Metabolomic biomarkers for benign conditions and malignant ovarian cancer: Advancing early diagnosis.

BACKGROUND: Ovarian cancer (OC) is a major global cause of death among gynecological cancers, with a high mortality rate. Early diagnosis, distinguishing between benign conditions and early malignant OC forms, is vital for successful treatment. This research investigates serum metabolites to find diagnostic biomarkers for early OC identification. METHODS: Metabolomic profiles derived from the serum of 60 patients with benign conditions and 60 patients with malignant OC were examined using ultra-performance liquid chromatography coupled with tandem mass spectrometry (UPLC-MS/MS). Comparative analysis revealed differential metabolites linked to OC, aiding biomarker identification for early-diagnosis of OC via machine learning features. The predictive ability of these biomarkers was evaluated against the traditional biomarker, cancer antigen 125 (CA125). RESULTS: 84 differential metabolites were identified, including 2-Thiothiazolidine-4-carboxylic acid (TTCA), Methionyl-Cysteine, and Citrulline that could serve as potential biomarkers to identify benign conditions and malignant OC. In the diagnosis of early-stage OC, the area under the curve (AUC) for Citrulline was 0.847 (95 % Confidence Interval (CI): 0.719-0.974), compared to 0.770 (95 % CI: 0.596-0.944) for TTCA, and 0.754 for Methionine-Cysteine (95 % CI: 0.589-0.919). These metabolites demonstrate a superior diagnostic capability relative to CA125, which has an AUC of 0.689 (95 % CI: 0.448-0.931). Among these biomarkers, Citrulline stands out as the most promising. Additionally, in the diagnosis of benign conditions and malignant OC, using logistic regression to combine potential biomarkers with CA125 has an AUC of 0.987 (95 % CI: 0.9708-1) has been proven to be more effective than relying solely on the traditional biomarker CA125 with an AUC of 0.933 (95 % CI: 0.870-0.996). Furthermore, among all the differential metabolites, lipid metabolites dominate, significantly impacting glycerophospholipid metabolism pathway. CONCLUSION: The discovered serum metabolite biomarkers demonstrate excellent diagnostic performance for distinguishing between benign conditions and malignant OC and for early diagnosis of malignant OC.

N-glycosylation of disease-specific haptoglobin for the early screening of diabetic retinopathy.

PURPOSE: Diabetic retinopathy (DR), as one of the microvascular complications of diabetes, is a leading cause of acquired vision loss. Most DR cases are detected in the advanced stage through fundoscopy, making molecular biomarkers urgently needed for early diagnosis of DR. EXPERIMENTAL DESIGN: Serum disease-specific haptoglobin-ß (Hp-ß) chains of 100 patients with type 2 diabetes mellitus (T2DM) and 156 T2DM patients with non-proliferative diabetic retinopathy (NPDR) were separated using polyacrylamide gel electrophoresis. After in-gel digestion and enrichment, the intact N-glycopeptides were detected by mass spectrometry. RESULTS: Fucosylation of Hp-ß was significantly increased and sialylation of Hp-ß was significantly decreased in background DR (BDR, an early-stage DR) patients compared with non-diabetic retinopathy patients (p < 0.05) and yielded area under curves (AUCs) of 0.801 and 0.829 in training and validation groups, respectively, which had an advantage over glycated hemoglobin A1c (AUC ≤ 0.691). Moreover, a significant increase in sialylated Hp-ß was found in severe NPDR patients compared with BDR patients and yielded an AUC of 0.828 to distinguish severe NPDR from BDR. CONCLUSION: Changes in Hp-ß glycosylation are closely related to DR, and may be used for early diagnosis and screening of DR.

Serum disease-specific IgG Fc glycosylation as potential biomarkers for nonproliferative diabetic retinopathy using mass spectrometry.

OBJECTIVE: To explore the potential of serum disease-specific immunoglobulin G (DSIgG) glycosylation as a biomarker for the diagnosis of nonproliferative diabetic retinopathy (NPDR). METHODS: A total of 387 consecutive diabetic patients presenting in an eye clinic without proliferative diabetic retinopathy (DR) were included and divided into those with nondiabetic retinopathy (NDR) (n = 181) and NPDR (n = 206) groups. Serum was collected from all patients for DSIgG separation. The enriched glycopeptides of the tryptic digests of DSIgG were detected using matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF MS). Patients were randomly divided into discovery and validation sets (1:1). The differences in glycopeptide ratios between the groups were compared by using Student's t-test or the Mann-Whitney U test. The predictive ability of the model was assessed using the area under the receiver operating characteristic curve (AUC). RESULTS: DSIgG1 G1FN/G0FN, G2N/G2, G2FN/G2N and DSIgG2 G1F/G0F, G1FN/G0FN, G2N/G1N, G2S/G2 were significantly different between NDR and NPDR patients (p < 0.05) in both the discovery and validation sets. The prediction model that was built comprising the seven glycopeptide ratios showed good NPDR prediction performance with an AUC of 0.85 in the discovery set and 0.87 in the validation set. CONCLUSION: DSIgG Fc N-glycosylation ratios were associated with NPDR and can be used as potential biomarkers for the early diagnosis of DR.

Disease-Specific haptoglobin N-Glycosylation in inflammatory disorders between cancers and benign diseases of 3 types of female internal genital organs.

BACKGROUND: N-glycosylation of the haptoglobin is closely related to pathological states. This study aims to evaluate the association of glycosylation of disease-specific Hp (DSHp) ß chain with different pathological states of the cervix, uterus, and ovary to explore differences in their inflammatory responses and to screen potential biomarkers to distinguish cancer from benign diseases. METHODS: DSHp-ß chains of 1956 patients with cancers and benign diseases located in the cervix, uterus, and ovary organs were separated from serum immunoinflammatory-related protein complexes (IIRPCs). The N-glycopeptides from DSHp-ß chains were detected using mass spectrometry, followed by an analysis of machine learning algorithms. RESULTS: 55 N-glycopeptides at N207/N211, 19 at N241, and 21 at N184 glycosylation sites of DSHp for each sample were identified. Fucosylation and sialylation of DSHp in cervix, uterus, and ovary cancer were significantly increased compared to their corresponding benign diseases (p < 0.001). The cervix diagnostic model, a combination of G2N3F, G4NFS, G7N2F2S5, GS-N&GS-N, G2N2&G4N3FS, G7N2F2S5, G2S2&G-N, and GN2F&G2F at N207/N211 sites, G3NFS2 and G3NFS at N241site, G9N2S, G6N3F6, G4N3F5S, G4N3F4S2, and G6N3F4S at N184 site), has shown a good diagnostic capability to distinguish cancer from benign diseases, with the area under curve (AUC) of 0.912. The uterus diagnostic model including G4NFS, G2S2&G2S2, G3N2S2, GG5N2F5, G2&G3NFS, and G5N2F3S3 at N207/N211 sites, and G2NF3S2 at N184 site, with an AUC of 0.731. The ovary diagnostic model including G2N3F, GF2S-N &G2F3S2, G2S&G2, and G2S&G3NS at N207/N211 sites; G2S and G3NFS at N241 site, G6N3F4S at N184 site, with an AUC of 0.747. CONCLUSIONS: These findings provide insights into differences in organ-specific inflammatory responses of DSHp for different pathological states among the organs of the cervix, uterus, and ovary.

Diagnostic potential of site-specific serotransferrin N-glycosylation in discriminating different liver diseases.

BACKGROUND: Altered glycosylation modulates the structure and function of disease-related proteins. The associations between serotransferrin (STF) N-glycosylation and liver diseases (LDs) have been revealed. However, how intact N-glycopeptides vary among different types of liver diseases remains unclear. METHODS: Intact STF N-glycopeptides from patients with chronic liver disease (CLD, n = 92), primary liver cancer (PLC, n = 123), metastatic liver cancer (MLC, n = 57), and healthy controls (HCs, n = 59) were determined using high-resolution mass spectrometry. RESULTS: Significant changes were displayed in STF glycosylation among 4 groups. The LD screening model, including Asn432 G1S/G2S, Asn432 G2S/G2S2, and Asn630 G2NS2/G2FNS2, was constructed to differentiate LDs from HCs, with a AUC of 0.92. The liver cancer (LC) diagnostic model, a combination of Asn432 G1-N/G1S-N, Asn432 G1/G2, Asn432 G2FS/G2FS2, and Asn630 G1S-N /G1S, showed good performance in discriminating LC from CLD (AUC = 0.93). Moreover, AFP-negative LC patients (93 %) were successfully predicted by the LC diagnostic model. Furthermore, the MLC triage model, composed of Asn432 G1/G2, Asn432 G3F/G3FS, Asn630 G2/G2S, Asn630 G2S2/G2NS2, and Asn630 G3FS/G3FS2, yielded an AUC of 0.98 between PLC and MLC. CONCLUSIONS: STF N-glycosylation is a potential biomarker for the accurate classification of different LDs.