Abstract Background: Cardiovascular disease (CVD) remains the most common cause of mortality in chronic kidney disease (CKD) patients. Several studies suggest that the Mediterranean diet reduces the risk of CVD due to its influence on endothelial function, inflammation, lipid profile, and blood pressure. Integrating metabolomic and proteomic analyses of CKD could provide insights into the pathways involved in uremia-induced CVD and those pathways modifiable by the Mediterranean diet. Methods: We performed metabolomic and proteomic analyses on serum samples from 19 patients with advanced CKD (aCKD) and 27 healthy volunteers. The metabolites were quantified using four different approaches, based on their properties. Proteomic analysis was performed after depletion of seven abundant serum proteins (Albumin, IgG, antitrypsin, IgA, transferrin, haptoglobin, and fibrinogen). Integrative analysis was performed using MetaboAnalyst 4.0 and STRING 11.0 software to identify the dysregulated pathways and biomarkers. Results: A total of 135 metabolites and 75 proteins were differentially expressed in aCKD patients, compared to the controls. Pathway enrichment analysis showed significant alterations in the innate immune system pathways, including complement, coagulation, and neutrophil degranulation, along with disrupted linoleic acid and cholesterol metabolism. Additionally, certain key metabolites and proteins were altered in aCKD patients, such as glutathione peroxidase 3, carnitine, homocitrulline, 3-methylhistidine, and several amino acids and derivatives. Conclusions: Our findings reveal significant dysregulation of the serum metabolome and proteome in aCKD, particularly in those pathways associated with endothelial dysfunction and CVD. These results suggest that CVD prevention in CKD may benefit from a multifaceted approach, including dietary interventions such as the Mediterranean diet. Keywords: metabolomic and proteomic analysis, chronic kidney disease, cardiovascular disease 1. Introduction Chronic kidney disease (CKD) is a complex disease with multiple etiologies, of which the most predominant are diabetes and hypertension. Patients with CKD have a 10–20-fold higher cardiovascular mortality rate compared to that of age-matched and sex-matched controls with normal renal function [[54]1], with an annual mortality rate of 15–20% that is largely attributable to cardiovascular disease (CVD). Endothelial dysfunction is well-characterized in CKD patients, regardless of the etiology, and is acknowledged as a major determinant for CVD when observed in this population [[55]2]. In addition to CKD, these patients are also exposed to other non-traditional uremia-related CVD risk factors, including inflammation, increased oxidative stress, anemia, and abnormal calcium-phosphorus metabolism [[56]3]. Several clinical trials have been conducted to analyze the impact of specific therapeutic agents, such as statins [[57]4], erythropoiesis-stimulating agents [[58]5], or AST-120 (Kremezin^®; Kureha Chemical Industry Co. Ltd., Tokyo, Japan) [[59]6], an oral, spherical activated carbon that can adsorb small-molecule uremic toxins, without identifying a clear benefit regarding the incidence of CVD or the progression to end-stage kidney disease (ESKD). Dietary modifications significantly impact patient outcomes and are postulated as an attractive complementary treatment alongside disease-specific drugs, without the constraints of secondary side effects. A growing body of evidence suggests that the Mediterranean dietary pattern (MD) exerts a significant impact on various physiological processes, including (i) a lipid-lowering effect, (ii) protection against oxidative stress, inflammation, and platelet aggregation, (iii) the inhibition of nutrient sensing pathways by specific amino acid restriction, and (iv) the gut microbiota-mediated production of metabolites that influence metabolic health. However, renal diets are quite restrictive regarding vegetable and bean consumption, due to their high concentrations of phosphorus and potassium. Nonetheless, given the benefits of DM on CVD, as well as the systematic demonstration that DM effectively preserves renal function and delays the progression of CKD [[60]7,[61]8,[62]9,[63]10,[64]11,[65]12], the KDOQI 2020 Clinical Practice Guideline for Nutrition in CKD [[66]13], as well as the current KDIGO CKD 2024 guidelines [[67]14], recommend considering the addition of a plant-based “Mediterranean-style” diet to lipid-modifying therapy to reduce CV risk. Kwon et al. have recently shown that MD is safe for patients with stage 3–4 CKD and may even contribute to preserving kidney function [[68]15]. Interestingly, to date, it is not fully understood what the individual impact is of each food or nutrient found in the Mediterranean diet, in terms of its antioxidant and anti-inflammatory effects [[69]16]. Several studies have identified many diet-associated metabolites [[70]17,[71]18,[72]19]. Furthermore, a 67-metabolite signature has been developed to reflect adherence and the metabolic response to the MD, which is associated with reduced cardiovascular risk [[73]20]. A better understanding of the altered pathways in advanced CKD (aCKD) patients, related to the risk of CVD, could help in the design of tailored Mediterranean diet schemes to optimize interventional strategies to reduce CVD risk in patients with CKD. To explore such mechanisms, we performed an integrative metabolomic and proteomic analysis on serum samples from patients with chronic kidney disease (CKD) and healthy volunteers, with the aim of identifying significantly altered metabolic pathways that are relevant to the development of CVD and, in particular, endothelial dysfunction. 2. Material and Methods 2.1. Study Population This was a cross-sectional study in which no intervention was employed. It was a nested case-control study within the project “Targeting endothelial dysfunction in highly prevalent diseases (PIE15/00027)”. Patients with an estimated glomerular filtration rate (eGFR), according to the CKD-EPI formula, of below 20 mL/min/1.73 m^2 were prospectively included for the period between July 2016 and February 2018. Patients with essential arterial hypertension or dyslipidemia (defined by statin treatment) were included. Patients who presented any of the following concomitant pathologies associated with endothelial dysfunction were excluded from this study: kidney-related hereditary diseases, diabetes mellitus, malignancies, ischemic cardiovascular diseases, lung diseases, liver diseases, thyroid disorders, and active virus infection. Healthy volunteers were included as controls. The study protocol was approved by the local ethics committee on 5 October 2016 (HCB/2015/0585) and participating patients provided their written informed consent. 2.2. Epidemiological Characteristics The following variables were collected at the time of patient inclusion; age, gender, systolic blood pressure, diastolic blood pressure, heart rate, respiratory rate, body mass index (BMI), and cardiovascular risk factors (dyslipidemia, hypertension, and chronic kidney disease). A blood test was performed to evaluate renal function (creatinine, eGFR), glucose, hemoglobin, Na, K, Ca, and P), inflammation (C-reactive protein), liver function (total bilirubin and direct bilirubin, AST, ALT, GGT, ALP), and lipid profile (total cholesterol, HDL, LDL, and triglycerides). 2.3. Sample Collection Patient sera samples were collected at the Hospital Clinic de Barcelona. Blood was taken with a serum tube from an antecubital vein and was aliquoted within 1 h, then stored at −80 °C until metabolomic and proteomic analysis. 2.4. Metabolomic Analysis Four different methods were employed for the extraction of hydrophobic lipids (including methanol extraction and the Folch method), the extraction of amino acids, and the extraction of polar metabolites in central carbon metabolism. Detailed methodology is provided as [74]Supplementary Materials. After metabolomic analyses, the lipidomic method, based on methanol extraction, provided semi-quantitative results for 204 lipids, the lipidomic method based on Folch extraction provided 119 lipids, the amino acid method provided semi-quantitative results for 49 amino acids and derivatives, and polar metabolites in the central carbon metabolism method provided 28 additional compounds. Thus, 400 unique metabolites were successfully quantified in the serum samples being analyzed and these can be found in [75]Table S1. Some compounds can be determined in more than one analysis, after which the results with the highest level of confidence can be used. 2.5. Proteomic Analysis Before the proteomic analysis, the seven most abundant serum proteins (albumin, IgG, antitrypsin, IgA, transferrin, haptoglobin, and fibrinogen) were depleted using a Human-7 Multiple Affinity Removal Spin (MARS) cartridge (Agilent Technologies, Santa Clara, CA, USA) to increase the number of identified proteins. Afterward, the samples were processed for tandem mass tag (TMT) before acquisition by nanoscale liquid chromatography coupled to mass spectrometry (nano LC-MS/MS) analysis, with an LTQ-Orbitran Velos Pro (Thermo Fisher Scientific, Waltham, MA, USA). Protein identification/quantification was performed using Proteome Discoverer software, version 1.4.0.288 (Thermo Fisher Scientific), by multidimensional protein identification technology. After proteomic analysis, a total of 273 proteins were identified in the samples analyzed ([76]Table S2). 2.6. Statistical Analysis 2.6.1. Data Pre-Processing In the initial metabolomic and proteomic analyses, the readers were blinded to the patient’s status. For the statistical analyses, only those metabolites and proteins that were present in ≥70% of the samples in at least one group were considered. In addition, log base 2 transformation was applied to the protein data. Finally, the data were mean-centered and Pareto-scaled. 2.6.2. Multivariate Statistical Analysis Initially, a multivariate statistical approach was performed using Metaboanalyst 4.0 ([77]http://www.metaboanalyst.ca/, accessed on 27 October 2024). The modeling process included the use of unsupervised methods, such as principal component analysis (PCA) and hierarchical clustering (HCA), and supervised methods, which included partial least-squares discriminant analysis (PLS-DA) and orthogonal projection to latent structures discriminant analysis (OPLS-DA). 2.6.3. Univariate Statistical Analysis For each metabolite or protein, a univariate test was performed. For the univariate tests, the data were not Pareto-scaled. Initially, the Control and aCKD samples were compared. A Kolmogorov–Smirnov test was carried out for each protein to check for distribution normality. Afterward, either a t-test or a Wilcoxon test was performed, depending on each protein’s distribution. In the case of a t-test, a test for equality of variance was performed prior to the analysis. The Benjamini–Hochberg method was used to adjust the p-values for multiple testing with the consideration of a 5% false discovery rate (FDR). The reported results included the means and standard deviations (SD) for each group, the fold change (FC), and the p and q (with p corrected for FDR) values. In addition, a ROC analysis was performed for each protein, and the area under the curve (AUC) and p-values are reported below. 2.6.4. Pathway Analysis The metabolites that were differentially presented in serum samples from aCKD patients were used to identify the differential pathways in an over-representation analysis (ORA), performed in Metaboanalyst using metabolite set enrichment analysis (MSEA). Significantly different proteins identified with the univariate analysis were used for the pathway analyses, employing a method based on protein–protein interaction (PPI) networks and enrichment analysis from the Search Tool for the Retrieval of Interacting Genes/proteins (STRING) database ([78]https://string-db.org/, accessed on 27 October 2024). In addition, an enrichment analysis was performed for each displayed network in STRING, testing a number of functional annotation spaces, including gene ontology (with three categories: biological process (GO-BP), cellular component (GO-CC), and molecular function (GO-MF)), the Kyoto Encyclopedia of Genes and Genomes (KEGG), Reactome, PFAM protein domains, and InterPro protein domains and features. The results of the enrichment are sorted according to their enrichment p-values, which were corrected for multiple testing using the Benjamini–Hochberg method. The integrated pathway analysis was performed with the module Join Pathway Analysis from Metaboanalyst, combining metabolomics and gene expression (proteomics) studies conducted under the same experimental conditions. ORA, based on hypergenometrics analysis, was also chosen. The topology analysis evaluated whether a given protein or metabolite plays an important role in a biological response, based on its position within a pathway. Degree centrality was also used to measure the number of links that connect to a node (representing either a protein or metabolite) within a pathway. 3. Results 3.1. Epidemiological Characteristics A total of 19 patients with advanced chronic kidney disease (aCKD) and 27 healthy volunteers were included in the study; the epidemiological characteristics are listed in [79]Table 1. The aCKD patients were older than the volunteers in the control group (58.5 vs. 48.8 years; p = 0.0245). There were no differences between groups in terms of gender distribution and body mass index (BMI). The aCKD patients had a higher prevalence of dyslipidemia (57.9%) and hypertension (68.4%) and higher systolic blood pressure than the control group (132.3 vs. 109.1 mmHg) (p < 0.05). Triglycerides were increased in the aCKD patients, without differences being noted in the cholesterol levels. Table 1. Epidemiological characteristics. General Information aCKD Control p-Value   Age 58.5 ± 14.3 48.8 ± 10.7 0.0245   Gender (Female; %) 8; 42.1 15; 55.6 0.5499   Systolic blood pressure 132.3 ± 12.2 109.1 ± 33.8 0.0013   Diastolic blood pressure 75.5 ± 10.0 67.1 ± 20.4 0.2509   Heart rate 75.5 ± 12.5 61.9 ± 20.0 0.0107   Respiratory rate 15.9 ± 1.3 14.9 ± 0.8 0.1812   Body Mass Index (BMI) 28.6 ± 6.2 25.2 ± 2.6 0.0814 Cardiovascular risk (n; %)         Dyslipidemia 11; 57.9 0; 0.0 <0.0001   Hypertension 13; 68.4 0; 0.0 <0.0001   CKD 19; 100.0 0; 0.0 <0.0001    Stage G4 10; 52.6 0; 0.0      Stage G5 9; 47.4 0; 0.0   Blood parameters         Creatinine mg/dL 3.96 ± 1.13 0.73 ± 0.14 <0.001   eGFR mL/min/1.73 m^2 14.9 ± 3.33 88.7 ± 3.77 <0.001   Na mEq/L 141.5 ± 4.2 140.9 ± 1.7 0.0582   K mEq/L 4.59 ± 0.46 4.35 ± 0.22 0.1125   Ca mg/dL 9.23 ± 0.53 9.53 ± 0.50 0.3292   P mg/dL 4.50 ± 0.55 3.45 ± 0.29 0.0022   Glucose 98.4 ± 9.58 90.2 ± 9.86 0.0475   Hb g/L 121.2 ± 12.8 144.5 ± 9.7 <0.001   C-reactive protein 0.66 ± 0.99 0.098 ± 0.19 0.0423   Total bilirubin mg/dL 0.40 ± 0.11 0.82 ± 0.31 0.001   Direct bilirubin mg/dL 0.15 ± 0.05 0.33 ± 0.13 0.0039   AST U/L 17.6 ± 2.72 21.1 ± 4.32 0.0325   ALT U/L 17.8 ± 9.62 20.9 ± 9.56 0.5512   GGT U/L 22.0 ± 1.73 18.9 ± 10.1 0.2854   ALP U/L 120.2 ± 23.4 65.8 ± 16.6 0.0035   Total Cholesterol 177.8 ± 28.2 190.0 ± 31.4 0.1884   HDL 44.2 ± 9.63 53.5 ± 16.8 0.0566   LDL 109.5 ± 17.7 122.3 ± 29.2 0.1136   Triglycerides 136.6 ± 38.6 108.9 ± 48.4 0.0388 [80]Open in a new tab aCKD, Advanced chronic kidney disease; eGFR, estimated glomerular filtrate rate; AST, aspartate aminotransferase; ALT, alanine aminotransferase; GGT, gamma glutamyl transferase; ALP, alkaline phosphatase. Fisher’s exact test and the Mann–Whitney test were performed. p < 0.05 indicates significantly different results when compared to the control group. Ten patients (52.6%) were classified as CKD stage G4, whereas nine patients (47.4%) were classified as stage G5. Regarding the etiology of primary renal disease, 6 (31.3%) patients had chronic hypertensive nephropathy, and 5 (26.3%) patients had IgA nephropathy, whereas the remaining 8 (42.1%) patients had CKD of unknown etiology. The analysis of blood parameters revealed that sodium, phosphorous, glucose, and C-reactive protein levels were higher in aCKD patients, while the hemoglobin concentration was lower in aCKD patients ([81]Table 1). Liver function was analyzed using AST, ALT, GGT, and ALP; all parameters remained in a non-pathological range, although the ALP value was higher in aCKD patients compared to the control group ([82]Table 1). 3.2. Metabolomic Analysis Of the total 383 metabolites considered, 135 of them were found to be differently presented in the serum from aCKD patients; of these, 54 metabolites had increased and 81 metabolites had decreased compared to the control serum samples ([83]Table S3A and S3B, respectively). The analysis of lipids showed remarkably different results between aCKD patients and healthy volunteers. The aCKD patients had increased levels for 5 out of the 7 analyzed diglycerides (DG) and 6 out of the 25 triglycerides (TG). Furthermore, 29/35 lysophosphocholines (LPC), 1/35 lysophosphoethanolamines (LPE), 10/12 lysophoshoinositoles (LPI), 1/15 cholesteryl esters (ChoE), 6/23 sphingomyelin (SM), and 8/45 phosphocholine (PC) were observed to show reduced concentrations in aCKD patients. 3.3. Supervised Analysis From the PLS-DA analysis, a clear separation between the control and aCKD patients could be observed. The best PLS-DA model included five components and showed a strong predictive ability (Accuracy = 1.0; R^2 = 1.0 and Q^2 = 0.81) ([84]Figure S1A). The most important features contributing to class separation are shown in [85]Table 2. The results from the OPLS-DA analysis showed similar results to those from the PLS-DA analysis ([86]Table S4). Therefore, a significant model with strong predictive ability was obtained (Q^2Y = 0.83, p < 0.001, 1 predictive + 3 orthogonal components), and the metabolites contributing to class separation can be identified in the corresponding S-plot ([87]Figure S1B). Table 2. The top metabolite biomarkers that are differentially expressed.   Name FC ROC Analysis AUC p-Value 1 Homocitrulline 5.88 1.000 4.12 × 10^−27 2 1-Methylhistidine 7.54 1.000 6.12 × 10^−24 3 Cystathione 11.59 0.994 5.57 × 10^−15 4 Acetyl-carnitine 5.86 0.996 2.33 × 10^−14 5 Sucrose 18.56 0.979 7.73 × 10^−14 6 BAIBA 6.40 0.998 1.48 × 10^−13 7 Kynurenine 2.55 0.979 1.90 × 10^−13 8 AIBA 2.32 0.984 2.11 × 10^−13 9 Cortisone −1.96 0.979 5.00 × 10^−13 10 2-methyl-butyryl-carnitine/Isovaleryl-carnitine/Valeryl-carnitine 2.61 0.988 8.59 × 10^−12 11 Homocystine 6.10 0.969 6.79 × 10^−10 12 3-Methylhistidine 5.85 0.899 1.15 × 10^−8 13 Methylglutaryl-carnitine/Adipoyl-carnitine 6.81 0.922 1.64 × 10^−8 14 EpOME-iso3 −3.57 0.922 2.46 × 10^−8 15 Octenoyl-carnitine 3.06 0.903 1.23 × 10^−7 16 Androstenedione 3.10 0.832 4.53 × 10^−5 17 LPI 14:0 −4.07 0.793 2.54 × 10^−4 18 LPC 18:3-sn2 −3.42 0.782 2.66 × 10^−4 19 LPC 18:3-sn1 −3.04 0.774 3.76 × 10^−4 20 12-HETE −3.48 0.743 2.83 × 10^−3 [88]Open in a new tab 3.4. Pathway Analysis The ORA resulted in only one significantly enriched pathway, that for linoleic acid metabolism ([89]Table S5). The aCKD patients had lower levels in 7 out of the 15 metabolites from this pathway (FDR = 8.62 × 10^−7) ([90]Figure 1 and [91]Table 3). Figure 1. [92]Figure 1 [93]Open in a new tab Linoleic acid metabolism, taken from the Kyoto Encyclopedia of Genes and Genomes (KEGG-hsa00591). The pathway analysis using metabolomic data showed seven downregulated metabolites (green dots). These metabolites were lecithin (phosphatidylcholine), linoleic acid, 12(13)-DiHOME, 12(13)EpOME, 9(10)-DiHOME, 9(10)EpOME, and 9-oxoODE. Furthermore, the integrative analysis showed two additional altered metabolites: 9-HODE/13-HODE and dihomo-γ-linoleic. Table 3. Altered metabolites from linoleic acid metabolism in aCKD patients. Lecithin (phosphatidylcholine), Linoleic acid, 9(10)EpOME, 12(13)EpOME, 9(10)-DiHOME, 12(13)-DiHOME, 9-oxoODE, 9-HODE/13-HODE, and dihomo-γ-linoleic. Name Control aCKD diff LOG2 FC q-Value (FDR) EpOME-iso3 −6.736 (0.865) −8.572 (0.960) 0.280 −3.570 0.000  EpOME-iso2 −7.156 (0.445) −7.644 (0.531) 0.713 −1.402 0.010 9(10)-DiHOME −7.045 (0.810) −8.020 (0.704) 0.509 −1.966 0.001 Linoleic acid-iso1 6.462 (0.627) 5.767 (0.735) 0.618 −1.619 0.008 9-HODE/13-HODE −5.588 (1.068) −6.238 (0.497) 0.637 −1.569 0.012 9-OxoODE −6.650 (0.843) −7.291 (0.609) 0.641 −1.559 0.003 12(13)-DiHOME −6.814 (0.550) −7.457 (0.653) 0.640 −1.562 0.006 N-Linoleoyl ethanolamide 3.658 (0.469) 3.250 (0.474) 0.754 −1.327 0.021 dihomo-gamma-linolenic acid −0.076 (0.415) −0.409 (0.410) 0.794 −1.260 0.032 Lecithin (PhosphatidylCholine)            PC 34:2 7.836 (0.271) 7.609 (0.227) 0.854 −1.170 0.018  PC 36:2 7.021 (0.316) 6.743 (0.323) 0.825 −1.213 0.020  PC 31:0 −2.186 (0.540) −2.648 (0.525) 0.726 −1.377 0.021  PC 38:3 e −2.770 (0.601) −3.265 (0.463) 0.710 −1.409 0.017  PC 36:2 e −1.158 (0.437) −1.656 (0.451) 0.708 −1.412 0.004  PC 30:0 −0.397 (0.769) −1.134 (0.637) 0.600 −1.667 0.016  PC 32:2 0.144 (0.799) −0.714 (0.686) 0.552 −1.813 0.014 [94]Open in a new tab 3.5. Proteomic Analysis A total of 164 proteins were considered, and 75 of them were found to be differentially expressed in the aCKD group compared to the control group. In addition, 38 proteins were upregulated compared to the control group, whereas 37 proteins were downregulated ([95]Tables S6A and S6B, respectively). Our data confirmed the increased abundance of eight well-characterized uremic toxins, including beta-2-microglobulin, prostaglandin-H2 D-isomerase, and cystatin-C ([96]Table S7). 3.6. Unsupervised Analysis The PCA analysis and unsupervised clustering heatmaps showed a clear separation between the control and aCKD groups, without the presence of any outliers ([97]Figure S2). 3.7. Supervised Analysis From the PLS-DA analysis, a clear separation between the control and aCKD samples could be observed in the first component. The best PLS-DA model based on Q^2 included two components and showed strong predictive ability (R^2 = 0.97 and Q^2 = 0.85). The most important features contributing to class separation are shown in [98]Table 4. The results from the OPLS-DA analysis showed similar results to those from the PLS-DA analysis ([99]Table S8). Therefore, a significant model with strong predictive ability was obtained (Q^2Y = 0.90, p < 0.001) with 1 predictive + 1 orthogonal component, and the proteins contributing to class separation can be identified in the corresponding S-plot ([100]Figure S3). Table 4. The top protein biomarkers that are differentially expressed. Swiss-Prot ID Name FC ROC Analysis AUC p-Value [101]P61769 Beta-2-microglobulin (B2M) 3.104 1 8.41 × 10^−21 [102]P01034 Cystatin-C (CST3) 2.433 1 1.71 × 10^−20 [103]P22692 Insulin-like growth factor-binding protein 4 (IGFBP4) 1.985 1 1.31 × 10^−18 [104]P22352 Glutathione peroxidase 3 (GPX3) −1.889 1 4.33 × 10^−17 [105]P06727 Apolipoprotein A-IV (ApoA4) 2.390 1 1.95 × 10^−17 [106]P36955 Pigment epithelium-derived factor 1.690 1 9.57 × 10^−16 [107]P61626 Lysozyme C (LYZ) 3.061 0.998 9.41 × 10^−16 [108]P00746 Complement factor D (CFD) 2.145 0.996 1.58 × 10^−16 [109]P41222 Prostaglandin-H2 D-isomerase (β-trace protein) 2.882 0.981 3.26 × 10^−14 [110]P16070 CD44 antigen 1.519 0.981 8.24 × 10^−11 [111]Q6EMK4 Vasorin 1.409 0.959 4.27 × 10^−11 [112]P02753 Retinol-binding protein 4 (RBP4) 1.876 0.975 1.03 × 10^−11 [113]P02760 Protein AMBP (AMBP) 1.331 0.963 1.58 × 10^−10 [114]P02749 Beta-2-glycoprotein 1 (ApoH) 1.449 0.986 1.27 × 10^−10 [115]Q6UXB8 Peptidase inhibitor 16 (PI16) 1.491 0.945 5.99 × 10^−9 [116]P43251 Biotinidase −1.302 0.926 4.61 × 10^−8 [117]P08697 Alpha-2-antiplasmin −1.244 0.924 4.09 × 10^−8 [118]P02652 Apolipoprotein A-II (ApoA2) −1.320 0.932 3.56 × 10^−8 [119]P05160 Coagulation factor XIII B (F13B) 1.235 0.930 1.75 × 10^−8 [120]P02751 Fibronectin (FN1) −1.504 0.889 6.91 × 10^−7 [121]Open in a new tab 3.8. Pathway Analysis The inclusion of the 75 differentially expressed proteins in STRING software reflects 315 edges between 75 nodes, with a node degree of 8.4, a local clustering coefficient of 0.566, and a PPI enrichment p-value below 1.0 × 10^−16 ([122]Figure 2). Figure 2. [123]Figure 2 [124]Open in a new tab Protein–protein interaction network for the 75 statistically different proteins obtained using the STRING database. Here, the nodes represent proteins and edge interactions between proteins. The thickness of the edge indicates the degree of confidence prediction of the interaction. Only interactions with a high confidence score (>0.7) were considered. Proteins included in the complement and coagulation cascades appear in red, proteins from the innate immune system are shown in green, and proteins involved in cholesterol metabolism are in blue. The enrichment analysis showed that the complement and coagulation cascades (KEGG, 21 of 80 proteins, FDR = 3.15 × 10^−24) ([125]Figure 3 and [126]Table 5), innate immune system (Reactome 26 of 1012 proteins, FDR = 1.023 × 10^−13) ([127]Table 6 and [128]Figure S4), and cholesterol metabolism (KEGG, 7 of 48 proteins, FDR = 4.67 × 10^−8) ([129]Table 7 and [130]Figure S5) were highly altered in aCKD patients. All the pathways that were altered are presented in [131]Table S9. Figure 3. [132]Figure 3 [133]Open in a new tab An illustration of the complement and coagulation cascades from KEGG-hsa04610. A total of 21 (out of 78) proteins were significantly enriched. Proteins that showed significantly lower and higher concentrations in aCKD are shown in green and red, respectively. The downregulated proteins were C3, C8G, CPB2, F12, F2, PLG, SERPINF2, VTN, and MASP1, whereas the upregulated proteins were C4B, C4BPA, C9, CFD, F11, F13B, F5, F9, LNG1, PROS1, SERPING1, and VWF. Table 5. Proteins from the complement and coagulation cascades that are differentially expressed in aCKD patients compared to the control group. Name Control aCKD diff LOG2 FC q-Value (FDR) Complement cascade           Complement factor D −0.288 (0.230) 0.813 (0.351) 1.101 2.15 0.000 Complement factor H-related protein 2 −0.218 (0.763) 0.471 (0.611) 0.689 1.61 0.000 Complement C4-B 0.024 (0.368) 0.477 (0.537) 0.453 1.37 0.005 C4b-binding protein alpha chain 0.021 (0.512) 0.474 (0.777) 0.453 1.37 0.028 Complement component C9 −0.132 (0.267) 0.308 (0.329) 0.44 1.36 0.000 Coagulation factor XIII B chain 0.089 (0.141) 0.394 (0.158) 0.305 1.24 0.000 von Willebrand factor −0.682 (0.320) −0.433 (0.392) 0.249 1.19 0.049 Coagulation factor XI 0.152 (0.333) 0.369 (0.234) 0.217 1.16 0.041 Vitamin K-dependent protein S 0.136 (0.275) 0.315 (0.200) 0.179 1.13 0.033 Coagulation factor V 0.077 (0.229) 0.236 (0.166) 0.159 1.12 0.031 Coagulation factor IX 0.039 (0.161) 0.185 (0.174) 0.146 1.11 0.013 Plasma protease C1 inhibitor (SERPING1) 0.184 (0.138) 0.329 (0.156) 0.145 1.11 0.006 Kininogen-1 0.025 (0.159) 0.135 (0.124) 0.11 1.08 0.036 Coagulation cascade           Carboxypeptidase B2 0.133 (0.174) −0.021 (0.224) −0.154 −1.11 0.029 Complement component C8 gamma chain 0.110 (0.204) −0.043 (0.204) −0.153 −1.11 0.038 Plasminogen 0.140 (0.155) −0.029 (0.129) −0.169 −1.12 0.001 Prothrombin 0.188 (0.131) 0.022 (0.140) −0.166 −1.12 0.001 Complement C3 0.090 (0.150) −0.072 (0.219) −0.162 −1.12 0.012 Coagulation factor XII 0.082 (0.339) −0.185 (0.334) −0.267 −1.20 0.011 Vitronectin −0.069 (0.258) −0.364 (0.157) −0.295 −1.23 0.000 Alpha-2-antiplasmin (SERPINF2) 0.205 (0.164) −0.110 (0.150) −0.315 −1.24 0.000 Fibronectin 0.008 (0.365) −0.581 (0.298) −0.589 −1.50 0.000 [134]Open in a new tab Table 6. Proteins from the innate immune system that are differentially expressed in aCKD patients compared to the control group. Several pathways from the innate immune system contain these proteins: the complement cascade, neutrophil degranulation, antimicrobial peptide, and Toll-like receptor cascade. Name Control aCKD diff LOG2 FC q-Value (FDR) Complement cascade           Carboxypeptidase N catalytic chain 0.104 (0.159) −0.142 (0.193) −0.246 −1.19 0.000 Carboxypeptidase N subunit 2 0.138 (0.178) −0.108 (0.211) −0.246 −1.19 0.001 Complement C3 0.090 (0.150) −0.072 (0.219) −0.162 −1.12 0.012 Complement C4-B 0.024 (0.368) 0.477 (0.537) 0.453 1.37 0.005 C4b-binding protein alpha chain 0.021 (0.512) 0.474 (0.777) 0.453 1.37 0.028 Complement component C8 gamma chain 0.110 (0.204) −0.043 (0.204) −0.153 −1.11 0.038 Complement component C9 −0.132 (0.267) 0.308 (0.329) 0.44 1.36 0.000 Carboxypeptidase B2 0.133 (0.174) −0.021 (0.224) −0.154 −1.11 0.029 Plasma protease C1 inhibitor (SERPING1) 0.184 (0.138) 0.329 (0.156) 0.145 1.11 0.006 Vitronectin −0.069 (0.258) −0.364 (0.157) −0.295 −1.23 0.000 Complement factor D −0.288 (0.230) 0.813 (0.351) 1.101 2.15 0.000 Prothrombin 0.188 (0.131) 0.022 (0.140) −0.166 −1.12 0.001 Neutrophil degranulation           Alpha-1-acid glycoprotein 1 −0.101 (0.372) 0.520 (0.460) 0.621 1.54 0.000 Alpha-1-acid glycoprotein 2 0.080 (0.334) 0.449 (0.248) 0.369 1.29 0.001 Leucine-rich alpha-2-glycoprotein −0.276 (0.369) 0.333 (0.412) 0.609 1.53 0.000 Cystatin-C −0.287 (0.258) 0.996 (0.261) 1.283 2.43 0.000 Complement factor D −0.288 (0.230) 0.813 (0.351) 1.101 2.15 0.000 Lysozyme C −0.443 (0.407) 1.171 (0.485) 1.614 3.06 0.000 CD44 antigen −0.164 (0.232) 0.439 (0.246) 0.603 1.52 0.000 Gelsolin 0.119 (0.149) 0.378 (0.250) 0.259 1.20 0.001 Lipopolysaccharide-binding protein −0.180 (0.435) 0.201 (0.340) 0.381 1.30 0.008 Alpha-1B-glycoprotein −0.074 (0.127) −0.225 (0.135) −0.151 −1.11 0.001 Beta-2-microglobulin −0.490 (0.299) 1.144 (0.357) 1.634 3.10 0.000 Antimicrobial peptide           N-acetylmuramoyl-L-alanine amidase 0.135 (0.244) −0.037 (0.217) −0.172 −1.13 0.017 Lysozyme C −0.443 (0.407) 1.171 (0.485) 1.614 3.06 0.000 Toll-like receptor cascade           Monocyte differentiation antigen CD14 −0.048 (0.244) 0.239 (0.282) 0.287 1.22 0.002 Lipopolysaccharide-binding protein −0.180 (0.435) 0.201 (0.340) 0.381 1.30 0.008 [135]Open in a new tab Table 7. Proteins from cholesterol metabolism that are differentially expressed in aCKD patients compared to the control group. Furthermore, the integrative analysis showed 7 additional altered metabolites: glycocholic acid iso-4, TG52:1, TG52:2, TG52:3, TG54:2, TG54:3, and TG54:4. Name Control aCKD diff LOG2 FC q-Value (FDR) Proteins            Apolipoprotein A-IV (APOA4) −0.216 (0.249) 1.041 (0.375) 1.257 2.39 0.000  Beta-2-glycoprotein 1 (APOH) 0.058 (0.248) 0.593 (0.151) 0.535 1.45 0.000  Apolipoprotein C-III (APOC3) −0.231 (0.380) 0.161 (0.407) 0.392 1.31 0.005  Apolipoprotein M 0.259 (0.280) −0.246 (0.284) −0.505 −1.42 0.000  Apolipoprotein L1 −0.158 (0.386) −0.756 (0.400) −0.598 −1.51 0.000  Apolipoprotein A-II (APOA2) 0.258 (0.200) −0.142 (0.201) −0.4 −1.32 0.000  Apolipoprotein E (APOE) −0.054 (0.348) −0.382 (0.425) −0.328 −1.26 0.016  Apolipoprotein A-I (APOA1) 0.098 (0.168) −0.150 (0.255) −0.248 −1.19 0.001  Apolipoprotein B-100 (APOB) 0.033 (0.295) −0.200 (0.273) −0.233 −1.18 0.024 Metabolites  TG54:3 7.419 (0.704) 8.258 (0.994) 1.789 1.789 0.010  TG52:1 4.279 (0.886) 5.090 (1.253) 1.754 1.754 0.039  TG54:2 4.805 (0.785) 5.484 (1.030) 1.601 1.601 0.042  TG54:4 7.961 (0.691) 8.621 (0.623) 1.580 1.580 0.010  TG52:2 9.512 (0.648) 10.124 (0.831) 1.528 1.528 0.024  TG52:3 9.487 (0.664) 10.012 (0.546) 1.439 1.439 0.023  Glycocholic acid-iso4 −0.708 (1.236) −2.079 (1.589) 0.387 −2.586 0.011 [136]Open in a new tab In addition, the enrichment analysis classified all proteins according to three categories from the Gene Ontology resource: biological process (GO-BP), cellular component (GO-CC), and molecular function (GO-MF). The altered proteins found in aCKD patients are involved in the regulation of proteolysis (32 proteins, FDR = 4.05 × 10^−22), in the inflammatory response (21 proteins, FDR = 3.20 × 10^−17), and in the response to stress (46 proteins, FDR = 8.82 × 10^−16), among other biological processes ([137]Table S10). In terms of the cellular component, 72 out of 75 of the altered proteins in aCKD (96%) are located in the extracellular region (FDR = 1.35 × 10^−57). Most of the proteins are found in the lumen of vesicles or the secretory granules ([138]Table S11). Regarding molecular function, most of the altered proteins have enzyme regulator/inhibitor activity. These proteins mainly participate in peptidase, endopeptidase, or carboxypeptidase reactions ([139]Table S12). 3.9. Integrated Pathway Analysis The inclusion of the differentially expressed metabolites and proteins in the integrative analysis increased the statistical impact of several pathways that include proteins and metabolites ([140]Table 8 and [141]Table S13). Although the more significantly altered pathways in aCKD patients remained the complement and coagulation cascades, linoleic acid metabolism, and cholesterol metabolism, other pathways that also reached the level of statistical differences included fat digestion and absorption, protein digestion and absorption, the biosynthesis of unsaturated fatty acids, and beta-alanine metabolism. Table 8. Integrative pathway analysis.   Total Hits Proteins Metabolites Expected Raw p −log(p) Holm Adjust FDR Impact Complement and coagulation cascades 80 21 21 0 0.656 9.5297 × 10^−27 59.915 3.1543 × 10^−24 3.1543 × 10^−24 0.731 Linoleic acid metabolism 57 9 0 9 0.467 7.3207 × 10^−10 21.035 2.4158 × 10^−7 1.2116 × 10^−7 0.683 Cholesterol metabolism 60 9 7 2 0.492 1.179 × 10^−9 20.559 3.879 × 10^−7 1.3009 × 10^−7 0.075 African trypanosomiasis 45 6 3 3 0.369 1.621 × 10^−6 13.332 0.00053169 0.00013414 0.028 Fat digestion and absorption 54 6 3 3 0.442 4.8446 × 10^−6 12.238 0.0015842 0.00032071 0 Protein digestion and absorption 142 8 1 7 1.160 2.0511 × 10^−5 10.795 0.0066865 0.0011315 0 Vitamin digestion and absorption 63 5 4 1 0.516 0.00016024 8.7388 0.052079 0.0075772 0 Amoebiasis 115 6 4 2 0.942 0.00035372 7.947 0.11461 0.014635 0.017 Biosynthesis of unsaturated fatty acids 79 5 0 5 0.647 0.00046354 7.6766 0.14972 0.017048 0.030 PPAR signaling pathway 81 5 4 1 0.664 0.00052024 7.5612 0.16752 0.01722 0.153 Pertussis 86 5 5 0 0.705 0.00068479 7.2864 0.21982 0.020606 0.176 Staphylococcus aureus infection 98 5 5 0 0.804 0.0012364 6.6956 0.39564 0.034104 0.442 beta-Alanine metabolism 63 4 1 3 0.516 0.0017406 6.3535 0.55527 0.041442 0.759 Central carbon metabolism in cancer 106 5 0 5 0.869 0.0017528 6.3465 0.5574 0.041442 0 [142]Open in a new tab 3.10. Other Metabolite and Protein Biomarkers We analyzed several metabolites and proteins that are considered to be TOP biomarkers; however, they were not involved in the enriched pathways described above ([143]Figure 4). The enzyme involved in the detoxification of hydrogen peroxide, glutathione peroxidase 3 (GPX3), was reduced in aCKD patients. Acetyl-carnitine and free carnitine were increased in the aCKD patients. Several amino acids and derivates (arginine, beta-alanine, cystine, kynurenine, proline, 1-methylhistidine, 3-methylhistidine, homocitrulline, homocysteine, and cystathionine) were increased in aCKD patients, whereas three (serine, tryptophan, and tyrosine) were reduced. Figure 4. [144]Figure 4 [145]Open in a new tab Quantification of the altered metabolites and proteins in aCKD patients. The Mann–Whitney test was performed. * Significantly different when compared to the control group (** p < 0.01; *** p < 0.001). 4. Discussion The metabolomic and proteomic analysis of serum samples showed that numerous pathways are disrupted in aCKD patients, all of which exert direct or indirect effects on the cardiovascular system. We propose an interaction model of these altered pathways to elucidate the impact of aCKD on the development of CVD ([146]Figure 5). Additionally, based on the literature, we identify potential pathways that could be modulated by a Mediterranean diet to reduce the risk of CVD in aCKD patients. Figure 5. [147]Figure 5 [148]Open in a new tab Altered pathways in aCKD patients, related to the risk of CVD development. The metabolites, proteins, and pathways identified in this study are accompanied by the symbols ⇅, indicating increases or reductions in aCKD patients. Mediterranean diet interventions that could reduce CVD risk in patients with CKD are shown in blue boxes. A Mediterranean diet pattern is built around vegetables, fruits, herbs, nuts, beans, whole grains, and seafood, but also includes moderate amounts of dairy, meat, and eggs. This plant-based dietary pattern is rich in anti-inflammatory nutrients, fiber, and phytochemicals. Previous studies have shown that a plant-based diet low in animal-based and ultra-processed foods may be helpful to slow the progression of CKD and delay the need for dialysis, via the reduction of cardiometabolic risk factors such as hypertension, CVD, diabetes, and obesity [[149]21,[150]22,[151]23,[152]24,[153]25]. Several randomized controlled clinical trials have confirmed that replacing a dietary saturated fat intake with vegetable polyunsaturated fats, as in the Mediterranean diet, reduces cardiovascular disease incidence by approximately 30%, which is similar to the decrease induced by statins [[154]26,[155]27]. In this study, we have demonstrated that in a high CVD-risk population, such as in those with chronic kidney disease (CKD), these changes will bring alterations in lipid metabolism, particularly in circulating apolipoproteins [[156]28]; in particular, a statistically significant decrease in ApoA-I, ApoA-II, ApoB-100, ApoE, ApoM, and ApoL1 and a statistically significant increase in ApoA-IV, ApoC-III, and ApoH. An imbalance in blood lipid metabolism is thought to contribute to an increased risk of CVD in aCKD patients, which primarily manifests as increases in plasma triglycerides and reductions in high-density lipoprotein (HDL) cholesterol, with little change in low-density lipoprotein (LDL) cholesterol [[157]29]. Nonetheless, atherosclerotic cardiovascular disease (ASCVD) remains in the context of chronic kidney disease (CKD), despite treatment with statins [[158]30]. A meta-analysis of 13 randomized controlled trials of statins in CKD found that the response to statins diminishes in the later stages of CKD [[159]31]. In our study, we observed a lipid imbalance despite more than 50% of patients being treated with statins, which highlights that there is obvious room for a greater reduction in ASCVD risk in CKD beyond the lowering of LDL-C with statins. The traditional Mediterranean diet contains less than half the amount of choline and L-carnitine molecules when compared to a typical Western diet [[160]32]. The production of tri-methylamine N-oxide (TMAO), a gut microbial metabolite, from dietary choline and L-carnitine has been demonstrated to enhance the likelihood of developing cardiovascular disease in both murine models and humans, operating independently of traditional cardiometabolic risk factors. Carnitine, on the other hand, has a pivotal role in fatty acid β-oxidation and energy production by transporting long-chain fatty acids from the cytoplasm to mitochondria [[161]33]. Both these molecules were also altered in our aCKD population, and these results are in accordance with those previously reported by other researchers [[162]34,[163]35]. A number of carnitine metabolites were identified in the 67-metabolite signature that was proposed for monitoring adherence to the Mediterranean diet [[164]20]. Altogether, these results highlight the potential positive impact that a reduction in the content of choline and L-carnitine in tailored aCKD diets may have on lipid metabolism in this population. The enrichment analysis performed with metabolomic data on serum samples revealed that linoleic acid metabolism is altered in aCKD patients. Specifically, nine metabolites were reduced in aCKD patients compared to the control group. The use of statins enhances the conversion of linoleic acid to long-chain polyunsaturated fatty acid derivatives [[165]36], which could explain the reduction of linoleic acid and derivates in aCKD patients. However, we performed an analysis according to the use of statins in aCKD patients ([166]Supplementary Figure S6), grouped according to whether they receive statins or not; they showed reduced levels of each metabolite in linoleic acid metabolism compared to the healthy group. Recently, Szczuko et al. demonstrated that the course of inflammation in CKD occurs through a decrease in polyunsaturated fatty acids (PUFA) and the synthesis of monounsaturated fatty acids (MUFA) [[167]37]. In particular, the index C18:3n6/C22:4n6 (gamma linoleic acid/docosatetraenoate) was defined as a new marker in the progression of the disease [[168]37]. Furthermore, dietary conjugated linoleic acid treatment reduces the inflammation observed in chronic kidney disease animal models, due to the inhibition of COX2, and, furthermore, the reduction of prostanoid levels [[169]38,[170]39]. Huang et al. demonstrated, using Swedish dialysis patients, that the proportion of plasma phospholipid linoleic acid was inversely associated with inflammation (IL-6) and all-cause mortality [[171]40], while the low dietary consumption of linoleic acid has been correlated with an increased risk of diabetic kidney disease [[172]41]. Altogether, these results indicate that aCKD patients could benefit from an increased intake of vegetable oils, the primary source of linoleic acid in the Mediterranean diet. The Mediterranean diet is rich in MUFA from olive oil and low in saturated fats from meat and dairy products. The plasma metabolome of patients on the Mediterranean diet is characterized by an increase in unsaturated lipid metabolites [[173]20]. For that reason, the Mediterranean diet is widely recognized for its benefits in reducing cardiovascular risk factors by improving the lipid profile [[174]8]. Endothelial disease and cardiovascular disease progression are largely immune-mediated [[175]42]. It is recognized that CKD is characterized by a remarkable increase in pro-inflammatory cytokine levels, in particular, tumor necrosis factor α (TNFα) and interleukin 6 (IL-6) [[176]43]. Interestingly, CKD has been associated with an increased risk of infection, due to the dysregulation of the innate immune system [[177]44]. In our study, several components from the innate immune system were altered in aCKD patients, including the complement cascade, neutrophil degranulation, Toll-like receptor cascade, and antimicrobial peptide. The uremic milieu alters the ability of endothelial cells to control the alternative complement pathways, which, in turn, amplifies endothelial injuries [[178]45]. These factors are likely to explain the observed imbalances in complement protein expression that were observed in our aCKD population, with some showing an increase while others presented reduced protein expression. There are compelling data regarding the efficiency of the Mediterranean diet in modulating the gut microbiota. This modulation results in increased microbial diversity and alterations in the proportions of certain bacterial species [[179]46]. Furthermore, a diet comprising a high proportion of fiber-rich foods has also been shown to have a beneficial impact on the integrity of the intestinal barrier [[180]47]. Over the last few years, a group of uremic toxins that are generated in the gut, which potentially connect CKD and the occurrence of CVD, has been described as a risk factor for CKD [[181]48]. The thrombolome is a phenomenon in which CKD-associated dysbiosis and its effect via the generation of gut microbial metabolites induces the prothrombotic phenotype [[182]49]. Our data confirmed the increased abundance of eight well-characterized uremic toxins, including beta-2-microglobulin, prostaglandin-H2 D-isomerase, and cystatin-C [[183]50,[184]51], as well as an upregulation of vWF, in aCKD patients. Other deregulations in the gut microbiota may play a role in vascular and bone disease in CKD [[185]52,[186]53]. These data highlight the relevance of the development of controlled intervention studies on gut microbiota composition and activity, to evaluate the impact of a Mediterranean diet on aCKD patients’ thrombolome. Oxidative stress and inflammation are features of CKD and drivers of CKD progression, as well as the related cardiovascular and other complications [[187]54,[188]55]. Oxidative stress has been proposed to play a major role in the development of endothelial dysfunction through the production of radical oxygen species (ROS), which activate the intracellular signaling pathways [[189]56]. In particular, glutathione peroxidase catalyzes the reduction of hydrogen peroxide and other organic hydroperoxides into water. Therefore, this enzyme protects cell membrane lipids, proteins, and DNA against oxidative stress. In our study, glutathione peroxidase 3 (GPX3) was significantly decreased in aCKD patients, and this change was positively associated with eGFR [[190]57]. Recently, the activation of the glutathione peroxidase pathway in vitro, by using antioxidant enzyme mimetics (ebselen, glutathione peroxidase mimetic; EUK-134 and EUK-118, both superoxide dismutase mimetics) or N-acetylcysteine reduced not only oxidative stress but also the inflammatory process induced by the uremic milieu on the endothelium [[191]58]. The endothelium, under uremic conditions, exhibits a proinflammatory phenotype with an increased expression of adhesion molecules, such as the above-mentioned vWF and tissue factor, and the production of proinflammatory cytokines, which have been reported as key processes in endothelial activation and damage [[192]59,[193]60]. Antioxidant drugs have been shown to have renoprotective effects in animal studies but have not shown significant effects in clinical trials. The evidence indicates that a diet rich in antioxidants, particularly the Mediterranean diet, protects cells and tissues from oxidation and prevents or delays the development of cardiovascular diseases, thereby reducing mortality risk [[194]61]. The myeloperoxidase (MPO) pathway is a major contributor to oxidative stress in aCKD. The MPO pathway and urea deamination could cause carbamylation, a non-enzymatic reaction during which a carbamoyl moiety is added to proteins, peptides, or amino acids. Carbamylation is involved in the pathogenesis of various diseases, such as atherosclerosis, thrombus formation, infections, autoimmune diseases, and kidney diseases, with homocitrulline being a well-recognized biomarker, and the degree of carbamylation was identified to be an important risk factor for cardiovascular events and mortality in patients with aCKD [[195]62,[196]63,[197]64]. We have also identified homocitrulline levels to be increased in aCKD patients. Nutritional therapy, in particular the Mediterranean dietary pattern, induces a decrease in urea levels that has been associated with the reduction of protein carbamylation in CKD [[198]65]. Amino acid supplementation has been postulated as one of the several strategies to reduce carbamylation in aCKD patients. Considering this phenomenon, we have carefully analyzed protein and amino acid metabolism. From the integrative analysis, we identified that beta-alanine metabolism and the protein digestion and absorption pathways were altered in aCKD patients. Four amino acids and derivates (arginine, beta-alanine, cystine, kynurenine, and proline) were increased in aCKD patients, whereas three (serine, tryptophan, and tyrosine) were reduced. The reduced levels of tryptophan and the accumulation of toxic tryptophan metabolites, especially kynurenine, have been described previously in CKD patients [[199]35,[200]66]. Kynurenine may promote atherosclerosis in aCKD by activating oxidative stress and by leukocyte activation in endothelial and vascular smooth muscle cells [[201]67]. Dahabiyeh et al. have suggested that inhibition of the kynurenine pathway could be a promising target to delay progression from CKD to ESRD, with kynurenine being a potential prognostic biomarker to monitor the progression of CKD [[202]68]. Kynurenine may promote atherosclerosis in ESRD by activating oxidative stress and leukocyte activation in endothelial and vascular smooth muscle cells [[203]59]. In addition, altered tryptophan metabolism may precipitate fatigue in patients with CKD, due to a deficit of melatonin [[204]69] and the mitochondrial dysfunction observed in skeletal muscle under uremic conditions [[205]70,[206]71]. Razquin et al. demonstrated that tryptophan-kynurenine pathway metabolites were associated with high risk of heart failure and atrial fibrillation in a 2 case–control study nested within the PREDIMED trial [[207]71,[208]72]. Moreover, an alteration in the impact of the Mediterranean diet, notably when combined with extra virgin olive oil, was evident in the correlation between kynurenine-related metabolites and heart failure. This indicates that the adverse consequences of these metabolites were confined to the control group [[209]63]. Kynurenine is among the metabolites that constitute the distinctive metabolic profile associated with adherence to the Mediterranean diet. In patients following a Mediterranean diet, the levels of this metabolite were diminished [[210]20]. In our study, homocysteine and cystathionine levels were increased in aCKD patients, which is in line with previous studies [[211]73,[212]74]. Hyperhomocysteinemia undoubtedly has a central role in such a prominent cardiovascular burden, promoting atherosclerosis through increased oxidative stress, impaired endothelial function, and the induction of thrombosis. Homocysteine is the key factor in the pattern of nucleic acid methylations in the genome and epigenetic landscape [[213]75]. It has been reported that the machinery that controls methyl transfer reactions is significantly influenced by the metabolic alterations found in the uremic state, since the uremic toxins themselves have been proven to be related to methyl metabolism and sulfur amino acid metabolism [[214]76]. Frailty and sarcopenia are both linked to a reduction in muscle quantity and quality, due to a catabolic state with elevated muscle protein turnover. Plasma 3-methylhistidine (3-MH) has been defined as a biomarker to display elevated muscle protein turnover and as a biomarker for frailty status [[215]77]. 3-MH is formed in the muscle by the post-translational methylation of histidine residues in actin and myosin. During muscle degradation, 3-MH is released, then is not further metabolized, leaving the body through quantitative excretion in the urine. However, patients with CKD accumulate 3-MH in plasma; aCKD patients showed increased levels of 3-MH. Further analysis should be performed to determine if 3-MH could be useful for identifying frailty in CKD patients. It is important to acknowledge the limitations inherent in this research. The cross-sectional design limits our ability to assess longitudinal changes or disease progression in patient outcomes. Furthermore, this is not an interventional clinical trial. The sample size was limited, and the aCKD patient characteristics were highly restrictive, excluding patients with other conditions associated with endothelial dysfunction and CVD. It is essential to remember that most patients with CKD have other medical conditions that may influence the results. 5. Conclusions Our findings reveal significant dysregulation of the serum metabolome and proteome in aCKD, particularly related to inflammation, the innate immune system, oxidative stress, uremic toxins, dyslipidemia, and acidosis; most of these pathways are associated with CVD. Previous studies in other cohorts demonstrate that high adherence to MD led to profound changes in the metabolome that are associated with favorable cardiometabolic health. MD-induced changes could reverse altered CVD pathways in patients with aCKD; however, further prospective interventional clinical trials would be beneficial in confirming the potential of the Mediterranean diet in preventing or ameliorating the progression of CVD in aCKD patients. Acknowledgments