Disclosure: The authors declared no conflict of interest.
Funding agencies: This work was funded by Korea Yakult Co., Ltd, and the Basic Science Research Program (Center for Food & Nutritional Genomics: grant number 2012-0000644) of the National Research Foundation (NRF) of Korea funded by the Ministry of Education, Science and Technology.
Department of Food Science and Nutrition, Center for Food & Nutritional Genomics, Kyungpook National University, Republic of Korea
Objective: To determine the effects of naturally derived probiotic strains individually or combination on a short-term diet-induced obesity model.
Design and Methods: C57BL/6J mice (n = 50) were randomly divided into five groups, then fed a high-fat high-cholesterol diet (HFCD), HFCD and Lactobacillus plantarum KY1032 (PL, 1010cfu/day), HFCD and Lactobacillus curvatus HY7601 (CU, 1010cfu/day), HFCD and in combination with PL+CU (1010cfu/day), or a normal diet (ND) for 9 weeks.
Results: PL and CU showed distinct and shared metabolic activity against a panel of 50 carbohydrates. Fat accumulation in adipose tissue and liver was significantly reduced by probiotic strains CU or PL+CU. Probiotic strains CU or PL+CU reduced cholesterol in plasma and liver, while PL+CL had a synergistic effect on hepatic triglycerides. Probiotic strains PL+CU combination was more effective for inhibiting gene expressions of various fatty acid synthesis enzymes in the liver, concomitant with decreases in fatty acid oxidation-related enzyme activities and their gene expressions.
Conclusions: Multi-strain probiotics may prove more beneficial than single-strain probiotics to combat fat accumulation and metabolic alterations in diet-induced obesity.
The mammalian gut itself harbors between 1013 and 1014 microorganisms, which are collectively termed the gut microbiota . Although we are far from understanding the function of the bacterial species residing in the mammalian gut, studies in recent years have established that the gut microbiota composition is altered in obesity [2, 3] and related metabolic disorders [4, 5]. In addition, probiotics can also modify the gut microbiota and therefore may help protect against weight gain and the metabolic consequences of obesity; however, there appears to be wide variations in the functional effects of different probiotic strains . Probiotics with multiple or individual bacteria strains have been widely tested in diet-induced obesity models to determine whether they exert any protective effects against the pathophysiological changes associated with obesity in vivo .
Multiple strains of probiotics are assumed to be more effective than that of individual strains, although in vivo evidence is lacking to determine whether combining individual strains is more effective in improving the host health. Several underlying mechanisms have been suggested for the lipid-lowering effect of some probiotic strains regulating energy harvest and inflammation or lipid metabolism genes and enzyme activities [2, 13]. Despite these observations, it remains to be established whether specific probiotic strains individually or in combination exert any effects on lipid metabolism-associated gene expression and lipid metabolism-related enzyme activity in liver.
The objective of this study was to determine the individual and combination effects of two probiotic strains Lactobacillus plantarum KY1032 and Lactobacillus curvatus HY7601 on fat accumulation in adipose tissue and in the liver of diet-induced obese mice. In addition, we examined whether these two probiotic strains exerted any individual or combinatorial effects on lipid metabolism or inflammation.
Isolation and preparation of probiotic constituents
L. plantarum KY1032 (PL) and L. curvatus HY7601 (CU; Korea Yakult Co., Ltd, Gyeonggi, Republic of Korea) were isolated from Kimchi, which is a fermented cabbage widely consumed in the Korea, which contains over 1000 Lactobacillus strains .
PL and CU were grown anaerobically in deMan–Rogosa–Sharpe medium at 37°C for 18 hours. The cells were collected by centrifugation, washed with sterilized PBS, and lyophilized for storage at −20°C until use. The identity of both PL and CU was confirmed by DNA sequencing of the 16S rRNA. The carbohydrate metabolism profile of PL and CU was characterized using the API 50CHL system (BioMerieux, Marcy l'Etoile, France), which consists of 49 different carbohydrates (Supporting Information Table 1).
Table 1. Effects of probiotic supplementation on plasma, hepatic, fecal lipids, and fecal total bile acids concentrations in HFCD-fed mice
***Significant differences between HFCD versus ND are indicated at p < 0.001. abcMeans in a row not sharing a common letter are significantly different between the HFCD and probiotic supplement groups (PL, CU, PL+CU) at p < 0.05. ND, normal diet group; HFCD, high-fat high cholesterol diet group; PL, L. plantarum KY1032 1010/day + HFCD group; CU, L. curvatus HY7601 1010/day + HFCD group; PL+CL, L. plantarum KY1032, and L. curvatus HY7601 5 x 109/day each + HFCD group.
Plasma lipids (mmol/L)
4.75 ± 0.12
6.25 ± 0.19***,a
6.10 ± 0.14ab
5.54 ± 0.18bc
5.41 ± 0.028c
3.34 ± 0.11
4.33 ± 0.23***,a
4.08 ± 0.15ab
3.79 ± 0.17ab
3.66 ± 0.25b
1.24 ± 0.06
1.13 ± 0.06b
1.35 ± 0.081
1.28 ± 0.06ab
1.25 ± 0.07ab
Free fatty acid
1.02 ± 0.07
1.20 ± 0.06
1.06 ± 0.04
1.12 ± 0.05
1.11 ± 0.04
Hepatic lipids (mg/g liver)
2.80 ± 0.15
12.46 ± 0.74***,a
10.20 ± 0.98ab
7.67 ± 0.54b
9.66 ± 1.07b
9.25 ± 0.28
14.75 ± 0.37***,a
14.22 ± 0.641
12.35 ± 0.75b
6.64 ± 0.51c
Free fatty acid
2.46 ± 0.10
3.99 ± 0.17***,a
4.07 ± 0.121
3.60 ± 0.11b
3.26 ± 0.13b
Fecal parameters (mg/day)
301.2 ± 11.5
302.3 ± 12.4c
353.6 ± 8.4b
442.2 ± 10.31
435.5 ± 14.11
0.24 ± 0.04
2.73 ± 0.15***,b
3.51 ± 0.421
3.89 ± 0.371
3.65 ± 0.211
0.03 ± 0.01
0.10 ± 0.01
0.12 ± 0.01
0.14 ± 0.01
0.11 ± 0.02
Total bile acid
15.85 ± 0.87
14.60 ± 1.08
13.71 ± 0.97
16.39 ± 1.07
15.79 ± 0.78
Animals, diet, and experimental design
Fifty male 4-week-old C57BL/6J mice were purchased from Jackson Laboratories (Bar Harbor, ME, USA). All mice were individually housed at a constant temperature and humidity (22 ± 1°C, 55 ± 10%) with a 12 h light/12 h dark cycle. After allowing 1 week for adaptation, the mice were randomly divided into five groups (n = 10 per group) and fed a normal diet (ND), high-fat high-cholesterol diet (HFCD), HFCD and probiotic PL alone (1010 colony-forming units, cfu/day), HFCD and probiotic CU alone (1010 cfu/day), or HFCD and probiotic PL+CU combination (5 × 109 cfu/day each) for 9 weeks. The composition of the diets was formulated based on the AIN-76 semi-synthetic diet (Supporting Information Table 2). To prepare the probiotic supplements, we resuspended lyophilized bacteria in sterilized PBS. Probiotic supplements were prepared and mixed with the diet (2.5 g per mouse) every day immediately before feeding; afterwards mice were given free access to food. An equivalent amount of PBS was mixed with the diet for the control groups (ND, HFCD). Food intake and body weight were measured every day and once a week, respectively. After 9 weeks, the mice were fasted for 12 h and then killed under anesthesia. Feces were collected during the final 4 days, and dried feces were used to determine the fecal lipid content. The experimental design was approved by the Ethics Committee at Kyungpook National University for the care and use of laboratory animals.
Table 2. Effects of probiotic supplementation on hepatic lipid-regulating enzyme activities in HFCD fed mice
***Significant differences between HFCD versus ND are indicated at p < 0.001. abcMeans in a row not sharing a common letter are significantly different between the HFCD and probiotic supplement groups (PL, CU, PL+CU) at p < 0.05. FAS, fatty acid synthase; ME, malic enzyme; G6PD, glucose-6-phosphate dehydrogenase; CPT, carnitine palmitoyltransferase. ND, normal diet group; HFCD, high-fat high cholesterol diet group; PL, L. plantarum KY1032 1010/day + HFCD group; CU, L. curvatus HY7601 1010/day + HFCD group; PL+CL, L. plantarum KY1032 and L. curvatus HY7601 5 × 109/day each + HFCD group.
Fatty acid synthesis (nmol/min/mg protein)
16.29 ± 1.10
17.97 ± 1.20a
11.99 ± 0.73b
9.77 ± 0.97b
11.13 ± 0.95b
166.66 ± 7.66***
68.69 ± 3.14a
59.42 ± 3.89ab
49.55 ± 3.16bc
48.19 ± 3.83c
20.94 ± 1.68***
6.86 ± 0.47
7.69 ± 0.48
8.24 ± 0.35
8.43 ± 1.00
Fatty acid oxidation
13.08 ± 0.70
11.24 ± 0.59a
12.27 ± 1.30a
11.62 ± 0.72a
8.72 ± 0.60b
19.26 ± 2.49
19.76 ± 1.41
23.27 ± 2.17
24.30 ± 2.86
19.00 ± 1.67
Measuring of the amounts of PL and CU in the feces
Total DNA was isolated from the feces of the mice with a fast spin-column procedure using the QIAamp DNA stool mini kit (Qiagen, Valencia, CA, USA) according to the manufacturer's instructions. The amounts of PL and CU were quantified using qPCR (thermocycler, MJ Research, USA) with 200 ng DNA of each of fecal sample under the following conditions: 94°C for 5 min and 20 cycles of denaturing at 94°C for 30 s, annealing at 55°C for 45 s, and elongation at 72°C for 1 min 30 s. A standard curve was created from a serial dilution of DNA from PL and CU containing known copy numbers of the template. The primer sequences are shown in Supporting Information Table 3.
Hepatotoxicity biomarkers were measured including plasma aspartate transaminase (AST), alanine transaminase (ALT), and gamma-glutamyltransferase (GGT) using commercial enzymatic kits (Asan Co., Seoul, Korea).
Liver and epididymal white adipose tissue samples from each mouse were rinsed with sterilized PBS, fixed in 10% v/v formalin/PBS, and embedded in paraffin for staining with hematoxylin and eosin (H&E). The stained area was viewed using a microscope (Motic, Germany) and the adipocyte size was measured using the Motic image plus 2.0 software (Motic) at a magnification of ×200.
Profile of lipids and total bile acid analysis
Plasma triglyceride (TG), total cholesterol (total-C), and HDL-cholesterol were enzymatically determined using a commercial kit (Asan Co., Seoul, Korea). Free fatty acid (FFA) was measured in plasma using a commercial kit (Wako, Osaka, Japan). LDL-cholesterol was calculated using the formula by Friedewald et al.’s formula . Hepatic and fecal lipids were extracted as previously described . The hepatic and fecal lipid concentrations were measured with the same commercial kits as used in the plasma analysis. Total bile acids were extracted using previously established methods  and measured following instruction manual (BQ Kits Inc., San Diego, CA, USA).
Plasma inflammation markers
Three inflammation markers (tumor necrosis factor-alpha, TNFα; Interleukin-1β, IL-1β; Interleukin-6; IL-6) were measured using multiplex detection kits and analyzed with a Bio-Plex Suspension array system (Bio-Rad, Hercules, CA, USA).
RNA isolation and RT-qPCR
Total RNA was isolated from whole livers using TRIZOL reagent (Invitrogen, Carsbad, CA, USA) and reverse-transcribed to cDNA using a QuantiTect reverse transcription kit (Qiagen). RNA expression was quantified by real-time quantitative PCR, using the QuantiTect SYBR Green RT-qPCR kit (Qiagen). The mouse primer sequences were designed to detect Ppar a, Nr1h3, Fasn, Me1, Acaca, Cpt1a, Cpt2, Hmgcr, Acat1, Acat2, and Gapdh (Supporting Information Table 3). Gapdh was used as an internal control and relative gene expression was calculated using the 2-ΔΔCt method.
Hepatic lipid-regulating enzyme activities
Fatty acid synthase (FAS), Malic enzyme (ME), Glucose-6-phosphate dehydrogenase (G6PD), Carnitine palmitoyl transferase (CPT), and Fatty acid β-oxidation activities were assayed with a spectrophotometric method as previously described .
All data are presented as the mean ± standard error. Significant difference between the HFCD and ND groups were determined using Student's t-test. Significant differences between the HFCD and three experimental groups (PL, CU, PL+CU) were assessed using Duncan's multiple-range test. The values were considered statistically significant when p < 0.05.
The amounts of PL and CU in the feces
To determine whether probiotic strains passed through the mammalian gut, we examined their relative abundance in fecal matter following supplementation. L. plantarum KY1032 or L.curvatus HY7601 was detected in the feces of HFCD-fed mice administered with the probiotics PL or CU alone (Figure 1A and B). Probiotic supplementation with PL+CU combination increased the relative abundance of both species recovered in the feces, reflecting the relative concentration of each (5 × 105 cfu) in the combination probiotic supplement. PL and CU showed a shared metabolic activity against ten carbohydrates including ribose, galactose, glucose, fructose, mannose, α-methyl-D-glucoside, N-acetyl-glucosamine, salicin, maltose, and sucrose (Supporting Information Table 1).
Food intake and hepatotoxicity
Food intake and food efficiency ratio was not adversely affected by probiotic supplementation with PL, CU, or PL+CU combination compared to the untreated HFCD-fed mice (Figure 1C and D). AST was unchanged by probiotic supplementation with CU alone, or PL+CU combination, but was increased by PL alone (Figure 1E). ALT was unchanged by probiotic supplementation with PL or CU alone, but was decreased by the combination of PL+CU, while GGT remained unchanged (Figure 1E).
Effect of probiotic strains on fat accumulation in adipocytes and the liver
Final body, liver, and adipose tissue weights were significantly increased in the HFCD-fed mice (Figure 2A-C). However, probiotic supplementation with CU or PL+CU combination effectively suppressed a gain in final body weight and reduced the liver and adipose tissue weights. Furthermore, adipocyte size was significantly decreased by all the probiotic supplementation groups compared to the HFCD group (Figure 2D).
Enlarged hepatic lipid droplet and adipocyte was revealed in HFCD-fed mice, while all the probiotic supplementation groups were reduced hepatic lipid droplet accumulation and adipocyte size (Figure 2E and F).
Effect of probiotic strains on plasma, hepatic, and fecal lipid levels
Plasma cholesterol and hepatic cholesterol storage were all significantly increased in HFCD-fed mice, but significantly decreased by probiotic supplementation with CU or PL+CU, but not by PL alone (Table 1). LDL-C was significantly increased in the HFCD-fed mice, but significantly decreased by probiotic supplementation with PL+CU. However, fecal weight and fecal cholesterol were significantly higher in probiotic supplementation group compared to the HFCD group.
Plasma TG level was not altered by the high-fat diet  or probiotic supplements except for marginally increase plasma triglyceride levels in the PL group (Table 1). Hepatic FFAs and triglycerides were significantly increased in the HFCD-fed mice, but reversed by probiotic supplementation with CU or PL+CU. Furthermore, the combination of PL+CU was significantly more effective than both PL and CU alone for lowering hepatic triglyceride levels. However, fecal triglyceride excretion was significantly elevated in all HFCD-fed mice regardless of probiotic supplementation.
Effect of probiotic strains on hepatic lipid metabolism-related gene expression and enzyme activity
Probiotic supplementation with PL+ CU together or PL alone led to a marked reduction in the Ppara mRNA levels (Figure 3). Probiotic supplementation with PL+CU together, but neither PL nor CU alone, led to a reduction in the Nr1h3 mRNA levels. In relation to fatty acid synthesis, the Fasn and Acaca mRNA levels were decreased in the HFCD-fed mice, and further decreased by probiotic supplementation with PL+CU together, but neither PL nor CU alone. Fasn enzyme activity was lower after probiotic supplementation with PL+CU combination, and also with PL or CU alone (Table 2). Me1 mRNA levels were lower in the HFCD-fed mice and after probiotic supplementation with PL, CU, or PL+CU (Figure 3). Although the ME activity was significantly lower in the HFCD-fed mice, only probiotic supplementation with CU or PL+CU suppressed ME activity further (Table 2). In relation to fatty acid oxidation, Cpt 1a and Cpt 2 mRNA levels were reduced in the HFCD-fed mice, and further decreased by probiotic supplementation with PL+CU (Figure 3).
In agreement with the above results, probiotic supplementation with PL+CU suppressed CPT enzyme activity, although neither PL nor CU was effective alone (Table 2). However, probiotic supplementation with PL, CU, or PL+CU did not result in a significantly detectable difference in the β-oxidation activity in the liver (Figure 3). Hmgcr was reduced by HFCD, but largely unchanged by probiotic supplementation, although in the PL group. Acat1 mRNA levels were reduced by probiotic supplementation with PL+CU, but not with PL or CU alone. Although Acat 2 mRNA levels were reduced by probiotic supplementation with PL+CU or PL alone but not with CU alone. Cyp7a1 mRNA levels were not significant differences between normal-diet-fed mice and all HFCD-fed mice.
Effect of probiotic strains on proinflammatory and anti-inflammatory biomarkers
TNFα and IL-1β were significantly decreased by probiotic supplementation with PL but with neither CU nor PL+CU combination (Figure 4A and B). IL-6 was largely unchanged by probiotic supplementation (Figure 4C).
Past studies indicate some probiotics containing single or multiple strains can suppress body weight gain in diet-induced obese mice [9, 23], while some probiotic strains have little effect  or promote weight gain [12, 24]. Multiple strain probiotics might be more effective than single-strain probiotics against diet-induced obesity, but this has been rarely tested in vivo. The main findings of this study were first that probiotic strain L. curvatus HY7601 and L. plantarum KY1032 combination reduced fat accumulation in both the liver and adipose tissue, although the L. curvatus HY7601 strain appeared to be marginally more effective. Second, we showed that administration of probiotics with two strains may prove more beneficial for overall host-health than administration of probiotics with a single strain. Fat accumulation in the liver is an early stage in the development of nonalcoholic fatty liver disease , which is associated with dysregulation of lipid and glucose. Both probiotic strains combination resulted in lowered expression of Pparα and Nr1h3, which are transcriptional regulators of lipid metabolism genes involved in fatty acid transport and fatty acid β-oxidation . Similarly, fatty acid synthesis-related genes (Fasn, Me1, and Acaca) and fatty acid β-oxidation genes (Cpt1α and Cpt2) were significantly lowered by both probiotic strains combination, which was largely consistent with changes in the enzyme activity, although no changes in β-oxidation were detectable. Lipoprotein lipase, which encodes a rate limiting enzyme involved intracellular lipid transport, also appeared to be reduced by probiotic supplementation not only in the adipose tissue, but also in the colon (data not shown). Lipogenesis, fatty acid synthesis, fatty acid oxidation, and lipid transport are all responsive to dietary fatty acid supply , yet hepatic FFA was markedly reduced following probiotic supplementation with both strains together or L. curvatus HY7601 alone despite the HFCD intake. Therefore, it appears that probiotic strains must exert some effect within the gut, which reduces the FFA supply to the liver resulting in decreased fat accumulation in the liver. Gut microbials play an important role in extracting energy from the diet . Within the gut, probiotic strains compete for nutrients as well as secret metabolic byproducts and in some cases, antimicrobial proteins; therefore, probiotic strains can alter the gut microbiota community in several ways. The importance of nutrient supply for gut microbials is evident from the shift in gut microbiota composition caused by prebiotic supplementation, resulting in reduced fat accumulation and improvements in the host health [27, 28]. We show here that L. curvatus HY7601 and L. plantarum KY1032 can metabolize a broad range of carbohydrates and hence may increase competition for nutrients and reduced mircrobial diversity within the gut microbial community similar to other multi-species probiotic . Several lines of evidence support the view that probiotic-induced competition for nutrients, which reduces the gut microbial diversity, leads to reduced fat accumulation and improvements in the host health.
First, transplantation of conventional microbiota to gnotobiotic mice increases the gut microbial diversity and energy harvest from the diet leading to excessive weight gain [7, 8]. Second, administration of antibiotics reduces the gut microbial diversity and improves host health in obese mice [30, 31]. Third, some probiotic strains isolated from kimchi have been shown to secrete antimicrobial proteins . Taken together, this evidence supports the view that probiotic L. curvatus HY7601 and L. plantarum KY1032 may decrease fat accumulation partly by increasing the competition for nutrients in the gut or secreting antimicrobial proteins, which both promote reduced gut microbiota diversity and reduced energy harvest from the diet.
Elevated plasma cholesterol levels are commonly associated with obesity and dyslipidemia that progressively develop in diet-induced obesity models . Probiotic with the combination of L. curvatus HY7601 and L. plantarum KY1032 lowered cholesterol in both plasma and liver. Lactobacillus strains and other probiotic strains have been reported to lower the plasma cholesterol concentration , but the fate of the excess cholesterol is not well established. Studies suggest some probiotic strains can promote fecal cholesterol excretion in animals fed a combined high fat and cholesterol diet [10, 34]. In this study, our probiotic combination also led to elevated fecal cholesterol concentrations compared to the HFCD group as shown in Table 1. In general, cholesterol not absorbed in the small intestine can be metabolized by the resident intestinal microbiota. It has been suggested that cholesterol-to-coprostanol conversion by the intestinal microbiota may facilitate the removal of body cholesterol [35, 36]. However, the major factors that modulate the cholesterol-lowering microbial activity have not yet been elucidated. Probiotic strains did not alter the expression of hepatic Hmgcr mRNA; however, it reduced the hepatic Acat mRNA. Although decreased formation of cholesterol ester can be associated with increased expression of Cyp7a1 , the probiotic strains we used exhibited no effect on Cyp7a1 expression as well on the fecal bile acid levels. These findings suggest that the probiotics had a beneficial effect on plasma cholesterol through two mechanisms. First, probiotic supplementation inhibited the absorption of dietary cholesterols, which led to the promotion of fecal cholesterol excretion based on the present data. Second, probiotics decreased hepatic Acat mRNA expression, which can lead to down-regulated cholesterol storage in the liver. Accordingly, supplementation with the combination of L. planatrum KY1032 and L. curvatus HY7601 or L. curvatus HY7601 alone can possibly modulate cholesterol metabolism, which results in lower plasma and hepatic cholesterol concentrations.
Obesity is commonly associated with chronic low-grade inflammation and probiotic strains can modulate the immune response . In this study, probiotic supplementation with L. plantarum KY1032 led to a marked decrease in plasma TNFα and IL-1β levels, although neither L. curvatus HY7601 alone was not effective. Experimental manipulation of the gut microbiota affects the activation of the innate immune system [7, 39]. Some gut microbial families release endotoxins, which are taken up from the gut with chylomicrons or can leak through the gut wall. Some reports suggest probiotic strains may protect the gut wall from microbial endotoxin release, and alternative probiotic strains may reduce innate immune system activation through increased competition with bacteriocin producing microbial families . However, inflammation was not a major characteristic of this short-term diet-induced obesity model, primarily because of the short duration and dietary fat content (35% kcal). Nevertheless, it may be worthwhile to determine whether L. plantarum KY1032 or L. curvatus HY7601 influences inflammation in other chronic inflammatory disease models.
We demonstrated a novel therapeutic probiotic consisting of L. plantarum KY1032 and L. curvatus HY7601, which protects against fat accumulation in adipose tissue and the liver. Multi-strain probiotics may prove more beneficial than single-strain probiotics, but it is essential that the effects of individual probiotic strains are evaluated. Probiotics with L. plantarum KY1032 and L. curvatus HY7601 may provide a natural way to promote weight-loss and/or body fat reduction and reduce inflammation, but further human trials are necessary.