Blend of essential oils can reduce diarrheal disorders and improve liver antioxidant status in weaning piglets

Objective This study was to assess the effects of different doses of an essential oil blend (EOB) on growth performance, diarrhea occurrence (DO), hematological and blood biochemical profile, intestinal morphometry, morphology and microbiology, relative weight and length of organs, digestive content pH, and liver antioxidant status in weaning piglets. Methods A total of 135 barrows (7.09±0.29 kg body weight) were allotted randomly in a randomized complete block design based on body weight with nine replications and three animals per pen. Dietary treatments were a negative control (NC): basal diet; positive control (PC): NC plus 125 mg performance-enhancing antibiotic (enramycin 8%)/kg diet; NC plus 100 mg EOB/kg diet (EO100); NC plus 200 mg EOB/kg diet (EO200); and NC plus 400 mg EOB/kg diet (EO400). Diarrhea occurrence was monitored daily, and performance at the end of each phase. Results Gain to feed ratio was greater (p<0.05) in starter II pigs fed EO400 and EO200 than in those fed EO100. Pigs fed EO400 had lower (p<0.05) DO than those fed NC and EO100 in the total period. Pre-starter II pigs fed NC had (p<0.05) lower serum total protein and plasma protein than pigs fed PC. Pigs fed EO100 showed smaller (p<0.05) mean corpuscular volume (MCV) than pigs fed EO400. Starter II pigs fed EO400 had (p<0.05) greater MCV and lower mean corpuscular hemoglobin and erythrocytes than those fed EO100. There was a greater concentration (p<0.05) of band cells for PC, similar to EO400 and EO200. Performance-enhancing antibiotic and EOB to diets increased (p<0.05) liver superoxide dismutase activity. Conclusion Adding 200 and 400 mg EOB/kg diet decreased DO and was advantageous to hematological and blood biochemical profile and liver antioxidant status without being detrimental to growth performance and gastrointestinal health in nursery pigs.


INTRODUCTION
Immediate post-weaning is a critical phase for weaning piglets because of the transition from a liquid (milk) to a solid diet while the gastrointestinal tract (GIT) and immune system are immature. Furthermore, other stressing factor such as sow separation, transportation, and new social and environmental interactions can be additional challenges for weaning piglets.
The occurrence of intestinal disorders that affect growth performance is common in weaning piglets. This can be aggravated by the prevalence of pathogenic microorganisms [1]. To alleviate these disorders and optimize animal productivity, sub-therapeutic doses of antimicrobials are added to diets continuously acting as performance-enhancing antibiotics. Conversely, the wide use of performance-enhancing antibiotics (PEA) in animal production coupled with the possible presence of their residues in animal food and in the environment has become a great concern. This is based on the emergence of resistant bacteria with the potential to cause diseases in animals and humans [2].
The use of some antimicrobials as PEA has been banned as food additives throughout the European Union on January 1st, 2006. This has led to the use of other still permitted, such as enramycin, a polypeptide produced during Streptomyces fungidicus fermentation. It has antibiotic action against gram-positive bacteria, such as Clostridium perfringens [3]. Even so, there is strong pressure for a complete prohibition of PEA use and hence research has focused on studying alternative substances, among which are phytogenic feed additives. These are plant-derived substances that can be classified according to their origin or processing [4].
Essential oils (EO) are phytogenic substances composed of volatile and lipophilic compounds. Essential oils can be extracted from plants through different processes, such as cold expression or steam distillation [4]. These additives have antimicrobial potential due to their ability to change the structure of bacteria cell membranes. Because of their hydrophobic characteristic, EO can associate with proteins present in the membrane and change their functions, affecting cell permeability and resulting in loss of viability. Essential oils are also able to positively affect immune response through modulation of the activity of cellular proteins responsible for releasing inflammatory molecules. Phenolic compounds present in EO can contribute to the body' s antioxidant system through the upregulation of signaling pathways involved in the expression of antioxidant enzymes. Furthermore, EO have been previously reported to increase digestive enzymes secretion in GIT epithelium [5].
Essential oils in their free form when orally administered are rapidly absorbed in GIT (mainly stomach and proximal small intestine). Encapsulated EO can show a slower rate of absorption. Once in the bloodstream, EO by-passe body tissues and most are metabolized by kidneys and excreted in the urine, with a half life of approximately two hours. Thus, there is little possibility of accumulation in animal tissues [6].
There is a wide variety of EO that can be extracted from plants, each with its biological singularity, but few studies have focused on the combinations of more than a couple of isolated compounds. Therefore, we hypothesized that an administration of an EO blend (EOB) (up to 400 mg/kg diet) as an alternative for conventional PEA improves the response of pigs and hence decreases diarrhea, increases liver antioxidant capacity, and growth performance. Based on this line of reasoning, this study was conducted to assess the effects of different doses of EOB on growth performance, diarrhea occurrence (DO), hematological and blood biochemical profile, intestinal morphometry, morphology and microbiology, relative weight and length of organs, digestive content pH, and liver antioxidant status in weaning piglets.

MATERIALS AND METHODS
The study was conducted in the swine center of the experimental farm (Professor Antônio Carlos dos Santos Pessoa) belonging to Unioeste, Brazil.

Animal care
All animal procedures were approved by the Ethics Committee on the Use of Animals in Experimentation at the Universidade Estadual do Oeste do Paraná -Unioeste (protocol no. 02/2020-CEUAP).

Animals, experimental design, and housing
A total of 135 crossbreed barrow piglets (Landrace×Large White, Agroceres♂ and DanBred♀) weaned at 25 d were used. Initial body weight (BW) was 7.09±0.29 kg. Pigs were allotted randomly to one of the five treatments in a randomized complete block design with three replications in each of three batches over time resulting in a total to nine replicates per treatment and three animals per pen.
Pigs were weighed, identified with an ear tag, and housed in a masonry nursery room with raised slatted plastic floor pens (1.5 m 2 ) equipped with gutter feeders and nipple drinkers. Pens were washed daily (in the afternoon) with running water. The experiment lasted 42 d.
All diets were corn and soybean meal-based and were offered as mash. Diets were divided into four phases (prestarter I, pre-starter II, starter I, and starter II) and formulated taking into account the nutritional requirements limits proposed by Rostagno et al [7]. Pigs were allowed ad libitum access to feed and water throughout the experiment.

Growth performance and diarrhea occurrence
The average final BW (kg), the average daily gain (ADG, g/d), the average daily feed intake (ADFI, g/d), and gain to feed ratio (G:F, g/g) were evaluated to address growth performance. Pigs were weighed (UL-50 Digital Scale; DIGI-TRON, Curitiba, Brazil) at the beginning and the end of each phase, as well as the feed offered and leftovers. Average daily gain was determined by the difference between initial and final BW divided by the days of each phase. Average daily feed intake was calculated by subtracting the leftovers from the amount of feed offered in each phase. Gain to feed ratio was calculated by dividing ADG by ADFI. Diarrhea occurrence was observed throughout the experiment. Feces of animals from each treatment were scored daily (1000 h) on a 4-point scale (0 = formed feces, 1 = soft feces, 2 = semi-solid feces, and 3 = liquid feces) as previously described by Huang et al [9]. Data on fecal scores were transformed into binary traits as follows: 0 = no diarrhea (fecal scores = 0 and 1) and 1 = presence of diarrhea (fecal scores = 2 and 3). The frequency of DO in each phase was calculated by dividing the sum of DO by the total number of observations in each treatment and then multiplying by 100%. Results were expressed as observed percent (%).

Hematological and blood biochemical profile
At the end of pre-starter II and starter II, 18 pigs per treatment were selected for blood collection based on the closest BW relative to the average BW in each experimental unit.
Pigs fasted for 10 h and then blood samples (≅ 20 mL) were withdrawn from vena cava (0800 h) using 0.70×30 mm needles. Blood was collected in tubes containing ethylenediaminetetraacetic acid for determination of complete blood count, potassium fluoride for determination of glucose (GLU), heparin for determination of urea (URE), alanine aminotransferase (ALT), and alkaline phosphatase (AP). Tubes with no anticoagulant were used for the determination of albumin (ALB) and total protein (TP). Blood samples were placed on ice inside of a cooler (4°C) to be sent to the laboratory.
Samples for biochemical evaluation were centrifuged at 3,000 g (80-2B; Centrilab, Hangzhou, China) for 10 minutes at the blood analysis laboratory of Unioeste. Then, approximately 3 mL of supernatants were collected and duplicates were stored in microtubes at -20°C. Albumin and TP were analyzed via colorimetry, ALT via the kinetic method, and AP, URE, and GLU via enzymatic-colorimetry, using commercial kits (Gold Analisa Diagnóstica Ltda., Belo Horizonte, Brazil) and a Bel SPECTRO S05 spectrophotometer (Bel Engineering, Monza, Italy).

Intestinal microbiology
At the end of the experimental period (42 d), six piglets per treatment were slaughtered for GIT sampling. Pigs were selected based on the closest BW relative to the final average BW of all replicates. Pigs were slaughtered in a commercial abattoir after fasting for approximately 10 hours. All euthanasia procedures were performed by electronarcosis following Brazilian guidelines (Resolution No. 37 of February 15, 2018, CONCEA).
Samples of jejunum, ileum, and colon contents were collected and individually placed in sterile plastic containers. Then, at the microbiology lab from Unioeste, 1 g of each sample was serially diluted in 1% peptone water. Each dilution was vortexed (K45-2810; KASVI, Taizhou, China) for 30 seconds. A 100 μL aliquot of each diluted sample was spread evenly on the surface of eosin-methylene blue agar plates for Enterobacteriaceae (ETB) counting. A 1 mL of diluted sample was inoculated on tryptose-sulfite-cycloserine and de Man, Rogosa and Sharpe agar plates for sulfite-reducing clostridia (SRC) and lactic acid bacteria (LAB) counting, respectively.
Then, plates were incubated at 37ºC (EL 202; Eletrolab, São Paulo, Brazil) aerobically for 24 h (ETB) and anaerobically for 48 h (SRC and LAB) as previously described by Da Silva et al [10]. After incubation, bacterial colonies were counted and data were multiplied by the respective dilution and log-transformed (Log 10 ). Results were expressed in log colony forming unit/g.

Intestinal morphometry and morphology
Samples (≅ 3 cm) of jejunum (150 cm cranial to ileocecal junction), ileum (15 cm cranial to ileocecal junction), and colon (100 cm caudal to ileocecal junction) were taken. Fragments collected were washed with saline (0.9% sodium chloride) and stored in sterilized plastic containers with a 10% buffered formalin solution.
Samples were sent to a commercial laboratory (Mercolab, Cascavel, PR, Brazil) where they were processed in paraffin, and stained with hematoxylin and eosin to prepare slides as previously reported by Prophet et al [11]. A total of 10 villus height (VH) and their crypt depth (CD) were measured. Then, VH to CD ratio was calculated as previously described by Kisielinski et al [12]. Intestinal morphology was also assessed via infiltrate, congestion, desquamation, coccidiosis, bacterial lumps, band cells, cysts, mucus, necrosis, and edema observations as previously described by Kraieski et al [13]. Histological analyses were performed using an optical microscope (CX31RTSF; Olympus, Tokyo, Japan) and a computer system (ToupView ×86).

Relative weight and length of organs and pH of the digestive contents
After slaughter, the contents of the stomach, jejunum, ileum, cecum, and colon of the pigs were sampled and stored in plastic containers. Then, pH was measured using a digital pH meter (TEC-2 mp; TECNAL, Piracicaba, Brazil). The liver and gallbladder, heart, spleen, kidneys, empty stomach, small intestine (duodenum, jejunum, and ileum), and large intestine (cecum, colon, and rectum) were weighed using a digital scale (MX-111; Maxon, Beijing, China). Relative organ weight to final BW was then calculated. Small and large intestine length was measured using a measuring tape.
Liver samples were homogenized in potassium phosphate buffer solution (pH 6.5) with 10-fold dilution for GSH, SOD, CAT, and LP analyses, or 30-fold dilution for GST analysis. Samples were then centrifuged (NT 805; Nova Técnica, Piracicaba, Brazil) at 10,000 g at 4°C for 20 min.
The activity of CAT was measured as previously described by Aebi [14]. Briefly, the reaction was performed using 5 mM hydrogen peroxide and 50 mM phosphate buffer (pH 7.0) solution with the cytosolic protein. The reaction was continuously monitored at 240 nm for 60 sec using a microplate reader (Synergy HT; Biotek, Winooski, VT, USA). The extinction coefficient was 41 mM/cm. Superoxide dismutase activity was determined based on inhibition of the autoxidation of pyrogallol [15]. Briefly, samples (60 μL) were diluted in 1,327.5 μL Tris-HCl buffer (0.4 M, pH 8.9) and vortexed (QL-901; Biomixer, Rancho Cucamonga, CA, USA). Then, 75 μL of pyrogallol (15 mM) were added to diluted samples and incubated for 30 min at room temperature. The reaction was stopped by adding 37.5 μL HCl (1 N) and then absorbance was measured at 440 nm using the microplate reader. The amount of enzyme that inhibited the reaction by 50% was defined as one unit of SOD. The enzyme activity was expressed as SOD unities per milligram of total LP (SOD U/mg of LP).
Glutathione S-transferase analysis was performed by adding 200 μL of a solution (containing 3 mM 1-chloro-2,4dinitrobenzene (CDNB), diluted in ethanol, and 3 mM GSH, diluted in potassium phosphate buffer) to 100 μL of sample supernatant. The linear increase of absorbance at 340 nm was monitored using an extinction coefficient of 9.6 mmolar/cm. Glutathione S-transferase catalyzes the conjugation of CDNB to GSH (reduced glutathione). The product of this reaction is a thioether that can be monitored by the increase in absorbance, as previously described by Habig et al [16].
Glutathione peptide concentration was measured by the method of Sedlak and Lindsay [17]. Briefly, 80 μL of trichloroacetic acid (12.5%) was added to 100 μL of the sample homogenate and centrifuged at 6,000 g at 4°C for 15 min. Then, 20 μL of the supernatant was mixed with 280 μL of Tris-HCl buffer (0.4 M, pH 8.9) and 5 μL of 5,5′-dithiobi-2-nitrobenzoic acid in methanol. Absorbance was measured at 415 nm using the microplate reader. A known GSH concentration solution was used as an external standard.
Liver protein was determined using bovine ALB as a standard as previously described by Bradford [18]. Briefly, 10 μL of the sample was added to 250 μL of Bradford solution in each microplate well. Absorbance was measured at 595 nm using the microplate reader. The LP results were used to calculate the previous liver variables and were expressed as mg of protein in liver homogenates.

Statistical analysis
Statistical analyzes were performed using the SAS University Edition software (SAS Inst. Inc., Cary, NC, USA). Residual error was evaluated for outliers based on the normal distribution curve via the Student test (RStudent). If studentized residuals met or exceeded three standard deviations, the sample was considered significant.
The normality of experimental errors and the homogeneity of variances among treatments were evaluated using Shapiro-Wilk and Levene tests, respectively. The non-influence (p>0.05) of treatments on the initial BW of the piglets, for indication as a covariate and correction of observed average values, was verified via analysis of variance (ANOVA).
Data on growth performance, blood biochemical profile, intestinal morphometry and microbiology, liver antioxidant status, relative weight and length of organs, and pH of the digestive content were analyzed via analysis of covariance or ANOVA. The model included treatment as a fixed effect and block and residual error as a random effect with an individual pen as the experimental unit. The factors included in the model used were Y ijk = μ+T i +b j +β (X ijk -Data on growth performance, blood biochemical profile, intestinal morphometry and microbiology, r antioxidant status, relative weight and length of organs, and pH of the digestive content were ... = overall average for the covariate X; ε ijk = random error of the plot associated with level i, block j, and replication k. A generalized linear model was fitted for each distribution and linkage function to analyze data on DO, intestinal morphology, and hematological profile. The treatment effect was verified via type III analysis. The Akaike information criteria was used to test the model fitting. Generalized linear model used was represented by the systematic portion: η = μ+T i +b j , wherein μ was the effect associated with the overall average; T i was the effect associated with i-th treatment class, for i = (1, 2, 3, 4, and 5) and b j was the effect associated with j-th block, for j = (1, 2, and 3).
The Student-Newman-Keuls test was used to compare pairs of treatment averages for growth performance, blood biochemical profile, intestinal morphometry and microbiology, liver antioxidant status, relative weight and length of organs, and pH of the digestive content. The Tukey-Kramer test was used to compare the averages for DO, intestinal morphology, and hematological profile. Significant differences were set at p<0.05. Results were reported as means with pooled standard error of the mean.

Growth performance and diarrhea occurrence
There was no effect of treatment on growth performance in pre-starter I, II, and starter I phases (Table 3). Starter II pigs fed EO400 and EO200 showed (p<0.05) greater G:F than those fed EO100. Pre-starter I pigs fed EO400 and NC had lower (p<0.05) DO than those fed EO100. In pre-starter II, piglets consuming EO400 exhibited (p<0.05) a reduction in DO than those on NC, EO100, and EO200. There was no effect of treatments on the DO in starter I phase. Starter II pigs fed EO400 showed (p<0.05) lower DO than those fed NC and EO100. A similar result was observed in the total experimental period.

Hematological and blood biochemical profile
Lower TP was observed (p<0.05) in pre-starter II pigs fed NC (Table 4). Pre-starter II pigs fed EO400 showed (p<0.05) greater MCV than pigs fed EO100 and NC. Pre-starter II pigs fed PC showed (p<0.05) greater PP than those fed NC. At the end of starter II, greater MCV (p<0.05) and lower ER (p<0.05) and MCH (p<0.05) were observed in pigs fed EO400 when compared to those fed EO100. Band cells concentration was greater (p<0.05) in pigs fed PC, with results similar to those fed EO200 and EO400.

Intestinal microbiology, morphometry and morphology
No treatment effect was observed on colony-forming units of ETB, SRC, and LAB in the different intestinal segments of the pigs (Table 5). There was no effect on intestinal morphometry and morphology in pigs ( Table 6).

Relative weight and length of organs and pH of the digestive contents
There was no treatment effect on relative weight and length of organs and pH of digestive content in pigs (Table 7).

Growth performance and diarrhea occurrence
In general, the different treatments did not affect growth performance in the present study. Kommera et al [19] suggested that a more controlled environment, such as experimental units, may affect the animals' response to the use of growthpromoting feed additives because there are fewer stressing factors and strict sanitation. The composition of the diets may have contributed to the results as well. It is worth mentioning that the nutritional support provided by the diets, associated with the absence of challenges, did not allow the tested additives to positively influence pig performance.
Weaning is characterized by environmental, social, and dietary changes which can negatively impact on the intestinal health and growth of nursery pigs and hence favoring DO [1]. In this study, supplementation of EOB to the diets, at the highest experimental dose, reduced DO. Similarly, Tian and Piao [20] observed lower DO in pigs fed 100 mg/kg of a feed additive containing thymol and cinnamaldehyde, and associated the results with small intestine morphology improvement. However, it was not observed in the present study, therefore we relate the antidiarrheal effect of the EOB to a possible GIT spasmolytic activity, mainly caused by cinnamaldehyde.
Diarrhea can be defined as the excretion of liquid feces along with water and electrolytes from the GIT which are expelled through abnormal intestinal muscle contractions. It can be associated or not with microorganism infections [21]. Cinnamaldehyde has been previously reported to decrease gastrointestinal motility in other species by blocking calcium channels, which are involved in muscle contractility, and by inducing adrenaline secretion, another spasmolytic agent [21,22].

Hematological and blood biochemical profile
Our observations suggest that DO may be related to the blood tests results because pigs fed EO100 showed higher ER and MCH, and lower MCV along with higher DO. Erythrocytes, MCV, and MCH are the number of red cells present in the sample, their volume, and the amount of hemoglobin contained in each one of them, respectively. In cases of dehydration, there may be a higher concentration of red cells in the sample due to a lower water concentration in blood. Thus, there may be an increase in hemoglobin in the sample, which has affected MCH, and a decrease in MCV because red cells with less intracellular water have a smaller volume [23].
Blood analysis also showed a treatment effect on TP and PP in pre-starter II pigs. Total protein analysis method differs from PP regarding its absence of fibrinogen in the sample. However, both analyses count the number of ALBs and globulins in the blood. Higher TP values coupled with no change in ALB may be related to an increase in the number of globulins [24]. Huang and Lee [25] mention that EO such as carvacrol, cinnamaldehyde and thymol, are able to increase blood levels of antibodies (e.g. globulins), and it is believed SEM, standard error of the mean; IBW, initial body weight; FBW, final body weight; ADG, average daily gain; ADFI, average daily feed intake; G:F, gain to feed ratio; DO, diarrhea occurrence. 1) NC, negative control: basal diet; PC, positive control: NC+125 mg performance-enhancing antibiotic (enramycin 8%)/kg diet; EO100, NC+100 mg essential oil blend/kg diet; EO200, NC+200 mg essential oil blend/kg diet; EO400, NC+400 mg essential oil blend/kg diet. 2) DO (%) = dividing the sum of DO by the total number of observations in each treatment and then multiplying by 100%. A,B Averages with different uppercase superscripts in the same row differ significantly by the Student-Newman-Keuls test (p < 0.05). a-c Averages with different lowercase superscripts in the same row differ significantly by the Tukey-Kramer test (p < 0.05).
to be through the modulation of mitogen-activated protein kinases and nuclear factor kappa B signaling, the two major pathways involved in immune responses. Regarding PP, globulins, and fibrinogen content in each sample may be associated with the results we observed. In starter II, there was a treatment effect on band cells which are neutrophils not fully developed and are part of the innate immune system. These cells play role in inflammatory responses to infections or tissue injuries. They are released into the bloodstream when the bone marrow is stimulated by pro-inflammatory substances [26]. Schepetkin et al [27] reported that EO can modulate neutrophils migration, thus improving the immune response. This explains the higher amount of band cells in pigs fed EOB. SEM, standard error of the mean. 1) NC, negative control: basal diet; PC, positive control: NC+125 mg performance-enhancing antibiotic (enramycin 8%)/kg diet; EO100, NC+100 mg essential oil blend/kg diet; EO200, NC+200 mg essential oil blend/kg diet; EO400, NC+400 mg essential oil blend/kg diet.

Intestinal microbiology, morphometry and morphology
Essential oils can improve intestinal epithelium and modulate its microbiota via an antimicrobial effect on pathogenic species [5]. In the study conducted by Wei et al [28], a lower count of Escherichia coli and a higher count of Lactobacillus in the jejunum of pigs fed a blend of 50 mg thymol/kg and 50 mg carvacrol/kg were observed. Tian and Piao [20] observed a higher duodenum VH and jejunal VH to CD ratio in pigs (8.1±1.4 kg BW) fed 100 mg thymol and cinnamaldehyde/kg when compared to pigs fed a control diet. However, no effect of treatments on intestinal morphometry, morphology, and microbiology was observed in the present study. A possible explanation for this is that EOB compounds were not microencapsulated. This is corroborated by Michiels et al [6] who reported that non-encapsulated compounds, such as carvacrol and thymol, are mostly absorbed in the stomach and hence less absorbed in the intestine. Furthermore, as they are rapidly absorbed by GIT epithelium there is little time to affect microbiota in the intestinal lumen.

Relative weight and length of organs and pH of the digestive contents
No treatment effect was observed for the relative weight and length of organs and pH of digestive contents. This concurs with the findings of Gois et al [29] and Cairo et al [30] who studied different doses (500; 1,000; 1,500 mg/kg) of Schinus terebinthifolius Raddi EO in pigs (5.65±0.78 kg BW) weaned SEM, standard error of the mean. 1) NC, negative control: basal diet; PC, positive control: NC+125 mg performance-enhancing antibiotic (enramycin 8%)/kg diet; EO100, NC+100 mg essential oil blend/kg diet; EO200, NC+200 mg essential oil blend/kg diet; EO400, NC+400 mg essential oil blend/kg diet.