Effects of Dietary Fermented Garlic on the Growth Performance , Relative Organ Weights , Intestinal Morphology , Cecal Microflora and Serum Characteristics of Broiler Chickens

I Department of Animal Science and Technology, College of Animal Bioscience and Technology, Konkuk University, 120 Neungdong-ro, Gwangjin-gu, Seoul 143701, Republic of Korea II Animal Resources Research Center, Konkuk University, 120 Neungdong-ro, Gwangjingu, Seoul, 143-701, Republic of Korea III Department of Physiology and Anatomy, College of Veterinary Medicine, 120 Neungdong-ro, Gwangjin-gu, Seoul 143701, Republic of Korea


INTRODUCTION
Garlic (allium sativum) has long history of use as culinary or medicinal supplement.As garlic contains a sulfur volatile active component that has antibacterial, anti-inflammatory and antioxidant biological properties (Wilson and Demming-Adams, 2007), it has been explored as a potential alternative to antibiotics in poultry production.Nonetheless, inconsistent effects of garlic on the growth performance, egg production, gut physiology, nutrient digestibility, and lipid metabolism have been reported (Khan et al., 2012).Garlic inclusion levels widely varied in those studies, being added at levels as low as 0.125% to as high as 5% to broiler diets.
Several attempts have been recently made to enhance garlic functionality and stability using fermentation with specific bacteria strains which are resistant to garlic, and are able to ferment it.The bacterial strains commonly used for garlic fermentation are Lactobacillus plantarum, Bacillus subtilis, Weissella koreensis, and Leukonostoc mesenteroides.Garlic fermented with either W. koreensis or L. mesenteroides equally produced higher sulfur-related metabolites, such as 3-vinyl-[4H]-1,2,-dithiin, all ylmethyl trisulfide, diallyl disulfide,

Effects of Dietary Fermented Garlic on the Growth Performance, Relative Organ Weights, Intestinal Morphology, Cecal Microflora and Serum Characteristics of Broiler Chickens
diallyl trisulfide, but lower alliin compared with fresh garlic (Yan et al., 2012;Hussain et al., 2014).
Recent feeding trials have been tested the effect of fermented garlic on pig and broiler performance (Ao et al., 2010(Ao et al., , 2011;;Kang et al., 2010;Wang et al., 2011;Yang et al., 2012;Yan & Kim, 2013;Hossain et al., 2014Hossain et al., , 2015;;Chun et al., 2015).The addition of graded levels (e.g., dietary inclusion levels of 0.05, 0.1, 0.2%) of L. mesenterodes-fermented garlic in the diets of broiler chickens increased body weight and feed intake, but did not affect feed efficiency, when compared with the control-diet fed chickens (Hossain et al., 2014(Hossain et al., , 2015)).These results corroborated with the study of Kang et al. (2011), who used L. plantarum for garlic fermentation and added higher levels (1%) to a broiler diet.In contrast with the studies of Kang et al. (2011) and Hossain et al. (2014, 2015), dietary W. koreensis-fermented garlic failed to affect the growth performance of broiler chickens, albeit it enhanced host immunity, retarded meat lipid oxidation, and lowered serum total cholesterol and triglyceride concentrations compared with the control chickens (Ao et al., 2011).
In order to further increase our understanding of the efficacy of fermented garlic, we tested whether garlic fermented by L. citreum SK2556, a novel patented strain isolated from an aged garlic product, could affect the growth performance of broiler chickens, and if indeed so, whether the FG-induced improvement in production performance was related to changes in intestinal morphology.To our knowledge, no studies on the effect of FG on the intestinal morphology of broiler chickens have been conducted to date.It was reported that dietary non-fermented garlic exhibited antibacterial activity and improved the intestinal morphology of broiler chickens (Abdullah et al., 2010;Ham et al., 2010).Thus, the present study aimed at investigating the effects of the dietary supplementation of FG on the growth performance, intestinal morphology and microbiota, and serum characteristics of broiler chickens.

Preparation of the fermented garlic powder
The peeled garlic bulbs used in this study were purchased from a local market and ground in a mill (Philips HR 2860, Netherlands) to obtain a homogenized garlic juice.The garlic juice was then mixed with a buffer (K 2 HPO 4 0.1 g/L, MgCl 2 0.01 g/L, NaCl 1 g/L, CH 3 COONa 1g/L) at the ratio of 30:70.The garlic and buffer mixture was sterilized by autoclaving at 121ºC for 15 min and cooled to room temperature.The mixture was then inoculated with a starter culture of 1% Leuconostoc citreum SK2556 in a shaking incubator at 37ºC for 24 h.The pH of the fermented garlic was 4.47.The L. citreum SK2556 culture contained a 1× 10 8 cfu/mL.The L. citreum SK2556 strain used in this study was originally isolated from pickled garlic food, characterized and patented due to its ability to grow and ferment in medium containing garlic (Kim et al., 2012).

Experimental design, animals and diets
A total of 250 feather-sexed male broiler chicks (Ross 308) were purchased from a local hatchery.Upon arrival, birds were individually weighed and randomly distributed into 25 floor pens with rice husks as a bedding material.
Five dietary treatments with five replicates (10 chicks per replicate, n=50 chicks/treatment) were applied.Starter and finisher diets based on corn and soybean meal were formulated (Tables 1 and 2) and used as non-medicated control diets (negative control, NC).The experimental diets were formulated by mixing the basal diet with either 10 ppm avilamycin (positive control, PC) or three levels of fermented garlic at the concentrations of 0.1% (FG1), 0.3% (FG3), or 0.5% (FG5).All diets contained maduramicin (5 mg/kg of diet) as a coccidiostat.Diet and water were provided on an ad-libitum basis.The initial room temperature was set at 33ºC, gradually decreased to reach 25ºC at 15 days of age and kept constant thereafter (Kang and Kim, 2016).All animal care procedures were approved by Institutional Animal Care and Use Committee in Konkuk University.

Measurements
Body weight and feed intake were measured by pen on a weekly basis and used to calculate feed:gain ratio.On day 35, three birds per pen were randomly selected for blood sampling after euthanasia.Immediately after blood sampling, the liver, spleen, and bursa of Fabricius were excised, weighed and expressed as relative weight in grams per 100 g of body weight.For the morphological determination of the villi, 1-cmlong sections from mid-jejunum and mid-ileum were sampled.Cecal content was sampled and kept on ice until measurement of microflora on a same day of the sampling.

Intestinal morphology
Intestinal samples were fixed in 10% phosphatebuffered formalin for a minimum of 48 h, and 4.0μm sections were prepared as described elsewhere (Kim et al., 2015).The sections were stained with standard hematoxylin-eosin solution and villus height, villus width (mid-point), and crypt depth were measured at 100× magnification under light microscopy.Villus surface area was calculated using the formula = (2π) × (VW/2) × (VL), where VW = villus width and VL=villus length (Sakamoto et al., 2000).Twenty well-oriented villi per bird were measured and averaged for each bird.

Enumeration of the cecal microflora
Approximately 1 g of cecal contents were mixed with 9 volumes of sterilized ice-cold saline solution (w/v), vortexed, and serially diluted.A volume of 100 μL of cecal suspension was plated on the respective agar plates and incubated at 37ºC for 48 h.Colony was read and expressed as colony forming unit (cfu) per gram of cecal contents.The media used were MRS (Difco, USA), EMB (BBL™ Eosin Methylene Blue Agar, USA), and SS-agar (BBL™ Salmonella Shigella Agar, USA).No colonies were detected in SS-agar.Colonies in MRS and EMB agars were selected and confirmed by 16S rRNA by PCR analysis using commercial primers (forward primer 5'-AGA GTT TGA TCC TGG CTC AG-3' and reverse primer 5'-AAG GAG GTG ATC CAN CCR CA-3'; Lane, 1991).Nucleotide sequence from colonies in MRS exhibited 100% homology with Bacillus licheniformis and those in EMB with Escherichia coli.

Statistical analysis
Pen was considered an experimental unit.All data obtained was subjected to one-way analysis of variance (SPSS Ver.21, IBM Corp., USA).If a significant effect was observed, the difference between treatments was determined by Tukey's test.Results were expressed as treatment means with their pooled SEM.In addition, linear contrasts were used to assess the effect of FG levels.A p<0.05 was considered statistically significant.

RESULTS
During the experimental period, none of the production-related parameters, i.e., daily body weight gain, feed intake and feed:gain ratio, were affected by the dietary treatments (Table 2).However, average daily gain on day 21 linearly increased (p=0.024) with increasing FG levels.
On day 35, organ weights, when expressed as relative to body weight, were not affected by the dietary treatments (Table 3); however, relative bursa weight progressively declined with increasing FG levels.Jejunal villus height did not differ among dietary treatments (Table 4).Villus width linearly decreased as FG levels increased (p=0.17).Jejunal crypt depth was significantly lower (p<0.05) in the FG1 and FG3 groups compared with the NC group.Jejunal villus area was not affected by the FG treatments.Villus height:crypt depth ratio (p=0.056) in the FG-fed groups tended to be higher (p=0.0558)compared with the NC group.Villus height:crypt depth ratio linearly increased (p=0.0181) with increasing FG levels.In contrast, villus height and width, villus surface area, crypt depth and villus height:crypt depth ratio in the ileum were not   diet exhibited significantly higher total protein and cholesterol concentration compared with those fed the NC diet.

DISCUSSION
The present study aimed at investigating if garlic fermented by L. citreum SK2556, a novel patent strain isolated from prickled garlic product, could affect the growth performance and the intestinal morphology of affected by the dietary treatments.Neither PC or FG vs. NC affected the population of cecal microflora (Table 5).It was confirmed that the nucleotide sequence from colonies in MRS exhibited 100% homology with Bacillus licheniformis and those in EMB with Escherichia coli.None of dietary treatments affected blood parameters, such as albumin, total or direct bilirubin, GGT, uric acid, or glucose (Table 6).Creatinine, BUN, GPT, and GOT were not influenced by dietary FG compared with the NC group.Broiler chickens fed the FG5-added broiler chickens.Earlier, we isolated L. citreum SK2556 strain from pickled garlic food, characterized and patented due to its ability to grow and ferment in a medium containing garlic (Kim et al., 2012).In contrast with the strain used in this study, previously reported strains (i.e., L. plantarum, B. subtilis, W. koreensis, and L. mesenteroides) were not isolated from garlic, nor detailed information on their garlic-fermenting capacity was informed.However, earlier studies (Wang et al., 2011;Hossain & Kim, 2014) reported that both W. koreensis-and L. mesenteroides-fermented garlic produced equal amounts of active metabolites, such as alliin, allylmethyltrisulfide, diallyltrisulfide, diallyldisulfide, and 3-vinyl-[4H]-1,2-dithiin.
The addition of FG to the broiler diets linearly increased average weight gain between 1-21 days, which partially explains the FG-induced decrease in feed:gain ratio.The results of the present study confirm the findings of previous studies (Choi et al., 2010;Kang et al., 2010;Ao et al., 2011;Onyimonyi et al., 2012;Hossain et al., 2014Hossain et al., , 2015) ) showing that neither FG or non-fermented garlic adversely affected the feed intake in broiler chickens.Thus, the underlying mechanism of the FG-induced increase in daily weight gain is likely the consequence of improvements in nutrient digestibility, especially at early ages.The latter assumption is supported by Issa & Abo Omar (2012), who reported that dietary garlic powder at the levels of 0.2 or 0.4% increased CP and EE digestibility in broiler chickens, although they did not find any garlic-mediated increase in growth performance.It was previously reported that dietary FG significantly increased body weight gain and feed:gain ratio, without any effect on feed intake, when added at the levels of 0.2% (Hossain et al., 2014(Hossain et al., , 2015) ) or 1% (Kang et al., 2010)in broiler diets.
In this study, we found that dietary FG significantly reduced jejunal crypt depth compared with the NC group.The ratio of villus height:crypt depth in jejunum linearly increased with increasing FG levels.The decrease in crypt depth observed in this study may indicate that FG may reduce villus turnover rate, and therefore, that relatively less energy is used for new tissue synthesis (Abdullah et al., 2010), ultimately reducing energy requirements to maintain villus integrity.Given that villus:crypt ratio is an indicator of the digestive capacity of the small intestine (i.e., jejunum), the increased ratio observed in this study may provide the evidence that FG may play a role in digestion and absorption, as reported elsewhere (Adibmoradi et al., 2006).Therefore, the question is how dietary FG affects intestinal morphology?One  plausible explanation would be the effect of FG on gut microflora.Unfortunately, neither dietary FG or in-feed antibiotic (i.e., PC) affected the cecal microflora in this study.If this holds true for the jejunum, then the next plausible explanation is the direct impact of effectors present in the FG preparation on gut function, as also proposed by Abdullah et al. (2010).
To our surprise, we found that relative weight of bursa of Fabricius was linearly declined with increasing FG levels.This is in sharp contrast to the findings of Rahimi et al. (2011), who reported that dietary garlic significantly increased the relative weight of bursa of Fabricius compared with the non-medicated NC dietfed chickens.In addition, Hossain et al. (2014Hossain et al. ( , 2015) ) found no effect of L. mesenteroides-fermented garlic on the relative bursa weight of broiler chickens.At this stage, we do not know whether FG-mediated decrease in bursa weight impaired host immunity as we did not measure any immune-related parameters.Enhanced immune modulation by dietary garlic has been well established in broiler chickens (Ao et al., 2011;Kim et al., 2013) and it is considered the putative underlying mechanism in the garlic-mediated improvement of Eimeria-induced growth depression (Arczewska-Wlosek and Swiatkiewicz, 2013; Kim et al., 2013).Similarly, the enhanced host innate and/or acquired immunity by plant-derived phytochemicals is known to prevent Eimeria infection in broiler chickens (Lee et al., 2011;Kim et al., 2013).Further studies are warranted to see whether dietary FG can alter host immune response in broiler chickens or render chickens resistant or susceptible to enteric pathogens, such as Eimeria spp. or Clostridium perfringens.
Blood parameters were not different between the NC and PC groups.This was expected, as antibiotics are not absorbed in the intestine, therefore are less likely to exhibit direct systematic effect.On the other hand, the addition of FG to the diet altered various blood characteristics of the evaluated broilers.First, higher serum total cholesterol levels were determined in the FG5 vs. NC groups, which is contrary to the wellestablished hypocholesterolemic effect of garlic (Chi et al., 1982;Konjufca et al., 1997;Ao et al., 2010Ao et al., , 2011;;Khan et al., 2012).This contradictory result may be related to the added levels of garlic or to the fat source used.Chi et al. (1982) concluded that the effective garlic level to lower cholesterol in rats is approximately 2% when diets contain either exogenous cholesterol or 15% lard as a fat sources.Ao et al. (2010) also observed the hypocholesterolemic effect of fresh garlic added at the concentration of 2% to layer diets.In another report by Ao et al. (2011), the addition of W. koreensis-fermented garlic at 0.2% to a broiler diet, containing soybean oil as the single fat source, significantly lowered serum cholesterol compared with the NC group.On the other hand, adding garlic at 0.3% or 0.6% to broiler diets containing approximately 2% of soybean oil did not affect serum cholesterol levels in broiler chickens (Amouzmehr et al., 2012).Whether FG at low levels in diets containing either animal fat or plant oil as a single fat ingredient would exhibit identical effect on lipid metabolism needs to be clarified.The higher total serum protein levels in FG5 compared with NC seems secondary to the increased cholesterol levels.
In conclusion, dietary FG linearly increased average daily gain during the starter phase (1-21 days) and decreased the relative weight of the bursa of Fabricius of 35-d-old broiler chickens.Jejunal, but not ileal, crypt depth was significantly lower both in the FG1 and FG2 groups compared with the NC group.Consequently, villus height:crypt depth ratio in the jejunum linearly increased as FG levels increased.Finally, dietary FG significantly increased serum total protein and cholesterol concentrations compared with the NC group.Further studies are warranted to determine of dietary FG can modulate local or systemic immune responses in naïve chicken or, preferentially, in enteric disease chicken models, and affect lipid metabolism in the presence or absence of exogenous cholesterol in the diet.

Table 1 -
Ingredients and composition of basal diets

Table 4 -
Effect of fermented garlic on the intestinal morphology of broiler chickens1,2 2 Mean values are presented (n=5, each group).