Effect of weaning age and pace on blood metabolites, cortisol concentration, and the mRNA abundance of inflammation-related genes in gastrointestinal, adipose, and liver tissue of Holstein dairy calves

The objective of this study was to evaluate the effects of weaning age and pace on blood metabolites, cortisol concentration, and mRNA abundance of inflammation-related genes in Holstein dairy calves. Seventy-one day-old calves [38.8 ± 4.4 kg, body weight ( BW ) ± sd] blocked by gender and birth BW, were randomly assigned to a 2 × 2 factorial arrangement of treatments. The first factor was weaning age [6 weeks (early) vs . Eight weeks (late)], the second factor was weaning pace [abrupt (4 step-down over 3 d, the initial milk replacer was 7.6 L, which was reduced 1.9 L in each step-down) vs . gradual (7 step-down over 14 d, the initial milk re-placer was 7.6 L, which was reduced 1.09 L in each step-down)], generating early-abrupt ( EA ), early-gradual ( EG ), late-abrupt ( LA ), and late-gradual ( LG ) treatments. All treatments had 10 female and 8 male calves, except EA that had 1 fewer male calf. Milk replacer ( MR ; 24% CP, 17% fat) was bottle-fed, up to 1,200 g/d, twice daily (0600h and 1800h). EA and EG calves received 46.2 kg MR while LA and LG calves received a total of 63 kg MR. The study had 2 cohorts (2020, n = 40; 2021, n = 31), and each cohort included all treatments. Blood was collected from the jugular vein at 0900h on d 3 and d 7 of age, a day before starting and a day after weaning completion; male calves were humanely killed a day post-weaning. Rumen, jejunum, large intestine, liver, omental adipose and perirenal adipose tissues were sampled to determine the mRNA abundance of inflammation-related genes. Weaning pace, age, and pace × age, birth BW, and sex were included as fixed and cohort was included as random effects in the model. Blood metabolites and cortisol were analyzed as repeated measure, and sampling day, pace × sampling day, and age × sampling day were also included as additional fixed effects. Significance were noted at P ≤ 0.05 and tendencies when 0.05 < P ≤ 0.10. EA calves showed a tendency to have the greatest non-esterified fatty acid ( NEFA ) concentration compared with all other treatments. There was a pace × day effect on serum NEFA and β-hydroxybutyrate ( BHBA ); calves weaned in abrupt pace had an increased NEFA post-weaning compared with that for gradual weaning. Calves weaned in the gradual pace showed the greatest serum BHBA post-weaning. Most mRNA abundance of inflammation-related genes affected by treatments showed a similar pattern; downregulated by the abrupt (liver IL-1β) and early weaning (jejunum TNF-α and ICAM), and in some cases the interaction intensified the effect, demonstrating a weakened immune response in calves experiencing more stressful conditions (EA: IL-6 in the liver and NF-κB in the perirenal adipose tissue). Overall, the downregulation of mRNA abundance of inflammation-related genes in EA calves may be attributed to the suppression of the immune system and an immature immune response. Furthermore, the greater NEFA in EA calves could be attributed to a reduced starter intake, less developed rumen, or shorter time during the weaning transition.


INTRODUCTION
Weaning is one of the more challenging/stressful events in dairy calves.Managed poorly, weaning stress can adversely affect several growth and health indicators such as feed intake, average daily gain, mammary mass, and indicators of stress response and negative energy balance such as vocalization and blood nonesterified fatty acids (NEFA) (Eckert et al., 2015;Steele et al., 2017;McCoard et al., 2019).According to the USDA report (USDA, 2021), approximately 40% of calves have at least one disease occurrence in their early life (from birth to weaning), of which approximately 50% are digestive illnesses, such as diarrhea and bloat.Together, this highlights the impact of poor immune function and capacity in young calves, and the need to develop low stress weaning strategies.
Weaning is a multi-factorial management process; principal among them are weaning pace and weaning age.Abrupt weaning includes a reduced number of days with little or no step-down (reduction of milk replacer), whereas gradual weaning consists of weaning animals with several step-downs over several days (Hickey et al., 2003;Steele et al., 2017;Wickramasinghe et al., 2022).Another important factor is the age at weaning, with a range in weaning age from 6 to 10 weeks for most US dairy producers (USDA, 2021).Weaning calves at an early age is a weaning strategy to reduce the cost of milk replacer programs as well as labor costs.Conversely, weaning calves at a later age can favor the animal's performance (Eckert et al., 2015).Gradual weaning over a 2-week period improves digestible energy intake of calves (Eckert et al., 2015), but the benefits of gradual weaning disappear if weaning begins before 4 weeks of age (Sweeney et al., 2010).Evidence shows the impact of weaning pace on digestible energy intake and growth changes as calves' age, with unknown impacts on weaning success and calf related metabolic challenges.
The relative contributions of age and pace of weaning to stress response and immune function in the dairy calf remain understudied.When animals are exposed to oxidative stress from abrupt weaning, there might be downregulation of some of the immune response in dairy calves (Lynch et al., 2010) including depressed neutrophil phagocytic and oxidative capacity (Lynch et al., 2010;Hulbert et al., 2011).Neutrophils play a major role as the first line of cellular defense against pathogens; therefore, decreasing their defense potential can put dairy calves in a weakened position to fight against pathogens, making the animals more susceptible to infections (Kelley, 1980).Whether the sensitivity to abrupt weaning decreases as calves mature is unclear.
Developing weaning strategies that mitigate accentuated challenges/stress response in dairy calves is needed.
To date, much literature on weaning focuses on either age or pace, leaving uncertainty as to how age and pace interact to affect inflammation and blood metabolites associated with animal health.We hypothesized that combining early and abrupt would put more stress on the animals than each factor individually.When evaluating the same weaning strategies at 2 ages, we hypothesized that calves would be more resilient to abrupt weaning stress at 8 weeks of age than calves were at 6 weeks of age.Therefore, the objective of this study was to evaluate the effects of weaning age [6 weeks (early) vs. Eight weeks (late)] and pace (abrupt vs. gradual) on selected blood metabolites, cortisol concentration, and mRNA abundance of inflammation-related genes in Holstein dairy calves.

Animals and Experimental Design
All animal procedures were approved by the Animal Care and Use Committee at the University of Idaho (#2020-12; #2021-51) and the University of Alberta (#3857).The experiment was conducted on the Palouse Research, Extension and Education Center -Dairy Unit at the University of Idaho (Moscow ID 83844).
One-day-old Holstein calves (n = 71, 38.8 ± 4.37 kg, body weight ± standard deviation) were blocked by sex and body weight at birth (light <38.5 kg and heavy ≥38.6 kg, for both male and female calves) and then randomly assigned to a 2 × 2 factorial arrangement of treatments, using PROC PLAN in SAS (version 9.4, SAS Institute Inc., Cary, NC).The first factor was weaning age (early vs. late), the second factor was weaning pace (abrupt vs. gradual), generating 4 treatments including early-abrupt (EA), early-gradual (EG), late-abrupt (LA), and late-gradual (LG).Each treatment consisted of 18 calves, except the treatment EG which had 17 calves that resulted from a calf health issue unrelated to the treatment.The treatments EA, LA, and LG consisted of 10 females and 8 males, and the treatment EG consisted of 10 females and 7 males.The study took place between September and November 2020 (n = 40 calves) and between October and December 2021 (n = 31 calves), constituting 2 cohorts.Two cohorts were used because of the space limitation of our facilities that were capable of fitting 40 animals in individual hutches at the same period of time.

Calf Management
All calves received colostrum replacer (Perfect Udder 4 Liter, DairyTech Inc., Windsor, CO) in the first hour of life, as soon as the calves were unloaded from the truck.The bags were tubed directly to the calf upon arrival.Serum total proteins (TP) were measured using a refractometer (Hernandez et al., 2016), at approximately 48 h of life, to ensure an effective transfer of passive immunity (TPI; total protein = 5.83 ± 0.94 g/ dL).According to the proposed categories suggested by Godden et al. (2019), 23 calves presented TP ≥6.2 (TPI category = excellent); 17 calves presented TP between 5.8 and 6.1 (TPI category = good); 12 calves presented TP between 5.1 and 5.7 (TPI category = fair); and 19 calves presented TP between <5.1 (TPI category = poor); where 3 calves presented TP lower than 4.6.Calves were housed individually in covered hutches (PolyDome -Model PD-1185; Litchfield MN), with free access to water.Calves were bottle fed twice a day in equal proportions at 0600h and 1800h with a milk replacer formulated to contain 24% of protein and 17% of fat (Nurture Professional 24-17, Provimi; Lewisburg OH).Milk replacer intake was initially set up to 600 g per day in the first week and 900 g per day in the second week.Subsequently, milk replacer intake was restricted up to 1,200 g of dry matter (DM) per day for all the treatments, which corresponded to 7.6 L. A commercial calf starter formulated to contain 18% of CP, 1.15% of crude fat, and 4.2% of crude fiber (Purina Calf Startena -Purina, Neenah WI) was offered to the calves ad libitum when animals reached 3 weeks of age.Chopped alfalfa hay was offered at 4 weeks of age.
The weaning transition in the EA was conducted from d 39 to 42 of age; EG from d 35 to 49 of age; LA from d 53 to 56 of age; and LG from d 49 to 63 of age.The abrupt pace consisted of 4 equal step-downs over 3 d, and the gradual pace consisted of 7 equal step-downs over 14 d.The milk replacer volume was reduced by approximately 1.9 L and 1.09 L at each step-down for the gradual and abrupt pace, respectively.Using this transition method, all the calves weaned at early age were offered a total of 46.2 kg of milk replacer, and the calves weaned at late age were offered a total of 63 kg of milk replacer (Figure 1).

Sampling
Milk replacer, commercial calf starter, and alfalfa hay were sampled every 30 d and combined to obtain one composite sample per item per cohort.Samples were stored at −20°C for later analysis.
Blood samples were collected from the jugular vein into Vacutainer tubes (Becton Drive, Franklin Lakes, NJ).All animals were sampled beginning at 0900h on d 3 and d 7 of age, one day immediately before the weaning process began (pre-weaning), and one day after weaning was completed (post-weaning).Samples were centrifuged at 1,100 × g for 10 min using a desktop centrifuge (Spiplus-8 digital centrifuge, Walter Products, Plymouth, MI) at room temperature (24°C).Aliquots of 1 mL of serum samples were placed in microcentrifuge tubes and stored at −80°C for later analysis of blood metabolites and cortisol concentration.
Male calves were humanely killed via captive bolt and exsanguination, one day post-weaning.Samples from rumen, jejunum, large intestine, liver, omental adipose, and perirenal adipose tissues were collected, immediately placed in liquid nitrogen, and stored at −80°C for RNA extraction.Samples from rumen were collected from the ventral rumen area, full thickness, whereas large intestine samples were collected from the midpoint of the colon, and samples from other tissues were collected from the midpoint of their respective locations.

Analysis
Milk replacer, commercial calf starter, and alfalfa hay were sent to a commercial laboratory (Dairy One Inc., Ithaca, NY) to determine the chemical composition of each item (Table 1).The DM content was determined using an oven according to method No. 930.15 (AOAC, 1990).Ash was determined by combustion according to method No. 942.05 (AOAC, 1990).Organic matter (OM) was determined by the difference of 100% and ash concentration.Neutral detergent fiber (NDF) was determined, according to Mertens et al. (2002), using thermostable α-amylase and sodium sulfite modified for Ankom Fiber Analyzer (Ankom Technology, Macedon, NY).Acid detergent fiber (ADF) was determined according to method No. 973.18 (AOAC, 1990).Starch concentration was determined using a YSI 2700 SELECT Biochemistry Analyzers (YSI Inc., Yellow Springs, OH).Fat concentration in the milk replacer was determined by bases hydrolysis according to method No. 932.06 (AOAC, 1996), whereas the fat concentration for commercial calf starter and alfalfa was determined using an ANKOM XT15 Extractor (Ankom Technology, Macedon, NY).Calcium and phosphorus concentrations were determined using Inductively Coupled Plasma Spectroscopy, according to the method described by Wolf et al. (2003).
Serum total protein was determined in samples from d 3 and 7 of age, pre-and post-weaning, using a refractometer as described by Hernandez et al. (2016).Serum cortisol concentration was determined in samples from d 7 of age, pre-and post-weaning with a commercial ELISA assay kit, following the manufacturer instructions (Arbor Assays, Ann Arbor, MI).Serum concentration of NEFA in samples from d 3 and 7 of age, pre-and post-weaning was determined using enzymatic colorimetric assays (Wako series NEFA-HR (2), FUJI-FILM Healthcare Americas Corp., Valhalla, NY), according to manufacturer instructions.Serum albumin concentration was determined in samples from d 7 of age, pre-and post-weaning using enzymatic colorimetric assays with Bromocresol Green (Sigma Aldrich, St. Louis, MO), according to manufacturer instructions.Serum thiobarbituric acid reactive substances (TBARS) concentration was analyzed according to the protocol provided in Appendix O of the Meat Color Measurement Guidelines (AMSA, 2012) in samples from d 3 and 7 of age, pre-and post-weaning.Total antioxidant capacity was determined in samples from d 3 and 7 of age, pre-and post-weaning using ABTS reduction assay, according to Janaszewska and Bartosz (2002), using solution of Trolox (Sigma Aldrich, St. Louis, MO) as the standard antioxidant solutions.Briefly describing, 2 µL of sample, water, or standard were added to each well in a 96 well plate, followed by 200 µL of radical ABTS.The absorbance was determined after 6 min of incubation.Concentration was expressed in µM of equivalent Trolox/min.
Serum glucose concentration was determined in samples from d 3 and 7 of age, pre-and post-weaning by an enzymatic colorimetric assay using peroxidaseglucose oxidase (PGO; Sigma Aldrich, St. Louis, MO) and o-Dianisidine dihydrochloride (Sigma Aldrich, St. Louis, MO), as described by the manufacture and slightly modified by Tsai (2021).Briefly, one capsule of PGO Enzyme Preparation was dissolved in 50 mL of deionized water, and o-Dianisidine dihydrochloride was dissolved in 20 mL of deionized water.Subsequently, the working solution was prepared by adding 1.6 mL of dissolved Dianisidine to 50 mL of dissolved PGO.The assay was carried out in 96 well plates, where 150 µL of the working solution were added to 5 µL of serum sample in each well, followed by 45 min of incubation at room temperature.A standard curve was prepared with a glucose calibrator.The absorbance was determined by spectrophotometry (SpectraMax i3x, Molecular Devices San Jose, CA) at 450 nm.Serum β-hydroxybutyrate concentration was determined in samples from d 3 and 7 of age, pre-and post-weaning by an enzymatic colorimetric assay using 3-hydroxybutyrate dehydrogenase, as described by Tsai (2021).Ketone body calibrator (FujiFilm Medical Systems, Lexington, MA) was used as the standard.Serum haptoglobin concentration was determined in samples from d 3 and 7 of age, pre-and post-weaning using colorimetric assay, according to the method developed by Eckersall (2002) and described by Tsai (2021).A standard curve was prepared with a haptoglobin calibrator (Tridelta Development Ltd., Kildare, Ireland).Intra-assay coefficients of variation (CV) were < 9.9% for all the blood metabolites and cortisol concentrations.
Total RNA was extracted from rumen, jejunum, large intestine, and liver tissues using NucleoSpin® RNA (Macherey Nagel, Düren, Germany) according to the methods described by the manufacturer.Total RNA was extracted from omental adipose and perirenal adipose tissues using RNeasy Lipid Tissue (Qiagen, Hilden, Germany).The RNA concentration was determined using NanoDrop ND-1000 spectrophotometer (NanoDrop Technologies, Rockland, DE).The purity of the RNA extracts was evaluated by 260/280 and 260/230 ratios.Samples with 260/280 ratio lower than 1.8 or greater than 2.2 were discarded and extracted again.Also, samples with 260/230 lower than 2.0 or greater than 2.2 were discarded and extracted again.RNA quality was assessed in a subset of randomly selected samples, representing all the treatments and tissues, utilizing the Fragment Analyzer System (Agilent Technologies, Santa Clara, CA).The final RNA quality index corresponded to 7.7 ± 1.9.The complementary DNA synthesis was carried out using Applied Biosystems High-Capacity Reverse Transcription Kit (Applied Biosystems, Foster City, CA) according to the method described by the manufacturer with adaptations from our laboratory (Scholte et al., 2017;Tsai et al., 2017).Briefly, the RNA concentration was adjusted by the proportion of sample and water to contain 100 ng of RNA per reaction in a 200 µL microcentrifuge tube, and to which 34 µL of master mix were added.The master mix consisted of 10 µL of 10× RT Buffer, 4 µL of 25× dNTP Mix, 10 µL of 10× RT Random Primers, 5 µL of MultiScribe Reverse Transcriptase, and 5 µL of RNase Inhibitor.Real-time reverse-transcribed PCR was carried out in a 7500 Fast real-time PCR system (Applied Biosystems, Foster City, CA) using commercial TaqMan FAM probes (Table 2) on targeted genes Interleukin-1beta (IL-1β), Interleukin-6 (IL-6), Interleukin-8 (IL-8), Tumor Necrosis Factor-α (TNF-α), Interferon-gamma (IFN-γ), Nuclear Factor Kappa B (NF-κB), and Intercellular Adhesion Molecule (ICAM1).Bovine Ribosomal protein S9 and Glyceraldehyde-3-phosphate dehydrogenase were used as endogenous controls (housekeeping genes) to adjust the cycle threshold (CT).Delta CT values (ΔCT) were calculated by the difference between CT of each gene of interest and the arithmetic average of CT from 2 housekeeping genes mentioned above.The reaction mixture included 2 µL of cDNA, 10 µL of TaqMan® Universal Master Mix II with Uracil-N-Glycosylase (Applied Biosystems, Foster City, CA, USA), 1 µL of Applied Biosystems 20X custom primer probe mixture (Applied Biosystems, Foster City, CA, USA), and 7 µL of RNase-free water.

Statistical Analysis
All statistical analyses were carried out using the MIXED procedure of SAS (version 9.4, SAS Institute Inc., Cary, NC).One calf died of pneumonia before starting any stage of treatment; therefore, it was not included in the data analysis.No other calves were diagnosed with any diseases or removed from the study.Weaning pace, age, pace × age interaction, body weight at birth, and sex were included as fixed effects, and cohort was included as random effects in the model.Blood metabolites and cortisol were analyzed as repeated measurements and day of sampling (d 7 of age, one day before starting weaning, and one day after weaning was completed), pace × day of sampling, and age × day of sampling were also included as fixed effects, besides the previously stated fixed and random effects.using ΔCT values.Covariance structures were tested for repeated measures and chosen based on the lowest Akaike's information criterion.Interactions were investigated using the slice option, and Least Squares Means were compared by Tukey t-test.Differences were declared significant at P ≤ 0.05 and tendencies when 0.05 < P ≤ 0.10.

Growth and health
Growth and health data are discussed in detail in a previous paper (Wolfe et al., 2023).Briefly, the average daily gain (ADG) was lower in LA than EA treatment, and gradually weaned calves had a greater ADG.Furthermore, calves weaned gradually had a greater starter intake than those weaned abruptly.Finally, blood hematocrit levels were greater in calves weaned abruptly.In the beginning of the weaning transition, body weight was 57.1 ± 4.2 kg (average ± standard deviation) for animals in early-abrupt, 56.2 ± 6.3 kg in early-gradual, 71.6 ± 8.3 kg in late-abrupt, and 73.0 ± 7.1 kg in late-gradual.

Chemical composition of milk replacer, commercial starter, and alfalfa
Chemical compositions from milk replacer, commercial starter, and chopped alfalfa hay are shown in Table 1.

Blood metabolites and cortisol
There was a tendency for age × pace interaction in serum NEFA concentration (P = 0.10, Table 3); where calves in the EA showed the greatest NEFA concentra-tion compared with all the other treatments.All the blood metabolites and cortisol concentrations were affected over time, and their least squares means by sampling time are presented in Table 4.
There was a pace × day of sampling interaction in serum NEFA (P < 0.01) and BHBA concentration (P < 0.01).Calves weaned under abrupt pace showed an increased serum NEFA concentration post-weaning compared with d 7 of age and pre-weaning and all the samples for gradual weaning, whereas post-weaning NEFA concentration of calves weaned under gradual pace did not differ from the other values besides abrupt pace at post-weaning (Figure 2.A).On the other hand, calves weaned under gradual pace showed the greatest serum BHBA concentration at post-weaning, followed by calves weaned under abrupt pace at post-weaning, with the lowest concentration observed on d 7 of age for both paces (Figure 2.B).There was a tendency for pace × day of sampling interaction in serum cortisol (P = 0.10) and total antioxidant capacity concentration (P = 0.10).Cortisol concentrations were not different between weaning pace (abrupt vs. gradual) at each time point and were not different between calves from gradual pace on d 7 of age and abrupt pace post-weaning; all the other means tended to differ (Figure 3.A).The total antioxidant capacity at post-weaning tended to be greater (Figure 3.B) in calves from gradual pace compared abrupt pace, whereas the concentration did not differ between weaning paces at d 7 of age and pre-weaning.
There was an age × day of sampling interaction effect on serum BHBA (P = 0.03).The greatest serum BHBA concentration was observed at post-weaning for early weaning and late age and the lowest at d 7 of age (Figure 4).

mRNA abundance of inflammation-related genes
Calves weaned at an early age showed a reduced mRNA abundance of TNF-α in the jejunum (P = 0.05; Table 5) and tended to have a reduced mRNA abundance of ICAM1 in the jejunum (P = 0.06) compared with calves weaned at late age.However, calves weaned at an early age tended to have an increased mRNA abundance of IL-1β in the liver (P = 0.10) compared with calves weaned at a late age.Regarding IL-1β in the liver, calves weaned at an abrupt pace had a lower mRNA abundance than calves weaned at a gradual pace (P = 0.05).There was a tendency in age × pace interaction for mRNA abundance of IL-1β and TNF-α in the large intestine (P = 0.06 and 0.07, respectively).Calves in EA tended to have a greater mRNA abundance than LA and EG for IL-1β in the large intestine.Calves in EG tended to show a lower mRNA abundance of TNF-α than calves in LG.In addition, there was a tendency for age × pace interaction for IL-6 in the liver (P = 0.06); calves in the EA tended to show a reduced mRNA abundance than EG.There was no detectable effect of the weaning age, pace, or interaction between the factors on the mRNA abundance of inflammationrelated genes tested in the rumen of calves (P > 0.15; > 0.28; and >0.16, respectively).
There was age × pace interaction for mRNA abundance of NF-κB in the perirenal and omental adipose tissue (P < 0.01; 0.02, respectively, Table 6); mRNA abundance of NF-κB in the perirenal adipose tissue was lower for calves in EA than EG, LA, but did not differ from LG. Further, calves in the LG did not differ from EG and LA.Regarding the mRNA abundance of NF-κB in the omental adipose tissue, calves in the EA showed a lower mRNA abundance than EG and LA.There was age × pace interaction in mRNA abundance of ICAM1 in the perirenal adipose tissue (P = 0.05); animals in the LA tended to have greater mRNA abundance than in LG.

DISCUSSION
The objectives of this study were to evaluate the effects of weaning age and pace on blood metabolites and mRNA abundance of inflammation-related genes in Holstein dairy calves.The results suggest that calves weaned under abrupt pace and early age may lead to downregulation of mRNA abundance of several inflammation-related genes.In addition, the weaning pace and age affected serum BHBA concentration at weaning.Wolfe et al. (2023) described in another publication from the present study that during the weaning period, the average daily gain (ADG) of calves in late-abrupt was lower than that in early-abrupt calves.Also, calves gradually weaned had a greater ADG.Late-weaned calves showed a faster respiration rate than early weaned calves.Additionally, heart rate was faster in calves weaned abruptly than gradually.

Blood metabolites and cortisol
The literature shows that the weaning methods can impact the health status and blood metabolites as well as indices of energy balance (Eckert et al., 2015;Steele et al., 2017;McCoard et al., 2019).In line with the literature, we also observed the weaning methods affecting several blood metabolites, as discussed below.
The increased serum BHBA concentration at the preweaning of calves weaned at late age rather than early age may be attributed to a greater starter fermentation that might be explained by the combination of 2 factors.The first factor is that late-weaned calves had a greater daily starter intake during pre-weaning than early-weaned calves (Wolfe et al., 2023).The second factor is that weaning at late age (wk 8 vs. wk 6) gives the rumen a greater propensity for more  efficient fermentation.A similar observation was noted by Wickramasinghe et al. (2022), where serum BHBA concentration was greater in calves weaned at late age vs. early age.
Regarding the effect of weaning pace, an increased concentration of BHBA in the serum of calves gradually weaned compared with abrupt at post-weaning was also observed by Steele et al. (2017) and Khan et al. (2007).Deelen et al. (2016) observed that blood BHBA concentration is strongly associated with starter, which is supported by a greater rumen fermentation and synthesis of butyrate in the rumen.According to our data (Wolfe et al., 2023), animals gradually weaned had a greater grain intake than abrupt, which can be the main factor affecting the serum BHBA concentration.Rumen butyrate is well known for playing an important role in the rumen papillae development (Sander et al., 1959).Therefore, weaning calves at a late age and at a gradual pace seems to facilitate rumen capacity for handling solid diet fermentation.
Nonesterified fatty acids are released from the adipose tissues through the enzymes involved in lipolysis, such as hormone-sensitive lipase that have a high activity during negative energy balance (Koltes and Spurlock, 2011).Animals in the EA had a greater negative energy balance because of a combination of several situations;  those animals had a reduced starter intake at an early age, possibly a less developed rumen, and reduced time during the weaning transition, as presented by Wolfe et al. (2023).Even though the literature does not have any previous evidence from the combination of weaning pace and age, Khan et al. (2007) and Steele et al. (2017) observed the effect of weaning pace on NEFA concentration and metabolizable energy.Khan et al. (2007) observed that animals weaned over 4 d had a greater NEFA concentration than animals weaned gradually over 21 d.In the same context, lower intake leads to lower metabolizable energy intake; as such Steele et al. (2017) observed that calves abruptly weaned had a reduced metabolizable energy intake than gradually weaned calves that negatively impacted the body weight of calves at an abrupt weaning pace.Serum total antioxidant capacity was greater at weaning in animals weaned under gradual pace than abrupt pace.Antioxidant has been used as a biomarker of the oxidative status in animal research (Castillo et al., 2005;O'Boyle et al., 2006;Putman et al., 2018) and has been associated with regulating the immune system in dairy cattle (Putman et al., 2018).Lactating dairy cows under stress of early-lactation show reduction in blood total antioxidant capacity (Putman et al., 2018).Another indicator of oxidative stress used in animal research is the oxidative stress index (Abuelo et al., 2013;Putman et al., 2018;Cuervo et al., 2021), which is calculated by the imbalance between oxidant and antioxidant compounds.Neonatal dairy calves have shown an association between oxidative stress index and the concentration of inflammatory markers in blood (Cuervo et al., 2021).Therefore, the greater serum total antioxidant capacity in the gradual pace weaning group after weaning might indicate that animals under gradual pace weaning were exposed to a lower stressful condition than abrupt weaning pace.
Although cortisol is an indicator of stress response in adult animals, the absence of the main effects of weaning age and pace on this variable may be explained by several factors.One of the factors is that in young calves the hypothalamic-pituitary-adrenal axis is not completely developed and responsive as in older animals (Mormede et al., 1982;Hulbert and Moisá, 2016).Additionally, cortisol is a rapid response hormone (Masmeijer et al., 2021), so the management of the calves during blood sampling as a stressor as well as the sample timing might have led to the absence of difference in cortisol among the treatment groups.In line with this notion, Coetzee et al. (2007) observed that cortisol concentration in castrated calves were greater than uncastrated calves at 6 h; however, there  was a cortisol regulation and no difference after 8 h, demonstrating that after peak stress, cortisol concentration in the blood was normalized in a short-period of time.Further, both Hickey et al. (2003) and Hulbert et al. (2011) did not observe the effect of weaning pace on plasma cortisol in calves.Cortisol secretion varies according to circadian rhythm (Gardy-Godillot et al., 1989) and that might be another factor contributing in the absence of difference among weaning strategies in the present study.Therefore, collecting samples at different times of the day should be considered in future studies.Finally, the lack of variation in the level of stress/metabolic challenge to which the animals were exposed in different treatments may not have been strong enough to lead to a difference in cortisol concentration.The tendency for Pace × day interaction in concentration of cortisol is not physiologically explainable, and it might be attributed to a random effect or n per treatment.
Calves exposed to different weaning strategies, combining age and pace, did not show an impact on acute phase proteins evaluated under the experimental conditions of the present study (haptoglobin and albumin).This observation is in agreement with Hickey et al. (2003), Hulbert et al., (2011), andWickramasinghe et al. (2022), who reported no effect of weaning methods on the haptoglobin concentration of calves.Further, other events that lead to a stress response in dairy calves, such as dehorning, showed no effect on the albumin concentration (Laden et al., 1985).
It is important to highlight that the experimental design of this study, combining weaning age and pace, resulted in animals in the same age group (early vs. late) that were in fact different ages at pre-and post-weaning sampling times.For example, EG calves finished weaning transition at 49 d while EA calves finished at 42 d, resulting in different ages at pre-and post-weaning sampling.However, both EA and EG calves consumed 46.2 kg of milk replacer from d 7 until end of the weaning transition.In this study, experimental design confounded age while keeping total lifetime volume of milk replacer intake consistent among treatment, whereas most weaning studies confounds total lifetime volume of milk replacer intake while keeping age at weaning consistent.

mRNA abundance of inflammation-related genes
Stressful/metabolically-challenging events in animal life, such as the weaning process, have been associated with suppression of certain processes within the immune system in animals (Lynch et al., 2010).A previous study has shown that oxidative stress can downregulate the expression of cytokines in young dairy calves, as observed in neonatal dairy calves (Cuervo et al., 2021).Cytokines are responsible for controlling the cell survival, death, and differentiation, which are processes essential for tissue development and homeostasis.Cytokines are associated with several infections and immune system-affecting disorders playing an important role in pro-and anti-inflammatory mechanisms (Monastero and Pentyala, 2017).The downregulation in cytokines expression by metabolically-challenging events may explain why NF-κB mRNA abundance in the perirenal and omental adipose tissues were lower in the EA group, when there is the combination of 2 stressful conditions, and why early weaned calves showed a lower mRNA abundance of TNF-α and ICAM1 gene in the jejunum, as well as abrupt pace downregulated the mRNA abundance of IL-1β in the liver.We believe the downregulation of inflammation-related genes was induced by greater weaning stress in the abrupt pace and early age because stress in calves has been related to suppressed neutrophil phagocytic or oxidative capacity (Lynch et al., 2010;Hulbert et al., 2011).Another factor that might lead to the downregulation of inflammation-related genes in early weaning is the reduced immune competence in young animals because they are still developing the immune system.However, there is some contradiction regarding the mRNA abundance of other inflammation-related genes in the stress response, and some results remain unexplained, as observed for IL-1β in the liver, where animals exposed to the early weaning tended to have a greater mRNA abundance.
Having cytokines evaluated under combinations of different weaning methods is an appealing point because this is one of the unique studies demonstrating the impact of weaning strategy on inflammation-related genes.Our observation shows the effect of a more metabolically challenging method was not as severe as expected.Even the least challenging weaning method impacted the inflammation-related genes evaluated, leading to a lack of statistical difference among weaning methods.

CONCLUSION
We conclude that most inflammation-related genes tested were not affected by weaning age or pace.However, some of them, for example, ICAM1 and TNF-α in jejunum, were downregulated when calves were weaned at an early, or IL-1β in the liver was downregulated when exposed to abrupt pace that was attributed to the suppression of the immune system and an immature immune response.In some cases, the combination of those strategies (early and abrupt) simultaneously intensified this effect through more than one downregulation factor, as observed for TNF-α in the large Agustinho et al.: Weaning strategies and inflammation in dairy calves intestine.Furthermore, combining early and abrupt weaning also triggered a more severe negative energy balance.Therefore, weaning calves early-abruptly might negatively affect their health and performance.The present study evaluated the response in the animals immediately after the weaning process; therefore, further investigations would be valuable to determine the persistence of those impacts in health measurements and mRNA abundance of inflammation-related genes during a longer period after weaning.

Figure 1 .
Figure 1.Organization of the experimental treatments used in the present study.
Agustinho et al.: Weaning strategies and inflammation in dairy calves

Figure 2 .
Figure 2. Pace × Day interaction for serum nonesterified fatty acids (NEFA; panel A) and β-hydroxybutyrate (BHBA; panel B) concentration on d 7 of age, pre-, and post-weaning in dairy calves weaned under different weaning strategies, including pace [abrupt (n = 35) vs. gradual (n = 36)].a-c Means with different superscripts differ by Tukey t-test (P ≤ 0.05).Error bars correspond to standard error of the means. 1 day = day of sampling.
Figure 3. Tendency of Pace × Day interaction for cortisol (panel A) and total antioxidant capacity (panel B) concentration on d 7 of age, pre-, and post-weaning in dairy calves weaned under different weaning strategies, including pace [abrupt (n = 35) vs. gradual (n = 36)].Error bars correspond to standard error of the means. 1 day = day of sampling, 2 TAC = total antioxidant capacity.

Figure 4 .
Figure 4. Age × Day interaction for β-hydroxybutyrate (BHBA) on d 7 of age, pre-, and post-weaning in dairy calves weaned under different weaning strategies, including weaning age [early (n = 35) vs. late (n = 36)].a-c Means with different superscripts differ by Tukey ttest (P ≤ 0.05).Error bars correspond to standard error of the means. 1 day = day of sampling.

Table 1 .
Agustinho et al.:Weaning strategies and inflammation in dairy calves Chemical composition of undiluted milk replacer, commercial calf starter, and chopped alfalfa hay ADF = acid detergent fiber, NDF = neutral detergent fiber, 2 chemical compositions of undiluted milk replacer, 3 chopped alfalfa hay, 4 Results corresponded to one composite sample per item per cohort.

Table 2 .
One day before starting weaning corresponded to d 38 of age to early-abrupt weaning; d 34 of age to earlygradual; d 52 of age to late-abrupt; and d 48 of age to late-gradual.One day after weaning corresponded to d 43 of age to early-abrupt weaning; d 50 of age to early-gradual; d 57 of age to late-abrupt; and d 64 of age to late-gradual.Concentrations from samples from d 3 of age were used as a covariate in the model of each respective blood metabolite, except for albumin and cortisol.Data of mRNA abundance were analyzed Agustinho et al.: Weaning strategies and inflammation in dairy calves Taqman® bovine primer/probe sets used for real-time polymerase chain reactions

Table 4 .
Agustinho et al.: Weaning strategies and inflammation in dairy calves Least squares means of serum metabolites of dairy calves on d 7 of age, pre-weaning, and post-weaning Arithmetic average of CT of Bovine Ribosomal protein S9 and Glyceraldehyde-3-phosphate dehydrogenase was calculated to determine the Delta CT value, and presented in the table. 1