The effect of fructose exposure on amino acid metabolism among Chinese community residents and its possible multi-omics mechanisms

The consumption of fructose has increased dramaticly during the last few decades, inducing a great increase in the risk of intrahepatic lipid accumulation, hypertriglyceridemia, hyperuricemia and cancer. However, the underlying mechanism has not yet been fully elucidated. Amino acid metabolism may play an important role in the process of the diseases caused by fructose, but there is still a lack of corresponding evidence. In present study, we provide an evidence of how fructose affects amino acids metabolism in 1895 ordinary residents in Chinese community using UPLC-QqQMS based amino acid targeted metabolomics and the underlying mechanism of fructose exposure how interferes with amino acid metabolism related genes and acetylated modification of proteome in the liver of rats model. We found people with high fructose exposure had higher levels of Asa, EtN, Asp, and Glu, and lower levels of 1MHis, PEtN, Arg, Gln, GABA, Aad, Hyl and Cys. The further mechanism study displayed amino acid metabolic genes of Aspa, Cndp1, Dbt, Dmgdh, and toxic metabolites such as N-acetylethanolamines accumulation, interference of urea cycle, as well as acetylated modification of key enzymes in glutamine metabolic network and glutamine derived NEAAs synthesis pathway in liver may play important roles in fructose caused reprogramming in amino acid metabolism. This research provides novel insights of the mechanism of amino acid metabolic disorder caused by fructose and supplies new targets for clinical therapy.


Amino acid targeted metabolomics analysis
Amino acid targeted metabolomics detection was performed on UPLC-QqQMS platform (waters TQSm).Mobile phase A, water (containing 0.1% formic acid); mobile phase B, acetonitrile; and the flow rate was 0.5 mL/min.The gradient conditions of the mobile phase were as follows: 0 min for A/B 98:2 (v/v), 4 min for A/B 78:22 (v/v), 4.1 min for A/B 5:95 (v/v), 4.4 min for A/B 0:100 (v/v), 4.6 min for A/B 98:2 (v/v), 6 min for A/B 98:2 (v/v).The temperatures of the column and autosampler were maintained at 45 °C and 4 °C, respectively.The injection volume is 2 μL.Triple quadrupole scans were operating in a positive ion mode.The ESI source temperature was 350 °C and the voltage was 3000 V.The scaning detection of ion pairs were depended on the optimized declustering potential and collision energy.Supplementary table 1 showed multiple reaction monitoring (MRM) parameters including parent ions, daughter ions, dwell time, cone voltage and collision voltage of 39 amino acids.Each amino acid was validated through corresponding isotope internal standards.The chromatographic peaks were corrected manually to ensure the accuracy of quantitative determination according to suitable calibration curve (r 2 > 0.99).

Serum fructose detection
Serum fructose was detected using microplate method (kit manufacturer: Shanghai Zhenke Biotechnology Co., Ltd.; product number: G20221010H).In brief, fructose in 20 µL serum was converted into glucose by the action of specific enzymes and then the glucose could increase the amount of NADPH under the action of hexokinase enzyme complexes.The content of fructose could be calculated by using the microplate reader to detect the increase of the NADPH at 340 nm.

Animals and sample collection
Twelve three-week-old male SD rats [Purchased from Hunan SJA Laboratory Animal Co.; Ltd, Changsha, China; Licence: SYXK(Hunan)-2019-0014] were maintained in temperature-and light-controlled (14 h/10 h light/ dark cycle) conditions at the Laboratory Animal Center of School of department of zoology, University of South China, a facility approved by the association for assessment and accreditation of laboratory animal care [Licence: SYXK(Hunan)-2020-0002].The animals were acclimated to the laboratory for 1 week prior to the start of the experiments and were randomized into 2 groups: control group, fructose exposing group (n = 6 rats/group).Animals in each group had free access to food and water.In fructose exposing group, the drinking water was replaced with 5% fructose water for 6 months (the dissolution ratio is: 5 g fructose in 100 mL water, freshly prepared every day).Two groups of rats were provided equal amounts of food with the same ingredients.And rats of fructose exposing group were provided the same volume fructose water as the control group daily.
At the end of the experiment, femoral artery blood was collected into 1.5 mL centrifugal tube after anaesthesia with 3-5% isoflurane.Blood samples were centrifuged for 10 min at 3000 rpm and 4 °C.Serum was collected into microtubes and stored at − 80 °C.All rats were euthanized by cervical dislocation, and the left lateral lobe liver of the rats of every group were separated and were cut into small pieces and mixed evenly, and then grinded into powder in pulverizer after a few minutes for liquid nitrogen freezing.These liver samples were stored at − 80 °C.
All procedures performed in this study were in accordance with ARRIVE guidelines and the ethical standards of the Ethics Committees of Affiliated Nanhua Hospital, University of South China (Ethical approval number: 2021-ky-173).www.nature.com/scientificreports/

Rats liver PCR Array analysis
We applied Rat Amino Acid Metabolism II RT 2 Profiler PCR Array to profile the expression of 84 key genes important in amino acid biosynthesis and degradation.Mature RNA was isolated using an RNA extraction kit according to the manufacturer's instructions.RNA quality was determined using a spectrophotometer and was reverse transcribed using a cDNA conversion kit.The cDNA was used on the real-time RT 2 Profiler PCR Array (QIAGEN, Cat.no.PARN-130Z) in combination with RT 2 SYBR® Green qPCR Mastermix (Cat.no.330529).In brief, total RNA was isolated from 50 mg of liver tissue and the isolated RNA was quantified by measuring the absorbance at 260 nm.RNA purity was assessed by the 260/280 nm ratio, and the integrity was checked using 1% agarose gel.Reverse transcription of RNA to cDNA was carried out using Quantiscript Reverse Transcriptase and RT Primer Mix.Reverse transcription reaction mixture (20 µL) contained 14 µL of total RNA, 4 µL of buffer, 1 µL of RT Primer Mix, and 1 µL of Quantiscript Reverse Transcriptase.The resulting mixture was initially incubated at 42 °C for 15 min, and then heating the mixture to 95 °C for 5 min to inactivate the reverse transcriptase.cDNA obtained was stored at − 80 °C until further use.
The final reaction mixture (2000 µL) contained 1000 μL of 2 × QuantiNova SYBR Green PCR Master Mix, 100 μL of cDNA, 900 μL of H 2 O. Before PCR Array test, add 20 µL of the final reaction mixture to each hole of the 96 well plate and the qRT-PCR program consisted of the following temperature profile: initial activation of DNA polymerase at 95 °C for 2 min, and then 45 cycles of denaturation at 95 °C for 5 s and annealing at 60 °C for 10 s.
C T values were exported to an Excel file to create a table of C T values.This table was then uploaded on to the data analysis web portal at http:// www.qiagen.com/ geneg lobe.Samples were assigned to controls and fructose exposing group.The data analysis web portal calculated fold regulation using delta delta C T method, in which delta C T was calculated between gene of interest (GOI) and an average of reference genes (HKG), followed by delta-delta C T calculations (delta C T (Test Group)-delta C T (Control Group)).Fold Regulation was then calculated using 2^ (−delta delta C T ) formula.

Rats liver non-targeted metabolomics analysis
Liver metabolites extraction 50 ± 5 mg of liver tissue sample was accurately separated and was added 200 μL ultrapure water for homogenization.Added 800 μL methanol: acetonitrile (v:v = 1:1) solution to the homogenate, and vortexed for 30 s, ultrasounded for 10 min (ice bath).The mixed solution was stored in a refrigerator at − 20 °C for 1 h and then was centrifuged at 12,000 rpm (4 °C) for 15 min.The supernatant was blew dry with N 2 .The powder was redissolved with 100 μL acetonitrile: water (v:v = 1:1) solution, and vortexed for 30 s, ultrasounded for 5 min (ice bath).And then the mixed solution was centrifuged at 12,000 rpm (4 °C) for 15 min.The supernatant was injected into the bottle for MS analysis.
The specific conditions for mass spectrum were as follows: Ion Source: TurboIonSpray; Curtain gas: 35; CAD gas: 9; Ion source gas 1 (psi): 55; Ion source gas 2 (psi): 55; Temperature (°C): 550.Differences in liver metabolite content between fructose exposing group and control group were described by fold regulation, P value generated from paired t-tests, and corrected p value generated from Benjamin Hochberg correction.

Rats liver acetylation proteomics analysis
Protein extraction and peptide enzymatic hydrolysis Powdered liver tissues were extracted using UA (8 M Urea, 100 mM Tris/HCl, pH 8.5) lysis method to extract proteins 16 , and then protein quantification was performed using the Bradford method.20 µg of protein from each sample was added to 5 × loading buffer solution, and the boiling water bath was used for 5 min.12.5% SDS-PAGE electrophoresis (constant flow 14 mA, 90 min) and Coomassie brilliant blue staining were performed.Added DTT to all samples to make the final concentration to 10 mM, placed them in a constant temperature mixer (600 rpm, 37 ℃) for 1.5 h, and cooled them to a room temperature.Later added IAA to the samples to make the final concentration to 50 mM and kept them in dark for 30 min.Later added 4 times the volume of 50 mM Tris HCl (pH 8.0), diluted the UA concentration to 2 M, and added trypsin in a 50:1 ratio of protein: trypsin by mass to digest proteins at 37 ℃ overnight (15-18 h).Added TFA to make the final concentration to 0.1%, and adjusted pH (≤ 3) by 10% TFA.Finally, applied C18SPECarrridge column to desalinate and freeze-dry peptide segments.www.nature.com/scientificreports/Acetylated peptide enrichment Freeze-dried peptide samples were redissolved by 1.4 mL of precooled IAP buffer, and then were added pretreated Anti-Ac-K antibody beans (PTMScan Acetyl-Lysine Motif (Ac-K) Kit, Cell Signal Technology).Later incubated at 4 ℃ for 1.5 h, centrifuged at 3000 rpm for 30 s, and the supernatant were discarded.Anti-Ac-K antibody beans were cleaned 3 times with 1 mL precooled IAP buffer and then were cleaned 3 times with 1 mL precooled water again.Added 40 μL 0.15% TFA twice to cleaned anti-Ac-K antibody beans at room temperature for 10 min.And then centrifuged at 3000 rpm for 30 s, and the supernatant was desalinated with C18 STAGE Tips.

LC-MS/MS data collection
Each sample was separated using a nanoElute HPLC system with a nanoliter flow rate.Buffer solution A was a 0.1% formic acid aqueous solution, while solution B was a 0.1% formic acid acetonitrile aqueous solution.The chromatographic column was equilibrated with solution A, and the sample was separated through the column (homemade column, 25 cm ID 75 μm, 1.9 μm, C18) at a flow rate of 300 nL/min and the column temperature was 50 ℃.
After chromatographic separation, the sample was subjected to the timsTOF Pro mass spectrometer.The detection method was positive mode, and the ionization source voltage was 1.5 kV, both MS and MS/MS used TOF for detection and analysis.The mass spectrometry scanning range was set to 100-1700 m/z.The data collection adopted the Flat Accumulated Serial Fragmentation (PASEF) mode, and after the first level mass spectrometry collection, 10 times of PASEF mode collection was performed, with a cycle window time of 1.17 s.The dynamic exclusion time of the tandem mass spectrometry scan was set to 24 s to avoid repeated scanning of parent ion of MS/MS spectra of the charge from 0 to 5.

Protein identification and quantitative analysis
The original data of mass spectrometry was raw file, and MaxQuant software (version 1.5.3.17) 17 was used for database checking, identification and quantitative analysis.Significantly modified peptide segments were screened and the up-regulated and down-regulated modified peptide segments were obtained on the criteria of fold regulation > 2 (up regulation greater than 2.0-fold or down regulation less than 0.5-fold) and Benjamin Hochberg corrected P value < 0.05.Then the GO and KEGG Pathway enrichment analysis were carried out according to the methods introduced in the references 18,19 . 15e correlation coefficients among the differential genes, metabolites, and acetylated proteins in rat liver were measured by Pearson correlation coefficient in R software (version: 4.3.1)and were visualized by heatmap in Heml software (Version: 1.0.3.7), which is a software dedicated to drawing heatmaps (https:// hemi.biocu ckoo.cn/).

Ethics, consent and permissions
All subjects gave their informed written consent and the study was done in accordance with the Declaration of Helsinki.The study was approved by the Ethics Committees of the Affiliated Nanhua Hospital, University of South China (Ethical approval number: 2022-KY-167).

Association between fructose exposure and amino acid metabolism in Chinese community residents
We screened total 1895 fasting serum samples from a community in southern Hunan province.Basic demography characteristics of the total population was shown in Table 1.Fructose can be detected in 1040 fasting serum samples (the detection rate of fructose was 54.88%) (Table 2).In the residents who can detect fructose, the descriptive statistical indicators such as the mean, the median, the highest and lowest of serum fasting fructose content were 116.96 μg/mL, 16.41 μg/mL, 1704.64 μg/mL and 0.11 μg/mL, respectively (Table 2).These data indicated individual exposure levels of fructose varied obviously.
We further screened 260 high fructose exposed population from these 1040 subjects by quartile method according to their fructose exposure level in serum, and 260 controls were strictly matched based on age and  3).We conducted amino acid targeted metabolomics testing on these 1895 subjects using UPLC-QqQMS.In high fructose exposing group, fasting blood glucose, triglycerides, heart rate, and amino acids of arginosuccinate (Asa), ethanolamine (EtN), aspartic acid (Asp), and glutamic acid (Glu) elevated significantly, while serum creatinine and amino acids of 1-methyl-l-histidine (1MHis), phosphorylethanolamine (PEtN), arginine (Arg), glutamine (Gln), gamma-aminobutyric acid (GABA), amino-adipic acid (aAd), 5-hydroxylysine (Hyl) and cystine (Cys) decreased significantly (Table 3).Amino acid metabolism was further tested with regard to gender differences (Supplementary table 3 and 4).Fructose exposure significantly increased the fasting blood glucose, heart rate, and Car, but decreased 1MHis, Arg, GABA, Hyl, and Cys in the male residents (Supplementary table 3).In the female residents, fructose exposure significantly increased Ans, Asa, EtN, Asp and Glu, but decreased 1MHis, PEtN, Gln, Aad and Hyl (Supplementary table 4).These findings suggested that the health damage caused by fructose may be related to gender.

Effects on liver acetylated proteome after exposure to fructose in rat
Using acetylated 4D Label free quantitative proteomics technology, we further evaluated the impact of fructose exposure on liver acetylated proteome.As shown in Fig. 1A, the mass deviation of all identified acetylated peptide segments is mainly distributed within 10 ppm, indicating that the identification results are accurate and reliable.Then, using the rigorous analysis tool Andromeda, the MS profile data was analyzed to obtain the score for each MS 2 .Figure 1B showed that the Andromeda score of MS 2 is relatively ideal, with over 49.6% peptides above 20 points, and the median peptide score is 59.78 points.These results indicated that our mass spectrometry data is relatively reliable.Figure 2 displayed bioinformatics analysis for liver acetylation proteomics analysis.A total of 403 peptide segments were upregulated and 129 peptide segments were downregulated (Fig. 2A).Among them, there were 306 proteins in the cytoplasm, 246 proteins in mitochondria, and 219 proteins in the nucleus with acetylation modifications (Fig. 2B).The protein domains with the most significant acetylated modified level were middle domain, C-terminal domain and N-terminal domain of acyl-CoA dehydrogenase (Fig. 2C, D).GO annotation analysis displayed biological processes of cellular process, metabolic process and biological regulation, molecular function of binding, catalytic activity and structural molecule activity, and cellular components of cell part, cell and organelle changed significantly (Fig. 2E).And KEGG pathway enrichment map displayed that valine, leucine and isoleucine degradation changed most and followed by propanoate metabolism and citrate cycle (TCA cycle) (Fig. 2F).These bioinfor-matics results provided valuable clues for further research in the mechanism of fructose interference in amino acid metabolism.

Discussion
Along with the changes of food consumption patterns toward high-calorie diets rich in sugar sweetened beverages and processed foods containing HFCS, fructose consumption has increased substantially during the last few decades 5 .Previous studies reported that the consumption of fructose is associated with intrahepatic lipid accumulation, hypertriglyceridemia and hyperuricemia 20,21 .Our recent study have shown that long-term exposure to common dose of fructose (the common soft drinks contain about 5% in common soft drinks) may lead to energy metabolic damage in adipose tissue and significant changes in serum important adipocytokines 22 .
Fructose is absorbed in the small intestine and metabolized in the liver where it stimulates fructolysis, glycolysis, lipogenesis, and glucose production 20 .In current study, we found fasting blood glucose and triglycerides significantly increased in high fructose exposing individuals.However, the uric acid levels of high fructose exposing individuals did not significantly increased.Previous animal experiments shown that fructose could cause the development of hypertension and increase the heart rates 23 .In current study, we found fructose significantly increased the heart rates but had no significant effect on blood pressure in community residents.AAs, the most common metabolites in body, are organic compounds that are derivatives of hydrocarbons, containing amino groups (-NH 2 ) and carbonyl groups (-COOH).AAs can effectively reflect the health status of human.Especially in recent years, amino acids profile analysis in human biological fluids using modern analytical methods with the ability of simultaneous quantification of dozens of amino acids has aroused great interest among scientists 24 .In current study, we quantified 39 amino acids in the serum of high fructose exposing and control populations by UPLC-QqQMS based targeted metabolomics and found several amino acid biomarkers such as Asa, EtN, Asp, Glu, 1MHis, PEtN, Arg, Gln, GABA, aAd, Hyl and Cys for fructose exposure.
Gln, a nonessential amino acid (NEAA) which has many roles in cell metabolism, participating in tricarboxylic acid (TCA) cycle supplementation.Gln is transported into the mitochondrial matrix through the SLC1A5 variant and subsequently converted to glutamate (Glu).As a decomposition product of Gln, Glu plays a central role in NEAA metabolism because it is crucial for the biosynthesis of Ala, Asp, Pro, and Ser, which are in turn used for the biosynthesis of Asn, Arg, cysteine, and Gly 25 .Previous study have found that fructose reprogrammed cellular metabolic pathways to facilitate Gln breakdown 14 .Present study displayed Glu and Asp significantly elevated while Arg and Gln significantly decreased among individuals with high fructose exposure in the community.We did not find fructose had significant effect in population on other NEAAs synthesized from Glu, such as Ala, Asp, Pro, Ser, Asn and Gly.
Glu generats α-ketoglutarate which enters the TCA cycle and thereby affecting the energy metabolism of cells, through several enzymes such as glutamic-oxaloacetic transaminase 2 (Got2), glutamic-pyruvate transaminase (Gpt) and phosphoserine aminotransferase 1 (Psat1).Fructose did not seem to affect the expression of enzymes such as Got1, Psat1 and Gpt which are involved in the liver Gln metabolism pathway.However, the acetylated modification level of lysine at position 302 of Got2 significantly increased after fructose exposure.Meanwhile, the acetylated modification level of SLC25A1 and SLC25A12, transporters on mitochondrial membranes for citrate and Asp, also significantly increased after fructose exposure.The known function of SLC25A1 consists of Both EtN and PEtN are raw materials for synthesis of phosphatidylethanolamine, which plays a very important role in cell morphology, metabolic regulation, signal transduction, and various physiological functions of cells.Ser can be metabolized into EtN and PEtN via serine decarboxylation.However, we did not find Ser changed significantly in high fructose exposing population.Significant increase in EtN and significant decrease in PEtN in high fructose exposing population indicated fructose significantly interfered with effective utilization of EtN.
Blood calcidiol (25-hydroxyvitamin D) can effectively reflect the lack of vitamin D in the body.Previous research reported dietary fructose inhibited intestinal calcium absorption and induced vitamin D insufficiency in chronic kidney disease 33 .We speculate that fructose exposure may cause compensatory increase of hepatic synthesis of calcidiol.However, further evidence is needed.PGF2a-ethanolamide, stearoylethanolamide and palmitoylethanolamide are N-acylethanol-amines which were reported to be related to hepatic cancer, steatosis and cirrhosis 34 .The serum EtN is an important source for liver synthesis of N-acetylethanolamines.The increase of N-acetylethanolamines in liver could explain why serum EtN increased.Taurodeoxycholic acid 3-glucuronide is a relatively new metabolite, which is rarely reported, and may serve as a new marker for liver response to fructose exposure.www.nature.com/scientificreports/ The liver is the most important organ for metabolizing amino acids.Amino acid, as one of the three major nutrients in the body (the other two are sugar and lipid), long-term exposure to fructose had led to significant changes in the metabolomics of mouse liver, indicating that fructose is highly likely to alter the amino acid metabolism of the liver.The significant increase of N-acetylethanolamines provided a valuable clue to the future study of the mechanism of metabolic damage caused by fructose in the body.
Cndp1 is responsible for encoding carnosine dipeptidase which decomposes Car (β-alanyl-l-histidine) into Ala and His.In present study, fasting serum Car did not changed significantly in high fructose exposing group.However, PCR array displayed 6 months fructose exposure caused liver expressed higher level of Cndp1 in rats.The expression of Dbt, which encodes the core component of branched-chain alphaketo acid dehydrogenase complex,was similar to Cndp1 in liver.We did not observe any changes in Dbt's substrates such as Leu, Ile and Val in serum.These inconsistent phenomena may be due to species differences.
Dmgdh encodes a mammalian flavin mitochondrial enzyme, which catalyzes the oxidative demethylation of dimethylglycine 35 .Dmgdh plays an important role in the utilization of methyl groups derived from choline.Sar, a product of the oxidative demethylation of dimethylglycine, is known as an important intermediate in one-carbon metabolism.In present study, the expression of Dmgdh decreased significantly in rat liver after fructose exposure.And in high fructose exposing population, Sar also displayed a downtrend (P < 0.05, corrected P value = 0.10).
Additionally, we found the effect of fructose on amino acid metabolism in the human population was different from that in the rats, of which the main reason may be species differences.However, we still found the decrease in rats of 6 amino acids such as Arg, Tyr, Aad, Hyl, Hyp and PEtN was consistent with human.These findings indicated the decrease in Arg, Tyr, Aad, Hyl, Hyp and PEtN induced by fructose exposure may be conserved among different species.
Fructose can enhance glycolysis, promoting processes to store rather than use energy.Therefore, fructose once assisted human ancestors in surviving under extreme conditions such as severe climate change, asteroid impacts, food shortage and droughts 36 .However, it is precisely due to this unique metabolic characteristic that fructose is more likely to induce glucose and lipid metabolism damage.Amino acids metabolism is crucial for health, not only because amino acids are the raw materials for synthesizing proteins in the body, but also because amino acid metabolism is an important link with glucose and lipid metabolism.However, less attention has previously been paid to the impact of fructose on amino acid metabolism.This study can provide reliable supports for comprehensive understanding the interference of fructose on amino acid metabolism in the body.

Conclusion
Population based amino acid targeted metabolomics study showed that people with high level of fructose had higher levels of Asa, EtN, Asp, and Glu, whereas they had lower levels of 1MHis, PEtN, Arg, Gln, GABA, Aad, Hyl and Cys.The further mechanism study in rats model displayed amino acid metabolic genes of Aspa, Cndp1, Dbt, Dmgdh, and toxic metabolites such as N-acetylethanolamines accumulation, interference of urea cycle, as well as acetylated modification of key enzymes in glutamine metabolic network and glutamine derived NEAAs synthesis pathway in liver may play important roles in fructose caused reprogramming in amino acid metabolism.

Figure 1 .Table 6 .
Figure 1.Quality control for liver acetylation proteomics analysis.(A) Distribution map of ion mass deviation for modified peptide segments; (B) distribution map of modified peptide ion scores.

Figure 3 .
Figure 3. Heat map of correlation coefficients between differential genes and metabolites.

Figure 4 .
Figure 4. Heat map of correlation coefficients between differential genes and acetylated proteins.

Table 1 .
Basic demography characteristics of 1895 community residents.a Values are presented as mean ± standard deviation.

Table 4 .
Metabolites identified by high-resolution mass spectrometry with significant changes caused by fructose exposure in liver.# P value with Benjamin Hochberg correction.

Table 7 .
2mino acid content in rat serum after fructose exposure.1Consistentwiththechanges in human serum.2Contrary to the changes in human serum.