English

• Over the past two decades, there has been a growing concern about the detrimental effects of severe air pollution. Air pollution is the leading environmental risk factor for allergies, diseases, and even mortality among humans. According to a report by the World Health Organization (WHO), approximately 90% of the global population resides in areas where air quality falls below WHO guideline standards [1]. Among the various pollutants, fine particulate matter (PM_2.5), with an aerodynamic diameter <2.5 µm, is particularly detrimental [2].

  PM_2.5 can quickly penetrate into the bloodstream, different tissues and organs (such as the aorta), and various cell types (such as endothelial cells and macrophages) after reaching the lung epithelium [3,4]. Epidemiological and laboratory studies have accumulated evidence linking PM_2.5 exposure to an increased risk of cardiovascular diseases (CVDs), particularly in vulnerable populations [5]. Some studies have shown a close association between PM_2.5 exposure and elevated incidence and mortality rates of various CVDs, including atherosclerosis, heart failure, and myocardial infarction [6]. The causality between PM_2.5 exposure and CVDs is therefore worth examining.

  Long-term consumption of a high-cholesterol diet (HCD) was proven to be associated with an increased risk of CVDs, particularly atherosclerosis, as demonstrated in epidemiological research [7]. HCD may heighten the susceptibility of the cardiopulmonary system to the adverse health effects associated with PM_2.5 exposure [8,9]. However, the precise molecular mechanisms underlying the compounded cardiovascular injury caused by the combination of PM_2.5 exposure and high cholesterol intake remain unclear. In recent years, metabolomics has emerged as a high-throughput tool for biomarker discovery, elucidating disease mechanisms, and investigating potential therapeutic targets [10,11]. By analyzing the variations in endogenous metabolites, which serve as the end products of cellular alterations, metabolomics provides valuable and more closely related insights into the phenotypic changes compared to other layers in the omics cascade.

  Through metabolomics analysis, we examined the alterations in small molecules induced by whole-body real-world PM_2.5 exposure in ApoE^−/− mouse aorta with or without HCD. Additionally, we validated these changes in low-density lipoprotein (LDL) oxidation levels and arachidonic acid (AA) metabolism using mouse aortic endothelial cells (MAECs). Our findings revealed that PM_2.5 exposure exacerbated inflammation and increased the formation of atherosclerotic plaque in response to disruptions in AA metabolism caused by HCD feeding. These findings provide important insights into the mechanism of PM_2.5-induced vascular toxicity.

  A total of 40 ApoE^−/− mice of approximately 4 weeks were randomly grouped into 4 groups. Based on the diet (i.e., normal diet or high-cholesterol diet) and air conditions (i.e., filtered air or PM_2.5 air), the four groups were named the control group, PM_2.5 group, HCD group, and HCD-PM_2.5 group. The exposure experiments were conducted in a previously established real-world PM_2.5 system [12] from November 2019 to April 2020. After 6 months of exposure, all mice were sacrificed. The serum and tissues were collected and snap-frozen immediately and stored at −80 ℃ for the following histochemical staining, ELISA assays, and untargeted metabolomics. During the exposure period, the PM_2.5 samples at the same location were collected for the following MAEC experiments. All housing and experimental protocols received ethical approval from the Research Ethics Committee of Hong Kong Baptist University.

  The statistical analyses were conducted with Prism 8.0 (GraphPad Software, California, USA). The data were presented as mean ± standard deviation (SD). To determine the presence of significant differences, ANOVA tests were applied. Ingenuity Pathway Analysis (IPA, Qiagen, Valencia, US) was used to predict canonical pathways and related molecules between metabolites and ox-LDL. More details for experimental methods can be found in the Supporting Information.

  The ApoE^−/− mice were provided with HCD and exposed to PM_2.5 in cages for 6 months. Compared to the control group, blood glucose was declined impacted by PM_2.5 (P < 0.01), while the trend in response to the HCD was the same in the mice (Fig. 1A). The absolute and relative (to body weight) aorta weights have similar changes. A significantly increased trend could only be observed due to the impacts of HCD (Figs. 1B and C, P < 0.001 in Fig. 1B, P < 0.01 in Fig. 1C). The Oil Red-O staining for arteries was a visual way to display atherosclerosis. As illustrated in the cross-sections of arcus aortae photographs, the aortic wall was thickened after the HCD feeding, reducing the vascular inner diameter (Figs. 1D and E). The quantitative results of ORO-stained areas identified PM_2.5 and HCD significantly increased the atherosclerotic plaque formation additively (Fig. 1D). Six cytokines exhibited significant alterations, including thymus and activation-regulated chemokine (TARC), platelet factor 4 (PF4), keratinocyte chemoattractant (KC), lipopolysaccharide-induced CXC chemokine (LIX), intercellular cell adhesion molecule-1 (ICAM-1), and T-cell activation protein 3 (TCA-3) (Fig. 1F). Upon HCD introduction, the serum levels of all 6 cytokines elevated in general. TARC had the highest rise among these cytokines, which showed a 4.15-fold increase compared to the control group with a P < 0.0001 (Fig. 1F). PM_2.5 exposure triggered an additional increase in TCA-3, which might further aggregate atherogenesis caused by HCD. Moreover, intriguing interaction between PM_2.5 and HCD has been evidenced by the changes in the other four cytokines (TARC, PF4, KC, and LIX). Under the HCD-feeding conditions, PM_2.5 exposure downregulated the serum levels of these chemokines, suggesting an interaction between dietary factors and air pollutants influencing immune responses. Both H&E and Masson's trichrome staining in the aorta distinctly revealed marked infiltration of inflammatory cells, a larger number of foam cells, the formation of cholesterol crystals, and a greater fibrotic area in the arcus aortae of the co-treated mice (Fig. 1G). The quantitative analysis of the fibrotic area ratio, derived from Masson's trichrome stain, showed that the aortic fibrotic area was impacted by both diet and PM_2.5 exposure. PM_2.5 exposure resulted in more significant fibrosis in the HCD mice compared to those on a normal diet (Fig. S1 in Supporting information). Together, these data revealed that PM_2.5 exposure had negative effects on the pathogenesis of atherosclerosis in mice, and HCD-fed mice were more susceptible to PM_2.5-caused damage.

  Figure 1

  

  Figure 1.  Adverse impact of PM_2.5 inhalation on atherosclerotic risk. (A) Blood glucose, (B) aorta weight, (C) weight ratio of aorta-to-body, (D) quantitative results of Oil-Red-O staining in arcus aortae. (E) Representative Oil-Red-O staining for whole arteries (upper) and arcus aortae at ascending aorta (lower). (F) Inflammatory cytokines in murine serum, (G) H&E- and Masson's trichrome-stained arcus aortae sections. *, P < 0.05; **, P < 0.01, ***, P < 0.001; ****, P < 0.0001 from Tukey's multiple comparisons test (on the top of the plot) and the two-way ANOVA (on the bottom).

  DownLoad: Full-Size Img PowerPoint

  We conducted untargeted metabolomics of murine aorta and serum to explore the molecular mechanisms. The score plot of partial least squares-discriminant analysis (PLS-DA) of metabolomic data indicated the metabolic difference between groups (Figs. 2A–D). To prevent model overfitting, the 5-fold cross-validation and a 100-time permutation test were used to evaluate the predictive abilities of the PLS-DA model (Fig. S2 in Supporting information). We identified 116 metabolites in the aorta and 205 in the serum. Among them, 39 and 88 metabolites exhibited significant alterations in at least one of the 5 pairwise comparisons (see methods in Section 1 in Supporting information), in the aorta and serum, respectively (Tables S2 and S3 in Supporting information). The identified metabolites indicated the alterations in 17 and 32 KEGG metabolism pathways in the aorta and serum, respectively (Figs. 2E and F).

  Figure 2

  

  Figure 2.  Influence of PM_2.5 exposure on untargeted metabolomic profiles. PLS-DA score charts of aorta metabolomics in positive (A) and negative (B) ionization modes. PLS-DA score of serum metabolomics displays in positive (C) and negative (D) ionization modes. Differential pathway mapping for altered aortic (E) and serum (F) metabolites. Figures were produced via MetaboAnalyst.

  DownLoad: Full-Size Img PowerPoint

  The disturbed AA metabolism highlighted in both the aorta and serum, whose change was considered a serious cardiovascular risk and linked to atherosclerosis [13], was brought to our attention. Our metabolomics quantification detected changes in a lot of metabolites of this pathway, including AA, prostaglandin (PG)F_2α, 20-hydroxy-(5Z,8Z,11Z,14Z)-eicosatetraenoic acid (20-HETE), 11,12-EET, thromboxane (TX)B₂, and 6-keto-PGF_1α (Table S3), while some of them (i.e., AA and 6-keto-PGF_1α) altered also in the aorta (Table S2). The correlation analysis between the alterations in eicosanoids caused by HCD and varied pathological indicators revealed significant associations. Specifically, 20-HETE and TXB₂ were found to be negatively correlated with blood glucose. Notably, PGF_2α was significantly correlated with several indicators, including TCA-3, aorta weight, aorta-to-body ratio, PF4, and KC (Fig. S3A in Supporting information). Blood glucose was negatively correlated with 20-HETE and TXB₂ as altered by HCD. Conversely, PM_2.5-induced AA changes in the context of the HCD background showed a positive relationship with blood glucose levels (Fig. S3B in Supporting information). The PM_2.5-caused differential metabolites were predominantly observed in the serum. A comparison between the HCD and HCD-PM_2.5 groups demonstrated significant shifts in multiple amino acids, evidencing the disruptions in metabolism pathways, such as biosynthesis and degradation of valine, leucine and isoleucine. Additionally, significant changes were observed in metabolites related to the eicosanoid pathway, including adrenic acid, 6-keto prostaglandin F_1α, thromboxane B₂, and N-arachidonoyl dopamine, suggesting PM_2.5 may aggravate HCD-triggered injury through disruption of prostaglandin metabolism.

  Aortic endothelial cells (AECs) are the primary producers of AA and its metabolites in response to inflammation [14,15]. The alterations of AA metabolites (PGF_2α, and 6-keto-PGF_1α) in the results of untargeted metabolomics suggested that AECs may play a significant role in the PM_2.5 and HCD impacts. Previous studies showed that these metabolites prefer to be upregulated in AECs when injured [16,17]. We, therefore, employed MAECs to validate the effects of the co-treatment at the molecular level. MAECs were first exposed to different gradients of concentration in PM_2.5 and LDL (i.e., the carrier of cholesterol and fat) to detect dose-dependent cytotoxicity using the cell viability MTT assay (Figs. S4A and B in Supporting information). We chose 25 µg/mL and 12.5 µg/mL as the assay doses for PM_2.5 and LDL, respectively, because applying dosages higher than these thresholds resulted in significant decreases in cell viability. Moreover, the ROS assay further confirmed the cytotoxicity of PM_2.5 and LDL at these levels (Figs. S4C and D in Supporting information).

  Targeted metabolomics results indicated that, in contrast to LDL, which enhances AA production in MAECs (Fig. 3A), PM_2.5 significantly inhibited AA production in both single and combined exposure scenarios. These results were consistent with those from the animal experiment, in which HCD and PM_2.5 showed significant impacts on the AA levels, respectively. We identified metabolites from 3 different enzyme systems (cyclooxygenases (COXs), lipoxygenases (LOXs), and cytochrome P450 (CYP)) central to AA metabolism. The COX enzyme metabolic pathway was the most affected by PM_2.5 and LDL, with a total of 5 substances (PGE₂, PGD₂, PGJ₂, PGA₂, and PGB₂) exhibiting changes in abundance. PGE₂, the principal product of the COX enzyme via oxygenation of AA, is strongly associated with inflammation and is a predictor of inflammation [14,18]. PM_2.5 exposure showed a significant increase in PGE₂ production in both single and combined exposure scenarios (Fig. 3B), implying inflammation. The abundance of the downstream products of PGE₂, PGA₂ and PGB₂, had similar changes (Figs. 3C and D). PGD₂, a predictor of CVDs involved in vasoconstriction, blood pressure regulation, and induction of ROS [19], behaved similarly to PGE₂. LDL did not lead to PGD₂ production by COX enzymes, while PM_2.5 induced PGD₂ production (Fig. 3E). Furthermore, PGD₂ abundance was higher in the co-treated group than in the control and LDL groups, suggesting that co-treatment with PM_2.5 and LDL exacerbated the detrimental effects on MAECs (Fig. 3E). PGJ₂, the downstream metabolite of PGD₂, exhibited the same alteration pattern (Fig. 3F). Within the LOX pathway, Leukotriene (LT) B₄, another key factor in cardiovascular risk and inflammation that increases vascular permeability, was significantly increased by PM_2.5 (Fig. 3G). Epoxyeicosatrienoic acids (EETs) have been reported to contribute to the regulation of vascular function and possess vasodilatory, cardioprotective, and anti-inflammatory activities [14]. Interestingly, in the CYP pathway, 14,15-EETs were reduced by sole PM_2.5 and combined LDL-PM_2.5 treatments (Fig. 3H).

  Figure 3

  

  Figure 3.  Metabolic variations in arachidonic acid pathway post PM_2.5 and LDL exposure: Concentration comparisons of arachidonic acid (A) and metabolites in COX pathway (B–F), LOX pathway (G), and CYP pathway (H). Values are mean ± SD for 6 samples. Significance denoted by *, P < 0.05; **, P < 0.01; ***, P < 0.001 were from Tukey's multiple comparisons test (on the top of plots) and two-way ANOVA (inserts).

  DownLoad: Full-Size Img PowerPoint

  Circulating ox-LDL is involved in early-stage atherogenesis by stimulating the formation of foam cells. Once accumulating in the vessel walls, LDL is oxidized in the presence of excess ROS. To investigate the conversion of LDL oxidation by damaged endothelial cells, we used an ELISA kit to determine the ox-LDL content in the cell supernatant of MAEC after exposure to PM_2.5 and LDL. The results showed that the existence of MAEC caused the oxidation of LDL, and the ox-LDL production was increased in the presence of PM_2.5 (Fig. 4A).

  Figure 4

  

  Figure 4.  Analyses of ox-LDL synthesis and LDL oxidation bioinformatics: (A) Quantification of ox-LDL synthesis levels, (B) canonical pathway differentiation between standard LDL and PM_2.5-modified LDL groups, (C) metabolite and ox-LDL interaction networks. Data in the subplot (A) are presented as mean ± SD for 6 samples. Statistical significance was determined by one-way ANOVA with Dunnett's post-hoc test, noted with different letters at P < 0.05. The threshold for selecting pathways in Ingenuity Canonical Pathway analysis was set as P < 0.05.

  DownLoad: Full-Size Img PowerPoint

  We used the software IPA to comparatively analyze the targeted metabolomics data in the LDL and LDL-PM_2.5 groups. The resultant top-10 signal pathways by which metabolites impact ox-LDL production were shown in Fig. 4B. The eicosanoid signaling and synthesis of PG and TX were the most notable pathways with P-values of 7.4 × 10⁻⁶ and 8.3 × 10⁻⁶, respectively. The changes in AA metabolites might affect the levels of ATF2, CREB1, PPARα, HMOX1, RXRA, SIRT1, palmitic acid, and cholesterol. These changes, in turn, could impact ox-LDL levels by inhibiting the heme signaling pathway (Fig. 4C). PGs may regulate LDL oxidation by impacting some modules and the heme signaling pathway. PGE₂ is related to heme signaling. A high level of PGE₂ can indirectly induce the increase in ATF2 and HMOX1, which leads to the inhibition of the pathway. However, PGE₂ may increase the CREB1 and decrease SIRT1 levels to activate the pathway. In addition, PGJ₂ may upregulate palmitic acid, PPARα, and HMOX1 to inhibit the heme signaling. The effect of PGD₂ is similar to PGE₂, which can induce PPARα, HMOX1, RXRα, and NFE2L2 to inhibit the heme signaling and increase the CREB1 activity. Compared to the metabolites in the COX pathway, LTB₄ directly increases PPARα to inhibit the heme signaling. Moreover, it is noteworthy that LOX constitutes a class of non-heme iron-containing dioxygenases.

  As one of the key intracellular oxidation enzymes, LOX directly catalyzes the oxygenation of polyunsaturated fatty acids [20]. For instance, 15-LOX can oxidize LDL through both direct enzymatic pathways and indirect non-enzymatic mechanisms in vitro [20]. PGs, exemplified by PGE₂, upregulate the expression of 15-LOX, thereby modulating the substrate flux and diverting it from LTB₄ synthesis towards the 15-LOX metabolic pathway [14]. The oxidation of LDL is a critical event in atherogenesis and multiple mechanisms co-regulate the LDL oxidation. Heme plays a key role in LDL oxidation. Free heme initially binds to the lipoproteins LDL and HDL which are highly susceptible to oxidation by ferriheme (FeHM) [21,22]. AA metabolites may inhibit heme signaling and regulate the ox-LDL levels via both direct enzymatic pathways and indirect non-enzymatic mechanisms in vitro, thereby promoting atherosclerotic plaques, foam cells and inflammation.

  Utilizing an animal exposure model grounded on realistic PM_2.5 concentrations, the present study elucidates the risk of atherosclerosis resulting from PM_2.5-induced inflammation, fibrosis, and vascular damage in the HCD-fed mice. The aortas of mice fed on HCD exhibited greater susceptibility to PM_2.5 than those of normal mice. Furthermore, metabolomic data underscored the pivotal role of altered AA metabolic homeostasis in the disease risk associated with PM_2.5. The results of ox-LDL assay experiments, together with physiological indices and metabolite changes, validated the AA metabolite-mediated adverse effects of PM_2.5 in atherogenesis. These findings significantly contribute to our understanding of how PM_2.5 exposure induces vascular injury in individuals with excessive cholesterol intake. The mechanistic connection between PM_2.5 toxicity and ox-LDL binding to endothelial cells in vivo is an interesting yet challenging issue that requires further investigation.

  One limitation of this study is the absence of detailed information on the particular toxicity contribution of PM_2.5 components. It has been documented that PM_2.5, being a complex cocktail of contaminants from various sources, can cause diverse harmful impacts depending on its compositions. The present study did not delineate the specific constituents of PM_2.5 responsible for the observed metabolic alterations, although the use of a real-world exposure setting did facilitate the assessment of the PM_2.5 toxicity under environmental conditions. Consequently, even though our findings indicate that PM_2.5 may increase atherogenesis risk through disruption of the prostaglandin pathway and oxidation of low-density lipoprotein, exploring the effective precautions against the PM_2.5-caused atherogenesis risks remains challenging. Also, this work was limited to murine models. Though six months of mouse experiments is, to some extent, equivalent to years of human exposure to ambient PM_2.5, other toxic mechanisms may be present in other species, including humans.

  Declaration of competing interest

  The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

  CRediT authorship contribution statement

  Xuyang Chai: Writing – original draft, Visualization, Validation, Methodology, Investigation, Data curation. Hiu-Lok Ngan: Visualization, Validation, Methodology, Investigation, Formal analysis, Data curation. Yuanyuan Song: Writing – review & editing, Project administration, Methodology, Investigation, Conceptualization. Zenghua Qi: Writing – review & editing, Validation, Project administration, Methodology. Lifang Zhao: Resources. Wenqi Chen: Resources. Shanshan Chen: Resources. Zhenhua Yang: Resources. Ruijin Li: Resources. Chuan Dong: Resources. Zhu Yang: Writing – review & editing, Supervision, Project administration, Methodology, Conceptualization. Zongwei Cai: Writing – review & editing, Project administration, Methodology, Funding acquisition, Conceptualization.

  Acknowledgment

  This work received support from the Hong Kong General Research Fund (Nos. 12302922 and 12303320).

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2024.110671.

  
  
• 1. [1]

     WHO, New WHO Global Air Quality Guidelines aim to save millions of lives from air pollution, 2021. https://www.who.int/news/item/22-09-2021-new-who-global-air-quality-guidelines-aim-to-save-millions-of-lives-from-air-pollution.

  2. [2]

     J. Marshall, Proc. Natl. Acad. Sci. U. S. A. 110 (2013) 8756. doi: 10.1073/pnas.1307735110

  3. [3]

     A. Sidwell, S.C. Smith, C. Roper, J. Toxicol. Environ. Health B 25 (2022) 422–444. doi: 10.1080/10937404.2022.2142345

  4. [4]

     C.C. Cho, W.Y. Hsieh, C.H. Tsai, et al., Int. J. Env. Res. Public Health 15 (2018) 1380. doi: 10.3390/ijerph15071380

  5. [5]

     P. Yin, J. Guo, L. Wang, et al., Environ. Sci. Technol. Lett. 7 (2020) 95–101. doi: 10.1021/acs.estlett.9b00735

  6. [6]

     Z. Qi, Y. Zhang, Z.F. Chen, et al., Ecotoxicol. Environ. Saf. 201 (2020) 110827. doi: 10.1016/j.ecoenv.2020.110827

  7. [7]

     M. Kratz, Dietary cholesterol, atherosclerosis and coronary heart disease, in: A. von Eckardstein (Ed.), Atherosclerosis: Diet and Drugs, Springer Berlin Heidelberg, Berlin, Heidelberg, 2005, pp. 195–213.

  8. [8]

     V.W. Zhong, L. Van Horn, M.C. Cornelis, et al., JAMA 321 (2019) 1081–1095. doi: 10.1001/jama.2019.1572

  9. [9]

     X. Li, J. Gu, Q. Xiao, et al., Chin. Chem. Lett. 34 (2023) 107483. doi: 10.1016/j.cclet.2022.04.081

  10. [10]

      C.H. Johnson, J. Ivanisevic, G. Siuzdak, Nat. Rev. Mol. Cell Biol. 17 (2016) 451–459. doi: 10.1038/nrm.2016.25

  11. [11]

      A. Tan, X. Ma, Chin. Chem. Lett. 35 (2024) 109276. doi: 10.1016/j.cclet.2023.109276

  12. [12]

      Y. Song, L. Zhao, Z. Qi, et al., J. Environ. Sci. 105 (2021) 64–70. doi: 10.1016/j.jes.2020.12.007

  13. [13]

      T. Sonnweber, A. Pizzini, M. Nairz, G. Weiss, I. Tancevski, Int. J. Mol. Sci. 19 (2018) 3285. doi: 10.3390/ijms19113285

  14. [14]

      B. Wang, L. Wu, J. Chen, et al., Signal Transduct. Target. Ther. 6 (2021) 94. doi: 10.15376/biores.17.1.94-108

  15. [15]

      N.V. Bogatcheva, M.G. Sergeeva, S.M. Dudek, A.D. Verin, Microvasc. Res. 69 (2005) 107–127. doi: 10.1016/j.mvr.2005.01.007

  16. [16]

      M.J. Vassar, C.J. Weber, J.W. Holcroft, Prostaglandins, Leukot. Essent. Fatty Acids 33 (1988) 129–135. doi: 10.1016/0952-3278(88)90152-4

  17. [17]

      T. Arnould, R. Thibaut-Vercruyssen, N. Bouaziz, et al., Am. J. Pathol. 159 (2001) 345–357. doi: 10.1016/S0002-9440(10)61701-4

  18. [18]

      F. Cipollone, Lupus 14 (2005) 756–759. doi: 10.1191/0961203305lu2215oa

  19. [19]

      D. Kong, Y. Yu, J. Mol. Cell. Cardiol. 167 (2022) 97–105. doi: 10.1016/j.yjmcc.2022.03.011

  20. [20]

      H. Yoshida, R. Kisugi, Clin. Chim. Acta 411 (2010) 1875–1882. doi: 10.1016/j.cca.2010.08.038

  21. [21]

      Y.I. Miller, N. Shaklai, Biochim. Biophys. Acta 1454 (1999) 153–164. doi: 10.1016/S0925-4439(99)00027-7

  22. [22]

      V. Jeney, J. Balla, A. Yachie, et al., Blood 100 (2002) 879–887. doi: 10.1182/blood.V100.3.879

• 

  Figure 1  Adverse impact of PM_2.5 inhalation on atherosclerotic risk. (A) Blood glucose, (B) aorta weight, (C) weight ratio of aorta-to-body, (D) quantitative results of Oil-Red-O staining in arcus aortae. (E) Representative Oil-Red-O staining for whole arteries (upper) and arcus aortae at ascending aorta (lower). (F) Inflammatory cytokines in murine serum, (G) H&E- and Masson's trichrome-stained arcus aortae sections. *, P < 0.05; **, P < 0.01, ***, P < 0.001; ****, P < 0.0001 from Tukey's multiple comparisons test (on the top of the plot) and the two-way ANOVA (on the bottom).

  下载: 全尺寸图片 幻灯片
  

  Figure 2  Influence of PM_2.5 exposure on untargeted metabolomic profiles. PLS-DA score charts of aorta metabolomics in positive (A) and negative (B) ionization modes. PLS-DA score of serum metabolomics displays in positive (C) and negative (D) ionization modes. Differential pathway mapping for altered aortic (E) and serum (F) metabolites. Figures were produced via MetaboAnalyst.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  Metabolic variations in arachidonic acid pathway post PM_2.5 and LDL exposure: Concentration comparisons of arachidonic acid (A) and metabolites in COX pathway (B–F), LOX pathway (G), and CYP pathway (H). Values are mean ± SD for 6 samples. Significance denoted by *, P < 0.05; **, P < 0.01; ***, P < 0.001 were from Tukey's multiple comparisons test (on the top of plots) and two-way ANOVA (inserts).

  下载: 全尺寸图片 幻灯片
  

  Figure 4  Analyses of ox-LDL synthesis and LDL oxidation bioinformatics: (A) Quantification of ox-LDL synthesis levels, (B) canonical pathway differentiation between standard LDL and PM_2.5-modified LDL groups, (C) metabolite and ox-LDL interaction networks. Data in the subplot (A) are presented as mean ± SD for 6 samples. Statistical significance was determined by one-way ANOVA with Dunnett's post-hoc test, noted with different letters at P < 0.05. The threshold for selecting pathways in Ingenuity Canonical Pathway analysis was set as P < 0.05.

  下载: 全尺寸图片 幻灯片
•