Scholarlink

Background <p>Significant interindividual variability in radiosensitivity poses a major challenge to conventional radiation protection and radiotherapy. Current prediction strategies relying on DNA damage or genomic analysis have inherent limitations, underscoring the need for minimally invasive serum biomarkers. While serum apolipoproteins are crucial regulators of lipid transport, metabolism, and cellular stress response, their role as biomarkers for radiosensitivity remains largely unexplored.</p> Methods <p>A 7.3&#xa0;Gy ⁶⁰Co γ-ray whole-body irradiation mouse model (with training and independent validation cohorts) was established to assess individual radiosensitivity. Pre-irradiation peripheral serum samples underwent high-throughput proteomics analysis to identify differential proteins (DEPs) linked to 30-day post-irradiation survival. KEGG and GO enrichment analyses were conducted to characterize DEP-associated pathways. An XGBoost machine learning model was built using candidate biomarkers, with SHAP analysis to define their predictive contributions; Cox proportional hazards and Pearson correlation analyses were applied to evaluate survival associations.</p> Results <p>DIA-based proteomics identified 580 DEPs in the training cohort and 449 in the validation cohort. KEGG and GO enrichment analyses confirmed that these DEPs were predominantly enriched in the cholesterol metabolism and reverse cholesterol transport pathways. The predictive model based on an apolipoprotein panel (ApoA1/ApoA2/ApoA4), established using the XGBoost algorithm, exhibited exceptional performance in the training cohort (AUC = 1) and maintained robust generalizability in an independent validation cohort (AUC = 0.833). Compared with non-survivors, survivors exhibited significantly elevated serum levels of ApoA1 and ApoA2 but markedly reduced levels of ApoA4. Cox proportional hazards regression analysis established ApoA1 and ApoA2 as independent protective factors, whereas high ApoA4 expression was an adverse prognostic indicator. Notably, ApoA4 levels also demonstrated a strong negative correlation with post-irradiation survival time.</p> Conclusion <p>The serum apolipoprotein profile (ApoA1/ApoA2/ApoA4) serves not only as a promising minimally invasive biomarker for predicting individual radiosensitivity in mice but also reveals a critical link between the cholesterol metabolic pathway and radiation response. This finding lays a theoretical foundation for translating predictive, cholesterol metabolism-related biomarkers to support radiation response assessments. Given the limitations of animal models, subsequent studies are required to validate the clinical applicability of this panel in human cohorts, with the aim of offering an effective tool for personalized radiation protection and precise radiotherapy.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Predictive value of serum apolipoprotein panel (ApoA1 / ApoA2 / ApoA4) as a biomarker for individual radiosensitivity

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年1月26日

查看全文
  • Na Huang,
  • Heming Wang,
  • Xiao Li,
  • Yuhong Xiang,
  • Ziteng Liu,
  • Yaqiong Li,
  • Hongmei Zhou,
  • Qi Wang,
  • Hongwei Zhou,
  • Zhenhua Qi,
  • Zhidong Wang

摘要

Background

Significant interindividual variability in radiosensitivity poses a major challenge to conventional radiation protection and radiotherapy. Current prediction strategies relying on DNA damage or genomic analysis have inherent limitations, underscoring the need for minimally invasive serum biomarkers. While serum apolipoproteins are crucial regulators of lipid transport, metabolism, and cellular stress response, their role as biomarkers for radiosensitivity remains largely unexplored.

Methods

A 7.3 Gy ⁶⁰Co γ-ray whole-body irradiation mouse model (with training and independent validation cohorts) was established to assess individual radiosensitivity. Pre-irradiation peripheral serum samples underwent high-throughput proteomics analysis to identify differential proteins (DEPs) linked to 30-day post-irradiation survival. KEGG and GO enrichment analyses were conducted to characterize DEP-associated pathways. An XGBoost machine learning model was built using candidate biomarkers, with SHAP analysis to define their predictive contributions; Cox proportional hazards and Pearson correlation analyses were applied to evaluate survival associations.

Results

DIA-based proteomics identified 580 DEPs in the training cohort and 449 in the validation cohort. KEGG and GO enrichment analyses confirmed that these DEPs were predominantly enriched in the cholesterol metabolism and reverse cholesterol transport pathways. The predictive model based on an apolipoprotein panel (ApoA1/ApoA2/ApoA4), established using the XGBoost algorithm, exhibited exceptional performance in the training cohort (AUC = 1) and maintained robust generalizability in an independent validation cohort (AUC = 0.833). Compared with non-survivors, survivors exhibited significantly elevated serum levels of ApoA1 and ApoA2 but markedly reduced levels of ApoA4. Cox proportional hazards regression analysis established ApoA1 and ApoA2 as independent protective factors, whereas high ApoA4 expression was an adverse prognostic indicator. Notably, ApoA4 levels also demonstrated a strong negative correlation with post-irradiation survival time.

Conclusion

The serum apolipoprotein profile (ApoA1/ApoA2/ApoA4) serves not only as a promising minimally invasive biomarker for predicting individual radiosensitivity in mice but also reveals a critical link between the cholesterol metabolic pathway and radiation response. This finding lays a theoretical foundation for translating predictive, cholesterol metabolism-related biomarkers to support radiation response assessments. Given the limitations of animal models, subsequent studies are required to validate the clinical applicability of this panel in human cohorts, with the aim of offering an effective tool for personalized radiation protection and precise radiotherapy.