Anti-obesity effect of kimchi fermented with Weissella koreensis OK1-6 as starter in high-fat diet-induced obese C57BL/6J mice

Authors


Correspondence

Youn-Soo Cha, Department of Food Science & Human Nutrition, Chonbuk National University, 664-14 Duckjin-dong, Jeonju, Jeonbuk 561-756, South Korea. E-mail: cha8@jbnu.ac.kr

and

Suk-Heung Oh, Department of Food Science and Biotechnology, Woosuk University, 443 Samnye-ro, Samnye-eup, Wanju-gun, Jeollabuk-do 565-701, South Korea. E-mail: shoh@woosuk.ac.kr

Abstract

Aims

In this study, we investigated the anti-obesity effects of kimchi (Korean traditional fermented vegetable) fermented either without starter culture or with a specific starter culture, Weissella koreensis OK1-6.

Methods and Results

C57BL/6J mice were divided into four groups (n = 7); normal diet, HF (high-fat diet), HF-KC (high-fat diet containing 3% kimchi manufactured without starter) and HF-KCO (high-fat diet containing 3% kimchi manufactured with the starter culture W. koreensis OK1-6). After 12 weeks of dietary intervention, the mice were killed, and serum and tissue samples were examined. Serum and hepatic lipid profile, insulin, leptin concentration and expression level of lipid anabolic genes like peroxisome proliferator-activated receptor γ, stearoyl-CoA desaturase-1, liver X receptor α and SREBP2 were significantly decreased (<0·05) along with body and epididymal fat pad weight in the HF-KCO group compared with the HF-KC and HF group.

Conclusions

These results suggested that kimchi fermented with the starter W. koreensis OK1-6 has anti-obesity effects in HF-induced obese mice.

Significance and Impact of the Study

These results may contribute to nutraceutical and food industries in developing functional food and probiotics based therapies for the treatment and prevention of obesity.

Introduction

Obesity is a complex disorder of appetite regulation and energy metabolism, which is controlled by interactions among genetic, environmental and psychosocial factors (Sharma and Padwal 2009; Kopelman 2000; Nguyen and El-Serag 2010). The prevalence of overweight and obesity has risen steadily in both adults and children for the past several decades. A primary reason for public health concern is that obesity results in adverse conditions and diseases such as cardiovascular disease, strokes, type 2 diabetes mellitus and many forms of cancer (Cope and Allison 2008; Dixon 2010). The prevalence of obesity and its complications is a worldwide problem with high costs associated with treatment and lost productivity, so there is increasing interest in alternative nutritional interventions for preventing obesity (Finkelstein et al. 2008; Cawley and Meyerhoefer 2011).

Kimchi is a traditional Korean fermented vegetable dish containing Asian cabbage, garlic, red pepper, green onion and ginger that is included in most meals on a daily basis (Lee et al. 2002; Lee and Kim 2010). Kimchi is rich in functional nutrients like vitamin C, vitamin K, dietary fibre, chlorophyll and phenols (Cheigh and Park 1994). Many reports have been published about the functional properties of kimchi, including anti-obesity (Kong et al. 2007; Kim et al. 2011), anti-cancer (Park 1995), anti-oxidative (Kwon et al. 1998), anti-diabetic (Islam and Choi 2009) and fibrinolytic effect (Jeong et al. 1995). It was postulated that dietary fibre, vitamins, calcium, capsaicin, allicin and niacin components of kimchi's ingredients could be key factors for the reductions in body weight, BMI and body fat (Sheo and Seo 2004; Kim et al. 2011). In addition, fermented kimchi intake showed better efficacy against systolic and diastolic blood pressures, fasting blood glucose, total cholesterol and body fat compared with fresh kimchi (Kim et al. 2011). However, studies on the anti-obesity effects of kimchi fermented with lactic acid bacteria (LAB) and on the possible involvement of metabolic products from LAB in the effects are lacking.

Kimchi is typically fermented by ambient bacteria in the local environment, and approx. 160 species of various lactic acid bacteria are known to contribute to kimchi fermentation (Cheigh and Park 1994). Study on the microbial population dynamics of kimchi showed that Weissella koreensis, a psychrophilic bacterium, is the most predominant species in kimchi, stored at −1°C (Cho et al. 2006). Recent studies reported that W. koreensis strains exhibit excellent ornithine-producing capacities and suggested that they may be useful as starters for fermented foods like kimchi or yogurt (Yu et al. 2009; Yu and Oh 2010).

Ornithine is an amino acid involved as an intermediary compound in the urea cycle. Ornithine has been proven to be effective in increasing muscle synthesis and rapid metabolization of body fat by stimulating growth hormone secretion (Edmonds et al. 1987; Zajac et al. 2010), improving insulin release (Blachier et al. 1989), boosting immune system (Kawai et al. 2000), enhancing wound healing response (Shi et al. 2002) and reducing blood ammonia concentrations associated with several hepatic diseases and brain oedema (Rose et al. 1999). Recently, we observed that W. koreensis OK1-6 was highly effective in inhibiting the intracellular lipid accumulation in the differentiating adipocyte (Cha et al. 2010; Moon et al. 2012).

Based on these observations, we hypothesized that kimchi fermented using ornithine-producing W. koreensis OK1-6 as a starter will be more beneficial in preventing weight gain than kimchi manufactured without added starter in a murine model of diet-induced obesity.

Materials and methods

Preparation of kimchi powder

Powdered kimchi samples were prepared for the study. Briefly, Napa cabbage (Brassica rapa ssp. pekinensis) was soaked in brine for 12 h and then drained for 12 h. The seasoning paste was made by combining the red pepper, chopped garlic, ginger, salt water, green onions and chives. The seasoning pastes were either inoculated with the starter (W. koreensis OK1-6) or left without any starter. The prepared cabbages and seasoning pastes were mixed and fermented at room temperature for 24 h. The two varieties of kimchi (KC kimchi, kimchi manufactured without a starter; and KCO kimchi, kimchi manufactured with W. koreensis OK1-6 as the starter) were then stored in a kimchi refrigerator (−1°C, DISO R-D305SJ; LG Electronics Inc., Seoul, Korea) for 14 days. After the refrigeration period, the characteristics of kimchi such as pH, acidity, salinity, the contents of reducing sugar and total number of lactic acid bacteria were analysed as previously described (Seok et al. 2008). The contents of amino acids such as arginine and ornithine were analysed by the HPLC method as described (Yu et al. 2009). Thereafter, the two types of kimchi powder (KC and KCO) were prepared for use by hot-air drying (60°C, 40 h) and grinding.

Animals and diets

Male C57BL/6J mice, aged 4 weeks, were purchased from Charles River Laboratories (Tokyo, Japan). The animals were maintained on a pellet diet (Research Diets, New Brunswick, NJ, USA) for 1 week, and then randomly divided into four groups: normal diet (ND group), high-fat diet group (HF group), high-fat diet containing 3% kimchi fermented without the starter (HF-KC group) and high-fat diet containing 3% kimchi fermented with W. koreensis OK1-6 (HF-KCO group). Animals were assigned to groups using a randomized block design (n = 7) such that the average weight in each groups was comparable. Animal diets were purchased from Research Diets (New Brunswick, NJ, USA). The compositions of the experimental diets are shown in Table 1. Animals were housed in a temperature-controlled environment with a 12 h of light and dark cycle. The animals were given free access to food and water during the entire experimental period. The food intake and body weight were measured daily and weekly basis respectively using a digital weighing scale (PL6001-S; Mettler Toledo, Columbus, OH, USA). Biomarkers selected for this study were based on their role in biological systems with known involvement in the adipose tissue physiology and lipid metabolism. The experimental protocol was approved by the Animal Care and Use Committee of Chonbuk National University.

Table 1. The composition of experimental diets
Ingredient (g)NDaHFbHF-KCcHF-KCOd
  1. a

    Normal diet (ND); AIN-93 Modified diet with 4% fat (10% fat calories) content.

  2. b

    High-fat diet (HF); AIN 93 Modified high-fat diet with 35 fat (60% fat calories) content.

  3. c

    High-fat diet containing 3% kimchi manufactured without the starter.

  4. d

    High-fat diet containing 3% kimchi manufactured with the starter of Weissella koreensis OK1-6.

Casein18·9625·8425·0725·07
Cystine0·280·390·380·38
Corn starch29·860·000·000·00
Maltodextrin3·3216·1515·6715·67
Sucrose33·178·898·628·62
Cellulose4·746·466·276·27
Soybean oil2·373·233·133·13
Lard1·9031·6630·7130·71
Mineral0·951·291·251·25
Dicalcium phosphate1·231·681·631·63
Calcium carbonate0·520·710·690·69
Potassium citrate1·562·132·072·07
Vitamin mix0·951·291·251·25
Choline bitartrate0·190·260·250·25
Kimchi3·003·00
Total (g)100·00100·00100·00100·00
kcal 100 g−1385524522522
kcal g−13·95·25·25·2

Collection of serum and tissue samples

Blood samples were collected after 12 h overnight fasting by orbital vein puncture and kept on ice for 1 h. Serum was separated from the blood by centrifugation at 1100 g for 15 min at 4°C (Micro 17R; Vision scientific Co., Seoul, Korea) and stored at −80°C until analysed. After blood collection, organs such as liver and epididymal fat were surgically removed, washed in phosphate-buffered saline solution, wiped with a paper towel and stored at −80°C until analysed.

Analysis of lipids in serum and liver

Triglyceride (TG), total cholesterol (TC) and high-density lipoprotein-cholesterol (HDL-c) concentrations in serum and liver tissue were measured by an enzymatic method using a commercial kit (Asan Pharmaceutical Co., Seoul, Korea).

Analysis of insulin and leptin

Serum insulin and leptin concentrations were measured by an ELISA method using a commercial Mouse Insulin ELISA Kit (Shibayagi, Shibukawa, Japan) and Quantikine Immunoassay kit (R&D System, Minneapolis, MN, USA), respectively.

Hepatic mRNA expression analysis of lipid-regulating genes

Total RNA was extracted from liver tissue using Trizol reagent (Invitrogen Life Technologies, Carlsbad, CA, USA) and the concentration measured spectrophotometrically. The extracted RNA was reverse transcribed into complementary DNA using a high capacity cDNA reverse transcription kit (Applied Biosystems, Foster city, CA, USA). Then the RNA expression level was quantified by a quantitative real-time PCR using SYBR Green PCR Master Mix (Applied Biosystems, Woolston, Warrington, UK) and the 7500 Real-Time PCR system (Applied Biosystems, USA) according to the manufacture's protocol. Gene specific primers used are given in Table 2. Relative quantification of gene expression with real-time PCR data was calculated relative to β-actin.

Table 2. Primers for RT-PCR
GeneSenseAnti-sense
  1. ACC, acetyl-CoA carboxylase; C/EBPα, CCAAT/enhancer-binding protein α; CPT-1, carnitine palmitoyltransferase 1; FAS, fatty acid synthase; LXRα, liver X receptor α; PPARγ, peroxisome proliferator-activated receptor γ; SREBP-1c, sterol regulatory element-binding protein-1c; SREBP-2, sterol regulatory element-binding protein 2; SCD-1, stearoyl-CoA desaturase-1.

ACC5′-CCAACATGAGGACTATAACTTCCT-3′5′-TACATACGTGCCGTCAGGCTTCAC-3′
C/EBPα5′-AGCAACGAGTACCGGGTACG-3′5′-TGTTTGGCTTTATCTCGGCTC-3′
CPT-15′-AAAGATCAATVGGACCCTAGACA-3′5′-CAGCGAGTAGCGCATAGTCA-3′
FAS5′-AGGGGTCGACCTGGTCCTCA-3′5′-GCCATGCCCAGAGGGTGGTT-3′
LXRα5′-GCGTCCATTCAGAGCAAGTGT-3′5′-TCACTCGTGGACATCCCAGAT-3′
PPARγ5′-GCTGTTATGGGTGAAACTCTG-3′5′-ATAAGGTGGAGATGCAGGTTC-3′
SCD-15′-CATCATTCTCATGGTCCTGCT-3′5′-CCAGTCGTACACGTCATTTT-3′
SREBP-1c5′-GCGGAGCCATGGATTGCAC-3′5′-CTCTTCCTTGATACCAGGCCC-3′
SREBP-25′-GGTGTGATTGTCTTGAGCGTCTT-3′5′-GCCGAATAGTTCGCCGAA-3′
β-actinF: 5′-GTGGGGCGCCCCAGGCACCAG-3′R: 5′-CTCCTTAATGTCACGCACGATTTC-3′

Statistical analysis

All values are expressed as mean ± SD. The data were analysed by the Student's t-test or one-way anova using spss ver. 12·0 (SPSS Inc., Chicago, IL). The differences among groups were assessed using Duncan's multiple range test. Statistical significance was considered at P-value < 0·05.

Results

Characteristics kimchi powders

The characteristics and amino acid levels of the two varieties of kimchi are shown in Table 3. Kimchi fermented with the W. koreensis OK1-6 starter had higher total lactic acid bacteria counts than the traditionally fermented kimchi using only ambient bacteria, and also had a lower pH (4·3 vs 4·8), slightly higher acidity and lower reducing sugar content. The ornithine content was 21·43 mg 100 g−1 and 117·06 mg 100 g−1 in kimchi without starter (KC) and with starter culture W. koreensis OK1-6 (KCO), respectively. However, the differences in acidity and reducing sugar content had little effect on flavour, and the two types of kimchi were almost indistinguishable by sensory evaluation (data not shown).

Table 3. Characteristics of kimchi harvested at 14th day of storage
SampleKC kimchiKCO kimchi
  1. KC kimchi, kimchi manufactured without the starter; KCO kimchi, kimchi manufactured with the starter of Weissella koreensis OK1-6.

  2. Values are mean ± SD.

  3. *Significantly at (P < 0·05) between two groups by the Student's t-test.

  4. F.W., fresh weight.

pH4·80 ± 0·024·30 ± 0·01*
Acidity (%)0·84 ± 0·170·88 ± 0·08
Salinity (%)1·70 ± 0·011·70 ± 0·01
LAB content (log cfu ml−1)7·15 ± 0·057·76 ± 0·08*
Reducing sugar (mg ml−1)16·80 ± 0·0116·24 ± 0·07*
Ornithine (mg 100 g−1 F.W.)21·43 ± 4·29117·06 ± 2·89*
Arginine (mg 100 g−1 F.W.)170·21 ± 14·1488·84 ± 1·33*

Body weight, food intake, food efficiency ratio and epididymal fat pad weight

C57BL/6J mice fed with HFs for 12 weeks developed obesity associated with both increased body weight and fat mass, although the initial body weights were not different among the groups. However, by the end of the study, body weights and body weight gains in the HF-KCO group were significantly lower (<0·05) than the HF group (Table 4). The feed efficiency ratio was also significantly higher (<0·05) in the HF group compared with the HF-KC and HF-KCO groups (Table 4). The epididymal fat pad weight in the HF-KCO and HF-KC groups were significantly lower (<0·05) than the HF group (Fig. 1).

Figure 1.

Epididymal fat pad weight. Values are mean ± SD. Values with different superscripts are significantly different by anova with Duncan's multiple range test (P < 0·05). ND, normal diet; HF, high-fat diet; HF-KC, high-fat diet containing 3% kimchi manufactured without the starter; HF-KCO, high-fat diet containing 3% kimchi manufactured with the starter of Weissella koreensis OK1-6.

Table 4. Body weight gain, feed intake and feed efficiency ratio
GroupsNDHFHF-KCHF-KCO
  1. ND, normal diet; HF, high-fat diet; HF-KC, high-fat diet containing 3% kimchi manufactured without the starter; HF-KCO, high-fat diet containing 3% kimchi manufactured with the starter of Weissella koreensis OK1-6. [Feed Efficiency Ratio = Total weight gain/Total feed intake].

  2. The table shows changes in the body weight feed intake and feed efficiency ratio between the different groups.

  3. Values are mean ± SD.

  4. Values with different superscripts are significantly different by anova with Duncan's multiple range test (P < 0·05).

Initial body weight (g)18·65 ± 0·3518·35 ± 1·6319·55 ± 0·3518·85 ± 0·64
Body weight gain (g)8·12 ± 0·02c20·42 ± 2·79a16·32 ± 1·77ab14·45 ± 0·13b
Feed intake (g day−1)2·71 ± 0·11a2·21 ± 0·15b2·34 ± 0·02b2·43 ± 0·03b
Feed efficiency ratio0·04 ± 0·00c0·11 ± 0·01a0·08 ± 0·01b0·07 ± 0·00b

Serum and hepatic lipid profile

The effects of KCO supplementation on serum and hepatic lipid profile are shown in Table 5. Serum triglyceride was not significantly different among the entire group however; KCO was effective in lowering the hepatic triglyceride, which was comparable with the ND group, whereas KC failed to demonstrate such an effect. Serum total cholesterol was significantly decreased (<0·05) in the HF-KCO group compared with HF group. However, the hepatic total cholesterol and serum HDL-cholesterol level did not showed any significant differences among the group. The serum HDL-cholesterol to total cholesterol percentage ratio was significantly increased (<0·05) in the HF-KCO group compared with the HF group.

Table 5. Lipid concentration in serum and liver
GroupsNDHFHF-KCHF-KCO
  1. ND, normal diet; HF, high-fat diet; HF-KC, high-fat diet containing 3% kimchi manufactured without the starter; HF-KCO, high-fat diet containing 3% kimchi manufactured with the starter of Weissella koreensis OK1-6; HDL-c, high-density lipoprotein-cholesterol.

  2. The table shows changes in the lipid profile between the different groups.

  3. Values are mean ± SD.

  4. Values with different superscripts are significantly different by anova with Duncan's multiple range test (P < 0·05).

Serum (mg dl−1)
Triglyceride136·08 ± 21·36144·37 ± 21·76145·78 ± 33·73124·66 ± 34·43
Total cholesterol152·41 ± 12·27b201·10 ± 34·18a162·64 ± 32·99b144·92 ± 25·59b
HDL-cholesterol88·54 ± 8·6687·83 ± 27·1977·25 ± 23·4680·42 ± 15·21
HDL-c/TC (%)50·43 ± 12·80a42·98 ± 9·55b46·66 ± 7·42ab56·19 ± 11·25a
Liver (mg g−1)
Triglyceride42·46 ± 6·66b68·17 ± 19·18a62·30 ± 14·82a45·28 ± 10·39b
Total cholesterol11·22 ± 1·9010·22 ± 1·4511·42 ± 1·249·98 ± 1·30

Serum leptin and insulin concentrations

The expression and secretion of leptin are highly correlated with body fat mass and adipocyte size (Houseknecht et al. 1996) and were significantly lower (<0·05) in HF-KC and HF-KCO groups (Fig. 2) demonstrating a reduced fat mass in both groups (Fig. 1). Serum insulin levels in the HF-KCO group were also significantly lower (<0·05) than in the HF group, suggesting that the HF-KCO group had greater insulin sensitivity due to decreased adiposity (Fig. 2).

Figure 2.

Serum leptin and insulin levels. Values are mean ± SD. Values with different superscripts are significantly different by anova with Duncan's multiple range test (P < 0·05). ND, normal diet; HF, high-fat diet; HF-KC, high-fat diet containing 3% kimchi manufactured without the starter; HF-KCO, high-fat diet containing 3% kimchi manufactured with the starter of Weissella koreensis OK1-6.

Hepatic mRNA expression of lipid-regulating genes

Relative quantitation measures the relative change in mRNA expression levels. It determines the changes in steady state mRNA levels of a gene across multiple samples and expresses it relative to the levels of another RNA (Marisa and Medrano 2005). In this study, liver X receptor α (LXRα), SREBP-1c, acetyl-CoA carboxylase (ACC) and SREBP2 were significantly down-regulated (<0·05) in both the HF-KCO and HF-KC groups compared with the HF group (Fig. 3). Stearoyl-CoA desaturase-1 (SCD-1) and peroxisome proliferator-activated receptor γ (PPARγ) were significantly down-regulated (<0·05) in the HF-KCO compared with the HF-KC and HF groups (Fig. 3). KCO supplementation also significantly down-regulated (<0·05) fatty acid synthase (FAS) expression when compared with the ND and HF group (Fig. 3).

Figure 3.

Hepatic mRNA expression of lipid metabolism regulating genes. Values are mean ± SD. Values with different superscripts are significantly different by anova with Duncan's multiple range test (P < 0·05). ND, normal diet; HF, high-fat diet; HF-KC, high-fat diet containing 3% kimchi manufactured without the starter; HF-KCO, high-fat diet containing 3% kimchi manufactured with the starter of Weissella koreensis OK1-6; LXR α, liver X receptor α; SREBP-1c, sterol regulatory element-binding protein-1c; SCD-1, stearoyl-CoA desaturase-1; ACC, acetyl-CoA carboxylase; FAS, fatty acid synthase; CPT-1, carnitine palmitoyltransferase-1; PPARγ, peroxisome proliferator-activated receptor γ; C/EBPα, CCAAT/enhancer-binding protein α; SREBP2, sterol regulatory element-binding protein 2.

Discussion

Anti-obesity effects of kimchi have been reported by several investigators (Sheo and Seo 2004; Kim et al. 2011). For example, Sheo and Seo (2004) demonstrated that rats supplemented with high calorie diet containing 10% of traditional kimchi had significantly lower (<0·05) body weight gain rate than control groups.

The C57BL/6J mice strain used in this study were highly susceptible to diet-induced obesity (West et al. 1992; Rossmeisl et al. 2003). Twelve weeks of HF intake was sufficient for inducing obesity (Hariri and Thibault 2010; Sharma and Fulton 2012); however, the animals in HF-KCO group showed lesser body weight gain as compared to the HF group (Table 4). The prevention of weight gain in HF-KCO group was due in part to decreased epididymal fat mass gain (Fig. 1). The results indicated that KCO is more effective for decreasing visceral and hepatic fat accumulation and preventing hepatic steatosis than the kimchi manufactured without the starter. KCO supplementation also reversed the HF-induced elevation of total cholesterol in serum, but no significant difference in the hepatic total cholesterol was observed between the groups. Also, no significant changes in the serum HDL-cholesterol was observed between the groups. However, the HDL-c/TC ratio was higher in the HF-KCO and ND group compared with the other groups.

The greatest difference between the kimchi fermented with W. koreensis OK1-6 and kimchi fermented traditionally using only ambient bacteria was the increased ornithine content at the expense of arginine in the W. koreensis OK1-6 kimchi (Table 3). Therefore, the enhanced ornithine contents in KCO kimchi might play a role in the anti-obesity activities of the kimchi. The estimated amounts of ornithine consumption were 7·5 mg kg−1 BW in KC mice and 42·5 mg kg−1 BW in KCO mice. Although the mechanism by which ornithine might cause the anti-obesity effect is unclear, there are several possibilities. The amount of ornithine (42·5 mg kg−1 BW) realized in this study might be sufficient to cause the anti-obesity effects since orally administered ornithine is known to increase tissue levels in the brain of rats where it has anti-anxiety activity which may improve eating behaviour (Kurata et al. 2011). Also, ornithine is easily converted to citrulline (Marini 2012), which has been demonstrated to reduce fatigue and increase exercise performance in mice (Takeda et al. 2011). Furthermore, ornithine itself plays a vital role in enhancing the secretion of growth hormone (Bucci et al. 1990). In a study on aged obese rats, it was observed that treating the animals with growth hormone aided in breakdown of adipose tissue and a transient decrease in voluntary food intake leading to decreased body weight (Malmlöf et al. 2002). In another study, supplementation of human growth hormone aided in bodyweight reduction, increased lipolysis and fat oxidation in genetically obese mice (Heffernan et al. 2001). In our study, it can be speculated that ornithine in KCO group could easily be converted to citrulline (Marini 2012) and then citrulline could further converted to arginine, which has also been shown to decrease fat mass and increase lean body mass (Nall et al. 2009). Thus, in our study, the reason for weight loss could be due to high ornithine content in KCO kimchi, and/or due to combined effects of the amino acids including ornithine metabolites such as citrulline and arginine leading to increased growth hormone that may lower adiposity in KCO group compared with the KC or HF group. One of the other possible factors that may be involved in the anti-obesity effect is lactic acid bacteria in kimchi such as W. koreensis OK1-6. Previously, it has been shown that lactic acid bacteria was effective in inhibiting weight gain in obese animals (Lee et al. 2007; An et al. 2011) and the cytoplasmic fraction of lactic acid bacteria in inhibiting intracellular lipid accumulation in differentiating adipocyte (Cha et al. 2010; Moon et al. 2012). Further, even heat-killed lactic acid bacteria confer health benefits in hosts including immune response stimulation effect in obese host animals (Yoda et al. 2012). In this study, all lactic acid bacteria in KC kimchi and KCO Kimchi samples used were found heat-killed by hot-air drying process (data not shown). Therefore, the difference in ornithine content, cell number not cell viability, and other unknown factors may affected the results in this study. As the kimchi samples used in this study were prepared by hot-air drying process, the effects of kimchi with live W. koreensis OK1-6 strain on weight loss will be examined in the future. In addition, kimchi also contains various kinds of components, which could also affect body weight, BMI and body fat (Sheo and Seo 2004; Kim et al. 2011); the possible involvement of other factors could not be excluded.

The anti-obesity effect of KCO kimchi was further supported by the decreased level of leptin and insulin in serum and the mRNA expression level of LXRα, SREBP2, SCD1, PPARγ and CPT1. As leptin is secreted by adipose tissue, it was expected that leptin levels would be decreased along with a decrease in overall body fat. The decrease in insulin levels suggest that insulin sensitivity was also improved, although the effects of the kimchi products on blood sugar regulation were not thoroughly evaluated in this study. We are unaware of any direct effect of ornithine on blood sugar regulation and assume that the lower insulin levels were due to improved body composition. Coutant et al. (1998) reported that ornithine reduces leptin level and raises growth hormone level in human subjects. The adipocyte size increases as a result of uptake and assimilation of extracellular fatty acids into cytosolic triacylglycerol-rich lipid droplets leading to a hypertrophic state (Gonzales and Orlando 2007). Skurk et al. (2007) reported that the circulating leptin level is directly correlated with the adipocyte hypertrophy. In our earlier in vitro study, it was observed that ornithine as well as ornithine rich W. koreensis OK1-6 cytoplasmic extract inhibits intracellular lipid accumulation in 3T3-L1 cells (Moon et al. 2012); therefore, it is possible that the KCO prevented adipocyte hypertrophy, thereby reducing fat accumulation and the secretion of leptin.

Intake of HF raises serum insulin level (Hao et al. 2012), which further leads to insulin resistance condition (Hancock et al. 2008). The circulating insulin level depends on the secretion and clearance rate. Lower levels of circulating insulin could be correlated with lower secretion and higher clearance rate. Insulin clearance takes place in both liver and kidney. It was reported that impaired hepatic insulin clearance is highly associated with the increased liver fat (Kotronen et al. 2007). In our study, the KCO group showed lower hepatic triglycerides compared with both the HF and KC group. The reduction in hepatic triglyceride might improve the hepatic insulin clearance leading to lower serum insulin level in the KCO group compared with the HF and KC group.

LXRα is a ligand-activated transcription factor that up-regulates lipogenic transcription factors such as SREBP-1c (Baranowski 2008). In this study, kimchi supplementation (KC and KCO) significantly down-regulated (<0·05) the hepatic LXRα expression compared with the HF group, this could be the reason behind the down-regulation of SREBP-1c expression in the HF-KC and HF-KCO group. SCD-1 overexpression is closely associated with obesity and hepatic steatosis and hepatic lipogenic gene expression (Ntambi et al. 2002; Dobrzyn and Ntambi 2005; Miyazaki et al. 2009). In this study, SCD-1 mRNA expression and liver TG concentration levels were significantly down-regulated (<0·05) in the HF-KCO group compared with the HF and HF-KC groups, which may in part explain the decreases weight gain in the HF-KCO group.

ACC catalyses the ATP-dependent carboxylation of acetyl-CoA to form malonyl-CoA (Tong and Harwood 2006). Malonyl-CoA is also a potent inhibitor of fatty acid oxidation through allosteric suppression of carnitine palmitoyltransferase 1 (CPT-1) (Ruderman et al. 2003; Zang et al. 2005). The HF group exhibited significantly highest level of hepatic ACC and FAS mRNA expression in the liver, whereas the CPT-1 expression was not significantly changed among the groups; however, it showed inclining tendency in the HF-KCO group compared with HF group. On the other hand, KCO supplementation significantly down-regulated (<0·05) ACC and FAS mRNA expression, but CPT-1 expression was tend to up-regulated.

Juvet et al. (2003) reported that LXRα expression in adipocytes was up-regulated in response to PPARγ agonists. Our data also showed parallel results of PPARγ and LXRα levels. Several previous studies showed that HF-induced obesity increases PPARγ expression in liver (Vidal-Puig et al. 1996; Inoue et al. 2005). Moreover, PPARγ knockout mice exhibit decreased weight gain, hepatic triglyceride and serum leptin and are protected against insulin resistance (Jones et al. 2005). Additionally, PPARγ can cooperate with CCAAT/enhancer-binding protein α (C/EBPα) in the promotion of adipogenesis. C/EBPα is also believed to be important for the differentiation of preadipocytes (Samuelsson et al. 1991; Lin and Lane 1994). In this study, KCO supplementation significantly down-regulated the expression level of hepatic PPARγ, while the C/EBPα did not showed any significant change among the groups. However, C/EBPα tends to decrease in the KCO group compared with the HF group. Sterol regulatory element-binding protein 2 (SREBP-2) is a transcription factor of target mRNAs for cholesterol-related genes (Brown and Goldstein 1997; Shimano 2001). In our study, hepatic SREBP2 mRNA levels were significantly lower (<0·05) in the HF-KCO group than the HF group, which was likely responsible for the lower total and LDL-C concentrations. Therefore, our data suggest that KCO decreases fat mass, body weight and hepatic steatosis by down-regulating the lipogenic process via LXRα, SREBP1c, FAS, SCD-1, ACC and PPARγ, while accelerating β-oxidation by increasing CPT-1 expression.

In conclusion, this study demonstrated that using W. koreensis OK1-6, a strain of lactic acid bacteria commonly found in Kimchi, as a starter produces a kimchi with enhanced anti-obesity and cardio-protective functionality comparing with kimchi produced without W. koreensis OK1-6. Importantly, sensory evaluation (data not shown) found that the KCO kimchi was almost indistinguishable from the traditionally fermented kimchi. Therefore, a very conservative modification of the manufacturing process resulted in a substantial value-added functional food with significant health-promoting benefits, if the results can be duplicated in humans. Kimchi fermented with the ornithine-producing lactic acid bacteria W. koreensis OK1-6 can be used as functional food with health-promoting effects for preventing obesity and obesity-induced diabetes and other disorders.

Acknowledgements

This research was supported by Technology Development Program for Agriculture and Forestry, Ministry for Food, Agriculture, Forestry and Fisheries, Republic of Korea, and partially supported by the research fund from Chonbuk National University for the year 2011.

Conflict of interest

We do not have any conflict of interest.

Ancillary