Ginger essential oil and citral ameliorates atherosclerosis in ApoE−/− mice by modulating trimethylamine-N-oxide and gut microbiota – npj Science of Food

GEO and citral protects against atherosclerosis and improve plasma lipidemic biomarkers

To examine the protective effect of GEO and citral against CVD and atherosclerosis, ApoE^−/− mice were fed GAN diet + water supplemented with 1.3% ʟ-carnitine (GC) to induce atherosclerosis, and treated with GEO (Low: 50 mg/kg bw and High: 100 mg/kg bw) or citral (20 mg/kg bw) daily (Fig. 1a). After 16 weeks, the mice were sacrificed and aortic lesions were visualized using oil red O staining (Fig. 1b). Aortic lesion formation significantly increased (p < 0.0001) by 212% (13.1 ± 2.4%) in the GC group compared with control (CON) group (4.2 ± 1.2%), indicating successful induction of atherosclerosis using GAN diet and ʟ-carnitine (Fig. 1c). However, treatment with low-dose GEO, high dose GEO, and citral significantly reduced the occurrence of aortic plaques by 23% (p = 0.0311), 20% (p = 0.0610), and 29% (p = 0.0043), respectively. Additionally, the plasma levels of lipidemic biomarkers, including total triglyceride, total cholesterol, high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), and oxidized low-density lipoprotein (ox-LDL), were examined (Fig. 1d–h). Compared with the CON group, plasma cholesterol (p < 0.0001), HDL-C (p = 0.0008), and LDL-C (p < 0.0001) levels were significantly elevated in the GC group, indicating that this intervention affected the circulation of lipids. However, there was no significant difference in plasma total triglyceride and ox-LDL levels between the CON and GC groups. Treatment with GEO and citral caused a decreasing trend in plasma cholesterol and LDL-C levels in ApoE^−/− mice compared with the GC group; moreover, high-dose GEO and citral increased plasma HDL-C levels (p = 0.0015 and p = 0.0054, respectively). However, there were no significant differences in the plasma levels of several biomarkers between low and high GEO groups, which may be because the effect of GEO on plasma biochemical indices peaked at 50 mg/kg bw/day and higher doses did not induce any significant change. Overall, these results showed the GEO and citral prevented atherosclerotic lesion formation and improved plasma lipid profile.

GEO and citral improves glucose and insulin homeostasis, hepatic function, and plasma pro-inflammatory cytokine levels

Furthermore, the effect of GC, GEO, and citral on glucose and insulin homeostasis, hepatic function, and plasma pro-inflammatory cytokine levels were examined. Homeostatic model assessment-insulin resistance (HOMA-IR) index was calculated to determine the effect of the treatments on insulin resistance. HFD has been shown to negatively affect glucose and insulin metabolism²¹. Plasma glucose (p = 0.0048) and insulin (p = 0.2019) levels and HOMA-IR index (p = 0.0554) were substantially higher in the GC group compared with the CON group (Fig. 2a–c). However, low-dose and high-dose GEO, and citral improved plasma glucose (p = 0.0480, p = 0.1733, and p = 0.0007, respectively) and insulin (p = 0.0270, p = 0.0298, and p = 0.0163, respectively) levels and HOMA-IR values (p = 0.0113, p = 0.0192, and p = 0.0018, respectively). Plasma levels of hepatic parameters, including aspartate aminotransferase (AST; p < 0.0001) and alanine aminotransferase (ALT; p = 0.0319), increased by 2.2- and 3.4-fold, respectively, in the GC group compared with the CON group (Fig. 2d, e). However, treatment with low-dose and high-dose GEO and citral substantially ameliorated the ALT levels (p = 0.1216, p = 0.1789, and p = 0.0491, respectively), but did not affect AST levels. Compared with the CON group, there was a significant increase in plasma levels of pro-inflammatory cytokines in the GC group, including tumor necrosis factor-α (TNF-α, p < 0.0001), interleukin 6 (IL-6, p = 0.0149), and interleukin-1β (IL-1β, p = 0.0638). However, treatment with low and high doses of GEO tended to lower plasma TNF-α and IL-6 levels and significantly reduce IL-1β levels (p = 0.0413 and p = 0.0835, respectively). Moreover, treatment with citral significantly reduced plasma TNF-α (p = 0.0212), IL-6 (p = 0.1014), and IL-1β (p = 0.0520) levels. Overall, these findings showed that both GEO and citral have the potential to alleviate systemic inflammation, hepatic damage, and insulin resistance.

Plasma levels of (a) glucose and (b) insulin; (c) homeostasis model assessment for insulin resistance (HOMA-IR); d aspartate aminotransferase (AST); e alanine aminotransferase (ALT); f tumor necrosis factor-α (TNF-α); g interleukin 6 (IL-6); and h interleukin-1β (IL-1β). Dot plots are expressed as the mean ± SD. Statistical analyses were performed using an unpaired two-tailed Student’s t-test, CON vs. GC groups; one-way ANOVA with Tukey’s range test for comparing GC, GC + GEO_Low, GC + GEO_High, and GC + CIT. CON: control diet group, GC: Gubra Amylin NASH diet [GAN diet] + ʟ-carnitine in drinking water [1.3%] group, GC + GEO_Low: GC + Ginger essential oil (GEO) [50 mg/kg bw/day] group, GC + GEO_High: GC + GEO [100 mg/kg bw/day] group, GC + CIT: GC + Citral [20 mg/kg bw/day] group.

GEO and citral remodels gut microbiota and suppresses meta-organismal ʟ-carnitine-TMAO metabolic pathway

TMAO is an atherosclerotic risk factor. The primary multistep meta-organismal (gut microbiota and host) pathway for TMAO production through carnitine metabolism is as follows: ʟ-carnitine → γ-butyrobetaine (γBB) → TMA → TMAO²². Compared with the CON group, plasma TMA (p = 0.0032), TMAO (p = 0.0135), γ-BB (p < 0.0001), and carnitine (p < 0.0001) increased by 11.2-, 4.1-, 43.3-, and 1.9-fold, respectively, in the GC group (Fig. 3a–d). However, low and high doses of GEO and citral significantly reduced plasma TMA and TMAO levels (p = 0.0398 and p = 0.0202 for low and high doses of GEO, respectively). Additionally, there was a decreasing trend in plasma γ-BB levels in the citral and GEO treatment group, with more obvious decrease observed in citral treatment group (p = 0.0663).

Plasma levels of (a) trimethylamine (TMA), (b) trimethylamine-N-oxide (TMAO), (c) γ-butyrobetaine (γ-BB), and (d) carnitine. e Observed amplicon sequencing variants (ASVs); f Shannon diversity index; g principal coordinate analysis (PCoA) plot based on Bray–Curtis dissimilarity with gut microbiota-associated vector (n = 7–8). Dot plots are expressed as the mean ± SD. Box plots display median, mean (+), quartiles (boxes), and range (whiskers). Statistical analyses were performed using an unpaired two-tailed Student’s t-test CON vs. GC groups; one-way ANOVA with Tukey’s range test for comparing GC, GC + GEO_Low, GC + GEO_High, and GC + CIT. Analysis of similarity (ANOSIM) was calculated to determine heterogeneity of the fecal microbiota among the groups in PCoA. Vectors in the PCoA plot indicated a significant genus (p < 0.001), and its length shows the strength of the correlation. CON: control diet group, GC: Gubra Amylin NASH diet [GAN diet] + ʟ-carnitine in drinking water [1.3%] group, GC + GEO_Low: GC + Ginger essential oil (GEO) [50 mg/kg bw/day] group, GC + GEO_High: GC + GEO [100 mg/kg bw/day] group, GC + CIT: GC + Citral [20 mg/kg bw/day] group.

Since gut microbiota plays a critical role in the development of CVD and atherosclerosis, we examined the effect of GC, GEO, and citral on gut microbiota. To achieve this, the microbiota compositions of fecal samples were examined using V3–V4 16 S rRNA sequencing technique. The raw reads were processed using the QIIME2 pipeline to obtain the amplicon sequence variants (ASVs), which were compared against the SILVA database (version 132) for taxonomic classification. A total of 1504 ASVs were generated, which were assigned to 190 species and 122 genera. Additionally, the α-diversity, including observed amplicon sequencing variants (ASVs) and Shannon diversity index, of fecal microbiota was calculated using the vegan package in R species. Compared with the CON group, there was a decrease in observed ASVs (p = 0.0003) and Shannon diversity index in the GC group, indicating that GAN diet and ʟ-carnitine decreased the abundance of specific gut microbes (Fig. 3e, f). However, GEO and citral treatments increased the Shannon diversity index, with the effect of citral being significant (p = 0.0041).

Furthermore, the β-diversity of the gut microbiota was examined based on Bray−Curtis distance (Fig. 3g). GC, GEO, and citral affected and remodeled the fecal microbiota. Principal coordinate analysis (PCoA) showed significant separation among all groups (ANOSIM: R = 0.6812, p < 0.001). Specifically, the GC treatment induced a significant microbiome shift from the CON group, especially at the X-axis (PCoA1; 35.14%), indicating that the GAN/ʟ-carnitine diet altered the fecal microbiota. GEO and citral also had a secondary impact on the gut microbiome. Citral and GEO treatments induced a dose-dependent shift in microbiome from the GC group, especially at the Y-axis (PCoA2; 12.77%), with a more obvious shift in gut microbiome observed in the dot of high-dose GEO group. Furthermore, the PCoA plot provided information on the degree of aortic lesion and TMA and TMAO levels, with a larger circle indicating more severe aortic lesions and deeper colors in the circle indicating higher levels of TMA or TMAO. Consistent with the results of the histological analysis (Fig. 1c), a larger circle with deeper colors were observed in the GC group compared with the CON group; however, the GEO and citral groups tended have smaller circle sizes with lesser color intensity. Additionally, the envfit function in R package was used to elucidate the association between the genus and distance structure of the gut microbiome. Vectors in the PCoA plot represented a significant genus (p < 0.001), and the lengths indicated the strength of association. The CON treatment was correlated with beneficial genera, such as Lactobacillus, Alistipes, Bifidobacterium, and other bacteria groups, including Eubacterium xylanophilum group, Ruminococcaceae UCG−014, Ruminococcaceae UCG−013, Desulfovibrio, Candidatus Saccharimonas, Candidatus Stoquefichus, Turicibacter, Family XIII UCG−001, Ruminiclostridium 5, Lachnospiraceae FCS020 group, and Clostridium sensu stricto 1. The GC treatment was associated with Fecalibaculum. The low-dose GEO treatment was associated with the beneficial mucin degrading genus Akkermansia and other bacteria, such as Enterococcus and Parasutterella; the citral treatment was associated with CVD negatively correlated microbiomes Allobaculum and Dubosiella, and other bacteria—the Coriobacteriaceae UCG−002, Lachnoclostridium, and Eubacterium coprostanoligenes group. Overall, these results showed that GEO and citral potentially re-shaped the gut microbiome and decreased the meta-organismal metabolism of ʟ-carnitine by both gut microbiota and host to form TMAO.

GEO and citral supplementation modulates gut microbiota at the genus level

Kruskal–Wallis test was performed to identify significantly different (p < 0.05) genera among the five experimental groups, which were illustrated in a heatmap (Fig. 4). A total of 47 significantly different genera were identified among the treatment groups. The left panel displaying the hierarchical clustering of gut microbiota at the genus level was divided into two major clusters indicating the differences between the CON and GC groups. These data indicated that diet (control diet or GAN diet with ʟ-carnitine) was the primary factor influencing fecal microbiota at the genus level. The GC cluster consisted of 29 genera, whereas the CON cluster consisted of 18. GEO and citral treatments formed sub-clusters within the GC-induced cluster, indicating that GEO and citral played a secondary role in altering gut microbiota. The top panel indicates the degree of aortic lesions and TMA and TMAO levels in the treatment groups. Compared with the CON group, aortic lesion formation and TMA and TMAO levels were significantly higher in the GC group; however, GEO and citral treatment reversed these parameters. Statistical analyses were performed to determine significant differences between the group, and the p values are displayed in the left panel.

Heatmap of the relative abundance of significantly different fecal microbiota using Kruskal–Wallis test (p < 0.05), and Spearman’s correlation analysis between gut microbiota components at the genus level and aortic lesion and gut microbiota metabolites. Pairwise statistical analyses were performed using an unpaired Wilcoxon signed-rank test CON vs. GC groups; Kruskal–Wallis test with Dunn’s multiple comparison test for comparing GC, GC + GEO_Low, GC + GEO_High, and GC + CIT. CON: control diet group, GC: Gubra Amylin NASH diet [GAN diet] + ʟ-carnitine in drinking water [1.3%] group, GC + GEO_Low: GC + Ginger essential oil (GEO) [50 mg/kg bw/day] group, GC + GEO_High: GC + GEO [100 mg/kg bw/day] group; GC + CIT: GC + Citral [20 mg/kg bw/day] group.

Compared with the CON group, the GC group was enriched in CVD-related bacteria, including Enterorhabdus, Romboutsia, Proteus, Eubacterium nodatum group, Escherichia-Shigella, Eubacterium coprostanoligenes group, Parasutterella, Muribaculum, and Enterococcus, and there was a decrease in the abundance of several beneficial microbiotas, such as Bifidobacterium and Alistipes. These data indicated that GAN diet and ʟ-carnitine negatively affected gut microbiota homeostasis by increasing CVD-associated microbiome and decreasing beneficial microbiome. Compared with the GC group, a total of 16, 13, and 13 significantly different genera were identified in the low-dose GEO, high-dose GEO, and citral groups, respectively. Interestingly, GEO treatment reduced the relative abundance of the CVD-associated bacteria Enterorhabdus and Proteus but increased the abundance of the beneficial bacteria Allobaculum. Additionally, citral treatment decreased the abundance of Proteus but increased Allobaculum and Dubosiella.

Furthermore, Spearman’s correlation analysis was performed to assess the relationship between the significant genus and CVD-related biomarkers, including aortic lesion, TMA, TMAO, GEO, and citral. Seven genera, including Eubacterium coprostanoligenes group, Parasutterella, Enterohabdus, Akkermansia, Romboutsia, Proteus, and Olsenella (a microbiome associated with CVD), were positively correlated with aortic lesions. Fifteen genera, including the beneficial bacteria Bifidobacterium and Alistipes, were negatively correlated with aortic lesions. Interestingly, CVD-related Enterohabdus was positively correlated with plasma TMAO levels. GEO was positively associated with a healthy microbiome consisting of Akkermansia, whereas citral was positively correlated with Allobaculum and Dubosiella. In summary, the heatmap and correlation data indicated that GC adversely affected gut microbiota, resulting in microbiota dysbiosis. However, GEO and citral exhibited a favorable effect and improved general gut microbiota composition, indicating that GEO and citral treatment restored gut microbiota and ameliorated atherosclerosis in GC ApoE^−/− mouse model.