Evaluation of the Cholesterol-Lowering Mechanism of Enterococcus faecium Strain 132 and Lactobacillus paracasei Strain 201 in Hypercholesterolemia Rats

Hypercholesterolemia can cause many diseases, but it can effectively regulated by Lactobacillus. This study aimed to evaluate the cholesterol-lowering mechanism of Enterococcus faecium strain 132 and Lactobacillus paracasei strain 201. These results showed that both the strains decreased serum total cholesterol (TC), low-density lipoprotein cholesterol (LDL-C), triglycerides (TG), liver TC and TG and increased fecal TC, TG and total bile acid (TBA) levels. Additionally, both strains also reduced glutamic-pyruvic transaminase (ALT), glutamic oxaloacetic transaminase (AST) and levels of tissue inflammation levels to improve the lipid profile, and they reduced fat accumulation partially by alleviating inflammatory responses. Furthermore, both strains regulated the expression of the CYP8B1, CYP7A1, SREBP-1, SCD1 and LDL-R gene to promote cholesterol metabolism and reduce TG accumulation. Interventions with both strains also altered the gut microbiota, and decreasing the abundance of Veillonellaceae, Erysipelotrichaceae and Prevotella. Furthermore, fecal acetic acid and propionic acid were increased by this intervention. Overall, the results suggested that E. faecium strain 132 and L. paracasei strain 201 can alleviate hypercholesterolemia in rats and might be applied as a new type of hypercholesterolemia agent in functional foods.


Introduction
In 2016, approximately 17.6 million deaths were caused by cardiovascular disease globally, which amounted to an increase of 14.5% from 2006 [1]. Individuals with hypercholesterolemia are three times more likely to have a heart attack than those with normal blood lipid levels [2]. Hypercholesterolemia remains one of the largest causes of cardiovascular disease-related death worldwide [3]. The excessive accumulation of triglycerides by hypercholesterolemia in the liver is also considered an important contributor to nonalcoholic fatty liver disease (NAFLD) [4]. People usually adopt a controlled diet to manage their cholesterol levels, but if the levels are too high, the diet is less effective. Statins are the most commonly prescribed cholesterol-lowering drug, and they have a strong efficacy. Lovastatin (Mevacor ® , Altoprev™) is the most widely used statin used to reduce serum low-density lipoprotein cholesterol (LDL-C) levels, and it consequently reduces the risk After incubation, the bacterial cells were pelleted by centrifugation at 4500× g at 4 • C for 10 min. The cholesterol content was determined using a total cholesterol assay kit (Mindray Co., Ltd., Shenzhen, China) in the Mindray BS-480 automatic biochemical analyzer and calculated using Equation (1). E. faecium strain 132 and L. paracasei strain 201 were screened with a high cholesterol-lowering rate and further investigated.
where P 0 (%) represents the cholesterol reduction rate, C represents the cholesterol content in the sample and D represents the cholesterol content in the blank control. Each isolate was subjected to three replicates, and the mean ± SD was determined.

Serum Biochemical Index Analysis
Blood from the rat heart, collected after a 16-hour period of fasting, was collected for serum biochemical analysis. The blood was centrifuged at 2500× g for 10 min at 4 • C, and the serum was collected. Low-density lipoprotein cholesterol (LDL-C), serum total cholesterol (TC), total bile acid (TBA), triglyceride (TG), high-density lipoprotein cholesterol (HDL-C) cholesterol and AST and ALT levels were measured using the BS-480 Reagents on a Mindray BS-480 fully automatic biochemical analyzer (Mindray Co., Ltd., Shenzhen, China) [27].

Liver and Faeces TC, TBA and TG Contents
Lipids were extracted from the separated liver and feces by the Lepercq method [28]. The extracted lipids were used for the determination of TC, TBA and TG levels included in 2.3. The content of cholic acid in feces was determined as described by Kajiura [29]. TC, TBA and TG were measured as in Section with 2.3.

Determination of Related Inflammatory Factors in Liver and Epididymal Fat
Tissue (50 ± 0.25 mg) was dissolved in 500 µL 1× protease inhibitor (Beyotime Biotechnology, Shanghai, China), which was grinded in a grinder (65 Hz, 45 s, −20 • C). The supernatant was collected after centrifugation at 4 • C at 13,000× g for 5 min. The inflammatory cytokines MCP-1, IL-1β, IL-6 and TNF-α in liver and epididymal fat were determined using commercially available ELISA kits (Solarbio life sciences Co., Ltd., Beijing, China) according to the manufacturers' instructions. Total protein content in the liver and epididymis fat was measured with a bicinchoninic acid assay (BCA) protein assay kit (Sangon Biotech Co., Ltd., Shanghai, China). The levels were calculated by a method using Equation (2): where O represents the relative values of inflammatory factors and P and Q represent the values of inflammatory factors and total protein content, respectively.

Histopathological Examination
The epididymal fat and liver were rinsed with PBS and fixed in 10% buffered formalin for 24 h. After fixation, the tissues were dehydrated using an alcohol gradient, paraffinembedded and sliced with a cut angle of 10 • . The thickness of liver slices were 3 µm and that of fat slices were 5 µm; tissues were then dewaxed with xylene and stained with haematoxylin, and the ScopeImage 9.0 system was used at a magnification of 200× for observation [30]. Steatosis and hepatocyte inflammation were scored based on non-alcoholic fatty liver disease (NAFS) scores [31].

Intestinal Microbial Diversity
Genomic DNA was extracted from fecal samples of rats treated for 6 weeks (see Section 2.4) using an M6 HiPure Stool DNA Kit (Megan, Guangzhou, China). The concentration and quality of the extracted DNA were measured using a Thermo Scientific Qubit 3.0 (Thermo Scientific Co., Ltd., Waltham, MA, USA). The 16S rRNA gene sequences were amplified with primers specific of the V4-V5 region (515F: GTGCCAGCMGCCGCGG, 907R: CCGTCAATTCMTTTRAGTTT). The Illumina MiSeq platform was used for sequencing, with a depth of 2 × 250 bp. In brief, the paired-end reads were joined after inverse multiplexing. The FASTA quality file and the metadata file representing the barcode sequence of each sample were used as input, and the samples were segmented according to the barcode. The minimum count per sample was 3383, and the average count per sample was 6405. Quality control, rarefaction curve drawing and diversity analysis were performed using the open-source platform QIIME2. The DADA2 method was used for noise removal and quality control, and rooted-trees and feature tables (ASV) were created.

Determination of Short Chain Fatty Acids (SCFAs)
First, fecal samples were homogenized in 0.01% vitriol and centrifuged at 13,000× g for 25 min. The supernatant was collected, and volatile free acids were used as an external standard. The SCFAs of the supernatant were used with an Agilent 7693A gas chromatograph (Agilent Technologies, Santa Clara, CA, USA) under the following conditions: 7.2452 psi pressure, 20 mL/min desolvation gas flow, 3 mL/min cone gas flow, 170 • C oven temperature and 250 • C for the flame ionization detector and injection port temperature. Nitrogen was used as the carrier gas. The retention times and peak areas of samples were confirmed using a standard mixture.

Statistical Analysis
The results were reported as the mean ± SD of experiments, which were performed in triplicates. Graph Pad 8.4.3 software was used to calculate the mean and SD values. The statistical significance of differences among groups was calculated with one-way analysis of variance (ANOVA) and multiple comparisons with Tukey's test. SPSS 25.0 and Graph Pad 8.4.3 software were used for drawing and data analysis, respectively. The Kruskal-Wallis test was used to analyze the alpha diversity index, and the PERMANOVA test was used for beta diversity index analysis. Linear discriminant analysis (LDA) effect size was used for the detection of biomarkers; the false discovery rate (FDR) was used to adjust for microbiota analyses. p-values < 0.05 were considered statistically significant.

Effects of Screened Strains on Serum Parameters in of Hypercholesterolemic Rats
Body weight gain was significantly different between the HC rat group and other treatment groups (p < 0.05; Table 2), and weekly weight gain is shown in Figure S1. The liver and epididymal fat index were significantly higher in the HC group than in the other groups (p < 0.05; Table 2). There was no significant difference in other organ weights in each group. Serum TC, TG, LDL-C and TBA values differed significantly between the HC and control groups (p < 0.05; Table 2). Compared to those in the HC group (TG: 0.6 ± 0.30 mmol L −1 , LDL-C: 0.93 ± 0.35 mmol L −1 ), serum TG and LDL-C levels were significantly decreased in the HC-E. faecium strain 132 (TG: 0.38 ± 0.08 mmol L −1 , LDL-C: 0.692 ± 0.12 mmol L −1 ) and HC-L. paracasei strain 201 (TG: 0.39 ± 0.03 mmol L −1 , LDL-C: 0.61 ± 0.06 mmol L −1 ) groups (p < 0.05). Serum TC levels were decreased, but serum HDL-C and TBA levels were not affected in the treatment groups, compared to levels in the HC group (Table 2). Organs index = (organ weight/total weight) ×100; p-values were determined using a one-way ANOVA with Tukey's test for post-hoc analysis, n = 6. Significant differences between the HC group versus control group are indicated as # p < 0.05, ## p < 0.01, ### p < 0.01, #### p < 0.0001. Significant differences between the HC-lovastatin, HC-E. faecium strain 132 or HC-L. paracasei strain 201 group versus from HC group are indicated as * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001. TBA, total bile acid; TG, triglycerides; LDL-C, low-density lipoprotein cholesterol; HDL-C, high-density lipoprotein cholesterol; TC, total cholesterol.

Liver and Faecal TC, TBA and TG Levels in Hypercholesterolemic Rats
The liver TC and TG values in the HC group (TC: 9.98 ± 2.35 mmol L −1 ) were significantly higher than those in the control group (TC: 2.24 ± 0.38 mmol L −1 ; p < 0.05). Liver TC levels in the HC-E. faecium strain 132 (6.12 ± 0.93 mmol L −1 ) group were significantly decreased compared to those in the HC group (p < 0.05), and the values were similar to those in the control group ( Figure 1A). Liver TBA levels increased but liver TG levels decreased in the other groups compared to those in the HC group ( Figure 1B,C). Fecal TC increased in the HC-E. faecium strain 132 group compared with that in the HC group (p < 0.05; Figure 1D). A significant increase in fecal TBA content was observed in the HClovastatin (452.28 ± 42.75 µmol L −1 ), HC-E. faecium strain 132 (472.71 ± 45.48 µmol L −1 ) and HC-L. paracasei strain 201 (466.51 ± 31.06 µmol L −1 ) groups compared to that in the HC group (401.93 ± 49.86 µmol L −1 ; p < 0.05; Figure 1E). Fecal TG was similar among all rats ( Figure 1F).

Screening Strains Improved Liver Injury in Hypercholesterolemia Rats
To study whether the screened strains could improve liver injury, in this study, we measured the levels of ALT and AST in serum, determined the content of inflammatory cytokines in liver cells by ELISA and analyzed morphology based on HE staining in the liver. Inflammatory factors are associated with many diseases and mainly cause inflammatory changes in the liver with hyperlipidemia. The MCP-1, IL-1β, IL-6 and TNF-α contents in the liver are shown in Figure 2. Compared with levels in the HC group, there was significant inhibition in the HC-lovastatin group (p < 0.05; Figure 2C). At the same time, liver IL-1β significantly decreased after treatment with E. faecium strain 132 and L. paracasei strain 201 compared to levels in the HC group (p < 0.05; Figure 2D). Furthermore, liver IL-6 was significantly decreased in the HC-lovastatin group, and TNF-α content was only significantly decreased in the HC-lovastatin and HC-E. faecium strain 132 groups (p < 0.05; Figure 2E,F). HE-stained liver sections of HCD rats showed intracellular vacuolation and significant lipid accumulation. Compared to that in the HC group, the NAFS score showed that the degrees of steatosis and hepatocyte inflammation were appreciably reduced in HC-lovastatin and HC-E. faecium strain 132 groups (p < 0.05; Figure 2G,H).

Screened Strains Improved the Inflammation of Epididymis Fat and Fatty Hypertrophy in Hypercholesterolemic Rats
Epididymal adipose levels of MCP-1 and IL-6 showed significant decreases in the treatment groups compared with those in the HC group (p < 0.05; Figure 3A,C). However, there was no significant difference in the content of IL-1β and TNF-α in epididymal adipose among groups ( Figure 3B,D). Epididymal fat was observed by HE staining. Compare to that in the HC group (134 ± 8), the number of adipocytes in the same visual field was significantly increased in the control (176 ± 5), HC-lovastatin (173 ± 3), HC-E. faecium strain 132 (207 ± 6) and HC-L. paracasei strain 201 (190 ± 3) groups (p < 0.05), and the diameter of adipocytes was significantly reduced ( Figure 3E-G).
in the control group ( Figure 1A). Liver TBA levels increased but liver TG levels decreased in the other groups compared to those in the HC group ( Figure 1B,C). Fecal TC increased in the HC-E. faecium strain 132 group compared with that in the HC group (p < 0.05; Figure  1D). A significant increase in fecal TBA content was observed in the HC-lovastatin (452.28 ± 42.75 μmol L −1 ), HC-E. faecium strain 132 (472.71 ± 45.48 μmol L −1 ) and HC-L. paracasei strain 201 (466.51 ± 31.06 μmol L −1 ) groups compared to that in the HC group (401.93 ± 49.86 μmol L −1 ; p < 0.05; Figure 1E). Fecal TG was similar among all rats ( Figure 1F). Data are expressed as the mean ± SD of triplicate tests. (n = 6 for weight gained, TC, TG and TBA in feces and liver). The pink bar represents the control group, the purple bar represents the HC group, the green bar represents the HC-lovastatin group, the blue bar represents the HC-E. faecium strain 132 group and the yellowish-brown bar represents the HC-L. paracasei strain 201group. p-values were determined using a one-way ANOVA with Tukey's test for post-hoc analysis. Significant differences between the HC group versus control group are indicated as # p < 0.05, #### p < 0.0001. Significant differences in the HC-lovastatin, HC-E. faecium strain 132 or HC-L. paracasei strain 201 group versus HC group are indicated as * p < 0.05, ** p < 0.01.

Screening Strains Improved Liver Injury in Hypercholesterolemia Rats
To study whether the screened strains could improve liver injury, in this study, we measured the levels of ALT and AST in serum, determined the content of inflammatory cytokines in liver cells by ELISA and analyzed morphology based on HE staining in the liver. Compared to those in the HC group (ALT: 69. Inflammatory factors are associated with many diseases and mainly cause inflammatory changes in the liver with hyperlipidemia. The MCP-1, IL-1β, IL-6 and TNF-α contents Data are expressed as the mean ± SD of triplicate tests. (n = 6 for weight gained, TC, TG and TBA in feces and liver). The pink bar represents the control group, the purple bar represents the HC group, the green bar represents the HC-lovastatin group, the blue bar represents the HC-E. faecium strain 132 group and the yellowish-brown bar represents the HC-L. paracasei strain 201group. p-values were determined using a one-way ANOVA with Tukey's test for post-hoc analysis. Significant differences between the HC group versus control group are indicated as # p < 0.05, #### p < 0.0001. Significant differences in the HC-lovastatin, HC-E. faecium strain 132 or HC-L. paracasei strain 201 group versus HC group are indicated as * p < 0.05, ** p < 0.01.

mRNA Expression Levels of Genes Related to Lipid Metabolism
To explore the effects of treatment with E. faecium strain 132 and L. paracasei strain 201 on lipid metabolism in rats, we mainly analyzed the transcript mRNA levels of the key hepatic fat metabolism genes CYP7A1, CYP8B1, LDL-R, PPARα, SREBP-1, SREBP-2, FXR and HMGR. The relevant data are shown in Figure 4. Data for the mRNA level of SREBP-2, FXR and HMGR showed no significant difference among the groups. Compared with those in the HC group, the expression levels of CYP8B1 and LDL-R were significantly increased, whereas SREBP-1 was significantly decreased in the HC-E. faecium strain 132 group (p < 0.05). The expression of CYP7A1 was also significantly increased (p < 0.05), whereas the expression of SREBP-1 and SCD1 was significantly decreased (p < 0.05) in the HC-L. paracasei strain 201 group. Furthermore, the expression of CYP7A1, LDL-R and PPARα was significantly increased in the HC-lovastatin group (p < 0.05), whereas the expression of SREBP-1 was significantly decreased (p < 0.05). E. faecium strain 132 and L. paracasei likely regulated those genes to improve cholesterol metabolism and reduce fat accumulation in the liver. strain 201 compared to levels in the HC group (p < 0.05; Figure 2D). Furthermore, liver IL-6 was significantly decreased in the HC-lovastatin group, and TNF-α content was only significantly decreased in the HC-lovastatin and HC-E. faecium strain 132 groups (p < 0.05; Figure 2E,F). HE-stained liver sections of HCD rats showed intracellular vacuolation and significant lipid accumulation. Compared to that in the HC group, the NAFS score showed that the degrees of steatosis and hepatocyte inflammation were appreciably reduced in HC-lovastatin and HC-E. faecium strain 132 groups (p < 0.05; Figure 2G,H). p-values were determined using a one-way ANOVA with Tukey's test for post-hoc analysis, n = 6. Significant differences between the HC group versus control group are indicated as # p < 0.05, ## p < 0.01, #### p < 0.0001. Significant differences in the HC-lovastatin, HC-E. faecium strain 132 or HC-L. paracasei strain 201 group versus HC group are indicated as * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001. p-values were determined using a one-way ANOVA with Tukey's test for post-hoc analysis, n = 6. Significant differences between the HC group versus control group are indicated as # p < 0.05, ## p < 0.01, #### p < 0.0001. Significant differences in the HC-lovastatin, HC-E. faecium strain 132 or HC-L. paracasei strain 201 group versus HC group are indicated as * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001. treatment groups compared with those in the HC group (p < 0.05; Figure 3A,C). However, there was no significant difference in the content of IL-1β and TNF-α in epididymal adipose among groups ( Figure 3B,D). Epididymal fat was observed by HE staining. Compare to that in the HC group (134 ± 8), the number of adipocytes in the same visual field was significantly increased in the control (176 ± 5), HC-lovastatin (173 ± 3), HC-E. faecium strain 132 (207 ± 6) and HC-L. paracasei strain 201 (190 ± 3) groups (p < 0.05), and the diameter of adipocytes was significantly reduced ( Figure 3E-G). p-values were determined using a one-way ANOVA with Tukey's test for post-hoc analysis, n = 6. Significant differences between the HC group versus control group are indicated as # p < 0.05, ## p < 0.01, #### p < 0.0001. Significant differences in the HC-lovastatin, HC-E. faecium strain 132 or HC-L. paracasei strain 201 group versus HC group are indicated as ** p < 0.01, *** p < 0.001, **** p < 0.0001. p-values were determined using a one-way ANOVA with Tukey's test for post-hoc analysis, n = 6. Significant differences between the HC group versus control group are indicated as # p < 0.05, ## p < 0.01, #### p < 0.0001. Significant differences in the HC-lovastatin, HC-E. faecium strain 132 or HC-L. paracasei strain HC-L. paracasei strain 201 group. Furthermore, the expression of CYP7A1, LDL-R and PPARα was significantly increased in the HC-lovastatin group (p < 0.05), whereas the expression of SREBP-1 was significantly decreased (p < 0.05). E. faecium strain 132 and L. paracasei likely regulated those genes to improve cholesterol metabolism and reduce fat accumulation in the liver. . The mRNA expression levels of genes related to lipid metabolism in the liver. FXR stands for Farnesol X Receptor, and HMGR is named for 3-hydroxy-3-methyl glutaryl coenzyme A reductase. p-values were determined using a one-way ANOVA with Tukey's test for post-hoc analysis, n = 6. Significant differences between the HC group versus control group are indicated as # p < 0.05, ## p < 0.01, ### p < 0.001. Significant differences in the HC-lovastatin, HC-E. faecium strain 132 or HC-L. paracasei strain 201 group versus HC group are indicated as * p < 0.05, ** p < 0.01, **** p < 0.0001.

Gut Microbiota Modulation by Selected Strains
When the rarefaction curve tends to be flat, this indicates that the amount of sequencing data is reasonable. As shown in Figure 5A, when the sequencing depth was deeper, the curve tended to flatten, indicating that the sequencing depth was reasonable. Similarly, with deeper sequencing, the Shannon index tended to be flat ( Figure 5B). In our study, we used the Chao1, Shannon and Simpson indexes to analyze the alpha diversity (richness and evenness), and the result showed that the three indexes in the HC-lovastatin and HC-E. faecium strain 132 groups were significantly improved compared with those in the HC group, but no significant difference was detected in the HC-L. paracasei strain 201 group (p = 0.60, p = 0.37, p = 0.37; Figure 5C-E), indicating that the HC-lovastatin and HC-E. faecium strain 132 groups showed high richness and diversity. The PCoA plot (unweighted UniFrac distance) showed that the gut microbiomes of different groups were well separated from each other, and each group was clustered in a different circle ( Figure 5F). Next, PERMANOVA multi-dimensional statistical analysis was performed; the control group  Figure 5G). The relative abundance of Proteobacteria showed no evident differences between the groups, but the proportion of Tenericutes was significantly higher in the control group than in the HCD group, and the relative abundance of Actinobacteria showed the opposite trend ( Figure 5G). At the class level, the proportions of Bacilli and Bacteroidia were significantly higher in the control group than in the HCD groups, but the abundance of Clostridia was higher in the HCD groups ( Figure S2A). At the order level, the bacterial abundance was similar to that at the class level ( Figure S2B). At the family level, after intervention with HC-lovastatin, HC-L. paracasei strain 201 or HC-E. faecium strain 132, the abundances of Veillonellaceae and Prevotellaceae were significantly lower than those in the HC group (p < 0.05), and bacteria belonging to the families Bacteroidaceae, Peptostreptococcaceae, Lactobacillaceae, Paraprevotellaceae, Coriobacteriaceae and Erysipelotrichaceae were also less abundant than those in that the HC group ( Figure 5H and Figure S2C). However, the abundances of Ruminococcaceae, S24_7, Lachnospiraceae, Spirochaetaceae and Corynebacteriaceae in the HC-lovastatin, HC-L. paracasei strain 201 or HC-E. faecium strain 132 groups were higher than those in the HC group. A high abundance of Veillonellaceae and Erysipelotrichaceae was observed in the fecal samples of the HC group ( Figure 6A,B, p < 0.05). At the genus level, the heat map indicated that CF231, which belongs to the family Paraprevotellaceae, was significantly enhanced in the control group compared to levels the other groups. Prevotella, Succinispira, Bacteroides, Ruminococcus, Paraprevotella, Collinsella and Eubacterium were decreased in the HC-lovastatin, HC-E. faecium strain 132 and HC-L. paracasei strain 201 groups compared to abundances in the HC group. Conversely, there was a lower abundance of Corynebacterium, Lactobacillus, Clostridium, Butyricicoccus and Lactonifactor in the HC group compared to that in the HC-lovastatin, HC-E. faecium strain 132 and HC-L. paracasei strain 201 groups ( Figure 5I and Figure S2D). Differences in the abundance of Prevotella and Butyricicoccus were statistically significant ( Figure 6C,D, p < 0.05). After FDR-adjustment, the abundances of Veillonellaceae, Eubacterium, Coriobacteriaceae, Erysipelotrichaceae, Prevotellaceae and Ruminococcus in the HC group were noticeably increased compared to their abundances in other groups (Figure 7). The LDA scores of CF231 and Paraprevotellaceae were significantly increased in the control group. The HC-L. paracasei strain 201 group showed a significantly higher proportion of Ruminococcaceae and Lachnospiraceae, and the abundance of Sporobacter was significantly increased in the HC-E. faecium strain 132 group (Figure 7).  After FDR-adjustment, the abundances of Veillonellaceae, Eubacterium, Coriobacteriaceae, Erysipelotrichaceae, Prevotellaceae and Ruminococcus in the HC group were noticeably increased compared to their abundances in other groups (Figure 7). The LDA scores of CF231 and Paraprevotellaceae were significantly increased in the control group. The HC-L. paracasei strain 201 group showed a significantly higher proportion of Ruminococcaceae and Lachnospiraceae, and the abundance of Sporobacter was significantly increased in the HC-E. faecium strain 132 group (Figure 7).

SCFA Content in Faeces in Hypercholesterolaemic Rats
To investigate the amount of SCFAs produced by bacteria in the gut, we used gas chromatography to determine fecal contents in this study. First, we measured total contents of acetic acid, propionic acid, isobutyric acid, butyric acid, isovaleric acid and valeric acid, and the results are shown in Table 3. The levels of isobutyric acid, butyric acid, isovaleric acid and valeric acid were not significantly different among the groups. However, the acetic acid and propionic acid levels were significantly increased in the HC-lovastatin and HC-E. faecium strain 132 groups, whereas the acetic acid level was not significantly different in the HC-L. paracasei strain 201 group compared to that in the HC group. These results suggested that acetic acid and propionic acid levels might be the differential metabolites. p-values were determined using a one-way ANOVA with Tukey's test for post-hoc analysis, n = 6. Significant differences in the HC-lovastatin, HC-E. faecium strain 132 or HC-L. paracasei strain 201 group versus HC group are indicated as * p < 0.05, ** p < 0.01.

Discussion
Lactic acid bacteria have been widely used to regulate lipid metabolism, inflammation levels, obesity and gut microbiota [32], which have aroused considerable attention Figure 7. Linear discriminant analysis (LDA) and effect quantity (LEfSe) were used to compare the microflora distribution among groups. LDA score > 4, FDR value < 0.1. The pink bar represents the control group, the purple bar represents the HC group, the blue bar represents the HC-E. faecium strain 132 group and the yellowish-brown bar represents the HC-L. paracasei strain 201group.

SCFA Content in Faeces in Hypercholesterolaemic Rats
To investigate the amount of SCFAs produced by bacteria in the gut, we used gas chromatography to determine fecal contents in this study. First, we measured total contents of acetic acid, propionic acid, isobutyric acid, butyric acid, isovaleric acid and valeric acid, and the results are shown in Table 3. The levels of isobutyric acid, butyric acid, isovaleric acid and valeric acid were not significantly different among the groups. However, the acetic acid and propionic acid levels were significantly increased in the HC-lovastatin and HC-E. faecium strain 132 groups, whereas the acetic acid level was not significantly different in the HC-L. paracasei strain 201 group compared to that in the HC group. These results suggested that acetic acid and propionic acid levels might be the differential metabolites. The result showed that E. faecium strain 132 displayed lowered serum TC levels, LDL-C, TG and liver TC levels and significantly higher faecal TC and TBA levels. In our cognitive scale, we know that CYP7A1 and CYP8B1 act as a rate-limiting enzymes in the classical pathway of bile acid anabolism [12]. Our results showed that the E. faecium strain 132 and L. paracasei strain 201 might up-regulate the expression of CYP8B1 and CYP7A1 genes, respectively, to promote the reduction in cholesterol levels in rats. Subsequently, high LDL-C levels reportedly correlate with a risk of coronary artery disease [33], and liver LDL-R receptors promote LDL transport. The expression of the LDL-R gene was increased after treatment with lovastatin and E. faecium strain 132, suggesting that this can accelerate LDL transport and reduce LDL-C levels in the body. However, because SREBP-1 can induce the expression of adipogenic genes and mediate the formation of TG and lipid accumulation, the SCD1 is a liposynthetase [34]. In this study, we discovered that the E. faecium strain 132 and L. paracasei strain 201 might reduce the expression of the SREBP-1 gene and that the L. paracasei strain 201 also reduces the expression of the SCD1 gene to suppress TG accumulation. These results suggest that E. faecium strain 132 and L. paracasei strain 201 might regulate the expression of genes related to lipid metabolism to promote cholesterol excretion and inhibit the accumulation of fat in the liver.
The liver is the main site of lipid metabolism, which is susceptible and vulnerable. In the process of hepatotoxicity, damaged hepatocytes release liver-specific enzymes, such as ALT and AST, into the bloodstream, causing an elevation of both in the serum [35]. Consistent with previous observations, interestingly, our biochemical analysis results revealed that the serum levels of AST and ALT returned to near-normal levels after treatment with E. faecium strain 132 and L. paracasei strain 201, suggesting that the two strains were more likely to improve liver damage. In adipose and liver tissue, MCP-1 is a highly representative chemokine, which plays a key role in the pathogenesis of liver diseases, and it is the major determinant of monocyte/macrophage recruitment to the site of tissue injury [36]. IL-1β acts primarily as a pro-inflammatory mediator, activating and recruiting white blood cells, especially neutrophils, into tissues [37]. IL-6 is a cytokine involved in hematopoiesis, immune response and the regulation of acute and chronic inflammation [38], but it also promotes the production of pro-inflammatory C-reactive protein (CRP) by macrophages and T cells, which is associated with an increased risk of diabetes, hypertension and cardiovascular disease [39]. TNF-α is the key mediator of liver injury induced by LPS [40]. However, the current study showed that the HC-E. faecium strain 132 and HC-L. paracasei strain 201 groups significantly decreased the liver IL-1β and the epididymal adipose MCP-1 and IL-6 levels. The HC-E. faecium strain 132 group also had significantly decreased the liver MCP-1 and TNF-α. Histopathological examination revealed that hepatocyte steatosis and adipocyte hypertrophy in the HC-L. paracasei strain 201 and HC-E. faecium strain 132 groups were expectedly alleviated compared to those in the HC group. The results of this study have shown that E. faecium strain 132 and L. paracasei strain 201 are likely to improve liver injury and obesity by regulating liver and epididymal adipose inflammatory cytokines.
Human gastrointestinal microbiota are not only a critical barrier to invasive substances, but they also regulate symbiotic homeostasis and normal physiological processes [41]. One study analyzed the correlation between the intestinal microbiota and abnormalities in 893 subjects and found that 34 species were significantly correlated with blood lipid levels, among which 6% were correlated with TC levels and 4% were correlated with HDL-C levels [42]. High-throughput sequencing of the 16S rRNA gene was performed using DNA extracted from feces samples for further microbiome analyses. From the alpha diversity index, treatment with lovastatin, L. paracasei strain 201 and E. faecium strain 132 could significantly improve the uniformity and complexity of the microbiota. From the beta diversity index, each group could be better separated by an unweighted UniFrac distance. In terms of the abundances of families, after intervention with lovastatin, L. paracasei strain 201 or E. faecium strain 132, those of Veillonellaceae and Erysipelotrichaceae and were significantly decreased compared to those in the HC group. The proportion of Veillonellaceae is positively correlated with all metabolic diseases caused by a high-fat diet [43]. In addition, studies have found that Erysipelotrichaceae can also cause abnormal lipid metabolism in the host [44]. At the genus level, the abundance of Lachnospiraceae and Butyricicoccus were significantly increased by treatment with L. paracasei strain 201 or E. faecium strain 132 compared to the HC group. The abundance of Prevotella was significantly decreased in the HC-lovastatin and HC-E. faecium strain 132 groups compared with that in the HC group. In studies on intestinal microbial transplantation, the abundance of Prevotella was positively correlated with plasma trimethylamine oxide (TMAO) level, indicating the influence of microorganisms on the formation of atherosclerotic lesions [41]. SCFAs not only serve as energy substrates, but also promote nutrient absorption, maintain microbial interactions and reduce the risk of cardiovascular events [45]. Studies have shown that both Lachnospiraceae and Butyricicoccus can produce short chain fatty acids (SCFAs), and butyrate and propionate have a protective effect on diet-induced obesity; propionate can especially reduce the synthesis of cholesterol in the liver and improve lipid metabolism [46][47][48]. The results have indicated that the E. faecium strain 132 or L. paracasei strain 201 might improve the microbiota to regulate the lipid-related metabolic balance. The inadequacy of our study is that we should have combined multiomics techniques to explore key substances and specific pathways.

Conclusions
In summary, oral administration of E. faecium strain 132 and L. paracasei strain 201, with their cholesterol-lowering excellent activity, ameliorated lipid metabolism, liver injury and gut microbiota. In terms of the cholesterol-lowering mechanisms, we found that genes were involved in cholesterol and lipid levels. CYP8B1, CYP7A1, SREBP-1, SCD1 and LDL-R were regulated in hypercholesteremia rats intervened by two strains. However, they might also alleviate liver injury by reducing inflammation and ultimately increasing the production of acetic acid and propionic acid in fecal samples to improve hyperlipidemia in rats. These findings support that E. faecium strain 132 and L. paracasei strain 201 could have promising applications for the treatment of hypercholesteremia.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/10 .3390/nu13061982/s1, Figure S1: The trend of weight gained in each group at 6 weeks, Figure S2: The species abundance of phylum, class, order, family and genus levels, Table S1: The cholesterol-lowing rate of isolates.
Author Contributions: Y.L., Q.W., X.X. and J.W. participated in the design of this study. Y.L., H.L., X.X. and L.Y. contributed to this work and performed statistical analyses. L.W., C.F., T.L., Y.X., S.Y., Y.D., M.C., L.X. and J.Z. gathered important background information. Y.L. drafted the manuscript. All authors have read and agreed to the published version of the manuscript.

Conflicts of Interest:
The authors state that the study was conducted with no commercial advertising or a financial relationship may be interpreted as a potential conflict of interest.