Blog

Nov 02, 2024

Circulating amino acid signature features urea cycle alterations associated with coronary artery disease | Scientific Reports

Scientific Reports volume 14, Article number: 25848 (2024) Cite this article 318 Accesses Metrics details Coronary artery disease (CAD) remains a leading cause of death worldwide and imposes a

Scientific Reports volume 14, Article number: 25848 (2024) Cite this article

318 Accesses

Metrics details

Coronary artery disease (CAD) remains a leading cause of death worldwide and imposes a substantial socioeconomic burden on healthcare. Improving risk stratification in clinical practice could help to combat this burden. As amino acids are biologically active metabolites whose involvement in CAD remains largely unknown, this study investigated associations between circulating amino acid levels and CAD phenotypes. A high-coverage quantitative liquid chromatography-mass spectrometry approach was applied to acquire the serum amino acids profile of age- and sex-coarsened-matched patients with CAD (n = 46, 66.9 years, 74.7% male) and healthy individuals (n = 120, 67.4 years, 74.7% male) from the COmPLETE study. Multiple linear regressions were performed to investigate associations between amino acid levels and (a) the health status (CAD vs. healthy), (b) the number of affected coronary arteries, or (c) the left ventricular ejection fraction. Regressions were adjusted for age, sex, daily physical activity, sampling, and fasting time. Urea cycle amino acids (ornithine, citrulline, homocitrulline, aspartate, and arginine) were significantly and negatively associated with CAD, the number of affected coronary arteries, and the left ventricular ejection fraction. Lysine, histidine, and the glutamine/glutamate ratio were also significantly and negatively associated with the CAD phenotypes. Overall, patients with CAD displayed lower levels of urea cycle amino acids, highlighting a potential role for urea cycle amino acid profiling in cardiovascular risk stratification.

Trial registration

The study was registered on https://www.clinicaltrials.gov (NCT03986892) on June 5, 2019.

Coronary artery disease (CAD) remains the leading cause of death worldwide and a significant contributor to disability1. Improving prevention and early detection of CAD could mitigate this socioeconomic burden2. In clinical practice, biochemical stratification of patients with or at risk of CAD still mainly relies on the measurement of glycated haemoglobin (HbA1c), low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C), and triglycerides3. These biomarkers, however, represent only partly the metabolic disturbances occurring with CAD4. Advances in mass spectrometry and bioinformatics recently enabled the acquisition of multiparametric metabolic profiles, including quantifying a wide range of circulating metabolites, which could improve cardiovascular risk stratification5.

While lipids have long been used as biomarkers of CAD, amino acids have sparsely been investigated in this context and are traditionally considered as building blocks of proteins. However, amino acids act as signalling molecules to regulate metabolic homeostasis and are important precursors to hormones and neurotransmitters6. Furthermore, amino acid and lipid metabolisms are closely related7. Indeed, amino acid catabolism produces acetyl-coenzyme A, a substrate of fatty acid synthesis, linking elevated amino acid concentrations to increased levels of circulating fatty acids and acylcarnitines8.

So far, methionine, histidine, glutamine, branched-chain amino acids, phenylalanine, and citrulline have been positively associated with the risk of incident cardiovascular events5,9,10. At the same time, homoarginine and tryptophan are believed to be atheroprotective11. Interestingly, arginine has been positively and negatively associated with CAD5. Building upon these promising findings, this study aimed to acquire the circulating amino acid profile of patients with CAD and age- and sex-coarsened-matched healthy individuals from the COmPLETE study12. Further, associations between circulating amino acid levels and CAD phenotypes were investigated. To acquire the most comprehensive panel possible of circulating amino acids, a novel high-coverage targeted metabolomics approach was applied7.

Subsets of the COmPLETE Heart (n = 46, 91.3% male) and COmPLETE Health (n = 183, 61.2% male) samples were investigated. The COmPLETE Heart subset comprised solely patients diagnosed with CAD, as verified by senior cardiologists. For this study, CAD was defined as "a pathological process characterised by atherosclerotic plaque accumulation in the epicardial arteries"13. Patients were included based on the diagnosis in the medical reports provided by treating cardiologists. In the case of clinically significant concomitant disease states (i.e. non-cardiovascular diseases), patients were excluded from the COmPLETE Heart study. As the aim was to investigate associations between CAD diagnosis, as evidence of a systematic atherosclerosis cardiovascular disease, and circulating amino acids, no difference was made between patients with or without stenting of coronary arteries and with or without left main coronary artery disease. As previously reported in the study protocol, "patients with unstable angina pectoris, uncontrolled brady- or tachyarrhythmia, permanent atrial fibrillation, severe valvular disease, acute myocardial infarction, transient ischemic attack, or stroke in the last three months were excluded"12,14. "Clinically healthy participants from the Basel area (Switzerland) who had no exercise-limiting chronic diseases and were non-smokers (or had quit at least ten years previously) were included in the COmPLETE Health sample"12,14. "This excluded participants with a history of CAD, stroke, heart failure, lower-extremity artery disease, any malignant tumour, diabetes, obesity, clinically apparent kidney failure, severe liver disease, chronic obstructive pulmonary disease (GOLD stadium II-IV), arterial hypertension (grades II-III according to the classification of the European Society of Cardiology), drug or alcohol abuse, exercise-limiting osteoporosis or musculoskeletal conditions and clinically manifest Alzheimer’s disease or dementia"12,14. The COmPLETE study conformed to the ethical guidelines of the Declaration of Helsinki and was approved by the Ethics Committee of North-Western and Central Switzerland (EKNZ 2017-01451). The study was registered on https://www.clinicaltrials.gov (NCT03986892) on June 5, 2019. All participants provided written informed consent.

All clinical data were gathered between January 2018 and December 2019 at the Department of Sport, Exercise, and Health, University of Basel. The methodology for participant recruitment and data acquisition was detailed within the study protocol12. Briefly, participants received instructions to maintain their usual dietary habits before clinical assessments (for the preceding 72 h), refrain from exercise and alcohol consumption (for the past 24 h), as well as abstain from caffeinated beverages (for at least the last 4 h) and food intake (for a minimum of 3 h). Participants were assigned to one of five available time slots (08:00, 10:00, 12:00, 14:00, and 16:00) based on their scheduling availability, with each assessment session lasting approximately 4 h.

Before the clinical examination, participants’ smoking status was evaluated via telephone. Physicians reviewed medical histories and current medications using an on-site questionnaire. Body fat composition was assessed through a four-segment bioelectrical impedance analysis (Inbody 720, Inbody Co. Ltd., Seoul, Korea). Trained physicians conducted transthoracic echocardiography in accordance with international standards15. Apical 2- and 4-chamber views were acquired using the Fukuda UF 760 ultrasound scanner (Fukuda Denshi, Tokyo, Japan) equipped with an SA16 (2–5 MHz) transducer (Fukuda Denshi, Tokyo, Japan). A trained cardiologist visually estimated left ventricular ejection fraction (LVEF) using the eyeballing method, demonstrating comparable accuracy to formal echocardiographic techniques for LVEF assessment16. The number of affected coronary arteries was determined based on patients’ most recent medical reports. Following 1 h of rest, blood samples were collected from participants in a fasting state (minimum of 3 h) via venipuncture of the cubital fossa (2 × 7.5 mL serum-gel, Monovette®, Sarstedt, Nümbrecht, Germany). Serum samples underwent gentle agitation for 30 min, followed by centrifugation (3000 rpm; 10 min; 20–23 °C), aliquoting, and storage at − 80 °C.

Physical activity levels were tracked for 14 days after the clinical assessment using a wrist-worn triaxial accelerometer (GeneActive Activinsights Ltd., Kimbolton, UK). Data analysis was conducted using the validated open-source Excel macro file "General physical activity" (version 2), with total and moderate-to-vigorous physical activity quantified in minutes per day (moderate defined as 4.00–6.99 Metabolic Equivalent of Task (METS) and vigorous as ≥ 7 METS)17.

Serum concentrations of total cholesterol, LDL-C, HDL-C, and triglycerides were analyzed using an Olympus AU680 automatic analyzer (Beckman Coulter, Brea, CA, USA), enzymatic reagents (DiaSys, Holzheim, Germany), and secondary standards (Roche Diagnostics, Mannheim, Germany). HbA1c levels were quantified from whole blood through high-pressure liquid chromatography employing D-10 (Bio-Rad, Hercules, CA, USA).

To quantify amino acid concentrations, 20 µL of blood serum was extracted using 250 µL of ice-cold methanol spiked with corresponding stable isotope-labelled standards. Subsequently, the solution was adjusted to 300 µL with 0.1% formic acid in water. After vortexing and centrifugation for 15 min at 4 °C and 2700g, the supernatants were transferred to liquid chromatography-mass spectrometry vials before injection.

Serum extracts underwent analysis utilizing hydrophilic interaction liquid chromatography coupled to high-resolution mass spectrometry (HILIC-HRMS) employing a Vanquish Horizon (Thermo Fisher Scientific) ultra-high performance liquid chromatography (UHPLC) system connected to a Q-Exactive™ Focus mass spectrometer interfaced with a heated electrospray ionization (HESI) source operating in positive mode7. Chromatographic separation utilized an Acquity BEH amide (1.7 μm, 100 mm × 2.1 mm I.D.) column (Waters, Massachusetts, US). The mobile phase consisted of A = 20 mM ammonium formate and 0.1% formic acid in water and B = 0.1% formic acid in acetonitrile. The gradient elution commenced at 95% B (0–2 min), decreased to 65% B (2–14 min), reached 50% B at 16 min, and was followed by an isocratic step (16–18 min) before a 4-min post-run for column re-equilibration. The flow rate was set at 400 μl/min, the column temperature at 25 °C, and the sample injection volume at 2 μl. The HESI source conditions in positive mode were as follows: sheath gas flow at 60, auxiliary gas flow rate at 20, sweep gas flow rate at 2, spray voltage at + 3 kV, capillary temperature at 280 °C, s-lens radio frequency level at 60, and auxiliary gas heater temperature at 300 °C. Full scan high-resolution mass spectrometry acquisition mode (m/z 50–750) was employed with parameters including a resolution of 70,000 full-width half maximum (FWHM), 1 µscan, 1e6 automatic gain control, and a maximum inject time of 100 ms.

Xcalibur 4.1 (Thermo Fisher Scientific) was used to process raw data files. The peak area of each amino acid was auto-integrated and manually corrected if necessary. Calibration curves and the stable isotope spike (or internal standard spike) at known concentration were used to determine the concentration of each amino acid concentration. A nine-point calibration curve was used to assess the linearity of the mass spectroscopy response of each amino acid. A human plasma standard reference material (Certificate of Analysis, NIST 1950) was evaluated within each batch of samples as an external quality control of the measurement accuracy.

Patients with CAD were age- and sex-coarsened-matched with clinically healthy individuals using the coarsened exact matching and average treatment effect (MatchIt R package, version 4.5.0)18. The coarsened exact matching method coarsened covariates into bins before applying exact matching. The balance ultimately comes from weighting the matched observations in the analysis. Thus, the benefit of coarsened exact matching over exact matching is that fewer participants are discarded while balancing between groups is maintained (even if it results in differences in group sizes)19.

Three multiple linear regressions were run to exploratively assess associations between circulating amino acids and CAD phenotypes. In addition to testing for associations between amino acids and the presence of CAD, associations with CAD severity were investigated. To this end, the number of affected coronary arteries, as determined by patients’ most recent medical reports, was used as a surrogate of CAD severity20,21,22,23. Finally, associations between amino acids and LVEF were investigated. LVEF is a functional parameter which has been reported to be associated with mortality and cardiovascular events in CAD patients24 and with subclinical atherosclerotic disease burden25. To determine for which variables regressions needed to be adjusted, a directed acyclic graph (DAG) was drawn (Supplementary Fig. 1)26. Amino acids served as dependent variables, whereas the variables of interest, age, sex, daily total physical activity, sampling time, and fasting time served as independent variables. The variable of interest was a binary variable CAD vs. healthy in the first regression, the number of affected coronary arteries in the second (0, 1, 2, or 3), and the LVEF in the third regression20. According to the current heart failure guidelines of the European Society of Cardiology, LVEF was divided into preserved (≥ 50%), mildly reduced (41–49%), or reduced (≤ 40%)27. During the review process, the authors were recommended to add for each set of regressions an additional model, in which participants with kidney failure (defined as an estimated glomerular filtration rate, abbreviated eGFR, lower than 60 ml/min/1.73 m2) were excluded, and eGFR was used as an additional variable for which regressions were adjusted28. eGFR was calculated with the CKD EPI formula using the ckd_epi function of the R package transplantr29. To estimate the effects (β coefficient) and the uncertainty of the effects (standard errors, 95% confidence intervals (CI), and p-values), the cluster-robust standard errors method was applied as recommended following matching30,31. This method accounts for dependence between observations within matched pairs (i.e. clusters)30,31.

Amino acid concentrations were log2-transformed before statistical analysis, and continuous dependent and independent variables were z-standardised32. Model assumptions were assessed graphically using residual plots. If the proportion of missing data remained below 5%, a complete case analysis was done. Otherwise, multiple imputations would have been considered. All p-values were adjusted using the Benjamini–Hochberg (BH) method33. The statistical analyses and all graphs were done with the software R (version 4.1.1)34. R-codes are available at https://github.com/JustinCarrard/Amino_acid_profiling_of_CAD_patients. The raw data are available in the Supplementary Table 1.

Participants’ characteristics are displayed in Table 1. One hundred twenty healthy participants (67.5 years (standard deviation 14.0 years), 68.3% male) could be age- and sex-coarsened-matched to 46 patients with CAD (65.6 (14.0) years, 91.3% male). The coarsened exact matching led to a balanced weighted distribution of age (mean of 67.2 years in the healthy participants vs. 66.9 years in CAD patients) and sex in both groups (74.7% of males in both groups). The healthy individuals showed a mean BMI within the normal range, while CAD patients were overweight on average. According to the European Society of Cardiology classification, mean systolic and diastolic blood pressure values were within the normal (healthy individuals and males with CAD) and high normal range (females with CAD)35. All CAD patients and 20.8% of the healthy individuals took antihypertensive drugs. Two-thirds of the healthy population never smoked, 15.6% of CAD patients quit less than ten years ago, and 13.0% were still smoking. As to cholesterol, all the CAD patients except one were taking lipid-lowering drugs (compared to 7.5% of healthy individuals). Thus, healthy individuals exhibited higher mean total cholesterol (5.8 (1.2) mmol/l) and LDL-C (3.2 (0.8) mmol/l) than CAD patients (total cholesterol: 4.1 (0.8) mmol/l, LDL-C: 2.2 (0.5) mmol/l). Antidiabetic drugs were taken by 21.7% of the CAD patients, with mean HbA1c values in the prediabetic range (6.0 ± 0.7%). Twenty-four (52.2%) CAD patients had an eGFR below 60 ml/min/1.72 m2, while 15 (12.5%) apparently healthy individuals showed a mildly or moderately impaired eGFR. Triple-vessel coronary artery disease was present in 39.1% of CAD patients (vs. 28.3% for single- and 26.1% for double-vessel coronary artery disease). The number of affected coronary arteries was missing in the medical report for three patients. Preserved LVEF was found in 47.8% of CAD patients (vs. 26.1% for mildly reduced and 26.1% for reduced LVEF). Finally, the mean fasting time before blood sampling was at least 6.7 (3.2) hours in each group.

Multiple amino acids displayed significantly different levels in patients with CAD compared to healthy participants, as displayed in Fig. 1. Specifically, amino acids implicated in the urea cycle, i.e., ornithine (β coefficient − 1.47, 95% CI [− 2.29, − 0.64], BH p-value ≤ 0.01), citrulline (β coefficient − 1.64 [− 2.28, − 1.00], BH p-value ≤ 0.0001), homocitrulline (β coefficient − 0.58 [− 0.99, − 0.17], BH p-value ≤ 0.05), aspartate (β coefficient − 1.18 [− 1.98, − 0.39], BH p-value ≤ 0.05) and arginine (β coefficient − 1.48 [− 2.09, − 0.87], BH p-value ≤ 0.0001), were all significantly and negatively associated with CAD. Glutamate (β coefficient 0.96 [0.32, 1.61], BH p-value ≤ 0.05), a precursor of ornithine feeding into the urea cycle, was significantly and positively associated with CAD. Although glutamine was not significantly different between CAD patients and healthy controls, the glutamine/glutamate ratio showed a significant and negative association with CAD (β coefficient − 0.92 [− 1.63, − 0.21], BH p-value ≤ 0.05). Additionally, lysine (β coefficient − 1.36 [− 2.07, − 0.65], BH p-value ≤ 0.01) and histidine (β coefficient − 1.43 [− 2.05, − 0.82], BH p-value ≤ 0.001) were significantly and negatively associated with CAD. Lastly, creatinine (β coefficient 0.68 [0.42, 0.93], BH p-value ≤ 0.0001), 2-aminoadipate (β coefficient 0.62 [0.42, 0.82], BH p-value ≤ 0.0001), kynurenine (β coefficient 0.52 [0.22, 0.82], BH p-value 0.01), valine (β coefficient 0.44 [0.16, 0.71], BH p-value 0.05) and isoleucine (β coefficient 0.39 [0.09, 0.69], BH p-value 0.05) were significantly and positively associated with CAD. All the results are available in the Supplementary Table 2.

Associations between circulating amino acid levels and the presence of coronary artery disease. This rainplot represents the results of the first set of regression, in which metabolites were used as dependent variables (vertical axis), while CAD phenotype (two-level variable opposing sickness vs health) and confounders served as independent variables (horizontal axis). The redder the dots, the higher the β coefficient and the bigger the dot, the smaller the adjusted p-value. Metabolites with similar β coefficients and adjusted p-values were clustered together (left side of the rainplot). BH Benjamini–Hochberg.

As shown in Fig. 2, the amino acids of the urea cycle (ornithine, citrulline, homocitrulline, aspartate, and arginine), as well as lysine and histidine, were also significantly and negatively associated with the number of affected coronary arteries. Importantly, the β coefficient tended to progressively decrease with the number of affected coronary arteries. Whereas glutamate was significantly and positively associated with double- and triple-vessel CAD, the glutamine/glutamate ratio showed a significant negative association with double- and triple-vessel CAD. Lastly, 2-aminoadipate was positively associated with single-, double- and triple-vessel CAD, while creatinine was positively associated with single- and triple-vessel CAD. Phenylalanine and valine were positively associated with double- and triple-vessel CAD. Kynurenine was positively associated with single-vessel CAD, β-alanine was positively associated with double-vessel CAD, isoleucine was positively associated with triple-vessel CAD, and pipecolate was negatively associated with triple-vessel CAD. All the detailed results are available in Fig. 2, and the Supplementary Table 3. β coefficients, 95% CI and BH p-values are intentionally omitted in this section to facilitate readability.

Associations between circulating amino acid levels and the number of affected coronary arteries. This rainplot represents the results of the second set of regression, in which metabolites were used as dependent variables (vertical axis). The number of affected coronary arteries (0, 1, 2 or 3) and confounders were independent variables (horizontal axis). The redder the dots, the higher the β coefficient and the bigger the dot, the smaller the adjusted p-value. Metabolites with similar β coefficients and adjusted p-values were clustered together (left side of the rainplot). BH Benjamini–Hochberg.

Ornithine, citrulline, homocitrulline, aspartate, arginine, lysine, and histidine were significantly and negatively associated with LVEF impairment (Fig. 3). 2-aminoadipate was significantly and positively associated with all grades of LVEF impairment. The glutamine-to-glutamate ratio was significantly and negatively associated with preserved and reduced LVEF. Hydroxyproline was significantly and positively associated with preserved and reduced LVEF, while proline was significantly and positively associated with reduced LVEF only. Glutamate, kynurenine, creatinine, valine, isoleucine, and phenylalanine were significantly and positively associated with preserved LVEF (≥ 50%). Finally, glycine and pipecolate were significantly and negatively associated with mildly reduced LVEF. All the detailed results are available in Fig. 3, and the Supplementary Table 4. β coefficients, 95% CI and BH p-values are intentionally omitted in this section to facilitate readability.

Associations between circulating amino acid levels and left ventricular ejection fraction impairment. This rainplot represents the results of the second set of regression, in which metabolites were used as dependent variables (vertical axis), while the level of left ventricular ejection fraction impairment (normal: healthy controls, preserved: LVEF ≥ 50%, mildly reduced: LVEF = 41–49%, and reduced: LEVF ≤ 40%). and confounders served as independent variables (horizontal axis). The redder the dots, the higher the β coefficient and the bigger the dot, the smaller the adjusted p-value. Metabolites with similar β coefficients and adjusted p-values were clustered together (left side of the rainplot). BH Benjamini–Hochberg, LVEF left ventricular ejection fraction.

Four of the five urea cycle amino acids (citrulline, homocitrulline, arginine, and ornithine) were significantly and positively associated with age in the three sets of multiple linear regressions. Aspartate was the only urea cycle amino acid which was not significantly associated with age. With the exception of citrulline, none of these amino acids was significantly associated with the interaction term age:male, which means that the associations observed between amino acids and age are valid for both females and males. None of the urea cycle amino acids showed any association with the male sex. Lysine, histidine, glutamate, the glutamine/glutamate ratio and kynurenine did not show any significant association with age, male or the interaction term age:male.

Methionine, tryptophan, asparagine, pipecolate, alanine, valine, phenylalanine, isoleucine, beta-alanine, proline, creatinine, and 2-aminoadipate were significantly and negatively associated with age but significantly and positively associated with the interaction term age:male. Asparagine, pipecolate, beta-alanine, and 2-aminoadipate were also significantly and negatively associated with the male sex. Finally, creatine was significantly and positively associated with daily physical activity, while alanine was significantly and negatively associated with daily physical activity. All the detailed results are available in the Supplementary Tables 1 to 3. β coefficients, 95% CI and BH p-values are intentionally omitted in this section to facilitate readability.

In these alternative models, 24 CAD patients (52.2%) and 15 apparently healthy individuals (12.5%) were excluded for having an eGFR below 60 ml/min/1.72 m2, leaving the total of included participants to 105 healthy participants (67.7% male) and 22 age- and sex-coarsened-matched CAD patients (86.3% male). The results obtained with these alternatives models are similar to the results obtained with the primary models. In particular, the circulating levels of the urea cycle amino acids (ornithine, citrulline, homocitrulline, aspartate, and arginine) remained negatively associated with the presence of CAD (Supplementary Fig. 2 and Supplementary Table 5), with the number of affected coronary arteries (Supplementary Fig. 3 and Supplementary Table 6) and with LVEF (Supplementary Fig. 4 and Supplementary Table 7).

As summarised in Fig. 4, the present study showed evidence that circulating levels of the urea cycle amino acids (ornithine, citrulline, homocitrulline, aspartate, and arginine) were negatively associated with CAD. There was also evidence that lysine, histidine, and the glutamine/glutamate ratio were negatively associated with CAD. Kynurenine, glutamate, valine, isoleucine, 2-aminoadipate and creatinine were positively associated with the presence of CAD. These findings remain valid regardless of the independent variable used to represent CAD phenotypes (a binary variable CAD vs. healthy, the number of affected coronary arteries, or the LVEF), highlighting their robustness.

(Graphical abstract): Findings of the present study illustrated in their biological context. The urea cycle occurs in the liver and, to a lesser extent, in the kidneys. Amino acid catabolism produces ammonia, the accumulation of which is toxic to the body. The urea cycle aims to detoxify ammonia to urea, which is then excreted by the kidneys. The conversion of ammonia to carbamoyl phosphate, as well as the conversion of ornithine and carbamoyl phosphate to citrulline, take place in mitochondria. The other reactions of the urea cycle happen in the cytosol. In addition to producing urea, the urea cycle is an important source of nitric oxide, which is key to maintaining endothelial homeostasis. The nitric oxide-producing amino acids were negatively associated with coronary artery disease in the present study, which might reflect the overconsumption of these amino acids to compensate for the increased nitric oxide demand in patients with coronary artery disease. By driving gluconeogenesis, glutamate is thought to promote lipid accumulation, insulin resistance and inflammation. Histidine can buffer reactive oxygen species. Amino acids highlighted in green were found to be negatively associated with coronary artery disease, while amino acids highlighted in red were positively associated with coronary artery disease. Molecules and pathways written in black are closely related to the urea cycle but were not quantified in the present study. Yellow boxes indicate physiological and pathophysiological consequences. Dashed arrows indicate potential pathways, as hypothesised by Zheng et al.76. The figure is based on37,77,78,79 and was drawn using https://www.mindthegraph.com (accessed on 08.02.23).

The urea cycle mainly occurs in the liver and has two critical roles in the human body36. First, it aims to detoxify ammonia, a product of amino acid catabolism, into harmless urea, which is then excreted by the kidneys37. Second, it produces nitric oxide, which is essential to the endothelial function of blood vessels37. Nitric oxide relaxes the smooth vasculature and dilates blood vessels37. This relaxation is impaired in patients with CAD due to reduced nitric oxide production and increased inactivation of nitric oxide37. As the nitric oxide-producing amino acids were negatively associated with CAD phenotypes, this might reflect the overconsumption of these amino acids to compensate for the increased nitric oxide demand.

Arginine levels were lower in patients with CAD compared to healthy controls. Through a transformation catalyzed by the nitric oxide synthetase, arginine is a precursor of nitric oxide and is, therefore, involved in regulating endothelial function38. Even though the clinical relevance of arginine as a limiting factor for the bioavailability of nitric oxide remains a matter of debate39,40. Low arginine levels have been previously associated with atherosclerosis, which aligns with the present findings41. Lysine, which also displayed lower levels in CAD patients, is another nitric oxide source. Indeed, the arginine-glycine amidino-transferase converts lysine to homoarginine, which the nitric oxide synthetase can then metabolize to produce nitric oxide42. As the nitric oxide synthetase holds a lower affinity for homoarginine than arginine, the observed lower lysine levels in patients with CAD might indicate an important nitric oxide demand. In addition, inflammation states, such as CAD, promote the carbamylation of lysine to protein-bound homocitrulline43. Carbamylation, a non-enzymatic binding of isocyanic, lowers circulating lysine and increases protein-bound homocitrulline levels in CAD patients44. While circulating homocitrulline was lower in patients with CAD, it must be pointed out that free and not protein-bound homocitrulline was measured. Therefore, a shift from free homocitrulline to protein-bound homocitrulline could occur in patients with CAD.

Glutamate was positively associated with CAD. This finding aligns with previous reports showing elevated glutamate levels in subjects with cardiovascular risk factors and diseases45,46. By driving gluconeogenesis, glutamate is thought to promote traditional cardiovascular risk factors such as obesity and insulin resistance47,48. There is evidence that circulating glutamate might also have pro-inflammatory effects49. As nutrition is a significant source of glutamate in humans (primarily through the consumption of animal proteins, including red meats, poultry, fish, and dairy products), it is believed that an elevated glutamate intake could favour the clustering of traditional cardiovascular risk factors and eventually the development of CAD50.

While elevated glutamine level was previously linked to lower cardiovascular risk46, no significant association between glutamine and CAD phenotypes was found in the present study. Therefore, the fact that the glutamine/glutamate ratio was found to be significantly and negatively associated with CAD is due to the elevated glutamate levels observed in patients with CAD. Nonetheless, glutamine is considered to be involved, as an indirect arginine precursor, in nitric oxide synthesis and the production of the anti-oxidative glutathione51.

The involvement of oxidative stress in the pathophysiology of CAD is well-recognized52. Elevated levels of reactive oxygen species inflict oxidative damage to macromolecules, impair vascular physiology, and ultimately promote CAD development52. Radical scavengers such as histidine can buffer reactive oxygen species53. Furthermore, histidine and alanine can form carnosine, another radical scavenger, especially in muscle tissue54. In that way, lower histidine levels in patients with CAD might sign an increased consumption of histidine to scavenge reactive oxygen species54. Lastly, in heart failure patients, in which oxidative stress also plays a pivotal role, histidine is negatively associated with B-type natriuretic peptide (BNP) but positively associated with serum albumin concentration, which might indicate that skeletal muscle wasting could also partly explain lower circulating histidine levels in cardiovascular diseases55.

Kynurenine is a product of tryptophan catabolism, mainly occurring in the liver56. Kynurenine and related metabolites are involved in maintaining vascular immune homeostasis57. Alterations in their metabolism have been reported in the pathogenesis of atherosclerosis56. Specifically, elevated blood kynurenine levels were shown to predict death and recurrent myocardial infarction in patients with CAD56,58 and acute myocardial infarction in patients with suspected stable angina pectoris59. Kynurenine levels were also associated with cardiovascular disease prevalence in end-stage renal disease patients60. These findings align with the present observation that CAD patients displayed higher kynurenine levels than healthy controls. Overall, the kynurenine pathway seems to be a promising research area regarding CAD pathophysiology.

2-Aminoadipate, a product of lysine catabolism, displayed elevated levels in patients with CAD compared to healthy controls in the present study61. Higher 2-aminoadipate levels have been associated with type 2 diabetes62, insulin resistance61, and coronary artery calcification63. Recently, elevated levels of 2-aminoadipate were linked to reduced high-density lipoprotein cholesterol and elevated triglyceride levels64,65, highlighting a potential mechanism through which elevated 2-aminoadipate could negatively influence cardiometabolic health.

Creatine is produced primarily in the liver from the methylation of guanidinoacetate and synthesized in kidneys from the amino acids arginine and glycine66. In skeletal muscle, creatine conversion to phosphocreatine forms creatinine, which diffuses from the cells and is excreted into the urine by glomerular filtration and partial tubular excretion66. The latter explains that creatinine is a classical marker of kidney function in clinical medicine67. Even though creatinine is considered to be a biologically inactive metabolite66, elevated creatinine levels have been associated with increased mortality in CAD patients68. Interestingly, plasma creatinine levels were related to arterial stiffness in untreated hypertensive subjects with preserved renal function69. Whether creatinine plays a biologically active role in atherosclerosis remains unknown and will require further investigation.

Valine and isoleucine belong to the branched-chain amino acids (BCAA). In humans, diet is the first provider of BCAA, even though a small part is synthesized by the gut microbiome10. Elevated levels of circulating BCAA have been observed in several cardiometabolic diseases, including atherosclerosis and CAD10. Interestingly, elevated circulating BCAA levels have been shown to be predictive of CAD severity70,71. This was also found in the present study, as there was evidence for positive associations with double- and triple-vessel CAD (valine) or even with triple-vessel CAD only (isoleucine). Molecular mechanisms linking BCAA and atherosclerosis or CAD remain largely unknown but could involve activation of the serine/threonine protein kinase mTOR, which leads to more reactive oxygen species production and endothelial dysfunction72, or activation of platelet-dependent pro-thrombotic pathways73. In the present study, there was no evidence that leucine, the third BCAA, was associated with any CAD phenotype.

Atherosclerosis is now recognized as a chronic inflammatory disease of the arterial wall driven by lipids and immune cells11. The CANTOS trial, which used canakinumab, a therapeutic monoclonal antibody, to block interleukin-1β, proved that targeting the immune system effectively reduces cardiovascular events74. As amino acids are highly involved in the immune regulation of endothelial, innate and adaptive immune cells, they could become future therapeutic targets11. Therefore, understanding amino acid metabolism in atherosclerosis is worthy of interest. For instance, pro- and anti-inflammatory cytokines act differently on the arginine pathway, promoting macrophage conversion into proinflammatory M1 or anti-inflammatory M2 types11. Nevertheless, the mechanisms through which amino acids influence the pathogenesis of atherosclerosis are largely unknown and require further research.

This study should be assessed considering its limitations. First, the cross-sectional nature of this study allows only for the investigation of associations, and not causality, between amino acids and CAD phenotypes. Second, investigating circulating metabolites does not reveal their cellular origin or destination. Thus, the present findings regarding hypothetical molecular mechanisms of action should be interpreted cautiously. Third, the small number of patients with triple-vessel CAD or LVEF ≤ 40% limits the extrapolation of the present results to severely sick patients. Fourth, as more than 90% of the included CAD patients were male, it is unclear to what extent the reported amino acid signature applies to female CAD patients. Finally, the findings reported in this manuscript should be replicated in larger population studies before any translation to clinical practice can be attempted.

It has been demonstrated that food intake and fasting only have minimal impact on circulating amino acid levels75. Nevertheless, serum samples were collected in a fasting state of at least 6.5 h on average, and regression analyses were adjusted for fasting time, further enhancing the robustness of the present findings. Overall, this cross-sectional study should be seen as an incentive to investigate amino acids further in the context of CAD pathophysiology and cardiovascular risk stratification.

The amino acids of the urea cycle, the antioxidant histidine and lysine, were negatively associated with CAD phenotypes, while glutamate was positively associated with them. We hypothesise that this circulating amino acid signature could feature pathophysiological changes commonly observed in CAD, including greater nitric oxide consumption, impaired endothelial function, and elevated oxidative stress. Overall, amino acid profiling seems a promising pathophysiology-based phenotyping tool to stratify cardiovascular risk. However, replication of the present findings in larger-scale and prospective population studies is required before amino acid profiling can be translated to clinical practice.

All data are available as supplements to the present manuscript.

Murray, C. J. L. et al. Global burden of 87 risk factors in 204 countries and territories, 1990–2019: A systematic analysis for the Global Burden of Disease Study 2019. Lancet 396, 1223–1249. https://doi.org/10.1016/S0140-6736(20)30752-2 (2020).

Article Google Scholar

Benziger, C. P., Roth, G. A. & Moran, A. E. The global burden of disease study and the preventable burden of NCD. Glob. Heart 11, 393–397. https://doi.org/10.1016/j.gheart.2016.10.024 (2016).

Article PubMed Google Scholar

Mach, F. et al. 2019 ESC/EAS Guidelines for the management of dyslipidaemias: lipid modification to reduce cardiovascular risk: The Task Force for the management of dyslipidaemias of the European Society of Cardiology (ESC) and European Atherosclerosis Society (EAS). Eur. Heart J. 41, 111–188. https://doi.org/10.1093/eurheartj/ehz455 (2019).

Article Google Scholar

Wu, Z. et al. Circulating cardiac biomarkers improve risk stratification for incident cardiovascular disease in community dwelling populations. eBioMedicine. https://doi.org/10.1016/j.ebiom.2022.104170 (2022).

Shah, S. H., Kraus, W. E. & Newgard, C. B. Metabolomic profiling for the identification of novel biomarkers and mechanisms related to common cardiovascular diseases: form and function. Circulation 126, 1110–1120. https://doi.org/10.1161/circulationaha.111.060368 (2012).

Article PubMed PubMed Central Google Scholar

Dai, Z., Zheng, W. & Locasale, J. W. Amino acid variability, tradeoffs and optimality in human diet. Nat. Commun. 13, 6683. https://doi.org/10.1038/s41467-022-34486-0 (2022).

Article ADS CAS PubMed PubMed Central Google Scholar

Teav, T. et al. Merged targeted quantification and untargeted profiling for comprehensive assessment of acylcarnitine and amino acid metabolism. Anal. Chem. 91, 11757–11769. https://doi.org/10.1021/acs.analchem.9b02373 (2019).

Article CAS PubMed Google Scholar

Green, C. R. et al. Branched-chain amino acid catabolism fuels adipocyte differentiation and lipogenesis. Nat. Chem. Biol. 12, 15–21. https://doi.org/10.1038/nchembio.1961 (2016).

Article CAS PubMed Google Scholar

Würtz, P. et al. Metabolite profiling and cardiovascular event risk: a prospective study of 3 population-based cohorts. Circulation 131, 774–785. https://doi.org/10.1161/circulationaha.114.013116 (2015).

Article PubMed PubMed Central Google Scholar

McGarrah, R. W. & White, P. J. Branched-chain amino acids in cardiovascular disease. Nat. Rev. Cardiol. https://doi.org/10.1038/s41569-022-00760-3 (2022).

Article PubMed PubMed Central Google Scholar

Nitz, K., Lacy, M. & Atzler, D. Amino acids and their metabolism in atherosclerosis. Arterioscler. Thromb. Vasc. Biol. 39, 319–330. https://doi.org/10.1161/ATVBAHA.118.311572 (2019).

Wagner, J. et al. Functional aging in health and heart failure: The COmPLETE Study. BMC Cardiovasc. Disord. 19, 180. https://doi.org/10.1186/s12872-019-1164-6 (2019).

Article PubMed PubMed Central Google Scholar

Knuuti, J. et al. 2019 ESC Guidelines for the diagnosis and management of chronic coronary syndromes: The Task Force for the diagnosis and management of chronic coronary syndromes of the European Society of Cardiology (ESC). Eur. Heart J. 41, 407–477. https://doi.org/10.1093/eurheartj/ehz425 (2019).

Article Google Scholar

Gander, J. et al. Metabolic impairment in coronary artery disease: Elevated serum acylcarnitines under the spotlights. Front. Cardiovasc. Med. 8. https://doi.org/10.3389/fcvm.2021.792350 (2021).

Lancellotti, P. & Cosyns, B. The EACVI Echo Handbook (Oxford University Press, 2015).

Gudmundsson, P., Rydberg, E., Winter, R. & Willenheimer, R. Visually estimated left ventricular ejection fraction by echocardiography is closely correlated with formal quantitative methods. Int. J. Cardiol. 101, 209–212. https://doi.org/10.1016/j.ijcard.2004.03.027 (2005).

Esliger, D. W. et al. Validation of the GENEA accelerometer. Med. Sci. Sports Exerc. 43, 1085–1093. https://doi.org/10.1249/MSS.0b013e31820513be (2011).

Article PubMed Google Scholar

Ho, D., Imai, K., King, G. & Stuart, E. A. MatchIt: Nonparametric preprocessing for parametric causal inference. J. Stat. Softw. 42, 1–28. https://doi.org/10.18637/jss.v042.i08 (2011).

Iacus, S. M., King, G. & Porro, G. Causal inference without balance checking: Coarsened exact matching. Polit. Anal. 20, 1–24. https://doi.org/10.1093/pan/mpr013 (2012).

Article Google Scholar

Özcan, C. et al. Coronary artery disease severity and long-term cardiovascular risk in patients with myocardial infarction: A Danish nationwide register-based cohort study. Eur. Heart J. Cardiovasc. Pharmacother. 4, 25–35. https://doi.org/10.1093/ehjcvp/pvx009 (2018).

Article PubMed Google Scholar

Sponder, M., Fritzer-Szekeres, M., Marculescu, R., Litschauer, B. & Strametz-Juranek, J. A new coronary artery disease grading system correlates with numerous routine parameters that were associated with atherosclerosis: A grading system for coronary artery disease severity. Vasc. Health Risk Manag. 10, 641–647. https://doi.org/10.2147/vhrm.S68919 (2014).

Article PubMed PubMed Central Google Scholar

Gensini, G. G. A more meaningful scoring system for determining the severity of coronary heart disease. Am. J. Cardiol. 51, 606. https://doi.org/10.1016/s0002-9149(83)80105-2 (1983).

Article CAS PubMed Google Scholar

Mortensen, M. B. et al. CAD severity on cardiac CTA identifies patients with most benefit of treating LDL-cholesterol to ACC/AHA and ESC/EAS targets. JACC Cardiovasc. Imaging 13, 1961–1972. https://doi.org/10.1016/j.jcmg.2020.03.017 (2020).

Article PubMed Google Scholar

Liu, Y. et al. Association of ejection fraction with mortality and cardiovascular events in patients with coronary artery disease. ESC Heart Fail. 9, 3461–3468. https://doi.org/10.1002/ehf2.14063 (2022).

Article PubMed PubMed Central Google Scholar

Maragiannis, D. et al. Association of left ventricular diastolic dysfunction with subclinical coronary atherosclerotic disease burden using coronary artery calcium scoring. J. Atheroscler. Thromb. 22, 1278–1286. https://doi.org/10.5551/jat.29454 (2015).

Article CAS PubMed Google Scholar

Textor, J., van der Zander, B., Gilthorpe, M. S., Liskiewicz, M. & Ellison, G. T. Robust causal inference using directed acyclic graphs: The R package “dagitty”. Int. J. Epidemiol. 45, 1887–1894. https://doi.org/10.1093/ije/dyw341 (2016).

Article PubMed Google Scholar

McDonagh, T. A. et al. 2021 ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure. Eur. Heart J. 42, 3599–3726. https://doi.org/10.1093/eurheartj/ehab368 (2021).

Article CAS PubMed Google Scholar

Stevens, P. E. et al. KDIGO 2024 clinical practice guideline for the evaluation and management of chronic kidney disease. Kidney Int. 105, S117–S314. https://doi.org/10.1016/j.kint.2023.10.018 (2024).

Article Google Scholar

transplantr: Audit and Research Functions for Transplantation v. 0.2.0 (2020).

Liang, K.-Y. & Zeger, S. L. Longitudinal data analysis using generalized linear models. Biometrika 73, 13–22. https://doi.org/10.1093/biomet/73.1.13 (1986).

Article MathSciNet Google Scholar

Abadie, A. & Spiess, J. Robust post-matching inference. J. Am. Stat. Assoc. 117, 983–995. https://doi.org/10.1080/01621459.2020.1840383 (2022).

Article MathSciNet CAS Google Scholar

Schielzeth, H. Simple means to improve the interpretability of regression coefficients. Methods Ecol. Evol. 1, 103–113. https://doi.org/10.1111/j.2041-210X.2010.00012.x (2010).

Benjamini, Y. & Hochberg, Y. Controlling the false discovery rate: A practical and powerful approach to multiple testing. J. R. Stat. Soc. Ser. B (Methodol.) 57, 289–300. https://doi.org/10.1111/j.2517-6161.1995.tb02031.x (1995).

R Core Team. R: A language and environment for statistical computing (R Foundation for Statistical Computing, 2020). https://www.R-project.org/

Williams, B. et al. 2018 ESC/ESH Guidelines for the management of arterial hypertension: The Task Force for the management of arterial hypertension of the European Society of Cardiology (ESC) and the European Society of Hypertension (ESH). Eur. Heart J. 39, 3021–3104. https://doi.org/10.1093/eurheartj/ehy339 (2018).

Article ADS PubMed Google Scholar

Matsumoto, S. et al. Urea cycle disorders—update. J. Hum. Genet. 64, 833–847. https://doi.org/10.1038/s10038-019-0614-4 (2019).

Article CAS PubMed Google Scholar

Farah, C., Michel, L. Y. M. & Balligand, J.-L. Nitric oxide signalling in cardiovascular health and disease. Nat. Rev. Cardiol. 15, 292–316. https://doi.org/10.1038/nrcardio.2017.224 (2018).

Article CAS PubMed Google Scholar

Stuehr, D. J. Enzymes of the L-arginine to nitric oxide pathway. J. Nutr. 134, 2748S–2751S. https://doi.org/10.1093/jn/134.10.2748S(discussion 2765S–2767S) (2004).

Wilson, A. M., Harada, R., Nair, N., Balasubramanian, N. & Cooke, J. P. L-arginine supplementation in peripheral arterial disease: no benefit and possible harm. Circulation 116, 188–195. https://doi.org/10.1161/circulationaha.106.683656 (2007).

Article CAS PubMed Google Scholar

An, P. et al. Micronutrient supplementation to reduce cardiovascular risk. J. Am. Coll. Cardiol. 80, 2269–2285. https://doi.org/10.1016/j.jacc.2022.09.048 (2022).

Popolo, A., Adesso, S., Pinto, A., Autore, G. & Marzocco, S. l-Arginine and its metabolites in kidney and cardiovascular disease. Amino Acids 46, 2271–2286. https://doi.org/10.1007/s00726-014-1825-9 (2014).

Article CAS PubMed Google Scholar

Tsikas, D., Bollenbach, A., Hanff, E. & Kayacelebi, A. A. Asymmetric dimethylarginine (ADMA), symmetric dimethylarginine (SDMA) and homoarginine (hArg): The ADMA, SDMA and hArg paradoxes. Cardiovasc. Diabetol. 17, 1. https://doi.org/10.1186/s12933-017-0656-x (2018).

Article CAS PubMed PubMed Central Google Scholar

Wang, Z. et al. Protein carbamylation links inflammation, smoking, uremia and atherogenesis. Nat. Med. 13, 1176–1184. https://doi.org/10.1038/nm1637 (2007).

Article CAS PubMed Google Scholar

Jaisson, S. et al. Increased serum homocitrulline concentrations are associated with the severity of coronary artery disease. Clin. Chem. Lab. Med. 53, 103–110. https://doi.org/10.1515/cclm-2014-0642 (2015).

Article CAS PubMed Google Scholar

Cheng, S. et al. Metabolite profiling identifies pathways associated with metabolic risk in humans. Circulation 125, 2222–2231. https://doi.org/10.1161/circulationaha.111.067827 (2012).

Article CAS PubMed PubMed Central Google Scholar

Zheng, Y. et al. Metabolites of glutamate metabolism are associated with incident cardiovascular events in the PREDIMED PREvención con DIeta MEDiterránea (PREDIMED) Trial. J. Am. Heart Assoc. 5. https://doi.org/10.1161/jaha.116.003755 (2016).

Palmer, N. D. et al. Metabolomic profile associated with insulin resistance and conversion to diabetes in the Insulin Resistance Atherosclerosis Study. J. Clin. Endocrinol. Metab. 100, E463–E468. https://doi.org/10.1210/jc.2014-2357 (2015).

Article CAS PubMed Google Scholar

Cabrera, O. et al. Glutamate is a positive autocrine signal for glucagon release. Cell Metab. 7, 545–554. https://doi.org/10.1016/j.cmet.2008.03.004 (2008).

Article CAS PubMed PubMed Central Google Scholar

Lehn-Stefan, A. et al. Elevated circulating glutamate is associated with subclinical atherosclerosis independently of established risk markers: A cross-sectional study. J. Clin. Endocrinol. Metab. 106, e982–e989. https://doi.org/10.1210/clinem/dgaa898 (2021).

Article PubMed Google Scholar

Ma, W. et al. Dietary glutamine, glutamate and mortality: Two large prospective studies in US men and women. Int. J. Epidemiol. 47, 311–320. https://doi.org/10.1093/ije/dyx234 (2018).

Article PubMed Google Scholar

Durante, W. The emerging role of l-glutamine in cardiovascular health and disease. Nutrients 11. https://doi.org/10.3390/nu11092092 (2019).

Münzel, T. et al. Impact of oxidative stress on the heart and vasculature: Part 2 of a 3-Part series. J. Am. Coll. Cardiol. 70, 212–229. https://doi.org/10.1016/j.jacc.2017.05.035 (2017).

Brosnan, M. E. & Brosnan, J. T. Histidine metabolism and function. J. Nutr. 150, 2570S-2575S. https://doi.org/10.1093/jn/nxaa079 (2020).

Article PubMed PubMed Central Google Scholar

Thalacker-Mercer, A. E. & Gheller, M. E. Benefits and adverse effects of histidine supplementation. J. Nutr. 150, 2588S-2592S. https://doi.org/10.1093/jn/nxaa229 (2020).

Article PubMed Google Scholar

Hakuno, D., Hamba, Y., Toya, T. & Adachi, T. Plasma amino acid profiling identifies specific amino acid associations with cardiovascular function in patients with systolic heart failure. PLoS One 10, e0117325. https://doi.org/10.1371/journal.pone.0117325 (2015).

Article CAS PubMed PubMed Central Google Scholar

Anand, S. K. et al. Amino acid metabolism and atherosclerotic cardiovascular disease. Am. J. Pathol. 194, 510–524. https://doi.org/10.1016/j.ajpath.2023.12.006 (2024).

Article CAS PubMed Google Scholar

Baumgartner, R., Forteza, M. J. & Ketelhuth, D. F. J. The interplay between cytokines and the Kynurenine pathway in inflammation and atherosclerosis. Cytokine 122, 154148. https://doi.org/10.1016/j.cyto.2017.09.004 (2019).

Article CAS PubMed Google Scholar

Metghalchi, S. et al. Indoleamine 2,3-dioxygenase fine-tunes immune homeostasis in atherosclerosis and colitis through repression of interleukin-10 production. Cell Metab. 22, 460–471. https://doi.org/10.1016/j.cmet.2015.07.004 (2015).

Pedersen, E. R. et al. Associations of plasma kynurenines with risk of acute myocardial infarction in patients with stable angina pectoris. Arterioscler. Thromb. Vasc. Biol. 35, 455–462. https://doi.org/10.1161/atvbaha.114.304674 (2015).

Article CAS PubMed Google Scholar

Pawlak, K., Domaniewski, T., Mysliwiec, M. & Pawlak, D. The kynurenines are associated with oxidative stress, inflammation and the prevalence of cardiovascular disease in patients with end-stage renal disease. Atherosclerosis 204, 309–314. https://doi.org/10.1016/j.atherosclerosis.2008.08.014 (2009).

Article CAS PubMed Google Scholar

Lee, H. J. et al. 2-Aminoadipic acid (2-AAA) as a potential biomarker for insulin resistance in childhood obesity. Sci. Rep. 9, 13610. https://doi.org/10.1038/s41598-019-49578-z (2019).

Article ADS CAS PubMed PubMed Central Google Scholar

Wang, T. J. et al. 2-Aminoadipic acid is a biomarker for diabetes risk. J. Clin. Investig. 123, 4309–4317. https://doi.org/10.1172/jci64801 (2013).

Article CAS PubMed PubMed Central Google Scholar

Saremi, A. et al. Advanced Glycation end products, oxidation products, and the extent of atherosclerosis during the VA Diabetes Trial and follow-up study. Diabetes Care 40, 591–598. https://doi.org/10.2337/dc16-1875 (2017).

Article CAS PubMed PubMed Central Google Scholar

Shi, M. et al. Genetic architecture of plasma alpha-aminoadipic acid reveals a relationship with high-density lipoprotein cholesterol. J. Am. Heart Assoc. 11, e024388. https://doi.org/10.1161/jaha.121.024388 (2022).

Article CAS PubMed PubMed Central Google Scholar

Desine, S. et al. Association of alpha-aminoadipic acid with cardiometabolic risk factors in healthy and high-risk individuals. Front. Endocrinol. 14. https://doi.org/10.3389/fendo.2023.1122391 (2023).

Wyss, M. & Kaddurah-Daouk, R. Creatine and creatinine metabolism. Physiol. Rev. 80, 1107–1213. https://doi.org/10.1152/physrev.2000.80.3.1107 (2000).

Article CAS PubMed Google Scholar

Levin, A. et al. Executive summary of the KDIGO 2024 Clinical Practice Guideline for the Evaluation and Management of Chronic Kidney Disease: Known knowns and known unknowns. Kidney Int. 105, 684–701. https://doi.org/10.1016/j.kint.2023.10.016 (2024).

Article PubMed Google Scholar

Chen, S. et al. Predictive effect of estimated glomerular filtrate rate by creatinine or cystatin C on mortality in patients with coronary artery disease. Ren. Fail. 46, 2327494. https://doi.org/10.1080/0886022x.2024.2327494 (2024).

Article CAS PubMed PubMed Central Google Scholar

Gosse, P. & Safar, M. E. Arterial stiffness and plasma creatinine in untreated hypertensive patients. Am. J. Hypertens. 18, 1140–1145. https://doi.org/10.1016/j.amjhyper.2005.04.008 (2005).

Article CAS PubMed Google Scholar

Shah, S. H. et al. Association of a peripheral blood metabolic profile with coronary artery disease and risk of subsequent cardiovascular events. Circ. Cardiovasc. Genet. 3, 207–214. https://doi.org/10.1161/CIRCGENETICS.109.852814 (2010).

Bhattacharya, S. et al. Validation of the association between a branched chain amino acid metabolite profile and extremes of coronary artery disease in patients referred for cardiac catheterization. Atherosclerosis 232, 191–196. https://doi.org/10.1016/j.atherosclerosis.2013.10.036 (2014).

Article CAS PubMed Google Scholar

Zhenyukh, O. et al. Branched-chain amino acids promote endothelial dysfunction through increased reactive oxygen species generation and inflammation. J. Cell. Mol. Med. 22, 4948–4962. https://doi.org/10.1111/jcmm.13759 (2018).

Article CAS PubMed PubMed Central Google Scholar

Xu, Y. et al. Branched-chain amino acid catabolism promotes thrombosis risk by enhancing tropomodulin-3 propionylation in platelets. Circulation 142, 49–64. https://doi.org/10.1161/circulationaha.119.043581 (2020).

Article CAS PubMed Google Scholar

Ridker, P. M. et al. Antiinflammatory therapy with canakinumab for atherosclerotic disease. N. Engl. J. Med. 377, 1119–1131. https://doi.org/10.1056/NEJMoa1707914 (2017).

Bröer, S. & Bröer, A. Amino acid homeostasis and signalling in mammalian cells and organisms. Biochem. J. 474, 1935–1963. https://doi.org/10.1042/bcj20160822 (2017).

Article PubMed Google Scholar

Zheng, Y. et al. Metabolites of glutamate metabolism are associated with incident cardiovascular events in the PREDIMED PREvencion con DIeta MEDiterranea (PREDIMED) Trial. J. Am. Heart Assoc. 5, e003755. https://doi.org/10.1161/JAHA.116.003755 (2016)

Lundberg, J. O., Gladwin, M. T. & Weitzberg, E. Strategies to increase nitric oxide signalling in cardiovascular disease. Nat. Rev. Drug Discov. 14, 623–641. https://doi.org/10.1038/nrd4623 (2015).

Article CAS PubMed Google Scholar

Neill, M. A., Aschner, J., Barr, F. & Summar, M. L. Quantitative RT-PCR comparison of the urea and nitric oxide cycle gene transcripts in adult human tissues. Mol. Genet. Metab. 97, 121–127. https://doi.org/10.1016/j.ymgme.2009.02.009 (2009).

Pearson, D. L. et al. Neonatal pulmonary hypertension. N. Engl. J. Med. 344, 1832–1838. https://doi.org/10.1056/nejm200106143442404 (2001).

Article CAS PubMed Google Scholar

Download references

We thank Caren Mutschmann (Synlab Analytics, Berlin, Germany) and Winfried März (Synlab Academy, Mannheim, Germany) for quantifying HbA1c in the serum samples. We thank Hubert Scharnagl (Clinical Institute of Medical and Chemical Laboratory Diagnostics, Medical University of Graz, Austria) for analysing cholesterol and triglycerides in the blood samples. We are grateful to Christopher Klenk, Gilles Nève, and all the Bachelor and Master students (Department of Sport, Exercise and Health, University of Basel, Switzerland) who contributed to the data collection. Finally, we acknowledge using the Mind the Graph platform [https://www.mindthegraph.com (accessed on 08.02.23)] to create Fig. 4.

This study was funded by the Swiss National Science Foundation (grant nos. 182815 to AST and 316030_183377 to JI). The funding sources were not involved in any way in the study process, including designing the study, collecting, analyzing, and interpreting the data, writing the report, or submitting the article to the journal.

These authors contributed equally: Luisa Prechtl and Justin Carrard.

These authors jointly supervised this work: Julijana Ivanisevic and Arno Schmidt-Trucksäss.

School of Cardiovascular and Metabolic Health, University of Glasgow, 126 University Place, Glasgow, G12 8TA, Scotland

Luisa Prechtl

Division of Sports and Exercise Medicine, Department of Sport, Exercise and Health, University of Basel, Grosse Allee 6, 4052, Basel, Switzerland

Justin Carrard, Karsten Königstein, Jonathan Wagner, Raphael Knaier, Denis Infanger, Lukas Streese, Timo Hinrichs, Henner Hanssen & Arno Schmidt-Trucksäss

Metabolomics Platform, Faculty of Biology and Medicine, University of Lausanne, Quartier UNIL-CHUV-Rue du Bugnon 19, 1005, Lausanne, Switzerland

Hector Gallart-Ayala, Rébecca Borreggine, Tony Teav & Julijana Ivanisevic

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

Conceptualisation: JC, HGA, TH, HH, JI, AST; Data curation: LP, JC, HGA, RB, TT, DI; Formal analysis: LP, JC, DI; Funding acquisition: JI, AST; Investigation: JC, HGA, RB, TT, KK, JW, RK, LS, TH; Methodology: JC, HGA, DI, TH, HH, JI, AST; Project administration: JC, HGA, JW, RK, JI, AST; Resources: JI, AST; Software: LP, JC, HGA, RB, TT, DI; Supervision: JC, HGA, JW, RK, JI, AST; Validation: HGA, RB, TT, DI, JW, RK, JI, AST; Visualisation: LP, JC, DI; Roles/Writing—original draft: LP, JC; Writing—review and editing: LP, JC, HGA, RB, TT, KK, JW, RK, DI, LS, TH, HH, JI, AST.

Correspondence to Justin Carrard or Julijana Ivanisevic.

The authors declare no competing interests.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Figures. The file containing the supplementary Figures is still is track change mode. Please convert it into a clean version. If you are unsure if it is fine or not, please send a second proof to review. Thank you.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

Prechtl, L., Carrard, J., Gallart-Ayala, H. et al. Circulating amino acid signature features urea cycle alterations associated with coronary artery disease. Sci Rep 14, 25848 (2024). https://doi.org/10.1038/s41598-024-76835-7

Download citation

Received: 01 April 2024

Accepted: 17 October 2024

Published: 28 October 2024

DOI: https://doi.org/10.1038/s41598-024-76835-7

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative