A novel high-entropy TiVCrMoC(3)T(x) assisted LDI MS for serum metabolic fingerprint in rheumatoid arthritis.

A novel high-entropy TiVCrMoC3Tx assisted LDI MS for serum metabolic fingerprint in rheumatoid arthritis - PMC Skip to main content An official website of the United States government Here's how you know Here's how you know Official websites use .gov A .gov website belongs to an official government organization in the United States. Secure .gov websites use HTTPS A lock ( Lock Locked padlock icon ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites. Search Log in Dashboard Publications Account settings Log out Search… Search NCBI Primary site navigation Search Logged in as: Dashboard Publications Account settings Log in Search PMC Full-Text Archive Search in PMC Journal List User Guide PERMALINK Copy As a library, NLM provides access to scientific literature. Inclusion in an NLM database does not imply endorsement of, or agreement with, the contents by NLM or the National Institutes of Health. Learn more: PMC Disclaimer | PMC Copyright Notice J Pharm Anal . 2025 Sep 22;16(4):101456. doi: 10.1016/j.jpha.2025.101456 Search in PMC Search in PubMed View in NLM Catalog Add to search A novel high-entropy TiVCrMoC 3 T x assisted LDI MS for serum metabolic fingerprint in rheumatoid arthritis Zhilong Chu Zhilong Chu a The First Affiliated Hospital of Shandong First Medical University, Shandong First Medical University, Jinan, 250014, China Find articles by Zhilong Chu a, 1 , Xi Yu Xi Yu a The First Affiliated Hospital of Shandong First Medical University, Shandong First Medical University, Jinan, 250014, China Find articles by Xi Yu a, 1 , Xiao Wang Xiao Wang b Shandong Analysis and Test Center, Qilu University of Technology (Shandong Academy of Sciences), Jinan, 250014, China Find articles by Xiao Wang b , Wenqiang Zhang Wenqiang Zhang a The First Affiliated Hospital of Shandong First Medical University, Shandong First Medical University, Jinan, 250014, China Find articles by Wenqiang Zhang a , Yuming Li Yuming Li c First Clinical Medical College, Shandong University of Traditional Chinese Medicine, Jinan, 250355, China Find articles by Yuming Li c , Xinfeng Yan Xinfeng Yan a The First Affiliated Hospital of Shandong First Medical University, Shandong First Medical University, Jinan, 250014, China Find articles by Xinfeng Yan a , Weining Yan Weining Yan d Shandong Provincial Hospital Affiliated to Shandong First Medical University, Jinan, 250022, China Find articles by Weining Yan d , Hongzheng Meng Hongzheng Meng a The First Affiliated Hospital of Shandong First Medical University, Shandong First Medical University, Jinan, 250014, China Find articles by Hongzheng Meng a , Lingyu Li Lingyu Li e Shandong Agricultural University, Tai'an, Shandong, 271018, China Find articles by Lingyu Li e , Guanhua Zhang Guanhua Zhang f China Academy of Chinese Medical Sciences, Beijing, 100700, China Find articles by Guanhua Zhang f , Ruya Wang Ruya Wang g Jilin University School of Pharmaceutical Sciences, Changchun, 130021, China Find articles by Ruya Wang g , Miaomiao Li Miaomiao Li b Shandong Analysis and Test Center, Qilu University of Technology (Shandong Academy of Sciences), Jinan, 250014, China Find articles by Miaomiao Li b , Jun Li Jun Li b Shandong Analysis and Test Center, Qilu University of Technology (Shandong Academy of Sciences), Jinan, 250014, China Find articles by Jun Li b , Weiqiang Liang Weiqiang Liang a The First Affiliated Hospital of Shandong First Medical University, Shandong First Medical University, Jinan, 250014, China Find articles by Weiqiang Liang a, ⁎ , Chunxia Ma Chunxia Ma b Shandong Analysis and Test Center, Qilu University of Technology (Shandong Academy of Sciences), Jinan, 250014, China Find articles by Chunxia Ma b Author information Article notes Copyright and License information a The First Affiliated Hospital of Shandong First Medical University, Shandong First Medical University, Jinan, 250014, China b Shandong Analysis and Test Center, Qilu University of Technology (Shandong Academy of Sciences), Jinan, 250014, China c First Clinical Medical College, Shandong University of Traditional Chinese Medicine, Jinan, 250355, China d Shandong Provincial Hospital Affiliated to Shandong First Medical University, Jinan, 250022, China e Shandong Agricultural University, Tai'an, Shandong, 271018, China f China Academy of Chinese Medical Sciences, Beijing, 100700, China g Jilin University School of Pharmaceutical Sciences, Changchun, 130021, China ⁎ Corresponding author. [email protected] 1 Both authors contributed equally to this work. Received 2025 Feb 20; Revised 2025 Sep 19; Accepted 2025 Sep 22; Issue date 2026 Apr. © 2025 The Authors This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). PMC Copyright notice PMCID: PMC13092000  PMID: 42011256 Abstract Worldwide, rheumatoid diseases account for approximately 37.57/100,000 disability-adjusted life years (DALYs). Early diagnosis of rheumatoid arthritis (RA) can effectively reduce associated risks and improve quality of life. To achieve this, accurate and rapid advanced tools are required. Herein, TiVCrMoC 3 , a novel high-entropy two-dimensional (2D) carbide MXene that exhibits excellent electrical conductivity, is prepared and applied to surface-assisted laser desorption ionization mass spectrometry (SALDI-MS) analysis for the first time. TiVCrMoC 3 demonstrates superior performance in small-molecule detection compared with typical inorganic and organic substrates. Moreover, its exhibits ultrahigh sensitivity (limit of detection (LOD) at the 10 pg/mL level), excellent repeatability (coefficient of variation (CV) < 4%), excellent quantitative detection capability (coefficient of determination = 0.99987), clean background, and broad analyte coverage. Results showed that TiVCrMoC 3 -assisted laser desorption ionization MS (LDI-MS) is efficient in screening RA diseases, with the advantages of high throughput and low cost. It enables accurate and quantitative analysis of various low-concentration metabolites in 1 μL of biological fluid within seconds. Combined with machine learning, TiVCrMoC 3 -assisted LDI-MS is used to accurately diagnose multiple RA samples. A total of 8 potential different small-molecule metabolites between individuals with osteoarthritis (OA) and RA were identified based on relative quantification methods, and their associated metabolic changes were discussed. TiVCrMoC 3 -assisted LDI-MS provides a potential tool for the rapid diagnosis of RA and may pave the way for precision medicine. Keywords: TiVCrMoC 3 , SALDI-MS, Rheumatoid arthritis, Serum metabolic fingerprint, Machine learning Graphical abstract Open in a new tab Highlights • A novel high-entropy 2D Carbide, MXenes TiVCrMoC 3 , was first used as a substrate for SALDI-MS. • The prepared TiVCrMoC 3 served as a LDI-MS substrate to improve the detection performance of small metabolites. • TiVCrMoC 3 -assisted LDI-MS was used to encode the serum metabolic fingerprints. • Combined with machine learning, serum metabolic fingerprints were analyzed in depth. 1. Introduction Rheumatoid arthritis (RA) is an autoimmune disease that originates from the synovial membrane of joints and involves a variety of inflammatory cells, which can lead to decreased joint function and even disability [ 1 , 2 ]. Its global incidence is about 0.5%–1.0% [ 3 , 4 ]. The global disability-adjusted life years (DALYs) of RA reached about 39.57/100,000 population and increase slightly year by year [ 5 ]. The lack of typical clinical symptoms and specific diagnostic markers during the early stage of RA often leads to misdiagnosis and missed diagnosis. Currently, the serum markers of RA are mainly rheumatoid factor (RF) and anti-cyclic citrullinated peptide antibodies (anti-CCP antibodies); however, their specificity is not high, making early-stage diagnosis difficult. RF and anti-CCP antibodies are both large molecular proteins, while small-molecule metabolites specific to RA in serum are often overlooked [ 6 ]. Therefore, there is an urgent need to explore the early specific diagnostic markers of RA, develop new diagnostic techniques, and establish an early diagnosis and warning system of RA. Metabolomic analysis of tissues and cells can reflect the current physiological state of individuals, making it more suitable for personalized diagnosis and treatment in clinical practice than proteomic and genomic analysis. In particular, blood metabolomics, which reflects systemic metabolic adjustment caused by various diseases [ 7 , 8 ], exhibits great potential in the diagnosis of RA. However, traditional blood tests are not practical in clinical applications because human blood is a complex environment with low metabolite abundance [ 9 ]. A blood test that can reflect metabolic changes is greatly significant for the diagnosis of RA. Matrix-assisted laser desorption ionization mass spectrometry (MALDI-MS) is a novel MS technique with high throughput, high sensitivity, and accuracy [ 10 ]. It has been widely used in the analysis of high molecular weight compounds (mass-to-charge ratio ( m / z ) > 1000 Da) [ 11 ]. MALDI-MS can be performed to obtain the m / z information of a large number of known and unknown substances from biological samples. Moreover, it can be conducted to distinguish small differences between different molecules, providing the possibility to analyze small-molecule metabolites specific for RA. The m / z data, when combined with other techniques such as MS/MS, can aid in the identification and characterization of these metabolites [ 12 ]. However, its application in the analysis of low molecular weight compounds ( m / z < 700 Da) remains to be addressed [ 13 ]. Previous conventional substrates, such as α-cyano-4-hydroxycinnamic acid (CHCA) and 2,5-dihydroxybenzoic acid (2,5-DHB), display a background interference in the low molecular weight region, seriously hindering the analysis of low molecular weight compounds [ 14 ]. Consequently, the development of new substrates to overcome the above problems has become a research hotspot in recent years. Recent progress in laser desorption ionization MS (LDI-MS) highlights the importance of designing novel substrates, including reactive matrices, organic small molecules, and functional nanomaterials, to enhance ionization efficiency, reduce background interference, and enable sensitive metabolic profiling for biomedical and clinical applications. For instance, hydralazine was used as a reactive matrix to enable on-target derivatization and direct saccharide detection [ 15 ]. Organic metal chalcogenides and new organic matrices like 1,4-Dioxo-1,2,3,4-tetrahydrophthalazine-6-carboxylic acid (DTCA) have shown remarkable signal enhancement and improved reproducibility for metabolite analysis [ 16 ]. Hybrid materials such as sulfur-functionalized covalent organic framework (COF-S)@Au nanoparticles further integrate structural and optical advantages for high-performance LDI-MS detection [ 17 ]. Comprehensive reviews underscore the growing potential of nanomaterial-based LDI/SALDI-MS platforms in biofluid and exosome metabolic fingerprinting, especially when combined with machine learning for clinical diagnostics [ 18 ]. Surface-assisted laser desorption ionization MS (SALDI-MS), which is mediated by nanomaterials, greatly reduces the interference of background signals and makes it possible to analyze the low molecular weight compounds ( m / z < 500 Da). Nanomaterial-assisted LDI-MS determines the performance of LDI-MS from the four aspects of nanomaterial: i) the field-enhanced electron transport structure for photon-induced desorption [ 19 ]; ii) the element composition with low heat dissipation is suitable for heat-driven desorption [ 20 ]; iii) it possesses nanoscale surface roughness and good crystallization stability, facilitating the selective ionization of metabolites [ 21 ]; and iv) it is easy to synthesize and inexpensive, which is suitable for industrial production. Recently, a lot of new nanomaterial substrates have been reported, such as graphene [ 22 ], graphitic carbon nitrides [ 23 ], covalent organic frameworks [ 24 ], nanomaterials [ 25 ], and silicon-based materials, while only have some of these properties. A novel high-entropy material (HEM), which consists of a mixture of multiple (usually five or more) elements at approximately equal molar ratios, exhibits excellent thermal and optoelectronic properties that can satisfy all the characteristics of the above nanomaterial-assisted LDI-MS and is therefore able to function as an excellent matrix-assisted LDI-MS [ 26 ]. Unlike the traditional dominant position of one or two main elements, the proportion of each element in HEM is almost equal, thus giving it unique advantages such as hardness, oxidation resistance, and conductivity [ 27 ]. However, many of the previously reported HEMs, such as oxides, nitrides, or spinels, are often limited by their bulk morphology, low electrical conductivity, and restricted surface accessibility, which negatively affect their efficiency in LDI-MS [ 28 ]. In contrast, TiVCrMoC 3 , a novel two-dimensional (2D) high-entropy carbide MXene, composed of four transition metals (Ti, V, Cr, and Mo) in equimolar ratios, possesses an ultrathin layered structure, high electrical conductivity, excellent photothermal conversion efficiency, and abundant surface functional groups [ 29 ]. These properties promote enhanced analyte-substrate interaction and efficient laser energy absorption in the low-mass region, ultimately improving detection sensitivity and accuracy in small-molecule metabolomics. In this regard, TiVCrMoC 3 with unique structure is promising to address all of the aforementioned merits, but remain to be explored as efficient LDI-MS substrate. SALDI-MS combined with machine learning method can be used to extract the features and reduce the dimension of data and identify the key metabolite differences from a large number of MS data, which has been applied in diseases such as cancer to capture the specific metabolic characteristics of patients with cancer and provide the possibility of an early cancer diagnosis [ 30 ]. In this paper, the method is applied to the diagnosis of patients with RA, and it not only can automatically distinguish between patients with osteoarthritis (OA) and RA, but also can mine potential biomarkers in the data to enhance the accuracy and efficiency of diagnosis. The pattern recognition and classification algorithms of machine learning can process high-dimensional complex data to achieve efficient and automated disease prediction and classification. The combination of TiVCrMoC 3 -assisted LDI-MS analysis and machine learning demonstrates a strong application prospect in medical detection, which provides the possibility to screen the differences of serum small-molecule metabolites in patients with RA. In this study, TiVCrMoC 3 , a novel high-entropy 2D carbide MXene, was synthesized, found to exhibit good electrical conductivity, stability, and photothermal properties, and then used as a substrate for SALDI-MS analysis for the first time. TiVCrMoC 3 -assisted LDI-MS was performed to detect a variety of small-molecule substances, including amino acids, fatty acids, and drugs. It showed a peak profile with a clean background and high intensity. The performance of TiVCrMoC 3 was compared with conventional organic substrates and graphene to verify its advantages in the detection of low molecular weight compounds. As a novel substrate, TiVCrMoC 3 greatly enhances the ionization ability of SALDI-MS for small-molecule metabolites, showing ultrahigh sensitivity, low background noise, and wide coverage in the detection of small metabolites and body fluids. It possesses good stability and salt and protein tolerance, helping to obtain reliable detection results in complex environments. Moreover, it can quickly and accurately quantify the content of stearic acid in samples. The combination of TiVCrMoC 3 -assisted LDI-MS with machine learning enabled the excellent diagnosis in classifying RA disease groups from OA groups. Furthermore, the regulation of 71 key metabolites in OA and RA was observed, of which 8 were identified. This result showed that TiVCrMoC 3 -assisted LDI-MS is an effective tool to identify and quantify differential metabolites in serum, which provides the possibility for the diagnosis and screening of diseases and has good clinical prospects. Owing to its generalizability and robust performance in detecting small-molecule biomarkers, this technique also shows potential for broader use in the early diagnosis of other autoimmune or metabolic diseases. This result also showed that TiVCrMoC 3 -assisted LDI-MS is an effective tool to identify and quantify differential metabolites in serum, providing the possibility for the diagnosis and screening of diseases and demonstrating good clinical prospects. 2. Experimental 2.1. Chemicals and reagents Hydrofluoric acid (HF) and tetramethylammonium hydroxide (TMAOH) solution were purchased from Aladdin Co., Ltd. (Shanghai, China). Tryptophan, histidine, proline, threonine, isoleucine, serine, glycine-alanine, glycine-glutamic acid, glycine-aspartic acid, glycine-histone-lysine, stearic acid, heptadecanoic acid, nonadecanoic acid, undecanoic acid, docosaenoic acid, palmitic acid, arachidonic acid (AA), triglyceride, glucose, cholesterol (CHO), reduced glutathione, oxidized glutathione, asparagine, taurine, inositol, creatinine (Cre), citric acid, adenine, hypoxanthine, guanine, vitamin C (VC), azathioprine (AZA), methotrexate (MTX) hydrate, cyclosporine (CSA), acetylsalicylic acid (ASA), acetaminophen, dopamine hydrochloride (DA), and nifedipine (NF) were purchased from Shanghai Yuanye Biotechnology Co., Ltd. (Shanghai, China). All chemicals were used without further purification. Single-layer graphene oxide (GO), carbon nanotubes (CNTS), MXene, and TiVCrMoC 3 powder were obtained from Suzhou Beike Nano Technology Co., Ltd. (Suzhou, China). CHCA, 2,5-DHB, bovine serum albumin (BSA) (98%), NaCl (99.5%), and KCl (99.5%) were purchased from Sigma Aldrich (St. Louis, MO, USA). Stearic acid isotope internal standard (stearic acid-D3) was procured from McLean (Shanghai, China). Chromatographic-grade ethanol was purchased from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). Chromatographic-grade methanol (MeOH), acetone, and acetonitrile (ACN) were purchased from Merck (Darmstadt, Germany). All solutions were prepared using deionized water generated by a Milli-Q system (Millipore, Billerica, MA, USA). 2.2. Preparation method of TiVCrMoC 3 For a completely adequate etching of the material, 2 g of TiVCrMoAlC 3 MAX powder was slowly added to 20 mL of HF (48%) solution and stirred at 55 °C for four days. The etch solution was then centrifuged (4200 rpm, 5 min) and washed several times until the supernatant became neutral and finally obtaining a multilayer TiVCrMoC 3 precipitate. The multilayer TiVCrMoC 3 was added to 20 mL of 5% TMAOH solution, stirred for 6 h at room temperature to allow adequate intercalation, centrifuged, and washed three times to remove excess TMAOH. Finally, the precipitate was dispersed in 20 mL of deionized water, sonicated in an ice water bath for 30 min, and centrifuged. Subsequently, the supernatant was collected, concentrated, and freeze-dried to obtain a single solid layer of TiVCrMoC 3 for use. 2.3. Characterizations of high-entropy TiVCrMoC 3 The structure and elemental composition of TiVCrMoC 3 were analyzed through high-resolution transmission electron microscopy (HR-TEM), high-angle annular dark-field scanning TEM (HAADF-STEM), and energy spectrum X-ray spectroscopy (EDS) using a field-emission transmission electron microscope (FEI Tecnai G2 F20, FEI Company, Columbia, MD, USA). X-ray diffraction (XRD) measurements were conducted to evaluate the crystallinity of the samples by using Cu Kα radiation ( λ = 1.54056) in the 2θ range from 5° to 80° (Empyrean, Malvern PANalytical, Almelo, The Netherland). X-ray photoelectron spectroscopy (XPS) (AXIS Ultra DLD, Kratos Analytical Ltd., Manchester, UK) was performed to analyze the elemental composition of TiVCrMoC 3 . Fourier transform infrared attenuated total reflection (FTIR-ATR) spectra of TiVCrMoC 3 were evaluated using a Nicolet iS20 FTIR spectrometer (Thermo Fisher Scientific Inc., Waltham, MA, USA). The morphology of the samples was further analyzed through atomic force microscopy (AFM) (Bruker Dimension ICON, Bruker, Billerica, MA, USA). The instrument used for ultraviolet-visible (UV-vis) was a UV-vis spectrophotometer (Shimadzu Corporation, Kyoto, Japan), and the concentration of TiVCrMoC 3 was 0.25 mg/mL. Photothermal experiments were conducted using an 808-nm laser (1 W) and an infrared thermal imager with a TiVCrMoC 3 concentration of 1 mg/mL. 2.4. Preparation of small-molecule metabolite stocks for SALDI-MS analysis Tryptophan, threonine, stearic acid, stearic acid-D3, heptadecanoic acid, nonadecanoic acid, undecanoic acid, docosaenoic acid, palmitic acid, triglyceride, adenine, inosine, guanine, and CSA were dissolved in MeOH to prepare stock solutions at 10 mM concentrations. Histidine, proline, isoleucine, serine, alanine, glutamine, l -carnitine, uric acid, uracil, palmitic acid, citrulline, glycine-alanine, glycine-glutamate, glycine-aspartic acid, glycine-histidine-lysine, glucose, reduced glutathione, oxidized glutathione, asparagine, taurine, inositol, Cre, citric acid, VC, and DA were dissolved in pure water to prepare stock solutions at 10 mM concentrations. AA, AZA, ASA, and acetaminophen were dissolved in ethanol to prepare 10 mM stock solutions each. CHO and NF were dissolved in acetone and each prepared as a 10 mM stock solution. The stock solution was diluted in different proportions to prepare solutions of appropriate concentrations for each analyte. All stock solutions were stored at 4 °C before use. For quantitative analysis, stearic acid-D3 isotope was selected as an internal standard. Stearic acid was gradient diluted in MeOH to 0.25 mM–10 mM solutions for standard curves. A stock solution for quantitative analysis was prepared by dissolving stearic acid-D3 in MeOH at a concentration of 5 mM and mixed with stearic acid solution (1:1, v / v ). 2.5. Evaluation of the efficacy of TiVCrMoC 3 -assisted LDI-MS TiVCrMoC 3 was dissolved in pure water to obtain a 1 mg/mL solution, and 1 μL of the obtained TiVCrMoC 3 solution was drip added to the target plate as a substrate. The added drop was dried in a vacuum pump, and then 1 μL of analyte was drip added to the substrate precipitate. Inorganic substrates, including TiVCrMoC 3 , adopt the principle of “matrix first and sample second”, whereas organic substrates, such as CHCA and 2,5-DHB, adopt the principle of “sample first and matrix second”. A total of six samples were prepared for SALDI-MS. The first sample consists of tryptophan, histidine, proline, threonine, isoleucine, serine, glycine-alanine, glycine-glutamic acid, glycine-aspartic acid, glycine-histone-lysine, stearic acid, heptadecanoic acid, nonadecanoic acid, decanoic acid, docosaenoic acid, palmitic acid, AA, triglyceride, glucose, CHO, reduced glutathione, oxidized glutathione, asparagine, taurine, inositol, Cre, citric acid, adenine, hypoxanthine, guanine, and VC solutions and an amino acid mixture (including tryptophan, proline, threonine, phenylalanine, and arginine). The second sample was used for MS analysis of common small-molecule drugs and drug mixture, including AZA, MTX hydrate, CSA, ASA, acetaminophen, DA, and NF solutions and a mixture of six different drugs (acetaminophen, ASA, AZA, ciprofloxacin, MTX, and CSA). The third sample included a mixture of serine, histidine, and arginine, which was mixed with salt (NaCl and KCl, 0.5 mM) and protein (BSA, 5 mg/mL), respectively, to validate the detection efficiency of TiVCrMoC 3 -assisted LDI-MS in complex biological environments. The fourth sample was a mixture of stearic acid in different proportions, with stearic acid-D3 used for quantitative analysis to generate a standard curve. The fifth sample consisted of an AZA solution, which was used to detect the reproducibility of the assay. The sixth human blood sample was donated by 100 patients (50 patients with RA and 50 patients with OA) from the First Affiliated Hospital of Shandong First Medical University (Jinan, China) with patient consent form acquired. The skin of donors was disinfected after they fasted overnight for at least 8 h. Subsequently, 5 mL of peripheral venous blood was collected, stored in a purple vacuum anticoagulant collection tube, and transferred to a 15-mL centrifuge tube for centrifugation (4000 rpm, 10 min). The supernatant (serum) was the taken after centrifugation and stored in Eppendorf (EP) tubes at −80 °C. For this sample, the serum was diluted 50 times with MeOH, the supernatant was obtained through centrifugation as a SALDI-MS sample, and 10 spots were repeated for each sample. All the experimental samples in this study have been approved by the ethics committee of the First Affiliated Hospital of Shandong First Medical University (Jinan, China) (Approval No.: 2021donglunshenzi (S1123)). In addition, 15 small-molecule substances (histidine, glucose, alanine, serine, glutamine, l -carnitine, uric acid, asparagine, tryptophan, inositol, VC, taurine, uracil, palmitic acid, and citrulline) were selected for SALDI-MS/MS analysis. In addition, the stability of TiVCrMoC 3 was assessed using a mixture of 0.5 mM tryptophan and 0.5 mM glucose as the substrate under room temperature and 4 °C storage conditions. Measurements were conducted at six time points (1, 2, 4, 10, 22, and 46 days). For each time point, the signal-to-noise (S/N) ratio was recorded for three independently prepared batches ( n = 9) of TiVCrMoC 3 . In order to assess its advantages in enrichment and separation aspects, this enrichment protocol was adapted from a previously reported magnetic solid-phase extraction (MSPE) method [ 31 ]. Briefly, 30 mg of each material (TiVCrMoC 3 , Ti 3 C 2 , and GO) was added to 30 mL of a mixed metabolite solution (10 mM serine and tryptophan) prepared in phosphate-buffered saline. The mixture was vortexed for 1 min and then sonicated for 30 min to ensure thorough dispersion and interaction between the adsorbent and analytes. Following centrifugation, the concentration of residual metabolites in the supernatant was determined using liquid chromatography (LC)-MS. The adsorption capacity ( Q , mg/g) was calculated using the equation: Q = c 0 – c × V / m where c 0 and c (mg/L) are the initial and final metabolite concentrations, V (L) is the volume of the solution, and m (g) is the mass of the adsorbent. 2.6. SALDI-MS instruments SALDI-MS was performed on a rapifleX MALDI Tissuetyper™ TOF/TOF MS (Bruker Daltonics, Billerica, MA, USA), and experiments in all samples were performed in positive ion mode. The MS instrument uses a nitrogen laser, which is equipped with a 355-nm Smartbeam-II laser setup (Bruker Daltonics). The laser was run at a repetition rate of 5000 Hz, and the data acquisition range covered m / z of 80–1000. MS calibration was performed using a mixture of 2,5-DHB and CHCA matrices supplied by Bruker Daltonics. Prior to SALDI-MS detection of liquid samples, 1 μL of TiVCrMoC 3 solution was dropped onto the target plate and then drained in a vacuum pump. Approximately 1 μL of analyte solution was then dropped onto the TiVCrMoC 3 precipitate and drained again. The dried target plates were then detected through SALDI-MS. SALDI-MS peak maps were generated using FlexAnalysis software (Bruker Daltonics). 2.7. Machine learning analytics Raw data obtained by SALDI-MS were imported into this software using FlexAnalysis Version 4.0 (Bruker Daltonics) for peak extraction, baseline calibration, smoothing, peak alignment, and normalization, resulting in data called serum metabolic fingerprint (SMF). SMF obtained from SALDI-MS data were analyzed using machine learning algorithms. Unsupervised principal component analysis (PCA) was first performed to transform the data ( m / z signal) from MS analysis into linearly uncorrelated variables to reduce the data dimensionality and identify natural clusters. All machine learning algorithms were written in the R environment. Gradient boosting with component-wise linear model (glmBoost), random forest (RF), least absolute shrinkage and selection operator (LASSO), Naïve Bayes, and XGBoost algorithms were used to model the data, and the average performance of the model was evaluated using precision, recall, F1 score, accuracy, and area under the curve (AUC). 2.8. Data analysis SALDI-MS peaks were generated using FlexAnalysis software (Bruker Daltonics). Origin64 was used for the plot. IBM SPSS Statistics 26 was used to perform t -test (95% confidence interval (95%CI)) and logistic regression analysis of differential metabolites. SMF recognition was performed based on human metabolites database ( https://hmdb.ca/spectra/ms/search ) and MetFrag ( https://ipb-halle.github.io/MetFrag/ ) according to the m / z values. PCA was performed using MetaboAnalyst 5.0. The above software is also used for volcano maps, heat maps, blueprints, and chamber maps. 3. Results and discussion 3.1. Characterization of TiVCrMoC 3 High-entropy TiVCrMoC 3 was obtained by etching and intercalating TiVCrMoAlC 3 MAX powder ( Fig. 1 A). The morphology and structure of TiVCrMoC 3 were studied through TEM. Fig. 1 B shows the TEM results of TiVCrMoC 3 , which displays a thin sheet and irregular shape and no obvious defects, pores, or oxidized phase impurities on the sheet surface. The edge part of the sheet is slightly curled, indicating that the sheet exhibits better flexibility. This is consistent with previous reports [ 32 ]. The typical element distribution of TiVCrMoC 3 was investigated using EDS and inductively coupled plasma optical emission spectrometry (ICP-OES) ( Table S1 ). As shown in Fig. 1 B, the Ti, V, Cr, Mo, and C elements are uniformly distributed on their lamellar surface. Further AFM data ( Fig. 1 C) showed that TiVCrMoC 3 presented as a sheet with a lateral size of 0.5 μm and a thickness of about 0.25 nm. At the same time, XRD confirms the phase of TiVCrMoC 3 , as shown in Fig. 1 D. The presence of the (002) crystal plane reflection around 6° confirms the existence of a single MXene layer, while the characteristic peaks at 39° and 42.5° indicate the presence of multiple MXene layers. This further supports the transition from a multiple layer to a single-layer structure. The polymerized material, with its high permeability, demonstrates excellent electrical conductivity, which significantly contributes to the understanding of the absorption/ionization process. Fig. 1. Open in a new tab Fabrication and structural characterization results of TiVCrMoC 3 . (A) Fabrication of TiVCrMoC 3 and a schematized workflow for the use of TiVCrMoC 3 -assisted laser desorption ionization mass spectrometry (LDI-MS). (B) Scanning transmission electron microscopy (STEM) results of TiVCrMoC 3 as sheets. The combination of the high-angle ring dark field (HAADF) STEM images of TiVCrMoC 3 with the energy spectrum X-ray spectroscopy (EDS) results of TiVCrMoC 3 revealed a layered morphology and uniform atomic distribution of Ti, Cr, V, Mo, and C atoms. (C) Atomic force microscopy (AFM) results of TiVCrMoC 3 with a transverse size of about 0.5 μm and a thickness of about 0.25 nm. (D) X-ray diffraction (XRD) patterns of TiVCrMoC 3 and MXene. HF: hydrofluoric acid; TMAOH: tetramethylammonium hydroxide. The major components and chemical bonds within TiVCrMoC 3 were analyzed through XPS. Figs. 2 A and S1–S3 show the XPS HR spectral data of the top surface of TiVCrMoC 3 . The C1s HR spectrum indicates the presence of C–Mo/Ti–Tx (282.3 eV), C–Ti/MoTx (282.7 eV), C–C–CHX (285.5 eV), C–O (286.3 eV), and COO (287.6 eV) motifs in TiVCrMoC 3 . The presence of C–C–CHX indicates the presence of a small amount of graphene or hydrocarbon contamination in the material, the presence of C–O indicates the presence of hydroxyl or ether bonds on the lamellae surface, and the presence of COO indicates the presence of a small amount of carbonyl groups. The Ti2p spectrograph shows two major peaks, highlighting that Ti exists mainly as Ti–C (454.8 eV), while the small peak appearing at 459.3 eV (Ti 4+ ) is attributed to the formation of a small amount of oxidized TiO 2 phase on the sheet surface. The V2p region indicates that V is coordinated in V 2+ (513.6 eV)/V 4+ (515.4 eV) or V 2 O 3 (517.1 eV) state. However, the Mo3d region indicates that Mo is present in TiVCrMoC 3 as Mo 5+ (230.5 eV)/Mo 6+ (232 eV), C–Mo–Tx (229.3 eV), and Mo (228 eV). For TiVCrMoC 3 , Cr is present as Cr 2+ (574.8 eV) and Cr 0 (576.4 eV). This is consistent with previous reports [ 29 ]. These results indicate that the distribution of Ti, V, Cr, Mo, and C is relatively uniform across TiVCrMoC 3 films. The presence of mixed oxidation states, such as Mo 5+ /Mo 6+ , plays a critical role in enhancing the electron transfer and photothermal conversion efficiency of TiVCrMoC 3 during SALDI-MS [ 33 ]. Mixed-valence states introduce localized electronic states near the Fermi level, promoting charge delocalization and increasing electrical conductivity, which facilitates photon-induced electron transfer during laser desorption/ionization. Moreover, the coexistence of Mo 5+ and Mo 6+ implies the presence of oxygen vacancies and structural defects, which serve as efficient photon absorption centers [ 34 ]. These defects enhance light-matter interactions and enable stronger photothermal effects through increased non-radiative relaxation, generating localized heat under laser irradiation [ 35 ]. This promotes more efficient analyte desorption and ionization, resulting in higher sensitivity and lower detection limits. Furthermore, the incorporation of multiple transition metals with variable oxidation states (e.g., Cr 3+ /Cr 6+ and V 3+ /V 5+ ) creates a synergistic electronic environment that enhances energy confinement and transfer [ 36 ]. This high-entropy effect not only contributes to the structural stability of the 2D material but also optimizes the surface electronic configuration for improved laser-analyte interaction. Fig. 2. Open in a new tab X-ray photoelectron spectroscopy (XPS), optical and photothermal properties, and Fourier transform infrared attenuated total reflection (FTIR-ATR) characterization results of TiVCrMoC 3 . (A) XPS characterizations: C1s spectrum and Mo3d spectrum. (B) FTIR spectra. (C) Ultraviolet-visible (UV-vis) absorption spectrum. (D) Temperature evolution curve under identical 808 nm laser irradiation conditions (1 W cm −2 ). (E) Two-dimensional (2D) thermal images at different time points. (F) 3D surface temperature plots illustrating localized heating. GO: graphene oxide. FTIR measurements were used to characterize the functional groups associated with the TiVCrMoC 3 surface ( Fig. 2 B). The absorption peaks of 3434.71 and 1632.95 cm −1 were attributed to the stretching vibration of –OH. In addition, methylene –CH 2 , with absorption peaks of 2923.04 and 2853.07 cm −1 , was found in the spectra of TiVCrMoC 3 , which may be caused by the stretching vibration of methylene on the sample surface. The absorption peak at 1167.92 cm −1 is attributed to C–OH and C–N stretching. The absorption peak at 1738.89 cm −1 may be attributed to the stretching of the aldehyde group C=O. The absorption peak at 1462 cm −1 is attributed to the bending of methyl and methylene groups. The exceptional photothermal properties of TiVCrMoC 3 were verified through UV-vis absorption spectra and photothermal experiments. The UV-vis spectra demonstrate that TiVCrMoC 3 exhibits strong absorption in the UV region, which gradually decreases as the wavelength shifts into the visible range ( Fig. 2 C). Under 808 nm laser irradiation conditions (1 W cm −2 ), TiVCrMoC 3 showed an efficient photothermal conversion capability, with the temperature rapidly rising from room temperature to 65 °C within 500 s and eventually stabilizing around 69 °C. For comparison, the photothermal performance of TiVCrMoC 3 was evaluated alongside several well-established photothermal materials, including GO and Ti 3 C 2 MXene. As shown in Fig. 2 D, TiVCrMoC 3 exhibited the most significant temperature increase under identical 808 nm laser irradiation conditions (1 W cm −2 ), indicating its stronger photothermal response. The natural cooling curves after laser irradiation are presented in Fig. S4 , from which the heat dissipation characteristics were evaluated. This heating effect was further confirmed by thermal imaging, which showed a uniform temperature distribution across the sample surface ( Fig. 2 E). In addition, Fig. S5 displays 2D thermal images of GO and Ti 3 C 2 MXene at different time intervals to visualize localized heating effects. These comparative results clearly highlight the excellent photothermal performance of TiVCrMoC 3 . Additionally, a 3D surface map revealed a localized heating phenomenon, showing a distinct temperature peak that inte ( Fig. 2 F). These properties make high-entropy TiVCrMoC 3 an ideal substrate for UV laser excitation in SALDI-MS. Moreover, its strong UV absorption capacity facilitates efficient energy transfer to the analyte, while the rapid and controlled heating process enhances desorption and ionization processes, minimizing thermal degradation of thermosensitive biomolecules. The uniform temperature profile ensures consistent ionization across the sample surface, which has the potential to improve spectral quality and reproducibility in SALDI-MS applications. Taken together, these characterization results show that the prepared TiVCrMoC 3 films are homogeneous and stable. UV-vis absorption spectroscopy and photothermal experiments confirm the photothermal properties of TiVCrMoC 3 , further validating their high potential as excellent substrates. 3.2. Detection of small-molecule metabolites First, the analytical performance of TiVCrMoC 3 was compared with conventional organic substrates (2,5-DHB and CHCA), single-layer GO, CNTS, and MXene in positive ion mode. An amino acid mixture of tryptophan ( m / z 204.23), proline ( m / z 115.13), threonine ( m / z 119.12), phenylalanine ( m / z 165.19), and arginine ( m / z 174.2) was used as the analyte. Fig. S6 shows the mass spectra of mixed amino acids using TiVCrMoC 3 , 2,5-DHB, CHCA, single-layer GO, CNTS, and MXene as substrates. Strong clean mass peaks of [M+Na] + and/or [M+K] + with relatively high-signal intensity were detected for all five amino acids when the TiVCrMoC 3 matrix was used. CHCA did not detect obvious analyte peaks due to the strong matrix interference in the low molecular weight range. Only three amino acids were detected using 2,5-DHB, CNTS, and MXene, while four amino acids were detected using single-layer GO yet with a large background interference and low intensity. It can be observed that TiVCrMoC 3 can provide a cleaner background compared with single-layer GO. TiVCrMoC 3 produced a higher ionic strength compared with 2,5-DHB, MXene, and CNTS. In addition, comparison of the background peaks of five different substrates (TiVCrMoC 3 , Ti 2 N, Ti 3 C 2 , single-layer GO, and CNTS) revealed that TiVCrMoC 3 displayed a cleaner background peak, especially in the m / z < 500 ( Fig. S7 ). To further verify the suitability of the positive ion mode for TiVCrMoC 3 -assisted LDI-MS, a comparative experiment was conducted using a mixture of glucose and tryptophan under both ionization modes. As shown in Fig. S8 , the positive ion mode yielded significantly higher signal intensities for target analytes along with notably lower background noise compared to the negative mode. This result highlights the enhanced analytical performance of TiVCrMoC 3 under positive ion conditions. These results indicated that the proposed TiVCrMoC 3 substrate largely improved the detection efficiency of the target analyte as expected with a low substrate background. Detection of specific molecules or small-molecule metabolites in body fluids, such as blood, urine, and saliva, can aid in the clinical diagnosis or in monitoring disease progression. Amino acids are protein metabolites that can reflect the synthesis, degradation, and metabolic abnormalities of proteins. Fig. 3 A shows the SALDI-MS detection results for tryptophan ( m / z 204.23), histidine ( m / z 155.16), and isoleucine ( m / z 131.17). The [M+Na] + and [M+K] + peaks of these amino acids could be detected efficiently. MS results of the three amino acids showed low background noise, clear ion peak signal peaks, and strong signal peaks. Fig. 3. Open in a new tab Mass spectra of amino acids, fatty acids, glucose, and triglycerides: (A) tryptophan (Trp), histidine (His), isoleucine (Ile) and (B) stearic acid, heptadecanoic acid, palmitic acid, arachidonic acid (AA), triglyceride, and glucose (Glu). m / z : mass-to-charge ratio. Fatty acids and sugars are involved in energy metabolism and cell membrane structure of the body and are indicators reflecting the body's metabolic levels. High levels of saturated fatty acids (such as stearic acid) may increase the risk of coronary heart disease, while polyunsaturated fatty acids (such as docosahexaenoic acid (DHA) and eicosapentaenoic acid (EPA)) have a protective effect. Therefore, fatty acids such as stearic acid and sugars have the potential of becoming disease biomarkers by helping to identify the early risk of related diseases. Figs. 3 B and S9 show the TiVCrMoC 3 -assisted LDI-MS analysis results for stearic acid ( m / z 284.48), heptadecanoic acid ( m / z 270.45), palmitic acid ( m / z 256.42), AA ( m / z 304.47), triglyceride ( m / z 885.43), glucose ( m / z 180.16), nonadecanoic acid ( m / z 298.5), heneicosylic acid ( m / z 326.56), and behenic acid ( m / z 340.58). SALDI-MS results showed high-signal, clear [M+Na] + and [M+K] + ion peaks for all these fatty acids. Peptides play important roles in human metabolism, mainly in the construction of protein, activity of enzymes, regulation of hormone levels, and participation in signal transduction and immune response. Fig. 4 A shows the SALDI-MS results for glycine-alanine ( m / z 146.16), glycine-glutamic acid ( m / z 204.2), and glycine-histone-lysine ( m / z 340.42). Moreover, the [M+Na] + and [M+K] + ion peaks of these metabolites were clearly detected, indicating that TiVCrMoC 3 as a substrate was very accurate and sensitive for the recognition of multi-amino acid derivatives. Fig. 4. Open in a new tab Mass spectrometry (MS) analysis of glycine derivatives, other disease markers, and small molecule drugs: (A) glycine-glutamate (Gly-Glu), alanine-glycine (Ala-Gly), and Gly-histone-lysine (Gly-His-Lys), (B) cholesterol (CHO), taurine, inositol, creatinine (Cre), adenine, vitamin C (VC), (C) nifedipine (NF), paracetamol, and dopamine hydrochloride (DA), and (D) a mixture of five different drugs (acetylsalicylic acid (ASA), azathioprine (AZA), ciprofloxacin, methotrexate (MTX), and cyclosporine (CSA)). m / z : mass-to-charge ratio. In addition, many small-molecule metabolites in the human body, such as Cre, CHO, and VC, are recognized as important disease biomarkers. Fig. 4 B shows the SALDI-MS results for [M+Na] + and [M+K] + peaks of CHO ( m / z 386.65), taurine ( m / z 125.15), inositol ( m / z 180.16), Cre ( m / z 113), adenine ( m / z 135.13), and VC ( m / z 176.13). Fig. S10 shows the SALDI-MS results of other disease markers, such as reduced glutathione ( m / z 307.32), oxidized glutathione ( m / z 612.63), asparagine ( m / z 132.12), citric acid ( m / z 192.13), hypoxanthine ( m / z 136.11), and guanine ( m / z 151.13). The [M+Na] + and [M+K] + peaks of these small molecules were clearly detected. The limit of detection (LOD) and quantitative limit of tryptophan, serine, AZA, and stearic acid were determined to further evaluate the sensitivity of TiVCrMoC 3 -assisted LDI-MS for the analysis of small-molecule metabolites. Each of the four metabolites was diluted to a concentration range of 500 μg/mL–5 pg/mL for SALDI-MS analysis, and the concentration of analyte peak S/N of 10 was selected as the limit of quantification (LOQ). The concentration with a S/N ratio of 3 was determined as LOD. The results showed that the LOQ and LOD of tryptophan were 800 and 200 ng/mL, respectively. In comparison, previously reported substrates such as graphdiyne (LOD = 1021.15 ng/mL) [ 37 ] and SBA-15@3-aminopropyl)triethoxysilane (APTES)@metal-organic framework (MOF)) (LOD = 1 × 10 4 ng/mL) [ 38 ] exhibited significantly higher LOD values in detection of tryptophan. Although the LOQ values for these materials were not reported, our results suggest an improvement in sensitivity by approximately one to two orders of magnitude. The LOQ and LOD of serine were 1 ng/mL and 10 pg/mL, which is notably lower than the reported LOD of 1 × 10 3 ng/mL for SBA-15@APTES@MOF [ 38 ]. These findings highlight the strong potential of the TiVCrMoC 3 -based nanoplatform for high-performance, low-detection-limit metabolite analysis ( Table S2 ). The LOQ and LOD of stearic acid were 500 and 120 ng/mL, respectively. Furthermore, the LOQ and LOD of AZA were 1 ng/mL and 250 pg/mL, respectively. These results indicate that TiVCrMoC 3 -assisted LDI-MS can detect and quantify trace amounts of amino acids, fatty acids, and small-molecule drugs in human body fluids. Various types of small molecule metabolites, including amino acids, fatty acids, and sugars, were used in this study to evaluate the efficiency of TiVCrMoC 3 -assisted LDI-MS. Most of these small-molecule metabolites are important biomarkers in the human body that reflect the metabolic levels of the organism and are thereby widely used in the diagnosis of diseases. An efficient SALDI platform capable of high-throughput screening of various biomarkers becomes important. The above results indicate that TiVCrMoC 3 -assisted LDI-MS exhibits high sensitivity and high-throughput performance in screening small-molecule metabolites. This may be attributed to the fact that TiVCrMoC 3 exhibits a regular four-atom layer structure with high light absorption efficiency and its unique structure can promote the absorption and transfer of laser energy, thereby improving the sensitivity as a SALDI-MS detection matrix. Nonetheless, it has great potential as a novel diagnostic tool for screening diseases. 3.3. Detection of drugs Detection of drugs can characterize the concentration, time, and location of drug action, which aids clinicians in adjusting the subsequent treatment plan. TiVCrMoC 3 -assisted LDI-MS was used for drug analysis. NF ( m / z 346.34), paracetamol ( m / z 151.16), DA ( m / z 189.64), and a drug mixture (ASA, AZA, ciprofloxacin, MTX, and CSA) were analyzed through TiVCrMoC 3 -assisted LDI-MS ( Figs. 4 C and D). It shows that the [M+Na] + and [M+K] + peaks of all drugs were clearly detected with high intensity. These results indicated that TiVCrMoC 3 -assisted LDI-MS not only efficiently detects small-molecule metabolites but also exhibits high-intensity ion peaks of trace small-molecule drugs. AZA, MTX, and CSA are all immunosuppressive agents mainly used to treat RA and other rheumatic diseases. By inhibiting the activity of the immune system, they can reduce inflammation and pain while simultaneously reducing the number of white blood cells, red blood cells, and platelets, which may lead to decreased immune system function. Patients with RA often need to use analgesics to relieve pain (ASA and acetaminophen). Real-time and efficient detection of drug concentrations can prevent the occurrence of some side effects caused by analgesics [ 39 ]. Most drugs for the treatment of cardiovascular diseases need to be selected and adjusted based on their content in the blood. In previous studies, there is a lack of rapid methods to monitor the blood concentration of various small-molecule drugs. The experimental results showed that TiVCrMoC 3 -assisted LDI-MS can quickly, efficiently, and accurately detect the drug residues in body fluids, helping clinicians adjust the treatment plan in time and showing great potential to be used in monitoring blood drug concentration. 3.4. Evaluation of salt tolerance, linearity, and repeatability of TiVCrMoC 3 Blood is a very complex biological fluid containing a variety of components, such as proteins, salt ions, and cells. Therefore, nanomaterials with good salt and protein tolerance can be used to improve the sensitivity, specificity, and accuracy of detection of small-molecule metabolites in blood, so as to ensure their effective detection in complex biological environments. The salt and protein tolerance of the TiVCrMoC 3 substrate was analyzed in this study. Three amino acids were mixed with NaCl, KCl, and BSA to simulate a high-salt, high-protein environment. Three amino acids, i.e., serine, histidine, and arginine, each with a concentration of 0.5 mM, were mixed in high-concentration NaCl solution (0.5 M). As shown in Fig. 5 A, the sodium addition peak signals of [Ser+Na] + ( m / z 128.1), [His+Na] + ( m / z 178.1), and [Arg+Na] + ( m / z 197.2) could be detected sensitively. Meanwhile, three amino acids, i.e., serine, histidine, and arginine, each with a concentration of 0.5 mM, were mixed in high-concentration KCl solution (0.5 M). As shown in Fig. 5 B, the potassium addition peak signals of [Ser+K] + ( m / z 144.1), [His+K] + ( m / z 194.1), and [Arg+K] + ( m / z 213.2) could be detected sensitively. Three amino acids, i.e., serine, histidine, and arginine, each with a concentration of 0.5 mM, were mixed in a complex solution containing protein (5 mg/mL BSA) and salt (KCl solution, 0.5 M). As shown in Fig. 5 C, results can be sensitive to detect [Ser+Na] + ( m / z 128.1), [His+Na] + ( m / z 178.1), [Arg+Na] + ( m / z 197.2), [Ser+K] + ( m / z 144.1), [His+K] + ( m / z 194.1), and [Arg+K] + ( m / z 213.2). The results also confirm that TiVCrMoC 3 exhibits good salt and protein tolerance and can be used for the detection of small-molecule metabolites in complex environments such as blood. As shown in Figs. S11–S13 , the S/N ratios of TiVCrMoC 3 -assisted LDI-MS remained highly consistent across all six time points, with no observable downward trend. The low standard deviation among replicate measurements further supports the excellent reproducibility of the material. In addition, the comparison between room temperature and 4 °C storage conditions revealed no statistically significant difference in performance, indicating that TiVCrMoC 3 retains its activity and stability under varied storage environments. These results collectively confirm the material's robust long-term stability and batch-to-batch consistency. Fig. 5. Open in a new tab Salt tolerance, protein tolerance, repeatability and quantification of TiVCrMoC 3 assay. (A, B) Mass spectra of a mixture of three amino acids (including serine (Ser), histidine (His), and arginine (Arg), 0.5 mM each) in a 0.5 mM NaCl solution (A) and in a 0.5 mM KCl solution (B). (C) Mass spectra of the mixture in a solution of 5 mg/mL bovine serum albumin (BSA) mixed with 0.5 mM KCl. (D) In the positive ion mode, the signal-to-noise (S/N) test was performed on eight different points on three separate substrates with azathioprine (AZA) as the analyte. (E) S/N was tested for 10 different positions in a single spot. (F) Mass spectrum of a mixture of stearic acid and stearic acid-D 3 . Inset shows standard curves for stearic acid versus stearic acid-D 3 . Error bars correspond to the standard error of 15 mass spectra acquired in a point. m / z : mass-to-charge ratio; CV: coefficient of variation. For medical testing, the results of testing are related to the accuracy of diagnosis and the formulation of treatment plans. Therefore, the repeatability of analysis method is particularly important to reduce the error and variation in the detection as much as possible. To assess the repeatability of the TiVCrMoC 3 substrate, 8 spots were analyzed on three separate substrates, and 10 different positions were analyzed in a single spot of analyte loaded on one substrate ( Figs. 5 D and E). The interspot coefficient of variation (CV) value was less than 5%, and the CV value of homogeneity comparison between different positions in the spot was less than 3%, indicating that TiVCrMoC 3 exhibited good detection repeatability and ensuring the reliability of the test results. These results indicated that TiVCrMoC 3 has good repeatability for the detection of small-molecule substances and has good clinical application prospects. Moreover, as shown in Fig. S14 , TiVCrMoC 3 demonstrated significantly higher adsorption capacity for all tested metabolites compared to Ti 3 C 2 and GO. This superior enrichment performance is attributed to its high specific surface area, multi-metallic composition, and abundant surface functional groups (–OH and –O), which facilitate strong interactions with small molecules via hydrogen bonding, electrostatic attraction, and van der Waals forces [ 40 ]. Moreover, the high-entropy structure of TiVCrMoC 3 may introduce greater surface chemical heterogeneity, further enhancing its passive adsorption efficiency. These results suggest that, beyond its excellent laser desorption/ionization performance, TiVCrMoC 3 also possesses notable passive enrichment capability, which can enhance detection sensitivity, particularly for low-abundance metabolites in complex biological samples. Stearic acid is a saturated fatty acid that is recognized as one of the important metabolites in the human body. However, excessive intake of saturated fatty acids is associated with the risk of various chronic diseases, such as obesity and cardiovascular disease. Detection of stearic acid in blood not only assesses dietary fatty acid intake, but also provides important information regarding cardiovascular disease, metabolic health, and inflammatory status. To verify the quantitative detection ability of TiVCrMoC 3 -assisted LDI MS, this substrate was used for the quantitative detection of stearic acid levels in humans. The quantitative relationship between different concentrations of stearic acid and fixed concentration of internal standard was detected, and the standard curve was generated. Due to its similar ionization efficiency with stearic acid, stearic acid-D3 was used as an internal standard to improve the accuracy of quantitative detection. As shown in Fig. 5 F, stearic acid and stearic acid-D3 were detected as sodium adduct peaks at m / z 307.5 and m / z 310.5, respectively. Standard curves were plotted based on the ratio of ionic strengths of [stearic acid+Na] + and [stearic acid-D3+Na] + at different dilutions. As shown in Fig. 5 F, stearic acid was diluted in concentrations ranging from 0.25 to 10 mM and stearic acid-D3 in a concentration of 5 mM. These results showed that the correlation between the peak area ratio and the sample concentration proved that the quantitative linearity of stearic acid was excellent ( R 2 = 0.99987), which could meet the analytical requirements of quantitative detection of stearic acid in blood. 3.5. Machine learning analysis of SMFs RA was identified based on machine learning analysis of SMFs. Serum samples from 50 patients with OA and 50 patients with RA were analyzed using TiVCrMoC 3 -assisted LDI-MS, and the resulting SMF was analyzed through machine learning ( Fig. 6 A and Table S3 ). There was no significant difference in age and gender between OA and RA patients ( P > 0.05). Notably, we draw the blueprint of SMF and each serum metabolic fingerprinting contained m / z 227 features as extracted from native LDI MS results by the Otsu algorithm, demonstrating the reliability of the machine learning approach for identifying differential metabolites in SMFs ( Fig. 6 B). RF, LASSO, Naïve Bayes, orthogonal partial least-squares discriminant analysis (OPLS-DA), Elastic Net, and support vector machine (SVM) algorithms were used to analyze the MS data. The average performance of the model was evaluated using precision, recall, F1 score, accuracy, and AUC. The results showed that both the test and training groups exhibited high AUC (AUC = 1), indicating that these algorithms could distinguish serum samples from patients with OA and RA in all cases. Next, PCA was performed ( Fig. 6 C), which revealed highly differential metabolites in the serum of patients with OA and RA, and there were significant between-group differences. Fig. 6. Open in a new tab Metabolic fingerprinting of serum for diagnosis. (A) An overall schematic of serum metabolic fingerprints (SMFs) for rheumatoid arthritis (RA) diagnosis extracted by machine learning. (B) The SMFs were extracted from the raw mass spectrometry (MS) data of 50 osteoarthritis (OA) controls and 50 RA patients and contained a blueprint of 227 mass-to-charge ratio ( m / z ) signals. (C) Principal component analysis (PCA) analysis plot of SMF differential metabolites in RA patients. (D) Volcano plot of SMF differential metabolites in RA patients. LDI-MS: laser desorption ionization MS; PC: principal component. Double difference method, t -test, and LASSO were used to screen the data generated by machine learning. Statistical parameters such as fold change (FC) ≥ 1.2 or FC ≤ 0.83 and P < 0.05 were considered comprehensively. The top 71 m / z signals with significant differences in intensity between patients with OA and RA were identified and screened for metabolic analysis. As shown in Fig. 6 D, volcano plots were generated for the detected metabolites based on P -values and FC values, where red and blue signals represent upregulated and downregulated signal intensities, respectively. Metabolites represented by the upregulated and downregulated m / z signals were represented as serum metabolites specific for RA and a heat map was generated ( Fig. 7 A). Fig. 7. Open in a new tab Evaluation of differential metabolites in rheumatoid arthritis (RA). (A) A heat map of the serum metabolic fingerprint (SMF) matrix was drawn using 71 mass-to-charge ratio ( m / z ) signals. (B–I) The box-plots of eight differential metabolites with elevated expression: [acetic acid+K] + (B), [lactic acid+Na] + (C), [4-hydroxynonenal+H] + (D), [glutamine+K] + (E), [citrulline+K] + (F), [itaconic acid+Na] + (G), [glucose+K] + (H), and [palmitic acid+Na] + (I). ∗∗∗ P < 0.001. OA: osteoarthritis. Differential metabolites were annotated according to the Human Metabolite Database and MetFrag. Of these, 8 metabolites were identified out of 71. The FC value and P -value were obtained by the algorithm based on 8 differential metabolic mass spectrum data analyzed by machine learning ( Table S4 ). These 8 differential metabolites were classified into 5 metabolites increased and 3 metabolites decreased in the serum of patients with RA. These metabolites were represented by box plots in Figs. 7 B−I. First, several amino acids and dipeptides, such as citrulline ( m / z 175.19) and glutamine ( m / z 146.15), were among the metabolites that showed an increase in expression in RA. The increase of amino acids in the serum of patients is often related to the changes in the immune system and inflammatory response. Inflammatory factors, such as tumor necrosis factor α (TNF-α) and interleukin-6 (IL-6), stimulate the release of specific amino acids, resulting in elevated concentrations of specific amino acids in the serum [ 41 ]. Anti-CCP antibody is increased in the serum of patients with RA, thus also increasing its metabolite citrulline [ 42 ]. In addition, lactate ( m / z 90.08) was also elevated in serum, which may be related to the increase of anaerobic glycolysis and local tissue hypoxia triggered by inflammation. Persistent inflammation in RA produces large amounts of reactive oxygen species (ROS) and other free radicals, leading to elevated oxidative stress. Acetic acid ( m / z 60.05), a product of fatty acid β-oxidation and ethanol metabolism, is often elevated under pro-inflammatory and hypermetabolic conditions. It also acts as a short-chain fatty acid that modulates immune signalling via G protein-coupled receptors 43 and 41 (GPR43/41) pathways, indicating enhanced metabolic reprogramming in RA monocytes and macrophages [ 43 ]. The increase in 4-hydroxynonenal ( m / z of 156.22), a lipid peroxidation product, suggests intensified oxidative stress in RA, likely driven by the excessive production of ROS due to chronic inflammation [ 44 ]. Taken together, these metabolic alterations reflect the combined effects of inflammation, and immune activation in patients with RA. Furthermore, the serum level of palmitic acid ( m / z 256.4) in patients with RA was decreased, which may be related to the hypermetabolism caused by inflammation and enhanced fatty acid metabolism and glycolysis. This is also the reason why the levels of glucose ( m / z 180.1) is reduced in the serum of patients with RA [ 43 ]. Itaconic acid ( m / z 130.1), an immunometabolite derived from cis -aconitate via immune-responsive gene 1 ( Irg1 ) expression in inflammatory macrophages, typically exhibits anti-inflammatory properties. Its reduction in RA patients may reflect a disrupted feedback regulation of inflammation or a shift toward a more pro-inflammatory macrophage phenotype [ 45 ]. Presently, RF and anti-CCP antibodies are the main serum detection methods for RA in clinical practice; however, their detection specificity is low and cannot be used as a single diagnostic criterion. It is particularly important to design a serum assay that can accurately and rapidly diagnose patients with RA. By collecting the SMF of patients with RA, TiVCrMoC 3 -assisted LDI-MS combined with a machine learning data model was established to identify and diagnose the subsequent RA serum. A total of 15 metabolites were selected from the common metabolites in serum for secondary MS analysis. As shown in Figs. S15–S29 , in addition to displaying [M+Na] + ion peaks, fragment ion peaks can be observed for each analyte, which can verify the accuracy of the primary MS results. The above results showed that TiVCrMoC 3 -assisted LDI-MS has the potential to discover new serum markers of RA and to become a serum detection method for the clinical diagnosis of RA in the future. In addition, several detected metabolites (e.g., taurine [ 46 ], uric acid [ 47 ], Cre [ 48 ], VC [ 49 ], and glutathione [ 50 ]) have been reported in literature to be associated with the pathogenesis or progression of other autoimmune and inflammatory conditions, such as systemic lupus erythematosus [ 46 ], multiple sclerosis [ 47 ], and inflammatory bowel diseases [ 48 ]. This suggests that TiVCrMoC 3 -assisted LDI-MS may also be extended to biomarker screening in these diseases. 4. Conclusions In summary, TiVCrMoC 3 , a novel high-entropy 2D carbide MXene, was prepared and applied as a substrate for SALDI-MS. TiVCrMoC 3 exhibited excellent electrical conductivity, facilitating the ionization of analytes and providing an excellent surface for energy and electron transfer. It can accurately and sensitively identify small-molecule substances such as amino acids, sugars, and fatty acids, and small-molecule drugs in the positive ion mode. In addition, the TiVCrMoC 3 -assisted LDI-MS method has the advantage of accurately detecting metabolites in low concentration and complex environment when it is used to detect small-molecule metabolites. Furthermore, TiVCrMoC 3 -assisted LDI-MS can be used to quantitatively analyze serum stearic acid with speed and accuracy, indicating its clinical significance. A novel TiVCrMoC 3 -assisted LDI-MS approach was developed to extract SMFs, which were deciphered through machine learning to facilitate the identification and diagnosis of RA. Meanwhile, 8 potential differential metabolites were identified in the serum of patients with RA. The clinical application of serum small-molecule metabolites as diagnostic markers of RA in the future is possible. This approach featured for the following: i) quantification of selected biomarkers (e.g., stearic acid) at concentrations as low as 500 ng/mL, ii) capability to detect trace serum samples (1 μL) toward SALDI-MS, iii) identification and diagnosis of RA with the high sensitivity and specificity, and iv) rapid diagnosis, good reusability, and utilization for prognosis monitoring. This work contributes to the development of advanced metabolic analysis protocols, facilitating precision medicine and enabling personalized diagnostic tools based on serum small-molecule biomarkers for a range of diseases in the near future. CRediT authorship contribution statement Zhilong Chu: Writing – original draft, Visualization, Software, Investigation, Data curation. Xi Yu: Data curation. Xiao Wang: Resources. Wenqiang Zhang: Writing – review & editing, Funding acquisition, Conceptualization. Yuming Li: Software. Xinfeng Yan: Writing – review & editing, Conceptualization. Weining Yan: Investigation. Hongzheng Meng: Writing – review & editing, Funding acquisition, Conceptualization. Lingyu Li: Investigation. Guanhua Zhang: Investigation. Ruya Wang: Investigation. Miaomiao Li: Data curation. Jun Li: Data curation. Weiqiang Liang: Writing – review & editing, Funding acquisition, Conceptualization. Chunxia Ma: Writing – review & editing, Funding acquisition, Conceptualization. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgments This research was financially supported by the National Natural Science Foundation of China (Grant No.: 82202419), the Natural Science Foundation of Shandong Province, China (Grant Nos.: ZR2022QH174, ZR2022MH299, and ZR2021MH166), and the Young Elite Sponsorship Program of Shandong Provincial Medical Association, China (Grant No.: 2023_LC_0139). Footnotes Peer review under responsibility of Xi'an Jiaotong University. Appendix A Supplementary data to this article can be found online at https://doi.org/10.1016/j.jpha.2025.101456 . Appendix A. Supplementary data The following are the Supplementary data to this article: Multimedia component 1 mmc1.docx (2.8MB, docx) Multimedia component 2 mmc2.pdf (596.1KB, pdf) Multimedia component 3 mmc3.pdf (55.9KB, pdf) Multimedia component 4 mmc4.docx (10.5KB, docx) Data availability The datasets supporting this article have been uploaded as part of the ESI. References 1. Zhang F., Jonsson A.H., Nathan A., et al. Deconstruction of rheumatoid arthritis synovium defines inflammatory subtypes. Nature. 2023;623:616–624. doi: 10.1038/s41586-023-06708-y. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 2. McInnes I.B., Schett G. The pathogenesis of rheumatoid arthritis. N. Engl. J. Med. 2011;365:2205–2219. doi: 10.1056/NEJMra1004965. [ DOI ] [ PubMed ] [ Google Scholar ] 3. Venetsanopoulou A.I., Alamanos Y., Voulgari P.V., et al. Epidemiology of rheumatoid arthritis: genetic and environmental influences. Expert Rev. Clin. Immunol. 2022;18:923–931. doi: 10.1080/1744666X.2022.2106970. [ DOI ] [ PubMed ] [ Google Scholar ] 4. Gravallese E.M., Firestein G.S. Rheumatoid arthritis: common origins, divergent mechanisms. N. Engl. J. Med. 2023;388:529–542. doi: 10.1056/NEJMra2103726. [ DOI ] [ PubMed ] [ Google Scholar ] 5. Smolen J.S., Aletaha D., McInnes I.B. Rheumatoid arthritis. Lancet. 2016;388:2023–2038. doi: 10.1016/S0140-6736(16)30173-8. [ DOI ] [ PubMed ] [ Google Scholar ] 6. Aletaha D. Precision medicine and management of rheumatoid arthritis. J. Autoimmun. 2020;110 doi: 10.1016/j.jaut.2020.102405. [ DOI ] [ PubMed ] [ Google Scholar ] 7. Paglia G., Astarita G. Metabolomics and lipidomics using traveling-wave ion mobility mass spectrometry. Nat. Protoc. 2017;12:797–813. doi: 10.1038/nprot.2017.013. [ DOI ] [ PubMed ] [ Google Scholar ] 8. Thong E.P., Codner E., Laven J.S.E., et al. Diabetes: a metabolic and reproductive disorder in women. Lancet Diabetes Endocrinol. 2020;8:134–149. doi: 10.1016/S2213-8587(19)30345-6. [ DOI ] [ PubMed ] [ Google Scholar ] 9. Wishart D.S. Emerging applications of metabolomics in drug discovery and precision medicine. Nat. Rev. Drug Discov. 2016;15:473–484. doi: 10.1038/nrd.2016.32. [ DOI ] [ PubMed ] [ Google Scholar ] 10. Karas M., Hillenkamp F. Laser desorption ionization of proteins with molecular masses exceeding 10,000 daltons. Anal. Chem. 1988;60:2299–2301. doi: 10.1021/ac00171a028. [ DOI ] [ PubMed ] [ Google Scholar ] 11. Darie-Ion L., Whitham D., Jayathirtha M., et al. Applications of MALDI-MS/MS-based proteomics in biomedical research. Molecules. 2022;27:6196. doi: 10.3390/molecules27196196. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 12. Liang W., Yan W., Wang X., et al. A single atom cobalt anchored MXene bifunctional platform for rapid, label-free and high-throughput biomarker analysis and tissue imaging. Biosens. Bioelectron. 2024;246 doi: 10.1016/j.bios.2023.115903. [ DOI ] [ PubMed ] [ Google Scholar ] 13. Guo Z., Zhang Q., Zou H., et al. A method for the analysis of low-mass molecules by MALDI-TOF mass spectrometry. Anal. Chem. 2002;74:1637–1641. doi: 10.1021/ac010979m. [ DOI ] [ PubMed ] [ Google Scholar ] 14. van Kampen J.J.A., Burgers P.C., de Groot R., et al. Biomedical application of MALDI mass spectrometry for small-molecule analysis. Mass Spectrom. Rev. 2011;30:101–120. doi: 10.1002/mas.20268. [ DOI ] [ PubMed ] [ Google Scholar ] 15. Ouyang D., Huang H., He Y., et al. Utilization of hydralazine as a reactive matrix for enhanced detection and on-MALDI-target derivatization of saccharides, Chin. Chem. Lett. 2024;35 [ Google Scholar ] 16. Ouyang D., Wang C., Zhong C., et al. Organic metal chalcogenide-assisted metabolic molecular diagnosis of central precocious puberty. Chem. Sci. 2024;15:278–284. doi: 10.1039/d3sc05633c. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 17. Jin Y., Chen J., Xie W., et al. Gold-modified covalent organic frameworks-assisted laser desorption/ionization mass spectrometry for analysis of metabolites induced by triclosan exposure. ACS Appl. Mater. Interfaces. 2025;17:7056–7065. doi: 10.1021/acsami.4c16044. [ DOI ] [ PubMed ] [ Google Scholar ] 18. Jin Y., Yan J., Cai Z., et al. Advances in nanomaterials for surface-assisted laser desorption ionization mass spectrometry: applications in small molecule analysis over the past five years. Trac. Trends Anal. Chem. 2025;183 [ Google Scholar ] 19. Aslam U., Chavez S., Linic S. Controlling energy flow in multimetallic nanostructures for plasmonic catalysis. Nat. Nanotechnol. 2017;12:1000–1005. doi: 10.1038/nnano.2017.131. [ DOI ] [ PubMed ] [ Google Scholar ] 20. Song J., Yang X., Yang Z., et al. Rational design of branched nanoporous gold nanoshells with enhanced physico-optical properties for optical imaging and cancer therapy. ACS Nano. 2017;11:6102–6113. doi: 10.1021/acsnano.7b02048. [ DOI ] [ PubMed ] [ Google Scholar ] 21. Tsugawa H., Cajka T., Kind T., et al. MS-DIAL: data-independent MS/MS deconvolution for comprehensive metabolome analysis. Nat. Methods. 2015;12:523–526. doi: 10.1038/nmeth.3393. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 22. Min Q., Zhang X., Chen X., et al. N-doped graphene: an alternative carbon-based matrix for highly efficient detection of small molecules by negative ion MALDI-TOF MS. Anal. Chem. 2014;86:9122–9130. doi: 10.1021/ac501943n. [ DOI ] [ PubMed ] [ Google Scholar ] 23. Lin Z., Zheng J., Lin G., et al. Negative ion laser desorption/ionization time-of-flight mass spectrometric analysis of small molecules using graphitic carbon nitride nanosheet matrix. Anal. Chem. 2015;87:8005–8012. doi: 10.1021/acs.analchem.5b02066. [ DOI ] [ PubMed ] [ Google Scholar ] 24. Feng D., Xia Y. Covalent organic framework as efficient desorption/ionization matrix for direct detection of small molecules by laser desorption/ionization mass spectrometry. Anal. Chim. Acta. 2018;1014:58–63. doi: 10.1016/j.aca.2018.02.017. [ DOI ] [ PubMed ] [ Google Scholar ] 25. Li Z., Liu Q., Lu X., et al. Magnetic metal-organic framework nanocomposites for enrichment and direct detection of environmental pollutants by negative-ion matrix-assisted laser desorption/ionization time-of-flight mass spectrometry. Talanta. 2019;194:329–335. doi: 10.1016/j.talanta.2018.10.058. [ DOI ] [ PubMed ] [ Google Scholar ] 26. Hsu W.L., Tsai C.W., Yeh A.C., et al. Clarifying the four core effects of high-entropy materials. Nat. Rev. Chem. 2024;8:471–485. doi: 10.1038/s41570-024-00602-5. [ DOI ] [ PubMed ] [ Google Scholar ] 27. Zeng Y., Ouyang B., Liu J., et al. High-entropy mechanism to boost ionic conductivity. Science. 2022;378:1320–1324. doi: 10.1126/science.abq1346. [ DOI ] [ PubMed ] [ Google Scholar ] 28. Zou Q., Qiu A., He Y., et al. Fabrication of polydopamine-coated high-entropy MXene nanosheets for targeted photothermal anticancer therapy. Adv. Sci. 2025;12 doi: 10.1002/advs.202410537. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 29. Nemani S.K., Zhang B., Wyatt B.C., et al. High-entropy 2D carbide MXenes: TiVNbMoC3 and TiVCrMoC3. ACS Nano. 2021;15:12815–12825. doi: 10.1021/acsnano.1c02775. [ DOI ] [ PubMed ] [ Google Scholar ] 30. Galli M., Zoppis I., Smith A., et al. Machine learning approaches in MALDI-MSI: clinical applications. Expert Rev. Proteom. 2016;13:685–696. doi: 10.1080/14789450.2016.1200470. [ DOI ] [ PubMed ] [ Google Scholar ] 31. Li W., Zhang J., Zhu W., et al. Facile preparation of reduced graphene oxide/ZnFe2O4 nanocomposite as magnetic sorbents for enrichment of estrogens. Talanta. 2020;208 doi: 10.1016/j.talanta.2019.120440. [ DOI ] [ PubMed ] [ Google Scholar ] 32. Pazniak H., Benchakar M., Bilyk T., et al. Ion implantation as an approach for structural modifications and functionalization of Ti3C2Tx MXenes. ACS Nano. 2021;15:4245–4255. doi: 10.1021/acsnano.0c06735. [ DOI ] [ PubMed ] [ Google Scholar ] 33. Bilger J.B., Kerzig C., Larsen C.B., et al. A photorobust Mo(0) complex mimicking [Os(2,2′-bipyridine)3]2+ and its application in red-to-blue upconversion. J. Am. Chem. Soc. 2021;143:1651–1663. doi: 10.1021/jacs.0c12805. [ DOI ] [ PubMed ] [ Google Scholar ] 34. Lai S., Jin Y., Shi L., et al. Fluorescence sensing mechanisms of versatile graphene quantum dots toward commonly encountered heavy metal ions. ACS Sens. 2023;8:3812–3823. doi: 10.1021/acssensors.3c01295. [ DOI ] [ PubMed ] [ Google Scholar ] 35. Wen S., Shi Y., Zhang Y., et al. pH-activated Ce-doped molybdenum oxide nanoclusters for tumor microenvironment responsive photothermal and chemodynamic therapy. Langmuir. 2023;39:10145–10153. doi: 10.1021/acs.langmuir.3c01075. [ DOI ] [ PubMed ] [ Google Scholar ] 36. Li P., Gao M., Wang D., et al. Optimizing vanadium redox reaction in Na3V2(PO4)3 cathodes for sodium-ion batteries by the synergistic effect of additional electrons from heteroatoms. ACS Appl. Mater. Interfaces. 2023;15:9475–9485. doi: 10.1021/acsami.2c22038. [ DOI ] [ PubMed ] [ Google Scholar ] 37. Luo P., Wang L., Jiang L., et al. Application of graphdiyne in surface-assisted laser desorption ionization mass spectrometry. ACS Appl. Mater. Interfaces. 2021;13:1914–1920. doi: 10.1021/acsami.0c18280. [ DOI ] [ PubMed ] [ Google Scholar ] 38. Fan B., Zhou H., Wang Y., et al. Surface siloxane-modified silica materials combined with metal–organic frameworks as novel MALDI matrixes for the detection of low-MW compounds. ACS Appl. Mater. Interfaces. 2020;12:37793–37803. doi: 10.1021/acsami.0c11404. [ DOI ] [ PubMed ] [ Google Scholar ] 39. Li S., Dai J., Zhu M., et al. Implantable hydrogel-protective DNA aptamer-based sensor supports accurate, continuous electrochemical analysis of drugs at multiple sites in living rats. ACS Nano. 2023;17:18525–18538. doi: 10.1021/acsnano.3c06520. [ DOI ] [ PubMed ] [ Google Scholar ] 40. Huang Z., Zhao D., Shen L., et al. Mxenes for membrane separation: from fabrication strategies to advanced applications. Sci. Bull. 2024;69:125–140. doi: 10.1016/j.scib.2023.11.008. [ DOI ] [ PubMed ] [ Google Scholar ] 41. Zou M., Zeng Q., Nie J., et al. The role of E3 ubiquitin ligases and deubiquitinases in inflammatory bowel disease: friend or foe? Front. Immunol. 2021;12 doi: 10.3389/fimmu.2021.769167. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 42. Darrah E., Andrade F. Rheumatoid arthritis and citrullination. Curr. Opin. Rheumatol. 2018;30:72–78. doi: 10.1097/BOR.0000000000000452. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 43. Pucino V., Certo M., Varricchi G., et al. Metabolic checkpoints in rheumatoid arthritis. Front. Physiol. 2020;11:347. doi: 10.3389/fphys.2020.00347. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 44. Chang J.L., Bashir M., Santiago C., et al. Intrauterine growth restriction and hyperoxia as a cause of white matter injury. Dev. Neurosci. 2018;40:344–357. doi: 10.1159/000494273. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 45. Peace C.G., O’Neill L.A.J. The role of itaconate in host defense and inflammation. J. Clin. Investig. 2022;132 doi: 10.1172/JCI148548. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 46. Huang S., Cui Y., Zhang Y., et al. Taurine corrects lupus CD4+ T cell imbalance through inhibition of mTORC1 signaling. Nutr. Metab. 2025;22:41. doi: 10.1186/s12986-025-00936-x. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 47. Moccia M., Lanzillo R., Palladino R., et al. Uric acid: a potential biomarker of multiple sclerosis and of its disability. Clin. Chem. Lab. Med. CCLM. 2015;53:753–759. doi: 10.1515/cclm-2014-0744. [ DOI ] [ PubMed ] [ Google Scholar ] 48. Velayos F.S. Rise over runs: elevated creatinine and a new diagnosis of inflammatory bowel disease. Dig. Dis. Sci. 2023;68:1706–1708. doi: 10.1007/s10620-023-07852-1. [ DOI ] [ PubMed ] [ Google Scholar ] 49. Isola S., Gammeri L., Furci F., et al. Vitamin C supplementation in the treatment of autoimmune and onco-hematological diseases: from prophylaxis to adjuvant therapy. Int. J. Mol. Sci. 2024;25:7284. doi: 10.3390/ijms25137284. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 50. Shah D., Sah S., Nath S.K. Interaction between glutathione and apoptosis in systemic lupus erythematosus. Autoimmun. Rev. 2013;12:741–751. doi: 10.1016/j.autrev.2012.12.007. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Associated Data This section collects any data citations, data availability statements, or supplementary materials included in this article. Supplementary Materials Multimedia component 1 mmc1.docx (2.8MB, docx) Multimedia component 2 mmc2.pdf (596.1KB, pdf) Multimedia component 3 mmc3.pdf (55.9KB, pdf) Multimedia component 4 mmc4.docx (10.5KB, docx) Data Availability Statement The datasets supporting this article have been uploaded as part of the ESI. Articles from Journal of Pharmaceutical Analysis are provided here courtesy of Xi'an Jiaotong University ACTIONS View on publisher site PDF (9.4 MB) Cite Collections Permalink PERMALINK Copy RESOURCES Similar articles Cited by other articles Links to NCBI Databases Cite Copy Download .nbib .nbib Format: AMA APA MLA NLM Add to Collections Create a new collection Add to an existing collection Name your collection * Choose a collection Unable to load your collection due to an error Please try again Add Cancel Follow NCBI NCBI on X (formerly known as Twitter) NCBI on Facebook NCBI on LinkedIn NCBI on GitHub NCBI RSS feed Connect with NLM NLM on X (formerly known as Twitter) NLM on Facebook NLM on YouTube National Library of Medicine 8600 Rockville Pike Bethesda, MD 20894 Web Policies FOIA HHS Vulnerability Disclosure Help Accessibility Careers NLM NIH HHS USA.gov Back to Top