.
        Choline: An important nutrient in brain development, liver function and carcinogenesis..
        J. Am. Coll. Nutr. 1992; 11 (1452945): 473-481
        • Zhou Y.F.
        • Zhou Z.
        • Batistel F.
        • Martinez-Cortés I.
        • Pate R.T.
        • Luchini D.L.
        • Loor J.J.
        Methionine and choline supply alter transmethylation, transsulfuration, and cytidine 5′-diphosphocholine pathways to different extents in isolated primary liver cells from dairy cows..
        J. Dairy Sci. 2018; 101 (30316602): 11384-11395https://doi.org/10.3168/jds.2017-14236
        • Zhou Z.
        • Garrow T.A.
        • Dong X.
        • Luchini D.N.
        • Loor J.J.
        Hepatic activity and transcription of betaine-homocysteine methyltransferase, methionine synthase, and cystathionine synthase in periparturient dairy cows are altered to different extents by supply of methionine and choline..
        J. Nutr. 2017; 147 (27881594): 11-19
        • Zhou Z.
        • Vailati-Riboni M.
        • Trevisi E.
        • Drackley J.K.
        • Luchini D.N.
        • Loor J.J.
        Better postpartal performance in dairy cows supplemented with rumen-protected methionine compared with choline during the peripartal period..
        J. Dairy Sci. 2016; 99 (27638261): 8716-8732
        • Zhu J.
        • Wu Y.
        • Tang Q.
        • Leng Y.
        • Cai W.
        The effects of choline on hepatic lipid metabolism, mitochondrial function and antioxidative status in human hepatic C3A cells exposed to excessive energy substrates..
        Nutrients. 2014; 6 (25010553): 2552
        • Zhu Y.
        • Liu G.
        • Du X.
        • Shi Z.
        • Jin M.
        • Sha X.
        • Li X.
        • Wang Z.
        • Li X.
        Expression patterns of hepatic genes involved in lipid metabolism in cows with subclinical or clinical ketosis..
        J. Dairy Sci. 2019; 102 (30471902): 1725-1735
        • Zom R.L.G.
        • van Baal J.
        • Goselink R.M.A.
        • Bakker J.A.
        • de Veth M.J.
        • van Vuuren A.M.
        Effect of rumen-protected choline on performance, blood metabolites, and hepatic triacylglycerols of periparturient dairy cattle..
        J. Dairy Sci. 2011; 94 (21787937): 4016-4027

      Linked Article