Isolation and characterization of UCMSCs
Primary UCMSCs migrated from around the tissue block, adhered well to the plastic flask surface and showed a fusiform shape (Sup Fig. 1A); UCMSCs that showed vortex-like growth were removed and mixed with other cells when the confluence reached more than 90% (Sup Fig. 1B). At passage 3, UCMSCs exhibited a fibroblast-like shape (Sup Fig. 1C).
The UCMSCs successfully differentiated into adipocytes, osteoblasts, and chondroblasts (Sup Fig. 1D, E, F). Additionally, the UCMSCs differentiated into osteoblasts, which were positively stained with Alizarin red after 21 days of osteoblast induction (Sup Fig. 1D). The dyes bound calcium in the matrix and displayed a red colour. When UCMSCs were in the induction medium, which promoted adipocyte differentiation, lipid droplets gradually accumulated in the UCMSC cytoplasm and were stained with oil red O (Sup Fig. 1E). The differentiation of UCMSCs into chondroblasts was also recorded in vitro after the induction of MSCs for 28 days in induction medium. Overexpression and accumulation of proteoglycans and collagen I in these differentiated cells were evaluated by Alcian blue staining (Sup Fig. 1F).
At passage 3, UCMSCs showed expression of the common MSC markers CD73 (99.8±0.35%), CD90 (96.9±0.44%) and CD105 (92.0±0.34%) (Sup Fig. 1I, J, K). However, the UCMSCs were negative (or low) for the haematopoietic markers CD34 (1.6±0.32%) and CD45 (0.4±0.47%) (Sup Fig. 1G, H).
Establishment of a mouse model of AS
A mouse model of AS was successfully established, and genotype, aortic arch plaque formation, and blood hyperlipidaemia biochemical indicators were identified. Visualization on agarose gels in which lane 1 contained the DNA ladder used to indicate the size of the fragment showed that lanes 2-15 (the model mouse samples) contained 245-bp fragments, while lanes 16 and 17 (control C57 mouse samples) contained 155-bp fragments (Sup Fig. 2A, B). The target band in the wild-type mouse sample was 165 bp, and the band from the KO mice was approximately 245 bp, which suggested that mice in the model group in the experiment were ApoE-/- mice.
Movat staining of the aortic arch in the model group and control group suggested that aortic arch plaques had formed in the model group. Movat staining stains plaques red, but normal tissues are not stained. Normal control mice showed no irregular plaque formation in the aortic arch after high-fat diet feeding (Sup Fig. 2C), and irregular plaques appeared in the aortic arches of ApoE-/-mice (the plaques were stained red) (Sup Fig. 2D). We also compared the blood lipid levels of the normal control mice and ApoE-/-mice (Sup Fig. 2E), which showed that the serum total cholesterol (TC) and serum low-density lipoprotein cholesterol (LDL-C) levels in the ApoE-/-model mice were 17.19±0.93 mmol/L and 12.31±0.18 mmol/L, respectively, which were significantly higher than those in the control group (p＜0.01，n=3).
UCMSC treatment improved cardiac remodelling in mice
Compared with the model control group, the treatment groups showed a significantly shorter maximum cardiac transverse diameter (p <0.01, n = 5) (Fig. 1A) and shorter minimum cardiac transverse diameter. When the heart-to-weight ratio was defined as the ratio of the wet weight of the heart divided by body weight, the heart-to-weight ratio of the treatment group was significantly lower than that of the control group (p <0.01, n = 5) (Fig. 1B). The ventricular wall of the model control group was significantly thicker, as determined b-y HE staining, while the thickness of the ventricular wall of the treatment group was significantly thinner than that of the model control group (Fig. 1C).
UCMSC treatment reduced aortic arch plaque formation
Changes in aortic arch plaque were observed after UCMSC treatment. Compared with those in the model control group, the number and area of lipid plaques in the aortic arch inner wall in the treatment group were reduced, and no lipid deposition was observed in the aortic arch inner wall of the normal control group (Fig. 1D). Movat staining was applied as a reflection of the pathological state of the aortic arch plaque. The nuclei and elastic fibres were dyed black, the proteoglycans were dyed blue, the collagen fibres and mesh fibres were dyed yellow, the muscles were dyed red, and the fibrin and red blood cells were dyed bright red. In the model control group, the intima of the aortic arch vessel wall was thickened, foam cells had aggregated, the intima was incomplete, and elastic fibres in the vessel wall were broken. The degree of lesioning in the aortic arch vessel wall was reduced compared with that in the model group, the intima of the aortic arch was partly thickened, and the elastic fibres of the vessel wall had basically returned to normal. In the normal control group, the thickness of the blood vessel wall was uniform, and the intima and elastic fibres of the blood vessel wall were intact (Fig. 1E).
UCMSCs reduced serum lipid levels
The blood lipid levels of the three groups of mice were detected, which showed that the levels of TG, TC, and LDL-C in the UCMSC treatment group were significantly lower than those in the model control group (p <0.05, n = 15), while the HDL-C level in the UCMSC treatment group was higher than that in the model control group (p <0.05, n = 15) (Fig. 1F).
Changes in the levels of autophagy-related genes
Studies have shown that the molecular mechanism of AS is related to autophagy. In this study, we speculated that UCMSCs would regulate the metabolism of AS in mice by regulating the level of autophagy and ultimately treat AS. To test this hypothesis, we determined the transcription levels of autophagy-related genes in the heart and liver and found that autophagy-related molecules were significantly upregulated in the treatment group compared with the other two groups and that the overall autophagy level was increased upon treatment (Fig. 2).
To clarify how UCMSCs regulate lipid metabolism, the plasma and liver metabolites of the three groups were studied by metabolomics. The metabolic products were sorted in descending order according to the difference multiple, and then products whose secondary matching score was close to 1 were selected (Fig. 3). Hierarchical clustering analysis classified metabolites with similar or complementary characteristics into one group (Fig. 4) and finally screened out the metabolites with statistically significant differences, such as phosphatidylinositol (PI) and ceramide (Cer). In detail, the plasma PI and Cer levels of the treatment group were lower than those of the model control group, while the plasma PI and Cer levels of the model control group were higher than those of the normal control group (Fig. 5); the liver PI levels of the treatment group and normal control group were lower than those of the model control group (Fig, 5).
KEGG analysis was used to analyse the metabolic pathways enriched in the differential metabolites in mice to identify the pathways in which all differential metabolites participate. Through enrichment and topology analyses of the pathways that include the differential metabolites, enrichment analysis obtained the p value (Raw p), and the influence factor (Impact) was obtained from the topology analysis. The pathways were further screened according to p value, and influencing factors were ranked in ascending and descending order, which showed that the key pathways most highly correlated with the differentially abundant metabolites were glycerophospholipid metabolism and linoleic acid metabolism (Fig. 6).