Integrative Single-Cell and Machine Learning Analysis Develops a Glutamine Metabolism-Based Prognostic Model and...

Area of Science:

• Oncology
• Metabolic Pathways
• Systems Biology

Background:

• Osteosarcoma (OS) patient stratification lacks a glutamine metabolism framework.
• Metabolic pathways significantly impact disease progression and patient outcomes.

Purpose of the Study:

• To develop a glutamine metabolism-based risk model for osteosarcoma patient stratification.
• To identify key genes and pathways involved in OS glutamine metabolism.
• To establish a prognostic tool for personalized therapeutic strategies.

Main Methods:

• Integrated single-cell RNA sequencing with bulk cohorts.
• Quantified glutamine metabolism-related gene (GRG) scores using five algorithms.
• Developed a five-gene prognostic model via machine learning (Step-Cox + Ridge).
• Validated the model in independent cohorts and performed functional studies (MSMO1 knockdown).

Main Results:

• Identified pronounced intratumoral heterogeneity in glutamine metabolism, especially in osteoblastic cells.
• Developed a five-gene prognostic model (GPX7, COL11A2, CPE, MSMO1, SGMS2) with moderate performance in independent cohorts.
• MSMO1 knockdown suppressed OS cell proliferation, migration, invasion, and altered key metabolic and signaling pathways.
• The model stratifies OS patients into distinct molecular risk subgroups with different outcomes.

Conclusions:

• Established a single-cell-anchored, glutamine-coupled state in osteosarcoma.
• Introduced an externally validated prognostic tool with translational promise for OS.
• Positioned MSMO1 as a critical metabolic-signaling node for further investigation and potential therapeutic targeting.