Discovery of Potential Biomarkers for Postmenopausal Osteoporosis Based on Untargeted GC/LC-MS

Jun Kou, He Chunyang, Lin Cui, Zhengping Zhang, Wei Wang, Li Tan, Da Liu, Wei Zheng, Wei Gu, Ning XiaFrontiers in Endocrinology 2022 · DOI

TL;DR

This study leverages untargeted GC/LC-MS metabolomics on serum samples from postmenopausal women to identify potential biomarkers for osteoporosis (OP). By analyzing metabolic differences, the research pinpoints 18 metabolites linked to glucose, amino acid, and lipid metabolism, offering a promising avenue for early OP diagnosis and monitoring.

What Problem Does This Paper Address?

Osteoporosis (OP) is a significant global health concern, particularly for postmenopausal women, characterized by low bone mineral density and increased fracture risk. While dual-energy X-ray absorptiometry (DXA) is the standard diagnostic tool, it primarily reveals the results of OP rather than its underlying pathological processes, making early detection and intervention challenging. Current bone turnover markers, while useful, also have limitations in predicting primary OP. This gap highlights the need for novel, sensitive biomarkers that can capture early metabolic shifts associated with OP development.

Methods

The researchers employed an untargeted gas chromatography/liquid chromatography-mass spectrometry (GC/LC-MS) approach to analyze serum samples. This dual-platform strategy aims to maximize the coverage of metabolites by leveraging the strengths of both GC-MS (robust separation, suitable for volatile compounds) and LC-MS (analysis of thermally labile, non-volatile, and polar compounds).

Key steps included:

  • Sample Preparation: Serum samples were processed with internal standards, followed by extraction using methanol and acetonitrile, derivatization for GC-MS, and filtration for LC-MS.
  • Data Acquisition: GC-MS and LC-MS analyses were performed to generate comprehensive metabolic profiles.
  • Data Processing: Raw data underwent peak detection, deconvolution, alignment, and normalization. Missing values were handled by imputation.
  • Multivariate Data Analysis: Principal Component Analysis (PCA) was used for unsupervised exploration of data variance. Partial Least Squares Discriminant Analysis (PLS-DA) and Orthogonal Partial Least Squares Discriminant Analysis (OPLS-DA) were employed for supervised classification and identification of differentiating metabolites. These methods help to visualize group separation and identify key variables (metabolites) contributing to the differences.
  • Biomarker Identification: Metabolites with an absolute p-value < 0.05 and Variable Importance for Projection (VIP) > 1.0 were selected as potential biomarkers.
  • Pathway Analysis: The Kyoto Encyclopedia of Genes and Genomes (KEGG) database was used to identify metabolic pathways associated with the identified biomarkers.

Datasets

The study utilized serum samples from two groups of participants:

  • Case Group: 50 postmenopausal women diagnosed with osteoporosis (BMD T-score < -2.5).
  • Control Group: 50 age-matched healthy postmenopausal women with normal BMD.

The data collected was from these two distinct groups, and no new benchmark datasets were introduced.

Key Insights and Ideas

  • Synergistic Power of GC/LC-MS: Combining GC-MS and LC-MS provided a broader and more comprehensive view of the serum metabolome compared to using a single platform, which is a common limitation in previous studies.
  • Metabolic Signatures of OP: The study revealed significant differences in metabolic profiles between postmenopausal women with OP and healthy controls, suggesting that OP is indeed associated with distinct metabolic dysregulations.
  • Disrupted Core Metabolic Pathways: The identified biomarkers are primarily linked to disruptions in glucose metabolism, amino acid metabolism, and lipid metabolism. These pathways are fundamental to energy production and cellular function, and their dysregulation can impact bone health.
  • Potential for Early Detection: The identification of these differential metabolites offers a promising avenue for developing early diagnostic markers for OP, potentially allowing for intervention before significant bone loss occurs.
  • Complex Interplay of Metabolism and Bone Health: The findings underscore the intricate relationship between systemic metabolism and bone microenvironment, highlighting how metabolic shifts can influence bone remodeling and strength.

Key Results

  • 18 Potential Biomarkers Identified: Through multivariate analysis, 18 metabolites were identified as potential biomarkers for postmenopausal OP.
  • Significant Group Separation: PCA, PLS-DA, and OPLS-DA models demonstrated clear separation between the OP and control groups, indicating high sensitivity and specificity of the metabolomic data for distinguishing between the conditions. For instance, R2Y and Q2 values for PLS-DA were as high as 0.966 and 0.994 (LC-MS), and for OPLS-DA, 0.996 and 0.995 (LC-MS), indicating robust model performance.
  • Metabolic Pathway Involvement: The identified biomarkers are associated with pathways such as glucose metabolism, lipid metabolism, and amino acid metabolism. Specific examples include increased levels of isothreonic acid and lactobionic acid, and decreased levels of stearic acid, tartaric acid, and glyceric acid.
  • Limited Correlation with Bone Turnover Markers: Only three of the identified metabolites showed significant correlation with traditional bone turnover markers (TRACP5b and BAP), suggesting that these novel metabolic biomarkers may offer complementary or superior information for OP diagnosis, especially for primary osteoporosis.

Limitations and Future Work

  • Sample Size and Geographic Limitation: The authors acknowledge that the sample size is not very large, and the participants are primarily from Sichuan Province. They suggest that future studies should include a larger and more diverse cohort to validate these findings.
  • Serum as Sole Sample Type: The study focused solely on serum. Future research could benefit from analyzing other biological samples like bone marrow and urine to build a more comprehensive metabolic database.
  • Untargeted vs. Targeted Analysis: The untargeted approach may produce false positives. The authors propose that targeted metabolomics studies are needed in the next step to confirm the identified biomarkers and provide absolute qualitative and quantitative data.
  • Focus on Postmenopausal Women: The current findings are specific to postmenopausal women. Further exploration in OP male patients is recommended.
  • Clinical Validation: The findings are preliminary and require further validation studies to be applied in clinical practice.

Summary updated

Abstract

These findings can be applied to clinical work through further validation studies. It also shows that metabonomic analysis has great potential in the application of early diagnosis and recurrence monitoring in postmenopausal OP women.
BibTeX
@inproceedings{kou2022discovery,
  author = {Jun Kou and He Chunyang and Lin Cui and Zhengping Zhang and Wei Wang and Li Tan and Da Liu and Wei Zheng and Wei Gu and Ning Xia},
  title = {{Discovery of Potential Biomarkers for Postmenopausal Osteoporosis Based on Untargeted GC/LC-MS}},
  booktitle = {Frontiers in Endocrinology},
  year = {2022},
  doi = {10.3389/fendo.2022.849076},
}
Figure