Metabolomic profiling of adenine-induced CKD: pathway interconnections and kidney injury

Abstract

Chronic kidney disease (CKD) is acknowledged as one of the largest public health problems in the world, characterized by a complex and diverse pathogenesis. Adenine-induced CKD, a classical model with multiple injury mechanisms, has been extensively employed in CKD research. However, the complete elucidation of the mechanisms underlying adenine-induced CKD remains elusive. In this study, the impacts of adenine (200 mg/kg/day) intake on the urine metabolome of rats were initially investigated using non-targeted metabolomics, and then targeted metabolomics was used to quantitatively verify key metabolites on crucial metabolic pathways. Interestingly, the interconnectedness of two significant pathways was discovered and validated through molecular biology techniques. The results found that adenine can cause significant perturbations in purine metabolism and the biosynthetic pathways of phenylalanine, tyrosine, and tryptophan. Subsequent targeted metabolomic analysis revealed a significant reduction in amino acid and hypoxanthine and creatinine levels in the kidneys of CKD rats, accompanied by an increase in xanthine level. Further analysis found that purine pathway can increase ROS production and affect the level of aromatic amino acid transporter SLC7A5, thus influencing the biosynthesis pathway of phenylalanine, tyrosine and tryptophan, ultimately contributing to kidney injury. This discovery provides offers novel insights into the underlying pathological mechanism of adenine-induced CKD. The development of chronic kidney disease is induced by multiple pathways of aromatic amino acid metabolism and purine metabolism.

Graphical Abstract

Graphical Abstract.

Introduction

The kidneys play a crucial role in regulating fluid and acid–base equilibrium through the excretion of metabolic waste products.¹ Due to population ageing and lifestyle changes, the incidence of chronic kidney disease (CKD) is increasing every year, presenting considerable challenges to healthcare systems and exerting a significant impact on both individual well-being and overall mortality.² CKD is characterized by progressive and irreversible nephron loss, decreased glomerular filtration rate (GFR), uremia, proteinuria, renal fibrosis, and ultimately organ failure, which lead to inflammation, oxidative stress, and metabolic changes.^3,4 To enhance comprehension of CKD and advance strategies for prevention and management, researchers have devised numerous animal models to investigate the disease’s underlying pathophysiology. Among these models, the 5/6 nephrectomy model is often used to study CKD, as it exhibits the hallmarks of human CKD, such as uremia, fibrosis, sparse capillaries, and progressive renal decline.⁵ Nevertheless, the use of surgical animal models carries a significant mortality risk and requires a high level of technical expertise from researchers. Adenine simulates kidney crystallization and tubulointerstitial fibrosis, which is helpful to study the progression mechanism of CKD.

Adenine can be obtained through dietary sources in addition to being synthesized in vivo. When adenine is supplemented through the diet, it is typically metabolized to 2,8-Dihydroxyadenine (2,8-DHA), resulting in the formation of crystalline deposits in the renal tubules. This process can ultimately lead to kidney damage, progressing to CKD, which is characterized by a decline in kidney function, tubular damage and Inflammation and fibrosis.^6,7 The renal injury induced by adenine in rats is now widely recognized to be linked to elevated plasma uric acid levels.^8,9 Additionally, it is posited that adenine may enhance oxidative stress by activating xanthine oxidase, resulting in kidney damage.^10,11 Furthermore, the formation of 2,8-DHA crystals can also induce tubular cell damage and stimulate inflammatory responses, which further aggravate kidney damage.¹² These processes involved both upstream and downstream signal transduction abnormalities, as well as a series of interrelated factors. Consequently, further investigation is required to elucidate the potential mechanism underlying adenine-induced CKD.

Metabolomics is the study of metabolite changes in biological systems (cells, tissues, or biological organisms) after stimulation or interference in vivo and in vitro.¹³ Non-targeted metabolomics facilitates a comprehensively and high-throughput analysis of the composition information of all metabolites in the body, providing new clues for biological research and disease diagnosis, and implicating for metabolic abnormalities that could explain disease pathogenesis.^14,15 For example, serum metabolic biomarkers might exhibit significant discriminative ability in monitoring the progression of CKD to end-stage renal disease (ESRD), thereby enhancing diagnostic accuracy and advancing treatment strategies for CKD¹⁶. Targeted metabolomics, enables an in-depth examination of special metabolites and associated metabolic pathways.¹⁷ The combination of the two approaches enhances the discovery and precise quantification of differential metabolites¹⁸ and metabolic pathways,¹⁹ thereby improving the accuracy in exploring the mechanism underlying disease occurrence and development. Metabolomics has been applied increasingly to a variety of questions in nephrology research, including metabolites qualitative identification, quantitative and functional verification, helping to promote early intervention to prevent progression of kidney damage.^20–22

This study utilized adenine-induced CKD rats and non-targeted metabolomic analysis of urine samples to identify significant disruption in purine and aromatic amino acid biosynthesis pathways. Subsequent targeted metabolomic analysis was employed to validate the expected changes in related metabolites. Additionally, the detection of the level of oxidative stress and the expression of aromatic amino acid transporter, large neutral amino acids transporter small subunit 1 (SLC7A5), was conducted to further clarify the internal relationship between the two, and to provide a new metabolic perspective for the pathological mechanism of adenine-induced CKD.

Materials and methods

Chemicals and regents

Adenine (purity≥99%) was obtained from Sigma Aldrich (St. Louis, Missouri, USA). Solarbio Science & Technology (Beijing, CHN) provided sodium carboxymethylcellulose (CMC-Na). A water solution containing 5% CMC-Na was prepared with ultrapure water obtained from Millipore (Boston, Massachusetts, USA) to serve as a suspension agent for the drug. Thermo Fisher Scientific (Waltham, Massachusetts, USA) provided acetonitrile and formic acid (all LC–MS grade). High Density Lipoprotein cholesterol (HDL-C), Low Density Lipoprotein cholesterol (LDL-C), and Cystatin C (Cys-C) kits were purchased from Wuhan Colorful Gene Biological Technology Co., Ltd (Wuhan, HuBei, CHN). Rat SLC7A5 (bs-10125R) antibody was purchased from Beijing Bioss Biotechnology Co., Ltd (Beijing, CHN). Purine reference substance (hypoxanthine, xanthine, inosine) was collected from Target Molecule Corp. (Boston, Massachusetts, USA). L-Phenylalanine and L-Tyrosine were purchased from Sigma-Aldrich Corporation (St. Louis, Missouri, USA). L-tryptophan is obtained from the source leaf (Shanghai, CHN).

Study design

Fourteen adult male Sprague Dawley rats, aged 6 wk and weighed 200 ± 20 g were purchased from River Laboratories (Beijing, CHN). During the experiment, rats were maintained in conditions of constant temperature (24 °C) and humidity (50% humidity) during a 12-h light/12-h dark cycle. Following 1 wk of adapting period, 14 rats were divided into two groups: the control group (Con) and the model group (CKD). Three wk were required to administer 200 mg/kg body weight (BW) adenine suspension (freshly dissolved in 0.5% CMC-Na) to rats with CKD. A 4-wk observation period followed the end of molding for each group of rats. Figure 1 shows the experimental procedure.

Fig. 1.

Design diagram for an experimental study.

Experiments with animals were conducted in accordance with the Chinese National Guidelines and in accordance with the National Institute of Health Guide for the Care and Use of Laboratory Animals.

As part of the study, the rats were kept in metabolic cages for 24 h and allowed to drink freely to collect urine samples, which were stored at 80 degrees Celsius for metabolomic analysis. All rats were put to death for histopathological and biochemical examination of kidney tissue and serum samples.

Assessment of histological and biochemical characteristics

Parafin wax was used to embed the kidneys after they had been fixed with 10% neutral-buffered formalin for 48 h. We cut and stained sections with hematoxylin–eosin to evaluate the histopathology. Using Image J, semi-quantitative analysis is conducted on immunohistochemical images.

We monitored the body weights of all rats throughout the experiment. The clinical chemistry analysis of plasma was conducted for the measurement of biochemical parameters, including creatinine (CR), urea nitrogen (UREA), triglycerides (TG), and total cholesterol (TCHO), using a fully automatic biochemical analyzer. We measured HDL-C activity, LDL-C activity, and Cys-C activity in rat plasma using commercial enzyme-linked immunosorbent assays (ELISA).

Untargeted metabolomic analysis

Sample preparation

As described by Qian Feng,²³ urine samples were prepared as follows: We thawed frozen urine samples at 4 °C, diluted them with 150 mL and 50 mL of ultrapure water, mixed them vortex, and centrifuged the supernatant (13,000 rpm/min, 10 min, 4 °C) for analysis. Store samples at 4 °C before analysis.

LC–MS conditions

In order to perform the LC–MS analysis, an Exion LC™ ad UHPLC coupled with a Triple TOF 5600 + mass spectrometer was used from AB Sciex (Boston, Massachusetts, USA). Chromatographic separation was performed by using a Waters Acquity UHPLC HSS T3 column (2.1 mm × 100 mm, 1.7 μm) (Milford, Massachusetts, USA) at 40 °C. The injected volume was 5 μL. In the mobile phases, ultrapure water with 0.1% (v/v) formic acid (A) and acetonitrile (B) was used. A flow rate of 0.3 mL/min was programmed for the LC pump, and gradient elution was optimized as follows: 0–0.5 min, 0%–2% B; 0.5–12 min, 2%–55% B; 12–15 min, 55%–90% B; 15–15.5 min, 90%–98% B; 15.5–16.5 min, 98% B; 16.5–18 min, 98–2% B; 18–20 min, 2% B.

The parameter settings of the electron spray ionization (ESI) source, which was operated in the negative and positive ion modes, were as follows: ion spray voltage − 4,500 V (negative), 5,500 V (positive), turbo spray temperature 550 °C, nebulizer gas (GS1) 55 psi, heater gas (GS2) 55 psi, curtain gas (CUR) 30 psi, declustering potential (DP) 100 V. For characterizing the urine metabolites, the ESI (+) mass spectrometer was operated in TOF MS coupled with information-dependent acquisition (IDA) trigger product ion scan modes. Each MS cycle included one TOF MS survey scan (accumulation time: 0.25 seconds). The IDA criteria were set as follows: the intensity threshold was 100 cps, and dynamic background subtraction was switched on. The scan range of TOF MS was m/z 100–1,500.

Statistical analysis

XCMS-online (The Scripps Research Institute, USA) was used for chromatographic peak detection, comparison, and xls. Format tables containing molecular weight, retention time, and peak intensity were extracted. In the following step, SIMCA-P software version 14.1 (Swedish Umetrics AB) is used to establish a partial least squares discriminant analysis (PLS-DA) model. Identifiable metabolites are based on projections (VIP) and p-values (VIP > 1.0, P < 0.05). Biomarkers are identified by comparing MS, MS/MS fragments, and retention times of LC–MS with reference materials from the Human Metabolome Database (https://hmdb.ca/). Using Metabolanalyst 5.0 (https://www.metaboanalyst.ca), metabolic pathways were analyzed.

Targeted metabolomic analysis

Preparation of the reference substance

Due to the low solubility of purine reference and L-tyrosine residue in neutral and organic solvents, this reference is dissolved in a 0.01 mol/L sodium hydroxide solution. Dissolve L-phenylalanine and L-tryptophan in the acetonitrile-water (V∶V = 5∶100) mixture. Standard solutions are used for the preparation and methodological investigation of standard curves. 1 mg of 2-chlorophenylalanine was mixed with methanol in a 1 mL volumetric flask to prepare 1 mg/mL internal standard mother liquor (IS1), and 1 mg of galantamine was mixed with acetonitrile in a 1 mL volumetric flask to prepare 1 mg/mL internal standard mother liquor (IS2). In order to prepare mixed internal standards, IS1, IS2 and acetonitrile are mixed together. The concentration of the reference substance is shown in Supplementary Table A.1.

Methodological investigation

We evaluated linearity, precision, stability, matrix effects, and extraction recovery of kidney samples to confirm the robustness of the method. In the absence of blank kidney samples, a Phosphate Buffered Saline (PBS) solution was selected to simulate the physiological environment of kidney samples.

Intraday precision and interday precision were determined by measuring the concentration of at least 3 batches of Lower Limit of Quantification (LLOQ), Quality Control 1 (QC1), QC2, QC3. As shown in Supplementary Table A.2. We investigated stability by repeatedly freezing and thawing samples three times at −20 °C and placing them in an autosampler for 24 h at 4 °C. Matrix effects and extraction recoveries were also examined. It is required that each item’s relative standard deviation (RSD) not exceed 15% and that the LLOQ’s RSD not exceed 20%.

Sample preparation

Precision weigh 20 mg of kidney tissue and add 400 μL of acetonitrile-water (3∶1, v/v) and 20 μL internal standard, vortex mix at 12000 rpm, centrifuge at 4 °C for 10 min, take supernatant vacuum freeze concentration and dry. Add 100 μL of acetonitrile-water (1∶1, v/v) for reconstitution, centrifuge at 12,000 rpm for 10 min at 4 °C, and take the supernatant. UPLC-MS/MS is used to analyze the specimen.²⁴

UPLC-MS/MS analysis

The chromatographic separation system used Exion LC™ad from Shimadzu Corporation (Kyoto, JPN) and the column selected Acquity UPLC HSS T3 (2.1 mm × 100 mm, 1.8 μm) from Waters (Milford, Massachusetts, USA). 0.1 mL/min flow rate and a volume of 5 μL were used for injection. The optimal separation conditions were phase A (aqueous solution of 0.1% v/v formic acid) and phase B (acetonitrile of 0.1% v/v formic acid). The separation gradient elution is as follows: 0–2 min, 0% B; 2–5 min, 0%–1% B; 5–6 min, 1%–3% B; 6–6.5 min, 3%–10% B; 6.5–8 min, 10%–20% B; 8–9 min, 20%–35% B; 9–11 min, 35%–95% B; 11–12 min, 95% B; 12–14 min, 95–0% B; 14%–15 min, 0% B. Mass spectrometry was performed using the AB SCIEX API 3200MD equipped with an electrospray ionization source (Boston, Massachusetts, USA). The parameters are as follows: air curtain air, 40 psi; Spray voltage, 5,500 V; GS1, 50 psi, GS2, 0 psi; Collision gas, N2; Ion source temperature, 500 °C. Multiple reaction monitoring (MRM) acquisition mode was used to detect in positive ion mode.²⁴

Detection of reactive oxygen species (ROS) production

A commercial kit was used to detect ROS activity according to instructions provided by the Beijing Andy Huatai Technology Co., Ltd (Beijing, CHN).

Western blot analysis

A western blot is used to determine the expression of the amino acid transporter SLC7A5. Kidney tissue samples (n = 3) are homogenized in Radio Immunoprecipitation Assay Lysis buffer (RIPA) containing 1% Phenylmethanesulfonyl fluoride (PMSF) and incubated on ice for 60–80 min. Collect the supernatant after centrifuging the homogenate for 10 minutes at 4 °C at 12,000 rpm. The assay kit for Bicinchoninic Acid Assay (BCA) protein determines total protein concentrations. Using 10% Sodium Dodecyl Sulfate-Polyacrylamide Gel Electrophoresis (SDS-PAGE) gel protein isolate, block with Tris buffered saline (TBST) containing 5% (w/v) skim milk for 2 h and overnight at 4 °C under SLC7A5 (1∶300) conditions. After washing the membrane three times with TBST, incubate it for 2 hours at room temperature with Horseradish Peroxidase (HRP) conjugated antibody (1∶5,000). Scan the membrane with a gel imaging device after three washes in TBST.

Statistical analysis

Statistical analysis: experimental results were expressed as means±SD. GraphPad Prism (version 7.0, GraphPad Software Inc., LaJolla, CA, USA) was employed for statistical analysis. The between-group differences (P < 0.05) were compared using the student’s t-test.

Results

Adenine induces kidney damage

General data

Compared to the Con group, the group of rats with CKD exhibited a deceleration in weight gain following 3 wk of adenine feeding (Fig. 2A). In addition, the CKD group showed significantly elevated levels of serum creatinine (Scr) and blood urea nitrogen (BUN) (Fig. 2B and C). Notably, Tcho and TG were also markedly higher (Fig. 2D and E). The quantitative analysis of LDL and HDL further suggested that the model causes disorders of lipid metabolism (Fig. 2F–G). Furthermore, Cys-C, a sensitive and specific marker for early detection of chronic kidney disease, increased in CKD group indicating that the model is successfully established (Fig. 2H).

Fig. 2.

The effects of adenine on rat body weight and kidney function. A) Body weight, B)CR, C)UREA, D) TG, E) TCHO, F) HDL-C, G) LDL-C, H) Cys-C; n = 7. ^*P < 0.05, ^**P < 0.01, ^***P < 0.001 vs. con.

Histological findings

As shown in Fig. 3A, Hematoxylin–Eosin (HE) staining analysis revealed the presence of tubular atrophy, tubular dilation, flattening of the tubular epithelium, tubular epithelial degeneration, increased interstitial and tubular uric acid crystals, as well as more lymphocyte infiltration in the CKD group. Conversely, none of these symptoms were observed in the Con group.

Fig. 3.

Renal histopathological findings. A) HE staining. (arrows indicate different kidney damage: arrow 1, tubular atrophy; arrow 2, tubular dilation; arrow 3, tubular epithelial degeneration; arrow 4, urate deposition; arrow 5, interstitial connective tissue hyperplasia; arrow 6, lymphocyte infiltration). B) Immunohistochemical analysis of E-cadherin, Podocin, and Synaptopodin (scale bar: 100 μm, 50 μm). C) Semi-quantitative expression of marker protein (scale bar: 50 μm). ^*P < 0.05, ^**P < 0.01, ^***P < 0.001 vs. con.

Synaptopodin, a proline-rich linear protein closely associated with actin filaments, functions as a marker for podocyte differentiation and maturation. Mutations or deletions in synaptopodin can result in structural changes in podocytes, consequently causing proteinuria and glomerular sclerosis. The podocin protein, situated at the insertion site of the glomerular slit diaphragm, is a vital component in the maintenance and regulation of the structural integrity of the slit membrane. Damage to podocytes can disrupt the function of podocin protein, leading to alterations in its function within the glomerular filtration membrane. E-cadherin plays a pivotal role in mediating cell-to-cell adhesion, thereby facilitating the maintenance of cell polarity and the integrity of tissue structures. In the context of renal tubules, the normal function of E-cadherin is equally crucial for preservation the structural and functional integrity of the tubules. As depicted in Fig. 3B and C, the expressions of the marker proteins were decreased, suggesting that adenine can cause significant destruction of the glomeruli and tubules.

Metabolite and metabolic pathway analysis based on urine non-targeted metabolomics

Multivariate statistical analysis

To explore the disparities between Con and CKD groups, Principal Component Analysis (PCA) was conducted and a clear separation trend was observed between the two groups. The permutation test of PLS-DA (model parameters: B, R² = 0.99, Q² = 0.95; E, R² = 0.99, Q² = 0.94) demonstrated that the model was not overfitting, indicating that the established model was stable and reliable (Fig. 4B and E).

Fig. 4.

Results of multivariate analysis among con and CKD groups. A and D) PCA score plots; B and E) permutation test; C and F) S-plot plots. A, B, C: Positive ion mode; D, E and F). Negative ion mode.

Identification of differential metabolites

Based on the S-plot, VIP > 1, and P < 0.05, differential variables were screened. By comparing the parent ions and secondary fragment ions with the HMDB database, a total of 66 differential metabolites were identified. Detailed information on the identification of differential metabolites was presented in Table 1.

Table 1.

Information of identified metabolites in rat urine.

No.	metabolite name	m/z	Rt(min)	ionic mode	Con	CKD	Trend, compared with Con
1	Picolinic acid	124.0390	2.33	[M + H]+	0.546 ± 0.175	0.171 ± 0.058c	↑
2	Thymine	127.0502	1.82	[M + H]+	0.225 ± 0.150	0.596 ± 0.191b	↑
3	Hypoxanthine	137.0444	3.99	[M + H]+	0.345 ± 0.102	0.213 ± 0.039a	↓
4	Urocanic acid	139.0503	2.93	[M + H]+	0.053 ± 0.033	0.240 ± 0.167a	↑
5	Methylimidazoleacetic acid	141.0660	1.74	[M + H]+	0.094 ± 0.061	0.390 ± 0.299a	↑
6	Tiglylglycine	158.0812	5.31	[M + H]+	0.243 ± 0.118	0.107 ± 0.051a	↑
7	Acetylcysteine	164.0380	8.09	[M + H]+	0.159 ± 0.044	0.042 ± 0.011c	↓
8	N2-Methylguanine	166.0703	2.67	[M + H]+	2.957 ± 0.690	2.235 ± 0.487a	↓
9	Glucosamine	180.0870	4.23	[M + H]+	0.304 ± 0.208	0.095 ± 0.046a	↓
10	Phenylpyruvic acid	163.0400	6.83	[M + H]+	0.349 ± 0.213	1.279 ± 0.790a	↑
11	3-Methyluric acid	183.0510	4.26	[M + H]+	0.148 ± 0.017	0.314 ± 0.065c	↑
12	N1-Acetylspermidine	188.1758	1.81	[M + H]+	0.456 ± 0.147	0.132 ± 0.148b	↓
13	N6, N6, N6-Trimethyl-L-lysine	189.1600	1.40	[M + H]+	0.025 ± 0.036	0.734 ± 0.342b	↑
14	N-Acetylhistidine	198.0872	1.78	[M + H]+	0.380 ± 0.151	0.141 ± 0.051b	↓
15	Symmetric dimethylarginine	203.1501	1.87	[M + H]+	0.442 ± 0.285	1.403 ± 0.394c	↑
16	Nα-Acetyl-L-arginine	217.1295	2.27	[M + H]+	1.638 ± 0.546	0.895 ± 0.193a	↓
17	Isoleucylproline	229.1546	1.93	[M + H]+	0.320 ± 0.294	1.724 ± 1.438a	↑
18	Butyrylcarnitine	232.1540	5.76	[M + H]+	0.337 ± 0.303	0.036 ± 0.021a	↓
19	L-Pyridosine	255.1340	2.79	[M + H]+	0.061 ± 0.040	0.577 ± 0.289b	↑
20	1-(beta-D-Ribofuranosyl)-1,4-dihydronicotinamide	257.1134	4.12	[M + H]+	0.294 ± 0.094	0.083 ± 0.021c	↓
21	N2, N2-Dimethylguanosine	312.1302	5.04	[M + H]+	0.398 ± 0.263	0.128 ± 0.063a	↓
22	Leukotriene B4	337.2363	15.90	[M + H]+	0.093 ± 0.026	0.017 ± 0.009c	↓
23	14,15-DiHETrE	339.2530	9.99	[M + H]+	0.153 ± 0.034	0.037 ± 0.012c	↓
24	Tetrahydrocorticosterone	351.2533	8.48	[M + H]+	0.365 ± 0.100	0.507 ± 0.102a	↑
25	Epinephrine	184.0956	2.59	[M + H]+	0.100 ± 0.069	0.290 ± 0.186a	↑
26	L-Tyrosine	182.0807	3.67	[M + H]+	0.314 ± 0.097	0.136 ± 0.102a	↓
27	Xanthine	152.1109	2.26	[M + H]+	0.033 ± 0.017	0.051 ± 0.009a	↑
28	3-Methyluridine	259.0913	2.08	[M + H]+	0.434 ± 0.222	1.378 ± 0.893a	↑
29	N4-Acetylcytidine	286.1019	3.98	[M + H]+	0.156 ± 0.079	0.052 ± 0.019a	↓
30	Taurine	124.0081	1.76	[M-H]−	0.950 ± 0.484	0.289 ± 0.190a	↓
31	2-Oxoglutarate	145.0147	1.74	[M-H]−	0.162 ± 0.120	0.656 ± 0.207c	↑
32	cis-4-Hydroxycyclohexylacetic acid	157.0871	11.34	[M-H]−	1.139 ± 0.436	0.075 ± 0.0280c	↓
33	Isovalerylglycine	158.0824	6.22	[M-H]−	0.635 ± 0.384	1.263 ± 0.304b	↑
34	Hydroxyoctanoic acid	159.1027	8.30	[M-H]−	0.610 ± 0.210	0.256 ± 0.129b	↓
35	3-Hydroxyoctanoic acid	159.1027	8.23	[M-H]−	0.076 ± 0.023	0.027 ± 0.008c	↓
36	N-Acetylglutamine	187.0723	1.82	[M-H]−	0.300 ± 0.144	0.091 ± 0.051b	↓
37	Indole-3-propionic acid	188.0717	5.15	[M-H]−	0.042 ± 0.014	0.157 ± 0.096a	↑
38	Galactonic acid	195.0512	1.45	[M-H]−	0.097 ± 0.025	0.006 ± 0.004c	↓
39	L-Metanephrine	196.0980	9.84	[M-H]−	0.560 ± 0.272	0.956 ± 0.185b	↑
40	Indolelactic acid	204.0667	8.82	[M-H]−	0.222 ± 0.183	0.585 ± 0.170b	↑
41	N-Acetyl-L-phenylalanine	206.0822	8.40	[M-H]−	0.177 ± 0.087	0.355 ± 0.181a	↑
42	Hydroxyphenylacetylglycine	208.0617	5.46	[M-H]−	0.111 ± 0.042	0.023 ± 0.005b	↓
43	Indoxyl sulfate	212.0025	6.61	[M-H]−	6.186 ± 2.032	9.756 ± 3.021a	↑
44	3-Hydroxysebacic acid	217.1081	7.79	[M-H]−	2.130 ± 0.819	1.030 ± 0.680a	↓
45	Dodecanedioic acid	229.1446	7.59	[M-H]−	0.182 ± 0.064	0.085 ± 0.021b	↓
46	Uridine	243.0623	2.25	[M-H]−	0.310 ± 0.327	1.043 ± 0.462b	↑
47	Deoxyadenosine	250.0943	8.36	[M-H]−	0.063 ± 0.018	0.014 ± 0.008c	↓
48	Palmitic acid	255.2329	6.69	[M-H]−	0.076 ± 0.012	0.146 ± 0.024c	↑
49	3-Deoxy-D-glycero-D-galacto-2-nonulosonic acid	267.0741	4.03	[M-H]−	0.399 ± 0.191	0.134 ± 0.084b	↓
50	Urothion	324.0231	5.72	[M-H]−	0.101 ± 0.038	0.032 ± 0.023b	↓
51	Acetaminophen glucuronide	326.0879	6.19	[M-H]−	0.073 ± 0.031	0.028 ± 0.019b	↓
52	Cyclic AMP	328.0452	3.89	[M-H]−	0.192 ± 0.114	0.024 ± 0.025b	↓
53	Nicotinamide ribotide	334.0687	4.58	[M-H]−	0.089 ± 0.036	0.017 ± 0.011b	↓
54	Inosine	267.0727	3.95	[M-H]−	0.399 ± 0.191	0.134 ± 0.084b	↓
55	Undecylic acid	185.1545	10.72	[M-H]−	0.107 ± 0.095	0.018 ± 0.009a	↓
56	10E,12Z-octadecadienoic acid	279.2305	18.68	[M-H]−	0.006 ± 0.002	0.014 ± 0.003c	↑
57	isocitric acid	191.0202	1.84	[M-H]−	1.554 ± 0.984	4.383 ± 2.096a	↑
58	Citric acid	191.0195	3.57	[M-H]−	0.229 ± 0.138	0.501 ± 0.232a	↑
59	N-Acetyl-L-histidine	196.0723	1.83	[M-H]−	0.070 ± 0.065	0.006 ± 0.012a	↓
60	N-Acetyl-L-tyrosine	222.0802	6.75	[M-H]−	8.674 ± 5.230	1.111 ± 0.720b	↓
61	Deoxycytidine	226.0802	9.62	[M-H]−	0.102 ± 0.036	0.035 ± 0.029b	↓
62	Dodecanedioic aicd	229.1441	9.73	[M-H]−	0.309 ± 0.124	0.105 ± 0.023b	↓
63	Guanosine 3′,5′-cyclic monophosphate	344.0398	3.91	[M-H]−	0.091 ± 0.046	0.025 ± 0.028b	↓
64	Nicotinuric acid	359.0989	10.04	[M-H]−	0.617 ± 0.262	0.141 ± 0.047b	↓
65	12-Oxo-20-hydroxy-leukotriene B4	349.2001	10.94	[M-H]−	0.031 ± 0.017	0.012 ± 0.006a	↓
66	L-Tryptophan	203.0830	5.97	[M-H]−	0.073 ± 0.023	0.039 ± 0.024b	↓

^a p < 0.05

^b p < 0.01

^c p < 0.001 vs. Con.

Metabolic pathway analysis

According to the impact value and P value, six metabolic pathways were predicted, including purine metabolism, tricarboxylic acid cycle (TCA cycle) and amino acids (phenylalanine, tyrosine and tryptophan biosynthesis, phenylalanine metabolism, histidine metabolism) disorders (Fig. 5) in CKD rats.

Fig. 5.

The bubble chart of pathway analysis.

Targeted quantitative analysis of purines and amino acid metabolites

Methodological validation

The declustering voltage (DP) and collision voltage (ce) values for each metabolite were listed in Table 2. The specific verification results are shown in Supplementary Tables A.3–A.6, and all the results meet the analysis requirements of tissue samples, indicating that the method is reliable, accurate and stable.

Table 2.

Instrumental parameters for the analytes and internal standards obtained after collision-induced dissociation in MRM mode.

Metabolites	Molecular weight	Precursor ion (m/z)	Product ion (m/z)	Fragment voltage (V)	Collision energy (eV)	Ionic mode
Hypoxanthine	136.11	137.0	110.0	36	20	+
Xanthine	152.11	153.1	110.1	35	15	+
Inosine	268.23	269.1	137.1	21	20	+
L-Tyrosine	181.1	182.0	136.0	40	18	+
L-Phenylalanine	165.1	120.0	77.1	60	13	+
L-Tryptophan	204.2	205.1	188.1	67	12	+
IS1	199.6	200.0	154.1	10	20	+
IS2	287.3	299.1	213.0	20	20	+

Targeted metabolomic quantification

The UHPLC–MS/MS method was used to determine the contents of purine metabolites and amino acid metabolites in rat kidney samples. The results showed a significant reduction in amino acid content in the kidneys of CKD rats compared to the control group (Fig. 6). Additionally, hypoxanthine and creatinine levels were also significantly reduced, while xanthine levels were elevated.

Fig. 6.

Changes in the content of metabolites in rat kidneys. ^*P < 0.05, ^**P < 0.01, ^***P < 0.001 vs. con.

Adenine induces dysfunction of purine metabolism pathway, resulting in increased ROS levels

As shown in Fig. 7A, adenine-treated rats exhibited significantly elevated ROS levels in their kidneys compared to the control rats. This observation implies that adenine causes oxidative stress in the renal tissue.

Fig. 7.

Effects of adenine on renal ROS levels A) and the expression of SLC7A5 protein B). ^*P < 0.05, ^**P < 0.01, ^***P < 0.001 vs. con.

The increase of ROS led to a decrease in the expression of SLC7A5 protein

The expression of aromatic amino acid transporter SLC7A5 in the kidneys of CKD rats was significantly downregulated compared to the control group, suggesting a reduction in SLC7A5 protein expression during oxidative stress. At the same time, previous research has demonstrated that adenine can induce amino acid imbalances in the kidneys, particularly leading to the reduction in essential amino acids (Fig. 7B).

Discussion

Non-targeted metabolomics offers a comprehensive and systematical analysis all metabolites within a biological system, exhibiting an unbiased approach that aids in the discovery of novel biomarkers.²⁵ On the other hand, targeted metabolomics allows for precise quantification of specific metabolites under investigation, thereby facilitating deeper insights into their functions and mechanisms.¹⁷ In this study, non-targeted combined with targeted metabolomic techniques was utilized to investigate the concurrent perturbation of purine metabolism and aromatic amino acid biosynthesis pathways in chronic kidney disease, as well as the potential interplay between them.

Adenine, a purine base, plays an important physiological role in living organisms²⁶ and is produced through a series of enzymatic reactions from simple compounds.²⁷ In addition, protein-rich foods (such as meat, fish, eggs and dairy products) are also sources of adenine.²⁸ However, it should be noted that excessive intake of adenine can easily cause CKD²⁹. Several metabolic pathways, including adenine metabolism, tryptophan metabolism, creatinine metabolism and fatty acid metabolism, have been documented as being dysregulated in adenine-induced CKD models.³⁰ According to studies, oxidative stress was closely related to kidney damage.³¹ Adenine-induced kidney disease may occur through the conversion of inosine to hypoxanthine in the body and eventually to xanthine.³² The quantitative results of this study were supported this hypothesis,³³ as the interconversion of xanthine and hypoxanthine can induce inflammation and increases ROS production, which ultimately leads to kidney disease. In a study conducted by Keiichi Ohata et al., they observed a significant increase in XO activity levels in the kidneys of mice with an adenine-fed diet compared to those on a normal diet.³⁴ This finding provides partial validation to our own results. The SLC7A5 protein transporter primarily utilizes aromatic amino acids with strong affinity as substrates.³⁵ Studies have reported that SLC7A5 expression decreases in response to oxidative stress.³⁶ In this study, the oxidative stress generated by the purine pathway can reduce the expression of SLC7A5 protein, thereby affecting the content of essential aromatic amino acids.

Tryptophan, phenylalanine, and tyrosine are essential amino acids that are unable to be synthesized in the body and must be obtained through dietary sources. Our results demonstrated a reduction in the content of these three amino acids compared to the control group, aligning with previous findings.³⁷ Furthermore, reduced levels of tyrosine and phenylalanine have been reported in patients with CKD³⁸. Tryptophan metabolic pathways include the indole pathway, the kynurenine pathway, and the serotonin pathway.³⁹ Most tryptophan is metabolized by the kynurenine pathway to produce kynurenine, and the accumulation of kynurenine derivatives may lead to kidney injury and subsequent loss of function.¹⁶ The study showed that markers associated with changes in tryptophan metabolism were identified using targeted and non-targeted serum metabolites to detect early kidney injury, consistent with our findings that changes in tryptophan metabolites are associated with CKD⁴⁰. It has been observed that endogenous metabolites in the body can interact with intestinal microorganisms to produce metabolites that may pose a threat to host health.⁴¹ Tryptophan undergoes metabolism to form indole derivatives, such as indole sulfate (IS) and indole acetic acid (IAA) through the action of intestinal flora.⁴² The hepatic production of the uremic toxin indoxyl sulfate (IS) from the precursor indole is facilitated by organic anions and sulfotransferases.⁴³ IS is subsequently transported to the kidneys for excretion by an organic anion transporter (OAT-1/3). IS can cause kidney damage by promoting cytokine production,⁴⁴ induing oxidative stress,⁴⁵ and altering the expression of kidney-associated proteins.⁴⁶ IS is currently the most widely studied indole protein-binding uremic toxin. In addition, tryptophan can be degraded to kynurenine by indoleamine 2, 3-dioxygenase (IDO).⁴⁷ The activity of metabolic enzyme IDO has been shown to be markedly enhanced in CKD and is positively correlated with disease severity.⁴⁸ It can be speculated that CKD can be alleviated by modulation the activity of the metabolic enzyme IDO and subsequent regulation of the kynurenine pathway. However, its precise mechanism of action is unclear and warrants further investigation. Resent clinical studies have shown that serum aromatic amino acid levels are reduced in patients with kidney injury.⁴⁹ Furthermore, a notable decrease in urine tryptophan and tyrosine levels has been observed at the initiation of kidney disease.⁵⁰ Nephropathy rats treated with L-tryptophan were found a reduction in renal function.⁵¹ In a study of the renal metabolome of cisplatin-induced acute kidney injury rats, a marked decrease in tryptophan was noted upon the onset of nephropathy.⁵² Similarly, abnormal tryptophan metabolites produced by intestinal microorganisms were significantly associated with creatinine levels in both adenine-induced rats and patients with CKD⁵³. These results indicate a close relationship between the abnormal metabolism of tryptophan and the decline of renal function. In this study, the reduced tryptophan and elevated IS levels also supports this finding.

This study has a few limitations. Firstly, there have been many reports on the close correlation between metabolic disorders of aromatic amino acids, especially tryptophan, and nephropathy. However, this study only provides a possible explanation for the internal relationship between purine metabolism and aromatic amino acid metabolism, which is worthy of further investigation and needs to be verified by relevant molecular biological experiments. Secondly, the connection between other key metabolic pathways (such as TCA cycle, phenylalanine metabolism) and chronic kidney disease requires additional exploration and validation.

Conclusion

In this study, a chronic kidney disease model was established by adenine administration, and non-targeted metabolomics was employed to identify the key pathways. Further, targeted metabolomics was used to confirm alterations in key metabolites. The significant correlation was then elucidated between purine metabolism and aromatic amino acid metabolism disorders by analyzing the biochemical pathways and inherent metabolic mechanisms of the involved metabolites. It was determined that oxidative stress resulting from purine metabolism disorders can adversely affect the transport of aromatic amino acids, thereby inducing aromatic amino acid metabolism disorders and ultimately causing kidney injury. This study is expected to provide a new interpretation for the pathological mechanisms of adenine-induced chronic kidney disease. The research ideas are shown in Fig. 8.

Fig. 8.

Synergistic potential mechanisms of action of aromatic amino acids and purines in inducing nephropathy.

Compliance with ethical standards

All animal care procedures are carried out in accordance with the protocol approved by Shanxi University Animal Ethics Committee (SXULL2023072). All experiments complied with the Guide for the Care and Use of Laboratory Animals and the National Institutes of Health Guide for the Care and Use of Laboratory Animals.

Author contributions

Ai-Ping Li: Writing–review & editing, Writing–original draft, Supervision, Resources, Conceptualization. Xing-Xing Zhang: Writing–original draft, Investigation, Conceptualization. Qing-Yu Zhang: Writing–original draft, Formal analysis. Meng-Hiao Wang: Formal analysis. Zheng Ju: Formal analysis. Xiao-Yu Zhang: Formal analysis. Xue-Mei Qin: Conceptualization. Guang-Zhen Liu: Writing– review & editing, Supervision, Resources, Conceptualization.

Funding

Financial support for this research was provided by the National Natural Science Foundation of China (82204595, U23A20517), Fundamental Research Program of Shanxi Province (202303021221070), China Postdoctoral Science Foundation (340903), Key Laboratory of Effective Substances Research and Utilization in TCM of Shanxi Province (202105D121009), the Traditional Chinese Medicine Innovation Team of Shanxi Province (No. zyytd2024020).

 

Conflicts of interest. The authors declare that there are no conflicts of interest.

Data availability

Data will be made available on request.