Effect of pelleted vs. ground starter with or without hay on preweaned dairy calves

The objective of this study was to evaluate the effect of the physical form of starter and inclusion of hay in the diet of preweaning dairy calves on performance, digestibility, ruminal development, and mRNA expression of genes involved in ruminal metabolism. Holstein × Gyr crossbred male calves (n = 38 1day old) were assigned to 3 treatments for 9 weeks: Control (n = 13; pellet starter with 4 mm diameter and 18 mm length and 4% steam-flaked corn), Ground (n = 12; same starter of the control but ground pass through a 4.0 mm sieve), or Ground plus 5% chopped Tifton hay GH (n = 13). All calves were fed 4 L/d of whole milk up to 63 d of age and were abruptly weaned at 64 d of age. Water and diets were offered ad libitum. Samples of ruminal contents were obtained from all animals at 30, 45, and 60 d of age to evaluate pH, ammonia nitrogen, and volatile fatty acids (VFA). At 55 d of age, an apparent digestibility assay was performed using 18 animals (n = 6/ treatment). At 65 d of age, the 18 animals were euthanized to evaluate the development of the digestive tract. The physical form of starter and the dietary inclusion of hay did not influence starter intake (Control 326 g/d, Ground 314 g/d and GH 365 g/d), daily weight gain (Control 541g/d, Ground 531g/d and GH 606g/d), feed efficiency, apparent nutrient digestibility, energy partitioning, nitrogen balance, ruminal pH, ammonia nitrogen concentration, VFA, the development of the digestive tract and the mRNA expression of genes involved in AGV metabolism.


Introduction
According to the NRC 2001 [1], besides being palatable, calf starter must be made of good quality ingredients to ensure adequate supply of protein, energy, fiber, vitamins, and minerals while also having a coarse texture. Coarse-textured starter is indicated to stimulate ruminal a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 allocated in a barn with open sides. At 8 d of age, preventive oral treatment against coccidiosis (Baycox Ruminants, Bayer, Leverkusen, Germany) was performed, at 3 mL/10 kg of BW.
From 4 to 63 d of age, the total volume of whole milk (4 L/d) was divided into two equal meals (0800 and 1600 h) and provided to calves in buckets. Calves were weaned at 64 d of age. Starter and water were offered ad libitum throughout the experimental period. The amount of starter provided was enough to result in approximately 20% orts.

Intake and growth
Calves' growth, body measurements, and intake were monitored between 4 and 63 d of age. Intakes of whole milk, starter, hay, and water were measured once a day, by subtracting refusals from the amounts provided. Water intake was measured using a portable balance (WH-A04; WeiHeng, Yongkang, China).
The BW and body measurements were obtained once a week before the morning feeding in a flat location, allowing animals to remain with their limbs well set. The height at withers and hip were evaluated using a hipometer (Walmur, Porto Alegre, Brazil), and hip width was assessed with a measuring tape (cm scale) (Bovitec, São Paulo, Brazil). Feed efficiency was calculated using the ratio between ADG and total DMI (includes milk DM) [19].

Apparent nutrient digestibility, energy partitioning and nitrogen balance
Between 55 and 60 d of age, 6 animals per treatment were submitted to a total tract nutrient digestibility trial with 5 d total feces and 1 d urine collection in metabolic cages (dimensions of 1.50 m x 0.80 m; Intergado Ltda., Contagem, Brazil). During these 5 days, the fecal excretions of the animal were collected on the floor of the metabolic cage, and the total of feces was weighed and homogenized at the end of every 24 h. Equivalent quantities of the daily sub-samples were composited to one sample per animal and were stored at -20˚C for further analysis. Urine collections were performed for 24 h [20]. The urine tray was drained into 5 L containers held in ice-filled polystyrene thermal boxes [21]. The volume, weight, and density of the urine from each animal were measured, and a sample of 50 mL was taken after it was filtered in cheesecloth and then stored at -20˚C for gross energy and nitrogen analysis. For the energy balance, gross energy intake (GEI) was calculated by the difference between the gross energy content (GE) of each feed offered (milk, starter, hay) and those obtained in the respective orts. Digestible energy intake (DEI) was calculated by the difference between GEI and fecal energy excretion. Metabolizable energy intake (MEI) was derived as the difference between DEI and urine energy. The percentages of consumed energy lost as feces and urine (DE/GE) and the relationships between ME/GE (metabolizability) and ME/DE, as energy efficiency indexes, were also calculated. Nitrogen balance, or nitrogen retained, was calculated according to the following equation: N retained (g/d) = N ingested-(fecal N + urine N).

Nutrient analysis
During the entire experiment, milk samples were collected twice a day (morning and afternoon) and analyzed for total solids, crude protein (CP), lactose, and fat content using spectrophotometry (Bentley 2000; Bentley Instruments Inc., Chaska, MN). Samples of starter and hay were collected 3 d per week, composited biweekly and stored at −20˚C until analysis. All samples of feed (experiment and digestibility assay) and feces were oven-dried at 55˚C for 72 h, ground through a 1 mm mesh diameter sieve in a Wiley type mill (model 3, Arthur H. Thomas Co., Philadelphia, PA), and analyzed.
Samples of the starter and hay were analyzed for dry matter (DM) (Method 934.01), CP (Method 988.05), ether extract (EE) (Method 920.39), and ash (Method 942.05), according to the [22]. Neutral detergent fiber (NDF) and acid detergent fiber (ADF) were determined according to the method proposed by [23], with amylase, adapted to the conditions of the Ankom 220 apparatus, without the use of sodium sulphite and corrected for residual ash [24], ADF [25], and EE [25] (Method 920.39). For nitrogen compounds, the correction of NDF and ADF as well as the estimation of insoluble nitrogen in FDN and FDA, followed the recommendations of [26]. Gross energy (GE) was determined using an adiabatic calorimeter (IKA-C5000, IKA 1 Works, Staufen, Germany). All analysis were performed at the Laboratory of Food Analysis of Embrapa Dairy Cattle. metaphosphoric acid for VFA analysis and another sample to 5 mL was used for ammonia concentration analysis. Ruminal pH fluid was measured immediately after collection using a portable pH meter (DM-2-Digimed, São Paulo, São Paulo, Brazil). Contents were filtered through four layers of cheesecloth to separate the liquid and solid fractions. All samples were stored in plastic tubes at -20˚C for further analysis. The ruminal NH 3 N concentration was quantified after distillation of Kjeldahl with magnesium oxide and calcium chloride according to Method 920.03 [22]. Concentration of VFA was determined by HPLC in a Dionex Ultimate 3000 Dual Detector HPLC (Dionex Corporation, Sunnyvale, CA) coupled to a hodex RI-101 refractive index detector (ECOM Ltd., Prague, Czech Republic) maintained at 40˚C using a Phenomenex Rezex ROA ion exchange column (Phenomenex, Torrance, CA), 300 × 7.8 mm, maintained at 45˚C. Mobile phase was prepared with 5 mmol/L H2SO4, and the flow rate was 0.7 mL/min. Samples of 2 mL were defrosted at room temperature (22-25˚C) and centrifuged (1,800 × g, 10 min). Cell-free supernatants were treated as described by [29]. The following acids were used to calibrate the standard curve: acetic, succinic, formic, lactic, propionic, valeric, isovaleric, isobutyric, and butyric. These acids were prepared to a final concentration of 10 mmol/L except isovaleric acid (5 mmol/L) and acetic acid (20 mmol/L).

Slaughter, Rumen and Omasum measurements
At 65 d of age, seven calves from each treatment were euthanized after being deprived of liquid diet for 12 h. Pre-anesthetic medication with 2% xylazine hydrochloride (30 to 40 mg per 100 kg BW) was administered intravenously. After 20 minutes, 5% thiopental sodium (5 mg/kg BW) was directly administered into the foramen magnum, located in the occipital bone of the skull. Once the cardiorespiratory arrest was verified, bleeding was carried out by an incision made in the jugular furrow at the base of the neck.
After slaughter, the abdominal cavity was immediately opened, and each region of the gastrointestinal tract (reticulum-rumen, omasum, and abomasum) was isolated, tied off, and weighed. After samples were collected from the gastrointestinal tract, it was emptied, washed with running water (to remove contents), and weighed. Internal organs and viscera were also weighed. All variables were evaluated as a proportion of empty body weight (BW).
An area of 5 cm 2 from the ventral ruminal sac and omasum laminae were fixed in formalin and processed for paraffin embedding. The paraffin blocks were sectioned using an Olympus CUT 4055 manual microtomes (Olympus, Tokyo, Japan) into 5 μm-thick serial sections. For morphometric papillae analysis, slides were stained with Hematoxylin-Eosin as described by [17]. Images were captured using a light microscope (CX31; Olympus) attached to a camera (OSIS SC30; Olympus) using Cell-B software (Olympus). Morphometric analysis were performed using AxioVision 4.8.2-06/2010 software (Carl Zeiss Images Systems, Jena, Germany). The measurements taken were height of ruminal and omasal papillae [18]. To determine mitotic index, 2,000 cells from the basal layer of the rumen and omasum epithelium were counted, including those with nuclei presenting mitotic figures (using the light microscope, 400× magnification). The mitotic index was calculated as a ratio between the number of nuclei in division and the total number of nuclei [19].

Tissue collection and RNA extraction
Tissue samples were collected from the anterior portion of the ventral sac of the rumen [13,20,33] at 65 d after birth. The rumen epithelium was manually separated from the underlying muscular layer by gloved hands and rinsed in tap water to remove residual feed particles. Samples of rumen epithelial tissue were stored in RNAlater (Thermo Fisher Scientific, Waltham, Massachusetts, USA) and thereafter stored at -20˚C. The rumen tissue (30mg) was macerated using 600 μL of RLT buffer and 6 μL β-Mercaptoethanol using the Tissueruptor equipment (Qiagen, Hilden, Germany), and total RNA was extracted using RNeasy Mini kit (Qiagen, Hilden, Germany) and treated with DNase using RNase-Free DNase Set kit (Qiagen, Hilden, Germany), according to the manufacture's protocol. After that, the quantity and purity of RNA were determined by absorbance at 260 and 280 nm using the NanoDrop ND-1000 spectrophotometer (Thermo Fisher Scientific, Waltham, Massachusetts, USA). The total RNA integrity was evaluated using the 2100 Bioanalyzer (Agilent, Santa Clara, CA, USA).

Real-time quantitative PCR
For reverse transcription, 6 μL of total RNA was added to 1 μL Oligo(dT) 20 and 1 μL annealing buffer. Then, this mixture was incubated at 65˚C for 5 min and kept on ice for 3 min. After that, 10 μL 2X First-Strand Reaction Mix and 2 μL of SuperScript™ III (Invitrogen 1 ) were added to this mixture and then incubated at 25˚C for 5-10 min, 50˚C for 50 min, and 85˚C for 5 min.
Real-time PCR analysis were performed using the ABI 7300 real-time PCR System (Thermo Fisher Scientific, Waltham, Massachusetts, USA). For each gene, an optimum amplification condition (cDNA and primer concentration) was selected to achieve a standardized efficiency among all selected genes. For that, five different concentrations of cDNA were used in duplicate. Each reaction was composed of the best cDNA concentration, 0.1 μM of specific primer sets, 7.5 μL power SYBR 1 green PCR master mix (Thermo Fisher Scientific, Waltham, Massachusetts, USA), and nuclease-free water for a final volume of 15 μL. Amplification conditions consisted of 2 minutes at 50˚C, 10 min at 95˚C, followed by 40 cycles (95˚C for 15 sec and 60 C for 1 min). After each run of qPCR, the melt curve analysis was performed for each sample to confirm that a single specific product was generated.
Each sample was run in duplicate. The gene expression levels were calculated relative to the average mRNA levels of the internal control gene that was expressed more (GAPDH or ACTB), according to [12].

Statistical analysis
Data were analyzed with SAS version 9.0 (SAS Institute Inc., Cary, NC). Weekly averages of weight gain and body measurements were analyzed using a repeated measures mixed model (PROC MIXED), including calf as a random effect and treatment, week, and their interaction as fixed effects. Variables with a single measurement during the study were analyzed by using the GLM procedure, differences among treatments, were analyzed using a Tukey adjustment for P-values. Least squares means for each treatment are reported. The variables BW at birth and total serum protein were considered covariates for intake, performance, body frame development, and internal organ weight (% of empty BW). Significance was declared at P � 0.05.

PLOS ONE
Effects of pelleted vs. ground starter with or without hay on preweaned dairy calves genes were a dependent variable and differences among treatments were assessed using Pvalues.

Results and discussion
No difference (P> 0.05) was observed in the DM intake of the liquid diet among treatments (Table 3, 502.4 ± 13.2 g DM/d), probably because of the restricted volume offered for all treatments.
The physical form of the starter and the dietary inclusion of hay did not influence the intake of nutrients and water (P> 0.05; Table 3). However, as in the study by [23], a numerical difference in consumption and weight gain favorable to the GH treatment was observed. Improvements in starter intake and performance when forage was offered to preweaned calves were reported [24], [25]. The positive effects of forage on starter intake may be related to the improvement in rumination, ruminal environment and enhanced muscular development of the rumen [24], [26], and can justify the numerical differences observed. Starter intake (g DM/ d) increased linearly as calves grew older (P<0.01; Table 3), from 40 ± 33 g DM/d at 1 week old to 722 ± 282 g DM/d at 9 weeks of age.
The intake of hay in the GH treatment was 17 ± 17 g DM/d and increased linearly as calves grew older, with a maximum value of 42 ± 17 g DM/d at 9 weeks of age, corresponded to 4.6% of that offered (forage: starter ratio offered of 5:95).
There was no effect of treatments on body development measurements (P> 0.05; Table 3). No differences in body weight of calves receiving hay during preweaning were also reported [23].

PLOS ONE
Effects of pelleted vs. ground starter with or without hay on preweaned dairy calves Doubling birth weight at 56 d of age is suggested as a weaning criterion [27]. In our study, the initial and final BW were, on average, 33.7 ± 4.4 and 69.4 ± 8.3 kg, respectively. These values were similar between treatments, demonstrating that the diets supplied in all treatments were adequate to reach this criterion.
The apparent digestibility of DM, organic matter (OM), CP, and GE and energy partitioning parameters were similar between treatments (P> 0.05; Table 4), showing that the physical form of the starter and inclusion of 5% hay was not able to cause an increase or decrease in the digestibility of the diet. Generally, high fiber diets [28,29] compromise diet digestibility. However, in the present study this was not observed, possibly due to the restricted inclusion of hay in the diet (5%). Values of 80% of dry matter digestibility were reported for calves receiving different sources of forage [30], corroborating with our results.
Since the different treatments did not yield differences in nutrient digestibility, it is possible to conclude that the inclusion of 5% hay in the diet of calves during the pre weaning does not alter the digestibility of nutrients. This statement is reinforced by the results reported by [31], evaluating crossbred Holstein x Gyr calves fed a complete diet consisting of 6 L/d of liquid diet with 13.5% of total solids and a commercial starter, similar to the one used in the present study, verified digestibilities of 85, 88, 85, and 88% for DM, OM, CP, and GE, respectively, in a trial performed between 50 and 55 d of age.
The energy partitioning parameters were similar between treatments (P> 0.05; Table 4). Gross energy intake (Mcal/day) and fecal energy loss (Mcal/d) did not differ between

PLOS ONE
Effects of pelleted vs. ground starter with or without hay on preweaned dairy calves treatments, which led to similar intakes of DE (Mcal/d), GE, metabolizability (ME/GE) (P> 0.05), and the parameters relative to the N balance. However, the efficiency of N retention in relation to N ingested in this study was higher to that observed by [32], averaging 40%, in crossbred Holstein x Gyr calves fed 4L/d milk and ground starter. This difference may be related to the different protein content of the diets in the experiments, since increased N intake has a positive correlation with N retention [33]. The similarity in feed efficiency observed between all treatments corroborates with the results observed by [34], who reported that there was no difference in performance, DM intake, nor feed efficiency between calves fed different physical forms of starter (coarse starter with 7.5% bromegrass hay and ground starter with 15% hay). Other studies [35,36] found higher feed efficiency in calves fed pelleted starter compared to other types of processing. However, the dietary inclusion of forage (chopped straw, the average intake of 100 g/d, 72% NDF, 8% CP, and 92% DM) with a textured starter increased feed efficiency concerning the starter without straw [37].
The physical form of starter and the dietary inclusion of 5% hay did not affect ruminal parameters (P> 0.05; Table 5), with the exception of N-NH 3 . An interaction between week and N-NH 3 concentration was observed, with higher N-NH 3 values at week 8 compared to weeks 4 and 6 for calves fed ground starter. For calves fed Control treatment, there was no difference between the weeks evaluated; for the treatment GH, weeks 4 and 8 were similar, and the lowest values were verified at week 6, which also did not differ from week 4.
Molar proportions of acetate, propionate, butyrate and, the sum of three VFA were not different among treatments. An increase in propionate concentrations was observed between the weeks evaluated, with the highest value being observed at week 8, indicating active fermentation at that age. This increase in propionate concentrations led to a reduction in the propionate acetate ratio, and values close to 1.1. This relationship between acetate and propionate was also reported by [38] in the treatment that received a pelleted starter and [39] in the treatment that received a mixture of steam flaked corn and extruded soybeans.
The mean pH 5.3 ± 0.4 was not differing between treatments and weeks. This result suggests that the physical form of starter and the inclusion of 5% hay in the diet with ground starter was not able to increase salivation or the VFA absorption, leading to this similarity between the treatments for pH values. This pH value is associated with ruminal acidosis in adult animals [40]; however, we did not observe signs of bloating, liquid feces nor gas bubbles, which are clinical signs of acidosis. According to [41], calves are apparently able to tolerate lower rumen pH values in comparison to adult animals, which may explain the potential adaptation of ruminal epithelium to starter fermentation [12]. However, subacute ruminal acidosis in young calves has received scant attention, and the severity of rumen acidosis caused by different factors is not yet clear in dairy calves [42]. The empty BW was similar between treatments (P> 0.05; Table 6). These results demonstrate that the intake of up to 5% hay did not affect digestive tract fill nor the performance of animals, corroborating with the results of [23]. There were no effects of treatments on reticulorumen, omasum, abomasum, and liver weights (P> 0.05; Table 6).
Histological parameters of rumen and omasum were not affected by treatments (P> 0.05; Table 6). All treatments used the same starter with the same ingredients, differing only in their physical form, since the ground starter was obtained by grinding the Control starter. The absence of differences for all the performance and development parameters showed that the physical form of starter (Control, Ground, or GH) had no effect on the development and performance of calves. These results allow us to hypothesize that the differences frequently pointed out between studies may occur due to the different ingredients and nutritional values of feedstuffs used.
According to [43] the particle size and nutritional factor could affect the expression of genes involved in cellular growth, ion binding, cell proliferation and microbial fermentation. The highest mRNA abundance to monocarboxylate transporter MCT4 was observed on rumen epithelial cells cultured from calves that were stimulated with acetate and propionate [44].
The transporter of the family NHE on the cell membrane help maintain the acid-base balance through the exchange of Na + into and H + out of the cell, and it was upregulated by an increase in VFA concentration [45,46], while the family of the transporters MTC is able to transfer of monocarboxylic acid, such as lactic acid, pyruvic acid, and ketones bodies [47,48]. We did not find difference in mRNA expression level when it was comparing the treatments and the mRNA expression of ATPA1, BHD1, HMGCL, HMGCS1, LDHA, PPARA, PPARD, MTC1, MTC4, NHE1, NHE2, NHE3, AKT1, and mTOR genes ( Table 7). The analysis of gene expression occurred at 65 d of age, and some studies showed that calves were not producing ketones at a rate, similar to that of a mature ruminant until 60 d of age, which is only at 40% of the ketogenic rate of a mature ruminant at 30 d of age [49]. Highest mRNA abundance to monocarboxylate transporter MCT1 and 4, on rumen epithelial cells cultured, was observed in calves during postweaning (55 to 58 d old) than in the preweaning period (22 to 34 d old) [44].

Conclusions
The physical form of starter evaluated in this study (pellet with steam-flaked corn, ground, or ground + 5% chopped Tifton hay) did not influence performance, feed efficiency, nutrient partitioning, nor rumen development of dairy calves. We did not find a difference in mRNA expression of genes involved in rumen metabolism. In this way, starter pellet with steam-flaked corn, ground, or ground + 5% chopped Tifton hay can be used for suckling calves.