Comparative of Plasma Proteomics & Metabolomics in Coronary Artery Disease: Obstructive,Non-Obstructive, & No Lesions

INTRODUCTION The presence and clinical evolution of coronary atherosclerosis depend on several classic risk factors, such as age, sex, arterial hypertension, dyslipidemia, diabetes, smoking, family history, and others, including inflammatory markers, coronary calcium, genetic scores, ethnicity, and lifestyle.

However, the markers for disease progression in individual patients remain uncertain; what is currently available is the average progression, not the certainty of real, personalized evolution. This is especially true in individuals with non-obstructive lesions (<50%). Nevertheless, such patients frequently suffer coronary events, such as myocardial infarction and acute coronary syndrome. Moreover, they represent a large proportion of individuals who experience cardiovascular events.

Thus, the search for more specific prognostic biomarkers continues. The field of plasma proteomics offers a great opportunity for discovering these new markers. Specifically, chemical compounds related to inflammation, platelet function, or endothelial dysfunction may have distinct expression patterns in the aforementioned patients.

With the early detection of coronary artery disease (CAD), ideal and personalized therapies can be initiated. Metabolites can clarify pathogenicity and potential therapeutic targets. A combination of multiple small-molecule metabolites may provide excellent diagnostic value. Therefore, metabolomics, a rapidly expanding field in systems biology, can indicate metabolic alterations in response to disease progression.

For this purpose, plasma, often considered a pool of metabolites, has been an excellent source for determining the metabolic profile in certain diseases. This has also been demonstrated in Coronary Artery Disease (CAD).

Fan et al., in a Chinese cohort, showed statistically significant differences in the metabolism of glycerophospholipids, linoleic acid, purines, and sphingolipids between normal individuals and those with non-obstructive disease. Furthermore, sphingolipids, glycerophospholipids, valine, leucine, isoleucine, primary bile acid biosynthesis, and arginine and proline metabolism differed between obstructive and non-obstructive coronary disease. These differences were detected in 89 differential metabolites.

In the African-American cohort of the Jackson Heart Study, N-acylamides, leucine, and lipid species were associated with the incidence of coronary artery disease.

OBJECTIVE Therefore, the objective of this study is to evaluate differences in the plasma proteomic profile among three groups: Group I, without atherosclerotic lesions; Group II, with non-obstructive lesions (<50%); and Group III, with obstructive lesions (≥50%), based on findings from coronary computed tomography angiography (CCTA) or invasive coronary angiography (ICA), in order to assess their relationship with typical clinical events of coronary artery disease (CAD) and identify potential biomarkers associated with the prognosis of each group.

Thus, in this pilot project, we hypothesize that plasma proteomic and metabolomic profiles may differ among these three patient groups.

METHODS This is a cross-sectional cohort study with 66 patients from the Heart Institute (InCor) of the Hospital das Clínicas, Faculty of Medicine, University of São Paulo (HC-FMUSP). Patients were selected based on coronary computed tomography angiography (CCTA) or invasive coronary angiography (ICA) performed at InCor HC-FMUSP between 2011 and 2017 and divided into three groups: Group I (n = 22), without atherosclerotic lesions; Group II (n = 22), with non-obstructive lesions (<50%); and Group III (n = 22), with obstructive lesions (≥50%). Patients will be randomly recruited from the Brazilian Cohort Project on Non-Obstructive Coronary Disease: Lesions with Less Than 50% Obstruction - BARD. This project was previously approved by the Ethics Committee.

Patients will be invited to participate in the study after a selection process based on medical records from the atherosclerosis and chronic coronary disease outpatient clinics at InCor. During a single visit, they will receive detailed information about the study. Upon their agreement and signing of the informed consent form (ICF), they will be included in the study and undergo blood collection and demographic data assessment on the same day.Participation in the study involves risks related to blood collection and the confidentiality of their information. Access to the data will be restricted to study personnel, who are professionally committed to maintaining absolute confidentiality. Participants' names will be replaced by a code, and the data collected for the study will not contain any elements that could potentially allow personal identification.

For sample size calculation, two approaches will be used: the proportion of significant bins-predefined segments in the spectrum after processing metabolomic fingerprinting data-(untargeted analysis) or the proportion of significant proteins (targeted analysis). It is expected that these proportions will be statistically significant, which generally does not exceed 50%. Since this is a pilot project, and there is no prior knowledge of the proportion of significant proteins, Probabilistic Principal Component Analysis (PPCA) was chosen, meaning that an untargeted analysis approach was applied.

For this choice, the following numerical covariates were determined: age, blood glucose, HDL-cholesterol, LDL-cholesterol, total cholesterol, and triglycerides; and categorical covariates: coronary lesion (obstructive lesion, non-obstructive lesion, and no lesion), diabetes mellitus (DM), dyslipidemia (DLP), systemic arterial hypertension (SAH), and smoking. The false discovery rate (FDR) was set at 5%, bins at 200, and the proportion of significant differences at 20%. Thus, the sample size calculated using MetSizeR software version 2.0 was 22 individuals per group (total of 66 participants). The study variables will be presented using descriptive statistics according to variable type (qualitative or quantitative). For qualitative variables, frequencies and percentages will be calculated. For quantitative variables, mean, median, standard deviation, minimum, maximum, first quartile, and third quartile will be presented.

The comparison between two dependent measures will be performed using the non-parametric Wilcoxon test. For comparisons between independent groups, the non-parametric Mann-Whitney test will be used.

The significance level adopted for all hypothesis tests will be 5%. Analyses will be performed using SPSS v.25 for Windows ©.

PROTOCOL During the participant's visit, after obtaining informed consent, demographic characteristics will be recorded (age, sex, risk factors, blood pressure, heart rate, weight, and height), and blood samples will be collected for laboratory biochemical analysis (fasting blood glucose, glycated hemoglobin, troponin I, complete blood count, urea, creatinine, sodium, potassium, high-sensitivity C-reactive protein (hs-CRP), aspartate transaminase, alanine transaminase, total and fractionated bilirubin, alkaline phosphatase, gamma-glutamyl transferase, prothrombin time, total cholesterol and fractions, and triglycerides) and proteomic analysis. The biochemical profile will be analyzed at the Clinical Analysis Laboratory of InCor. Blood samples for the proteomic profile will be collected in standard tubes, coagulated at 4°C for 1 to 2 hours, then warmed to room temperature for 30 minutes before centrifugation (2500g/10 min). The plasma will be stored at -80°C in the freezer of the Vascular Biology Laboratory on the 9th floor of InCor and later sent in a single batch to the Neuroproteomics Laboratory at UNICAMP for analysis. Additionally, metabolomic analysis will be conducted at the Max Feffer Laboratory of Plant Genetics, Department of Genetics, "Luiz de Queiroz" College of Agriculture (ESALQ), University of São Paulo, Piracicaba. Plasma samples will be depleted of the most abundant proteins, and the low-abundance protein fraction will be digested with trypsin. The resulting peptides will then undergo liquid chromatography coupled with mass spectrometry (LC-MS) for proteomic and metabolomic analyses.

DATA ANALYSIS The data obtained by LC-MS will be processed using algorithms for protein identification and quantification, with searches conducted in databases such as the Uniprot Human Proteomic Database (http://www.uniprot.org/). For spectrum processing, the software Progenesis QI for Proteomics 4.0 will be used. The data generated will serve as the basis for protein identification and quantification using the label-free MSE method. In the software, samples will be categorized into the predefined study groups. Proteins will be identified using strict search parameters, requiring at least two unique peptides for protein identification. Differentially expressed proteins between groups will be considered based on a fold change cutoff of 1.5, for both upregulated and downregulated expression. The confidence score will be set at ≥3, and analysis of variance (ANOVA) will be conducted with a p-value threshold of ≤0.05.