Abstract
Aim: Branched-chain amino acids (BCAAs) and aromatic amino acids (AAAs) were suggested as potential biomarkers in liver disease. This study aimed to develop and validate a simple and rapid LC-MS/MS method to simultaneously measure serum BCAAs and AAAs levels in patients with liver injury, and further establish reference intervals of Chinese healthy adult populations.
Patients & methods: Samples were prepared by a one-step protein precipitation and analysis time was 4 min per run.
Results: The validation results showed good linearity (r2 >0.9969), satisfactory accuracy (94.44% – 107.75%) and precision (0.10% – 5.90%).
Conclusion: This method proved to be suitable for high-throughput routine clinical use and could be a valuable adjunct diagnosis tool for liver injury and other clinical applications.
Keywords: : aromatic amino acids, bioanalysis, branched-chain amino acids, fischer ratio, LC-MS/MS, liver injury, reference intervals
Plain language summary
Article highlights.
A simple and rapid LC-MS/MS method for simultaneous determination of BCAAs and AAAs in human serum was established and verified according to CLSI C62-A and Consensus of Method Development and Validation of Liquid Chromatography-Tandem Mass Spectrometry in Clinical Laboratories.
Simple sample preparation procedures (one-step protein precipitation) and a short running time (4 min), which is suitable for high-throughput and routine clinical monitoring.
The reference intervals for six BCAAs and AAAs were established in Chinese adult population by this LC-MS/MS.
The LC-MS/MS method was successfully applied in the serum analysis of 68 patients with liver injury.
1. Introduction
Branched-chain amino acids (BCAAs) are a group of neutral, aliphatic amino acids with a branched carbon chain structure and include valine (Val), leucine (Leu) and isoleucine (Ile). Phenylalanine (Phe), tryptophan (Trp) and tyrosine (Tyr) are the only three amino acids that contain benzyl-based aromatic groups and hence, are referred to as aromatic amino acids (AAAs). BCAAs and AAAs are essential amino acids for human beings, which are strongly associated with metabolic disease and provide energy to the liver, heart, skeletal muscle and brain [1]. In addition to these roles, they are considered biomarkers for a variety of diseases. For example, BCAAs and AAAs are risk factors for obesity [2], insulin resistance [3], metabolic syndrome [4] and cardiovascular disease [5]. However, they are more commonly used clinically in the field of liver disease [6]. BCAAs and AAAs levels could be used for the early detection of non-alcoholic fatty liver disease (NAFLD) [7,8]. Particularly, the ratio of BCAAs to Phe and Tyr (Fischer ratio) and the BCAAs/Tyr molar concentration ratio (BTR) are important for assessing liver metabolism, hepatic functional reserve and hepatic malformation severity. A study showed that a decrease in the Fischer ratio was seen in cirrhosis, acute and chronic hepatitis and hepatocellular damage and continued to decrease as Child-pugh classification increased, suggesting more severe liver disease [9,10]. Considering these important functions, routine clinical monitoring and evaluating of BCAAs and AAAs levels is necessary.
The methods for determining amino acid content in blood include amino acid analyzer [11], capillary electrophoresis [12], gas chromatography [13], high-performance liquid chromatography and liquid chromatography-tandem mass spectrometry (LC-MS/MS)[14,15]. Among them, the LC-MS/MS method was considered to be the best platform for amino acid determination based on the advantages of high selectivity, high specificity and high sensitivity. However, most of the reported LC-MS/MS methods, which were used for neonatal genetic metabolic disease screening or metabolomic analysis [16–18]. Although a large variety of amino acids were determined, the long run time and the lack of a dedicated method for the BCAAs and AAAs determination were not suitable for routine clinical high-throughput assays. In addition, some methods still require derivatization during sample preparation and the complexity of sample processing is also a limitation to the throughput of clinical applications [19]. To be applicable to routine analysis in a clinical setting, the development of a simpler and faster analytical procedure for simultaneous determination of six BCAAs and AAAs is necessary. Meanwhile, there is a lack of corresponding reference range data for the Chinese population. Reference intervals play an important role in guiding the interpretation of laboratory results and are the first step in clinical interpretation. Although there are some relevant studies based on foreign populations, direct references could lead to misjudgment of clinical diagnosis, treatment effect and prognosis due to the reference intervals may vary according to race, region, dietary habits and laboratory methods [20,21]. Thus, it is absolutely urgent to establish BCAAs and AAAs reference intervals for the Chinese population matching clinical laboratory method.
The aim of this study was to develop and validate a simple and rapid LC-MS/MS method for simultaneous detection of plasma Val, Leu, Ile, Phe, Trp and Tyr levels and establish healthy Chinese adults reference intervals for all analytes, Fischer ratio and BTR being based on a multi-center study. Furthermore, this method was used to measure BCAAs and AAAs levels in patients with liver injury, as well as further exploring the potential role of BCAAs and AAAs in the indication of liver injury.
2. Materials & methods
2.1. Study subjects & sample collection
This study prospectively recruited study subjects who came to the First Affiliated Hospital of Jinzhou Medical University and the Dalian Municipal Central Hospital Physical Examination Center from May 2023 to December 2023. Participants were asked to complete a questionnaire to collect basic information. Inclusion criteria were as follows: over 18 years old, healthy, non-alcoholic, non-smoker, no fever within 2 weeks, and no history of medication within 1 week. Those with acute and chronic infections, neoplasms, obesity (body mass index (BMI) ≥28 kg/m2), severe malnutrition (BMI <18.5 kg/m2), diabetes mellitus, cardiovascular and cerebral vascular diseases, immune system disorders and hepatic diseases were excluded. Ultimately, 492 apparently healthy adults were recruited for the establishment of reference intervals.
In addition, patients with a definite diagnosis of liver injury were collected based on clinical data of basic information about the patients and the results of liver function tests, including alanine aminotransferase (ALT), aspartate aminotransferase (AST) and albumin (ALB) levels. Meanwhile, those with the following diseases were excluded, including hyperthyroidism, diabetes mellitus and pneumonia. Eventually, the clinical residual fresh blood from 68 patients (ages range 30–84 years) with liver injury was used.
The study was conducted in accordance with the World Medical Organization Declaration of Helsinki and approved by the Ethics Committee of the First Affiliated Hospital of Jinzhou Medical University (KYLL2023157) and the Dalian Municipal Central Hospital (2023-042-10). This study was exempted from requiring informed consent from the subjects by the Institutional Review Board as it only used the residual fresh serum samples (samples remaining after routine clinical testing).
2.2. Standard solutions & reagents
Val, (purity: ≥99%), Ile, (purity: ≥99%), Leu, (purity: ≥99%) and Phe, (purity: ≥99%) were purchased from the National Institute of Metrology (Beijing, China). Tyr, (purity: ≥99%) was purchased from Hai'an Hongmeng Standard Substance Technology Co. Ltd (Beijing, China). Trp, (purity: ≥99%) was purchased from the South China National Centre of Metrology (Guangdong, China). Isotopic-labeled internal standard (IS) included tyrosine-d4 (Tyr-d4, purity: 99%), leucine-d3 (Leu-d3, purity: 99%), tryptophan-d5 (Trp-d5, purity: ≥99%) were purchased from BePure (Beijing, China). Standard reference material 1950 (SRM 1950) human frozen plasma was purchased from the National Institute of Standards and Technology (NIST, Gaithersburg, MD). The remaining reagents acetonitrile (ACN) (Fisher Scientific, Far Lawn, NJ, USA), formic acid (FA) (Aladdin, Shanghai, China), hydrochloric acid (HCl) (Kermel, Tianjin, China), ammonium acetate (Sigma-Aldrich, St. Louis, MO, USA) and purified water (Watsons, Hong Kong, China) were high-performance liquid chromatography (HPLC) grade or higher.
2.3. Preparation of stock solutions, calibration standards, & quality control samples
Based on the solubility of BCAAs and AAAs, 0.1 M HCl aqueous solution was chosen as diluent. Stock solutions were prepared by dissolving individually BCAAs, AAAs and ISs powder and stored at −80°C until used. The stock solutions of analytes were mixed as a mixed standard solution, then they were gradually diluted into a series of concentrations as calibration standards and quality control (QCL, quality control low; QCM, quality control medium; QCH, quality control high) working solutions. The working solution of IS was prepared by appropriate dilution of the prepared IS stock solutions. The concentrations of calibration standard and QC working solutions are shown in Supplementary Table S1.
2.4. LC-MS/MS analysis
The serum BCAAs and AAAs levels were analyzed by AB SCIEX Triple Quad™ 4500MD LC-MS/MS system (AB SCIEX, Toronto, Canada). Chromatographic separation was achieved using an Atlantis™ Premier BEH Z-HILIC column (2.1 mm × 100 mm with 2.5 μm particles, Waters) at 40°C. The mobile phase A consisted of 10 mM ammonium acetate and 0.2% FA in 20% ACN aqueous solution and the mobile phase B consisted of 1 mM ammonium acetate and 0.2% FA in ACN. The elution conditions applied were 0–4 min, 76% mobile phase B isocratic at 0.3 ml/min flow rate. The injection volume was set at 1 μl. The mass spectrometer was operated in positive ion mode with an electrospray ion source, using multiple reaction monitoring (MRM). The specific finishing parameters were set as follows: nebulizer gas and heater gas, 50 psi; curtain gas, 35 psi; collision gas, 9 psi; ion spray voltage, 5500 V; heated nebulizer temperature, 500°C. The MRM transitions and acquisition parameters for each analyte and Isotopic-labeled IS are detailed in Table 1. The structures of BCAAs and AAAs and fragmentation products are shown in Supplementary Figure S1. Data acquisition and quantification were performed using Analyst software version MD 1.6.3 and MultiQuant software version MD 3.0.2.
Table 1.
The specific parameters, linearity (n = 3) and LLOQ (n = 6) results for each analyte.
| Analyte | MRM transition (m/z) | Declustering potential (V) | Collision energy (V) | Linearity (μmol/l) | r2 | LLOQ, (μmol/l) |
|---|---|---|---|---|---|---|
| Val | 118.1 >72.1 | 50 | 30 | 20.0–4000.0 | 0.9972 | 20.0 |
| Ile | 132.1 >69.1 | 58 | 25 | 5.0–1000.0 | 0.9978 | 5.0 |
| Leu | 132.1 >43.1 | 58 | 34 | 10.0–2000.0 | 0.9969 | 10.0 |
| Phe | 166.1 >120.0 | 56 | 38 | 5.0–1000.0 | 0.9978 | 5.0 |
| Tyr | 182.1 >136.1 | 50 | 20 | 5.0–1000.0 | 0.9982 | 5.0 |
| Trp | 205.1 >188.0 | 67 | 18 | 2.5–500.0 | 0.9982 | 2.5 |
| Tyr-d4 | 186.1 >140.1 | 50 | 20 | NA | NA | NA |
| Leu-d3 | 135.1 >46.1 | 38 | 34 | NA | NA | NA |
| Trp-d5 | 210.1 >193.0 | 67 | 18 | NA | NA | NA |
Ile: Isoleucine; Leu: Leucine; Leu-d3: Leucine-d3; LLOQ: The lower limit of quantitation; MRM: Multiple reaction monitoring; NA: Not available; Phe: Phenylalanine; Trp: Tryptophan; Tyr-d4: Tyrosine-d4; Trp-d5: Tryptophan-d5; Tyr: Tyrosine; Val: Valine.
2.5. Sample preparation
Fasting blood samples were collected after an overnight fast into vacuum blood collection tubes containing a gel for serum separation and processed within 2 h of collection by centrifugation at 3000 rpm for 5 min at room temperature. 10 μl serum samples, 10 μl IS working solutions and 150 μl precipitant solution were precisely added to 1.5 ml microcentrifuge tube. The mixtures were vortexed for 5 min and centrifuged at 15,000 × g at 4°C for 5 min. Subsequently, 50 μl of the supernatant was diluted with 50 μl ACN and 50 μl of this solution was transferred to the autosampler mini vial for LC-MS/MS analysis.
2.6. Method validation
To meet the requirements of each evaluation criterion, this LC-MS/MS method was validated according to Clinical and Laboratory Standards Institute guideline C62-A (CLSI, Liquid Chromatography-Mass Spectrometry Methods; Approved Guideline)[22] and the Consensus of Method Development and Validation of Liquid Chromatography-Tandem Mass Spectrometry in Clinical Laboratories [23]. The validation procedure involves the evaluation of specificity and selectivity, linearity, lower limit of quantitation (LLOQ), matrix effect, carry-over, precision, accuracy, interference, stability and trueness.
2.6.1. Specificity & selectivity
To assess selectivity and specificity, six double blank (no analyte, no IS), zero concentration (no analyte) and LLOQ samples were prepared and analyzed in parallel. The background noise meets one of the following conditions: no background peak or background peak area less than 20% of the analyte peak area at LLOQ; less than or equal to 5% of the IS peak area at the expected retention time.
2.6.2. Linearity & LLOQ
Linearity was evaluated by analyzing the calibration curve of nine concentrations three-times on three non-consecutive days. The analyte peak area/IS peak area ratio versus nominal analyte concentration (x) was calculated. The linear regression was performed using a weighting factor of 1/x2, and correlation coefficients r2 >0.995 were considered acceptable. The LLOQ was defined as the lowest concentration of an analyte that can be accurately quantified by the assay under specified experimental conditions, provided that the laboratory's requirements for accuracy and precision are met. The signal-to-noise ratio for LLOQ should be at least 20:1.
2.6.3. Matrix effects
The relative matrix effect was evaluated according to matrix admixing experiments. Six serum samples from different people (Including three cases of normal serum; one case each of severe hemolysis, severe jaundice and severe lipemia) were taken as the native matrix. QCM samples were mixed with pure water as a pure solution matrix. The above two matrices were further mixed in the ratio of 3:7 and 7:3 as mixed matrix. Each matrix sample was processed three-times in parallel. The difference between the response values of the mixed matrix samples compared with the mean of the response values of the native matrix samples and the pure solution matrix samples should be less than 20% [24].
2.6.4. Carryover
Carryover was assessed by selecting high (H, the highest concentration point on the calibration curves) and Low (L, the lowest concentration point on the calibration curves) concentration samples and injecting the samples in different combinations. Repeat injection of L-L concentration samples and alternating H-L concentration samples five-times each. The criterion that the difference in concentrations between H-L and L-L converted samples is less than three-times the standard deviations (3SD) of the L-L converted samples should be met.
2.6.5. Precision & accuracy
Pooled human serum samples spiked with known concentrations of LLOQ, QCL, QCM and QCH to assess precision (within run and between runs) and accuracy. Six samples of each concentration in each batch were processed in parallel for three consecutive batches. The within-run and between-run precision was expressed by a coefficient of variation (CV). The accuracy was expressed using the recovery: Recovery (%) = [determined concentration / (endogenous concentration + spiked concentration)] × 100. Recovery within the ranges 85–115% (LLOQ: 80–120%), and CV below 15% (LLOQ: <20%) were considered acceptable.
2.6.6. Interference
Potential serum interferences including severe hemolysis, lipemia and icterus were evaluated by spiking known concentrations of endogenous substances (100 mg/ml for hemoglobin, 740 mmol/l for triglyceride, 6.84 mmol/l conjugated and unconjugated bilirubin) into low and high concentration pooled serum samples. The relative error (RE) of different interference types samples from normal serum samples should be within ±15%.
2.6.7. Sample stability
The stability study of serum samples containing short-term (room temperature and 2–8°C), long-term (−20°C and −80°C), freeze-thaw and processed samples (autosampler stability) could evaluate the influence of sample handling environment and sample storage environment on the concentration of the measured analytes. Three replicates were assayed at pooled serum samples with low and high concentrations at each storage condition. The RE% of the measured concentration should be within the acceptable ±15% range.
2.6.8. Trueness
The trueness of Val, Ile, Leu, Phe and Tyr were evaluated by measuring the NIST, SRM 1950. The criterion is that the bias of the result from the target value is less than the assigned uncertainty or less than 1/2 TEa (biological variation – established by the European Federation of Clinical Chemistry and Laboratory Medicine (EFLM) [25]).
2.7. Establishment of reference intervals & statistical analysis
The establishment of reference intervals procedure was carried out by CLSI EP28-A3 “Define and determine the reference intervals in clinical laboratory”[26]. Statistical analysis was performed using SPSS Statistics version 20.0 (IBM Inc, NY, USA) and GraphPad Prism software version 5.0 (Graph Pad Software Inc, CA, USA). The specific analysis process was as follows: Kolmogorov-Smirnov was used for normality test and p-value < 0.05 was considered statistically significant. Tukey and Dixon methods were used to remove outliers, and Z score were used to assess the need to divide the reference intervals by sex. Non-parametric method was recommended to determine the upper and lower limits of the reference intervals by setting the bilateral percentile values (2.5% and 97.5%). Data was presented as medians (P25, P75) for skewed distribution and mean ± standard deviation (SD) for normal distribution. To investigate the correlation of other factors (ALT, AST and ALB) with BCAAs and AAAs, Spearman's analysis was used.
3. Results
3.1. Optimization of LC-MS/MS conditions
To efficiently separate all analytes in a short time, especially for the separation of the isomers Leu and Ile, chromatographic conditions of the mobile phase and the chromatographic column were optimized in this study. With 0.2% FA ACN aqueous solution and 0.2% FA ACN as mobile phases A and B, respectively, we performed the screening of Waters HSS T3 column (2.1 mm × 100 mm, 2.5 μm) and Waters Atlantis™ Premier BEH Z-HILIC column (2.1 mm × 100 mm, 2.5 μm). Waters Atlantis™ Premier BEH Z-HILIC column (2.1 mm × 100 mm, 2.5 μm) had a better separation effect, and baseline separation was almost achieved for the isomers Leu (2.07 min) and Ile (2.30 min). Further, to improve the chromatographic peak shape, 10 mM and 1 mM ammonium acetate were added to mobile phase A and mobile phase B, respectively. Duo to the analysis was performed in positive ion mods, the stable buffer system is formed between ammonium acetate and FA to promote ionization. Meanwhile, keeping the pH in a stable state, which improves the column efficiency and the chromatographic peak shape. In addition, ammonium acetate has good solubility and less damage to the column. As shown in Figure 1, the satisfactory peak shape and separation were achieved under 4 min total run time.
Figure 1.
Representative multiple reaction monitoring (MRM) chromatograms for BCAAs and AAAs.
3.2. Methods validation
3.2.1. Selectivity & specificity
The response averages for all analytes and IS were calculated. The analyte interference responses in double blank and zero concentration samples were 0.46–18.49%; the IS responses were ≤0.31% (Supplementary Table S2). The results indicated satisfactory selectivity and specificity of the method.
3.2.2. Linearity & LLOQ
As listed in Table 1, the linearity ranges were 20–4000 μmol/l for Val, 5–1000 μmol/l for Ile, Phe and Tyr, 10–2000 μmol/l for Leu, 2.5–500 μmol/l for Trp. Each analyte had a good linear relationship in its linear range, and the r2 was greater than 0.9969. The LLOQs of Val, Ile, Leu, Phe, Tyr and Trp were 20, 5, 10, 5, 2.5 and 5 μmol/l, respectively at signal-to-noise ratio ≥20.
3.2.3. Matrix effects
The RE% of relative matrix effects for each analyte ranged from −18.52 to 5.54%, with CV% <5.2%, when corrected by isotopic-labeled IS. The acceptable matrix effect suggested that no significant ion suppression or enhancement effects in serum matrix and was reliable for bioanalysis (Supplementary Table S3).
3.2.4. Carryover
For all analytes, carryover was not observed as the difference in concentrations between H-L and L-L converted samples was less than 3SD of the L-L converted samples (Supplementary Table S4).
3.2.5. Precision & accuracy
The CV% of within-run and between-run precision ranged from 0.10 to 5.90%, and the recovery was 94.44–107.75% as in Table 2, which demonstrated that this method was accurate, reliable and reproducible.
Table 2.
Within-run (n = 6) and between-run (n = 18) precision and accuracy (n = 6) of the method.
| Analyte | Spiked sample | Spiked concentration | Determined concentration (mean ± SD) | Within-run precision, CV, % | Between-run precision, CV, % | Recovery, % |
|---|---|---|---|---|---|---|
| Val | NA | 0.00 | 255.61 ± 3.46 | NA | NA | NA |
| LLOQ | 20.00 | 272.06 ± 10.76 | 3.80 | 2.00 | 98.71 | |
| QCL | 50.00 | 312.32 ± 6.02 | 1.70 | 1.30 | 102.19 | |
| QCM | 400.00 | 644.29 ± 35.63 | 5.90 | 0.70 | 98.25 | |
| QCH | 3200.00 | 3718.36 ± 88.66 | 2.50 | 0.70 | 107.61 | |
| Ile | NA | 0.00 | 77.72 ± 0.26 | NA | NA | NA |
| LLOQ | 5.00 | 82.90 ± 3.87 | 5.00 | 0.30 | 100.21 | |
| QCL | 12.50 | 95.03 ± 2.98 | 1.80 | 3.20 | 105.33 | |
| QCM | 100.00 | 177.09 ± 9.19 | 5.50 | 0.40 | 99.64 | |
| QCH | 800.00 | 938.92 ± 24.62 | 2.70 | 0.80 | 106.97 | |
| Leu | NA | 0.00 | 153.58 ± 0.93 | NA | NA | NA |
| LLOQ | 10.00 | 160.20 ± 4.32 | 2.70 | 1.10 | 97.92 | |
| QCL | 25.00 | 177.61 ± 4.14 | 2.30 | 1.10 | 99.44 | |
| QCM | 200.00 | 341.05 ± 12.93 | 4.00 | 0.10 | 96.46 | |
| QCH | 1600.00 | 1787.30 ± 47.04 | 2.70 | 0.70 | 101.93 | |
| Phe | NA | 0.00 | 67.17 ± 1.20 | NA | NA | NA |
| LLOQ | 5.00 | 70.95 ± 3.24 | 3.80 | 3.40 | 98.33 | |
| QCL | 12.50 | 81.16 ± 1.10 | 1.40 | 0.40 | 101.88 | |
| QCM | 100.00 | 157.94 ± 8.08 | 4.80 | 3.00 | 94.44 | |
| QCH | 800.00 | 934.33 ± 24.60 | 2.80 | 0.30 | 107.75 | |
| Tyr | NA | 0.00 | 63.47 ± 0.35 | NA | NA | NA |
| LLOQ | 5.00 | 68.05 ± 2.11 | 3.30 | 0.30 | 99.41 | |
| QCL | 12.50 | 75.12 ± 2.33 | 3.00 | 1.60 | 98.88 | |
| QCM | 100.00 | 158.81 ± 6.15 | 3.80 | 1.80 | 97.13 | |
| QCH | 800.00 | 883.13 ± 33.18 | 3.10 | 2.80 | 102.28 | |
| Trp | NA | 0.00 | 49.48 ± 0.48 | NA | NA | NA |
| LLOQ | 2.50 | 50.85 ± 0.73 | 1.50 | 0.20 | 97.80 | |
| QCL | 6.25 | 55.42 ± 1.76 | 3.00 | 1.80 | 99.44 | |
| QCM | 50.00 | 96.03 ± 3.15 | 3.20 | 1.60 | 96.53 | |
| QCH | 400.00 | 463.44 ± 14.05 | 3.20 | 0.50 | 103.11 |
The unit of concentration is μmol/l.
Ile: Isoleucine; Leu: Leucine; LLOQ: The lower limit of quantitation; NA: Not available; Phe: Phenylalanine; QCH: Quality control high; QCL: Quality control low; QCM: Quality control medium; SD: Standard deviation; Trp: Tryptophan; Tyr: Tyrosine; Val: Valine.
3.2.6. Interference
As shown Supplementary Table S5, the RE% between samples of different interference types and normal serum samples within −1.8% – 9.6% for all analytes. This result indicated that severe hemolysis, lipemia and icterus did not affect the determination of BCAAs and AAAs concentrations in serum samples.
3.2.7. Sample stability
As displayed in Supplementary Table S6, the RE% were all within −11.0–12.3%. The BCAAs and AAAs in serum samples could be stably stored for 3 days at room temperature, for a week at 2–8 °C, for 20 days at −20°C, for 31 days at −70°C, and during three freeze-thaw cycles. After preparation, all analytes were stable for 48 h in autosampler at 4°C.
3.2.8. Trueness
Results for trueness are summarized in Supplementary Table S7. For Val, Phe and Tyr, the bias between the measured results and the target value ranged from −5.1% to 3.8%, which all were below the assignment uncertainty. For Leu and Ile, although the deviation between the results and the target values was greater than the assignment uncertainty, it is less than 1/2 TEa.
3.3. Establishment of reference intervals
A total of 492 healthy adults (203 males, 289 females) aged between 22 and 73 years were included in the analysis. Supplementary Table S8 summarized the descriptive characteristics of the study subjects. Compared with females, males had higher levels of BMI, ALB, ALT, AST, triglycerides (TG), low-density lipoprotein cholesterol (LDL-C) and glucose (GLU). Conversely, high-density lipoprotein cholesterol (HDL-C) was higher in females than in males. Even so, these metrics were within normal limits, which showed that the nutritional status, blood glucose, blood lipids and liver function of the study subjects were appropriate and the reference intervals developed in this way could accurately reflect the levels of BCAAs and AAAs in healthy adults. The Kolmogorov-Smirnov test showed non-normal distributions for serum BCAAs and AAAs levels except for Trp which was normally distributed. After removing outliers within sex-specific groups, the minimum sample sizes of each group achieved the expected 95% confidence interval (95% CI). The reference intervals for Val, Ile, Leu, Phe, Tyr, Trp, Fischer ratio and BTR were 168.27–354.85, 39.24–105.42, 85.46–224.26, 56.68–127.16, 37.94–90.44, 2.31–4.21 and 5.47–10.79 μmol/l, respectively. Table 3 presented the comparison of reference intervals of the current study with other studies in Mayo Clinic Laboratories [27], Quest Diagnostics [28] and LabCorp [29] established by LC-MS/MS method. These results showed the reference intervals established in this study tending to be consistent with those published.
Table 3.
Reference interval comparisons of BCAAs, AAAs, Fischer ratio and BTR in this study and published results.
| Analyte | This study | Mayo Clinic | Quest | LabCorp |
|---|---|---|---|---|
| Val | 168.27–354.85 | 136–309 | 132–313 | 133.0–317.1 |
| Ile | 39.24–105.42 | 36–107 | 34–98 | 32.8–88.3 |
| Leu | 85.46–224.26 | 68–183 | 73–182 | 32.8–88.3 |
| Phe | 56.68–127.16 | 35–80 | 40–74 | 35.8–76.9 |
| Tyr | 37.94–90.44 | 29–77 | 38–96 | 27.8–83.3 |
| Trp | 34.86–76.03 | 31–90 | 40–91 | 23.5–93.0 |
| Fischer ratio | 2.31–4.21 | NA | NA | NA |
| BTR | 5.47–10.79 | NA | NA | NA |
BTR: BCAAs/Tyr molar concentration ratio; Ile: Isoleucine; Leu: Leucine; NA: Not available; Phe: Phenylalanine; Tyr: Tyrosine; Trp: Tryptophan; Val: Valine.
3.4. Clinical application
3.4.1. Application to patients with liver injury
This method was applied to detect serum BCAAs and AAAs levels in patients with liver injury to validate the clinical applicability. The healthy control group consisted of 49 healthy people, age- and gender-matched to the liver injury group. Serum levels of BCAAs, AAAs and the Fischer ratio in patients with liver injury group and healthy control group were listed in Supplementary Table S9. As shown in Figure 2, the liver injury patients group had higher levels of Ile, Phe, Tyr, Val and Leu (p < 0.001, p < 0.001, p < 0.001, p = 0.015 and p = 0.013, respectively) and lower levels of the Fischer ratio (p = 0.013) compared with healthy control group. However, serum Trp level did not significantly difference between the liver injury patients and healthy control group (p = 0.171). In order to further observe the differences between the two groups, the supervised pattern recognition method OPLS-DA was used for analysis. The OPLS-DA score chart is shown in Supplementary Figure S2, and the two groups could be distinguished significantly. The results shown that the established model has good accuracy and predictability. In addition, the percentage of each indicator outside the reference intervals were Val (16%), Ile (21%), Leu (12%), Phe (32%), Trp (29%) and the Fischer ratio (35%).
Figure 2.
The serum BCAAs, AAAs and Fischer ratio levels in patients with liver injury and healthy controls.
3.4.2. Correlation of BCAAs & AAAs with liver enzymes
Since BCAAs and AAAs are closely associated with liver disease, we examined their correlation with liver enzymes (including ALT, AST, AST/ALT and ALB) for further explanation. The results of the correlation are shown in Table 4. First, ALT, as one of the enzymes related to liver function, had a variable effect on BCAAs and AAAs in healthy people. The correlations are as follows (from highest to lowest): Leu (r = 0.31), Phe (r = 0.308), Ile (r = 0.298), Tyr (r = 0.282), Val (r = 0.269), Trp (r = 0.229) (p < 0.001 in all). Second, our results have shown that ALB instead plays a very important role in the liver injury group. Tyr (r = -0.261) and Phe (r = -0.443) were negatively correlated with it, while the rest were significantly positively correlated, especially the Fischer ratio (r = 0.77, p < 0.001). In addition, the present study has also shown that age has no effect on the reference intervals of BCAAs and AAAs. Surprisingly, in the healthy group, the Fischer ratio was not affected by liver enzymes (|r| <0.2 in all cases), but in liver injury group, the results were completely opposite. The correlation coefficients are as follows: ALT (|r |= 0.214), AST (|r| = 0.486), ALB (|r| = 0.77).
Table 4.
Influence of liver enzymes on BCAAs and AAAs in healthy and liver injury patients groups.
| Healthy people | Liver injury patients | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Analyte | Age (years) | ALT (U/l) | AST (U/l) | ALB (g/l) | Age (years) | ALT (U/l) | AST (U/l) | ALB (g/l) | |
| Trp (μmol/l) | r | 0.077 | 0.229a | 0.119 | 0.184 | -0.352a | -0.021 | -0.210 | 0.594b |
| P | 0.172 | <0.001 | 0.034 | 0.001 | 0.003 | 0.867 | 0.086 | <0.001 | |
| Tyr (μmol/l) | r | 0.198 | 0.282a | 0.154 | -0.038 | 0.098 | 0.231 | 0.333a | -0.261a |
| P | < 0.001 | <0.001 | 0.006 | 0.503 | 0.427 | 0.058 | 0.005 | 0.032 | |
| Phe (μmol/l) | r | 0.254a | 0.308a | 0.188 | 0.068 | 0.235 | 0.265a | 0.389a | -0.443b |
| P | < 0.001 | <0.001 | 0.001 | 0.237 | 0.054 | 0.029 | 0.001 | <0.001 | |
| Ile (μmol/l) | r | 0.050 | 0.298a | 0.124 | 0.079 | -0.328a | 0.019 | -0.128 | 0.387a |
| P | 0.379 | <0.001 | 0.027 | 0.163 | 0.006 | 0.877 | 0.299 | 0.001 | |
| Leu (μmol/l) | r | 0.171 | 0.310a | 0.144 | 0.148 | -0.182 | 0.050 | -0.186 | 0.384a |
| P | 0.002 | <0.001 | 0.010 | 0.008 | 0.137 | 0.686 | 0.128 | 0.001 | |
| Val (μmol/l) | r | 0.065 | 0.269a | 0.126 | 0.138 | -0.208 | -0.038 | -0.233 | 0.500b |
| P | 0.248 | <0.001 | 0.025 | 0.014 | 0.089 | 0.757 | 0.056 | <0.001 | |
| Fischer ratio | r | -0.185 | 0.049 | -0.013 | 0.159 | -0.403b | -0.214 | -0.486b | 0.770c |
| P | 0.001 | 0.384 | 0.823 | 0.005 | 0.001 | 0.080 | <0.001 | <0.001 | |
indicates 0.2≤|r|<0.4
indicates 0.4≤|r|<0.6
indicates 0.6≤|r|
The listed values are standardized partial regression coefficients (r). |r| = 0.2 and p ≤ 0.05 was considered as a minimum effect size of practical significance.
AAA: Aromatic amino acid; ALB: Albumin; ALT: Alanine transaminase; AST: Aspartate transaminase; BCAA: Branched-chain amino acid; Ile: Isoleucine; Leu: Leucine; Phe: Phenylalanine; Trp: Tryptophan; Tyr: Tyrosine; Val: Valine.
4. Discussion
Previously, a study on measurement of serum BCAAs and AAAs by LC-MS/MS method has been published by Yang et al. [30]. In comparing this method with the analytical method reported previously based on the parameters, including types of analytes, sample type, sample volume, sample preparation procedure, detection system and total run time. Although the two methods are similar or less different in terms of sample type, sample preparation procedure and total run time, the sample volume used in our method is small, only 10 μl. Meanwhile, in the comparison of types of analytes, our method has the advantage of detection indicators for BCAAs and AAAs was more comprehensive. In terms of detection system, the method reported by Yang et al. used the Agilent 1200 series LC system and API4000 triple quadrupole MS system, whereas in our method, the clinically applicable AB SCIEX Triple Quad 4500MD LC-MS/MS system with the medical device registration certificate was employed. For IS selection, with the extensive development of LC-MS/MS quantitative detection techniques, the stable isotope-labeled IS has been recognized as an excellent choice in the field of LC-MS/MS quantitative analysis. Because stable isotope compounds and analytes have almost exactly the same molecular structure, chemical properties, chromatographic and MS behavior, ionization changes and matrix effects can be effectively eliminated. In this study, the Leu-d3, Tyr-d4 and Trp-d5 were selected as an isotope-labeled IS, which was proved to be feasible by the results of method verification. Five isotope IS (Val-D8, Ile-D10, Leu-D3, Tyr-D4, Phe-D5) were used in the published method. In contrast, our method could achieve clinical cost savings.
Reporting of the healthy adults reference intervals for BCAAs and AAAs widely varied among different studies [20,21], which could be due to differences in variable adjustments or differences in analyses and populations [31]. However, the reference intervals in this study agree well with those published by Mayo Clinic Laboratories, Quest Diagnostics and LabCorp, indicating that this method could accurately and reliably reflect the levels of BCAAs and AAAs in healthy adults. Otherwise, the reference intervals in this study were comparable with published findings from Japan [32] and Singapore [33]. Nevertheless, the reference intervals for the Fischer ratio and BTR were proposed and established for the first time in this study. To the best of our knowledge, this is the first time that reference intervals have been established in an Asian population following Fischer's proposal in 1972 [1]. All in all, it is necessary for clinical laboratory to establish reference interval matching with detection methods and populations to provide a reliable guarantee for clinical diagnosis.
As expected, the higher levels of Ile, Phe, Tyr, Val and Leu and lower levels of the Fischer ratio were found in the liver injury patients. These results are in accordance with previous studies [34,35]. Since the correlation of the Fischer ratio with liver enzymes was more significant in the liver injury group compared with healthy control group. We hypothesized that the Fischer ratio would play an important role in the indication and treatment of liver disease. Because the liver is the main metabolizing organ for AAAs, the serum concentration increases when liver function is impaired. In contrast, BCAAs are metabolized and provide energy in skeletal muscle, resulting in an imbalance in the Fischer ratio. Indeed, there is growing evidence that the Fischer ratio could be used to assess the severity of liver injury and help to evaluate the prognosis of patients with cirrhosis [10].
In this study, we found a positive correlation between ALT and BCAAs levels in a healthy group, which is almost the same as the results reported previously [36–42]. However, the mechanism and significance of this is not clear. Some basic studies suggest that it may be related to amino acid metabolic pathways and Krüppel-like factor 15 (KLF15) [43–45]. It is well known that BCAAs are converted to branched-chain α-keto acids (BCKA) catalyzed by branched-chain amino transferase (BCAT), together with the removal of α-amino acids [46]. α-Ketoglutarate receives this α-amino acid to produce glutamate and changes to alanine via ALT[47]. Research conducted by Jeyaraj [48] in 2012 on adult mice found that KLF15 upregulated ALT expression in skeletal muscle and found a significant increase in glutamate levels in the supernatant of hepatocyte cultures. Li et al. [49] showed that KLF15 can regulate BCAAs metabolism by mediating ALT. However, the underlying mechanisms of these associations remain to be established.
Interestingly, our study found that ALB substituted for ALT was significantly most associated with BCAAs and AAAs in the liver injury group. A group of Italian researchers found that BCAAs supplementation increased ALB in vivo and suppressed the occurrence of adverse events in a multicenter, randomized clinical trial of 174 patients with cirrhosis [50]. Our study extends these findings. The Supplementary Figure S3 showed the scatter plot of the correlation of ALB with BCAAs, AAAs and the Fischer ratio in liver injury group. As ALB rises, BCAAs (Val, Ile, Leu) levels gradually increase while AAAs (Tyr, Phe) levels gradually decrease, leading to an increase in the Fischer ratio. Although the mechanisms involved are not fully understood, supplements with different the Fischer ratio have been extensively studied and recommended for the treatment of liver disease to improve liver function and increase ALB levels [51].
5. Conclusion
In summary, this study firstly developed and validated a simple, rapid and accurate LC-MS/MS method for simultaneously quantifying serum BCAAs and AAAs levels, as well as established the reference intervals for all analytes, Fischer ratio and BTR in healthy Chinese adults. Furthermore, this method was applied to patients with liver injury, revealing significant increase for serum Val, Ile, Phe, Leu and Tyr levels and decrease for the Fischer ratio in the liver injury patients group. This method could be a powerful tool for clinical diagnosis of liver injury patients and further study of related liver disease mechanisms. Meanwhile, this method could provide a reference for the clinical determination of BCAAs and AAAs.
Supplementary Material
Acknowledgments
The authors would like to acknowledge the support by the High-Level Talents Innovation and Entrepreneurship Projects of the Dalian Science and Technology Bureau (grant number: 2021RT09); and Dalian Key Technology R&D Program (grant number: 2022YF15SN063), as well as the Dalian Municipal Central Hospital for supplying serum samples and clinical information.
Funding Statement
The authors would like to acknowledge the support by the High-Level Talents Innovation and Entrepreneurship Projects of the Dalian Science and Technology Bureau (grant number: 2021RT09); and Dalian Key Technology R&D Program (grant number: 2022YF15SN063), as well as the Dalian Municipal Central Hospital for supplying serum samples and clinical information.
Supplemental material
Supplemental data for this article can be accessed at https://doi.org/10.1080/17576180.2024.2387467
Author contributions
T Wang: Conceptualization, Formal analysis, Writing (original draft). Y Zhang: Methodology, Validation, Writing (original draft). L Jia: Formal analysis. Y: Data curation. L Wang: Investigation. Y Zhu: Investigation. Y Jiang: Supervision. F Zhao: Supervision. S Wang: Conceptualization, Supervision, Methodology, Writing (review & editing). D Song: Conceptualization, Formal analysis, Writing (review & editing).
Financial disclosure
The authors have no financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.
Competing interests disclosure
The authors have no competing interests or relevant affiliations with any organization or entity with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, stock ownership or options and expert testimony.
Writing disclosure
No writing assistance was utilized in the production of this manuscript.
References
Papers of special note have been highlighted as: • of interest; •• of considerable interest
- 1.Fischer JE, Funovics JM, Aguirre A, et al. The role of plasma amino acids in hepatic encephalopathy. Surgery. 1975;78(3):276–290. [PubMed] [Google Scholar]
- 2.Lee A, Jang HB, Ra M, et al. Prediction of future risk of insulin resistance and metabolic syndrome based on Korean boy's metabolite profiling. Obes Res Clin Pract. 2015;9(4):336–345. doi: 10.1016/j.orcp.2014.10.220 [DOI] [PubMed] [Google Scholar]
- 3.Lu Y, Wang Y, Liang X, et al. Serum amino acids in association with prevalent and incident Type 2 diabetes in A Chinese population. Metabolites. 2019;9(1):14. doi: 10.3390/metabo9010014 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Siomkajło M, Daroszewski J. Branched chain amino acids: passive biomarkers or the key to the pathogenesis of cardiometabolic diseases? Adv Clin Exp Med. 2019;28(9):1263–1269. doi: 10.17219/acem/104542 [DOI] [PubMed] [Google Scholar]
- 5.Magnusson M, Lewis GD, Ericson U, et al. A diabetes-predictive amino acid score and future cardiovascular disease. Eur Heart J. 2013;34(26):1982–1989. doi: 10.1093/eurheartj/ehs424 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Moriwaki H, Shiraki M, Fukushima H, et al. Long-term outcome of branched-chain amino acid treatment in patients with liver cirrhosis. Hepatol Res. 2008;38(Suppl. 1):S102–S106. doi: 10.1111/j.1872-034X.2008.00434.x [DOI] [PubMed] [Google Scholar]
- 7.Tricò D, Biancalana E, Solini A. Protein and amino acids in nonalcoholic fatty liver disease. Curr Opin Clin Nutr Metab Care. 2021;24(1):96–101. doi: 10.1097/mco.0000000000000706 [DOI] [PubMed] [Google Scholar]
- 8.Cheng S, Wiklund P, Autio R, et al. Adipose tissue dysfunction and altered systemic amino acid metabolism are associated with non-alcoholic fatty liver disease. PLOS ONE. 2015;10(10):e0138889. doi: 10.1371/journal.pone.0138889 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Dam G, Sørensen M, Buhl M, et al. Muscle metabolism and whole blood amino acid profile in patients with liver disease. Scand J Clin Lab Invest. 2015;75(8):674–680. [PubMed] [Google Scholar]
- 10.Yoshida T, Muto Y, Moriwaki H, et al. Effect of long-term oral supplementation with branched-chain amino acid granules on the prognosis of liver cirrhosis. Gastroenterol Jpn. 1989;24(6):692–698. doi: 10.1007/bf02774169 [DOI] [PubMed] [Google Scholar]; •• The clinical value of BCAAs and AAAs were described.
- 11.Smon A, Cuk V, Brecelj J, et al. Comparison of liquid chromatography with tandem mass spectrometry and ion-exchange chromatography by post-column ninhydrin derivatization for amino acid monitoring. Clin Chim Acta. 2019;495:446–450. doi: 10.1016/j.cca.2019.05.007 [DOI] [PubMed] [Google Scholar]
- 12.Jeong JS, Kim SK, Park SR. Amino acid analysis of dried blood spots for diagnosis of phenylketonuria using capillary electrophoresis-mass spectrometry equipped with a sheathless electrospray ionization interface. Anal Bioanal Chem. 2013;405(25):8063–8072. doi: 10.1007/s00216-013-6999-6 [DOI] [PubMed] [Google Scholar]
- 13.Hanff E, Ruben S, Kreuzer M, et al. Development and validation of GC-MS methods for the comprehensive analysis of amino acids in plasma and urine and applications to the HELLP syndrome and pediatric kidney transplantation: evidence of altered methylation, transamidination, and arginase activity. Amino Acids. 2019;51(3):529–547. doi: 10.1007/s00726-018-02688-w [DOI] [PubMed] [Google Scholar]
- 14.Nakayama A, Imaizumi A, Yoshida H. Methods for absolute quantification of human plasma free amino acids by high-performance liquid chromatography/electrospray ionization mass spectrometry using precolumn derivatization. Methods Mol Biol. 2019;2030:207–218. doi: 10.1007/978-1-4939-9639-1_16 [DOI] [PubMed] [Google Scholar]
- 15.Le A, Ng A, Kwan T, et al. A rapid, sensitive method for quantitative analysis of underivatized amino acids by liquid chromatography-tandem mass spectrometry (LC-MS/MS). J Chromatogr B Analyt Technol Biomed Life Sci. 2014;944:166–174. doi: 10.1016/j.jchromb.2013.11.017 [DOI] [PubMed] [Google Scholar]
- 16.Lehotay DC, Hall P, Lepage J, et al. LC-MS/MS progress in newborn screening. Clin Biochem. 2011;44(1):21–31. doi: 10.1016/j.clinbiochem.2010.08.007 [DOI] [PubMed] [Google Scholar]
- 17.Hosseinkhani S, Arjmand B, Dilmaghani-Marand A, et al. Targeted metabolomics analysis of amino acids and acylcarnitines as risk markers for diabetes by LC-MS/MS technique. Sci Rep. 2022;12(1):8418. doi: 10.1038/s41598-022-11970-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Piri-Moghadam H, Miller A, Pronger D, et al. A rapid LC-MS/MS assay for detection and monitoring of underivatized branched-chain amino acids in maple syrup urine disease. J Mass Spectrom Adv Clin Lab. 2022;24:107–117. doi: 10.1016/j.jmsacl.2022.04.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Harder U, Koletzko B, Peissner W. Quantification of 22 plasma amino acids combining derivatization and ion-pair LC-MS/MS. J Chromatogr B Analyt Technol Biomed Life Sci. 2011;879(7-8):495–504. doi: 10.1016/j.jchromb.2011.01.010 [DOI] [PubMed] [Google Scholar]; •• Current analysis methods of BCAAs and AAAs in blood.
- 20.Casado M, Sierra C, Batllori M, et al. A targeted metabolomic procedure for amino acid analysis in different biological specimens by ultra-high-performance liquid chromatography-tandem mass spectrometry. Metabolomics. 2018;14(6):76. doi: 10.1007/s11306-018-1374-4 [DOI] [PubMed] [Google Scholar]
- 21.Filee R, Schoos R, Boemer F. Evaluation of physiological amino acids profiling by tandem mass spectrometry. JIMD Rep. 2014;13:119–128. doi: 10.1007/8904_2013_265 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.CLSI . Liquid Chromatography-Mass Spectrometry Methods; Approved Guideline, CLSI Document C62-A. Wayne, PA: Clinical and Laboratory Standards Institute; 2014. https://www.clsi.org/standards/products/clinical-chemistry-and-toxicology/documents/c62/ [Google Scholar]; •• Guidelines for the LC-MS/MS method validation.
- 23.Medicine. CMSCCMDAoL . Consensus of method development and validation of liquid chromatography-tandem mass spectrometry in clinical laboratories. Laboratory Medicine. 2019;34:189–196. doi: 10.3969/j.issn.1673-8640.2019.03.001 [DOI] [Google Scholar]; • Guidelines for the LC-MS/MS method validation.
- 24.Cao Y, Zeng J, Huang T, et al. A sensitive, robust and high-throughput isotope dilution LC-MS/MS method for quantifying three folate forms in serum. Bioanalysis. 2023;15(5):249–258. doi: 10.4155/bio-2023-0007 [DOI] [PubMed] [Google Scholar]
- 25.QC W . Desirable biological variation database specifications. 2019. [cited 2024 July 15]. Available from: https://www.westgard.com/biodatabase1.htm
- 26.CLSI . Defining, Establishing, and Verifying Reference Intervals in the Clinical Laboratory. Approved Guideline-Third Edition, EP28-A3c. Wayne, PA: Clinical and Laboratory Standards Institute; 2010. https://clsi.org/standards/products/method-evaluation/documents/ep28/ [Google Scholar]; •• Guidelines for the reference interval establishment.
- 27.MAYO CLINIC LABORATORIES [Internet]. Hong Kong: Mayo Foundation for Medical Education and Research (MFMER); c1998-2024 [cited 2024 June 26]. Available from: https://neurology.testcatalog.org/show/AAQP [Google Scholar]
- 28.Quest Diagnostics [Internet]. Boydton: Quest Diagnostics Incorporated; [updated 2024; cited 2024 June 26]. Available from: https://testdirectory.questdiagnostics.com/test/test-detail/19779/amino-acid-analysis-for-msud-plasma?p=r&q=amino%20acid&cc=MASTER [Google Scholar]
- 29.labcorp [Internet]. San Francisco: Laboratory Corporation of America; [updated 2024; cited 2024 June 26]. Available from: https://www.labcorp.com/tests/700068/amino-acid-profile-quantitative-plasma [Google Scholar]
- 30.Yang R, Dong J, Guo H, et al. Rapid and precise measurement of serum branched-chain and aromatic amino acids by isotope dilution liquid chromatography tandem mass spectrometry. PLOS ONE. 2013;8(12):e81144. doi: 10.1371/journal.pone.0081144 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Bi X, Tey SL, Loo YT, et al. Central adiposity-induced plasma-free amino acid alterations are associated with increased insulin resistance in healthy Singaporean adults. Eur J Clin Nutr. 2017;71(9):1080–1087. doi: 10.1038/ejcn.2017.34 [DOI] [PubMed] [Google Scholar]
- 32.Yamamoto H, Kondo K, Tanaka T, et al. Reference intervals for plasma-free amino acid in a Japanese population. Ann Clin Biochem. 2016;53(Pt 3):357–364. doi: 10.1177/0004563215583360 [DOI] [PubMed] [Google Scholar]
- 33.Tan IK, Gajra B. Plasma and urine amino acid profiles in a healthy adult population of Singapore. Ann Acad Med Singap. 2006;35(7):468–475. [PubMed] [Google Scholar]
- 34.Askanazi J, Carpentier YA, Michelsen CB, et al. Muscle and plasma amino acids following injury. Influence of intercurrent infection. Ann Surg. 1980;192(1):78–85. doi: 10.1097/00000658-198007000-00014 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Jiménez Jiménez FJ, Ortiz Leyba C, Morales Mendez S, et al. Variations in plasma amino acids in septic patients subjected to parenteral nutrition with a high proportion of branched-chain amino acids. Nutrition. 1992;8(4):237–244. [PubMed] [Google Scholar]
- 36.Magkos F, Bradley D, Schweitzer GG, et al. Effect of Roux-en-Y gastric bypass and laparoscopic adjustable gastric banding on branched-chain amino acid metabolism. Diabetes. 2013;62(8):2757–2761. doi: 10.2337/db13-0185 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Kalhan SC, Guo L, Edmison J, et al. Plasma metabolomic profile in nonalcoholic fatty liver disease. Metabolism. 2011;60(3):404–413. doi: 10.1016/j.metabol.2010.03.006 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Batch BC, Shah SH, Newgard CB, et al. Branched chain amino acids are novel biomarkers for discrimination of metabolic wellness. Metabolism. 2013;62(7):961–969. doi: 10.1016/j.metabol.2013.01.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Newgard CB, An J, Bain JR, et al. A branched-chain amino acid-related metabolic signature that differentiates obese and lean humans and contributes to insulin resistance. Cell Metab. 2009;9(4):311–326. doi: 10.1016/j.cmet.2009.02.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Nagata C, Nakamura K, Wada K, et al. Branched-chain amino acid intake and the risk of diabetes in a Japanese community: the Takayama study. Am J Epidemiol. 2013;178(8):1226–1232. doi: 10.1093/aje/kwt112 [DOI] [PubMed] [Google Scholar]
- 41.Wang TJ, Larson MG, Vasan RS, et al. Metabolite profiles and the risk of developing diabetes. Nat Med. 2011;17(4):448–453. doi: 10.1038/nm.2307 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Adeva MM, Calviño J, Souto G, et al. Insulin resistance and the metabolism of branched-chain amino acids in humans. Amino Acids. 2012;43(1):171–181. doi: 10.1007/s00726-011-1088-7 [DOI] [PubMed] [Google Scholar]
- 43.Platell C, Kong SE, McCauley R, et al. Branched-chain amino acids. J Gastroenterol Hepatol. 2000;15(7):706–717. doi: 10.1046/j.1440-1746.2000.02205.x [DOI] [PubMed] [Google Scholar]
- 44.Sun H, Olson KC, Gao C, et al. Catabolic Defect of Branched-Chain Amino Acids Promotes Heart Failure. Circulation. 2016;133(21):2038–2049. doi: 10.1161/circulationaha.115.020226 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Fan L, Hsieh PN, Sweet DR, et al. Krüppel-like factor 15: regulator of BCAA metabolism and circadian protein rhythmicity. Pharmacol Res. 2018;130:123–126. doi: 10.1016/j.phrs.2017.12.018 [DOI] [PubMed] [Google Scholar]
- 46.Felig P. The glucose-alanine cycle. Metabolism. 1973;22(2):179–207. doi: 10.1016/0026-0495(73)90269-2 [DOI] [PubMed] [Google Scholar]
- 47.Perry RJ, Wang Y, Cline GW, et al. Leptin mediates a glucose-fatty acid cycle to maintain glucose homeostasis in starvation. Cell. 2018;172(1-2):234–248.e17. doi: 10.1016/j.cell.2017.12.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Jeyaraj D, Scheer FA, Ripperger JA, et al. Klf15 orchestrates circadian nitrogen homeostasis. Cell Metab. 2012;15(3):311–323. doi: 10.1016/j.cmet.2012.01.020 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49.Li H, An X, Bao L, et al. MiR-125a-3p-KLF15-BCAA regulates the skeletal muscle branched-chain amino acid metabolism in Nile Tilapia (Oreochromis niloticus) during starvation. Front Genet. 2020;11:852. doi: 10.3389/fgene.2020.00852 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50.Yatsuhashi H, Ohnishi Y, Nakayama S, et al. Anti-hypoalbuminemic effect of branched-chain amino acid granules in patients with liver cirrhosis is independent of dietary energy and protein intake. Hepatol Res. 2011;41(11):1027–1035. doi: 10.1111/j.1872-034X.2011.00864.x [DOI] [PubMed] [Google Scholar]
- 51.Plauth M, Cabré E, Riggio O, et al. ESPEN Guidelines on Enteral Nutrition: liver disease. Clin Nutr. 2006;25(2):285–294. doi: 10.1016/j.clnu.2006.01.018 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.