A Five Glutamine-Associated Signature Predicts Prognosis of Prostate Cancer and Links Glutamine Metabolism with Tumor Microenvironment

Hai Wang, Yuxiao Chen, Wei Zhao, Haolin Liu, Hongtao Tu, Zhongyou Xia, Rui Wang, Jinze Tang, Chuang Zhu, Rui Li, Xiaodong Liu, Peng GuJournal of Clinical Medicine 2023 · DOI

TL;DR

This study identifies a five-gene signature linked to glutamine metabolism that accurately predicts biochemical recurrence in prostate cancer patients. This signature also reveals distinct tumor microenvironment characteristics and potential drug sensitivities, offering a new avenue for personalized treatment strategies.

What Problem Does This Paper Address?

Prostate cancer (PCa) is a major health concern, and while radical treatments exist, a significant portion of patients experience biochemical recurrence (BCR) after surgery, leading to metastasis and death. Current methods for predicting BCR are not always satisfactory, highlighting the need for more precise prognostic tools. Glutamine, a crucial amino acid for cell survival and biosynthesis, plays a significant role in tumor progression, but its specific role and associated metabolic signatures in PCa prognosis and the tumor microenvironment (TME) remain unclear. This paper aims to address this gap by developing a glutamine metabolism-based predictive model for PCa prognosis and exploring its implications for the TME and treatment.

Methods

The researchers employed a multi-pronged approach:

  • Gene Identification and Model Construction: They started by analyzing gene expression data from The Cancer Genome Atlas (TCGA) to identify differentially expressed genes (DEGs) in PCa. These DEGs were then intersected with a list of glutamine-related genes obtained from the GSEA database to pinpoint differentially expressed glutamine-related genes (DEGRGs). Univariate Cox regression and Lasso regression analyses were used to select five key DEGRGs that were significantly associated with biochemical recurrence-free survival (BCRFS). A risk score was calculated for each patient based on the expression levels of these five genes and their respective coefficients, allowing for the stratification of patients into high-risk and low-risk groups.
  • Validation: The developed risk model was validated using an independent dataset from the Gene Expression Omnibus (GEO) database.
  • Tumor Microenvironment Analysis: Various algorithms (CIBERSORTx, ssGSEA, ESTIMATE) were used to analyze immune cell infiltration patterns, immune scores, stromal scores, and tumor purity in the high- and low-risk groups.
  • Drug Sensitivity Prediction: The study predicted the response of patients in different risk groups to various chemotherapeutic agents using the Genomics of Drug Sensitivity in Cancer (GDSC) database and the CellMiner database.
  • Experimental Validation: Key gene expressions were validated in clinical PCa samples using quantitative real-time PCR (qPCR). In vitro experiments involving siRNA-mediated knockdown of specific glutamine metabolism genes (GLUL and ASNS) in PCa cell lines were conducted to assess their impact on cell proliferation and survival, particularly under glutamine-deprived conditions.

Datasets

  • Training Set: TCGA-PRAD dataset (499 PCa samples and 52 controls).
  • Validation Set: GEO dataset (GSE70769) for PCa.
  • Clinical Information: UCSC Xena database.
  • Glutamine-Related Genes: GSEA database.
  • Drug Sensitivity Data: Genomics of Drug Sensitivity in Cancer (GDSC) and CellMiner databases.

The study did not introduce new benchmark datasets but rather utilized publicly available, well-established datasets for its analyses.

Key Insights and Ideas

  • Glutamine Metabolism as a Prognostic Indicator: The study strongly suggests that glutamine metabolism is not just a metabolic pathway but a critical determinant of PCa prognosis, specifically for biochemical recurrence.
  • TME Remodeling by Glutamine Metabolism: Aberrant glutamine metabolism appears to actively remodel the tumor microenvironment, influencing immune cell infiltration and potentially creating an immunosuppressive milieu that favors tumor progression.
  • Therapeutic Vulnerabilities: The identified risk groups exhibit differential responses to treatments, with high-risk patients potentially benefiting more from chemotherapy, while low-risk patients might be less responsive to anti-androgen therapy.
  • Targeting Glutamine for Therapy: The in vitro experiments provide direct evidence that inhibiting key glutamine metabolism genes (GLUL, ASNS) can significantly impair PCa cell growth, especially when glutamine is scarce, pointing to glutamine metabolism as a promising therapeutic target.

Key Results

  • Five-Gene Signature: A five-gene signature (ATCAY, GLUL, ASNS, CAD, FPGS) was identified as a robust predictor of biochemical recurrence-free survival (BCRFS) in PCa.
  • Predictive Accuracy: The model demonstrated good predictive efficacy in both the TCGA and GEO cohorts, with AUC values for 1-year and 3-year ROC curves reaching 0.697 and 0.72, respectively, in the TCGA cohort.
  • Immune Landscape Differences: High-risk patients showed significantly higher infiltration of M2 macrophages and Tregs, along with lower stromal, immune, and estimate scores, indicating a more immunosuppressive TME.
  • Drug Sensitivity: High-risk patients exhibited lower IC50 values for bicalutamide (suggesting less sensitivity) but appeared to benefit more from chemotherapeutic agents like docetaxel, doxorubicin, etoposide, and mitomycin C.
  • Experimental Validation: qPCR confirmed differential expression of the five genes in clinical samples, and siRNA knockdown of GLUL and ASNS significantly inhibited PCa cell growth under glutamine-deprived conditions.

Limitations and Future Work

The authors acknowledge several limitations:

  • Sample Size: The study utilized a relatively small number of PCa patients from TCGA (422 for analysis) and GEO (GSE70769), which might limit the generalizability of the findings.
  • Retrospective Analysis: The retrospective nature of the study and potential selection inaccuracies are noted.
  • Mechanism Exploration: While the study proposes mechanisms linking glutamine metabolism to TME remodeling and PCa progression, extensive in vivo and in vitro experiments are still needed to fully elucidate these pathways.

Future work should focus on:

  • Larger Cohorts: Training and validating the model in larger, more diverse patient cohorts.
  • In Vivo Studies: Conducting animal model studies to further investigate the role of glutamine metabolism in PCa progression and its interaction with the TME.
  • Therapeutic Strategies: Developing and testing targeted therapies that exploit the vulnerabilities of high-glutamine-metabolizing PCa cells, potentially in combination with immunotherapy or chemotherapy.
  • Clinical Translation: Exploring the clinical utility of this signature for patient stratification and personalized treatment selection.

Summary updated

Abstract

Glutamine has been recognized as an important amino acid that provide a variety of intermediate products to fuel biosynthesis. Glutamine metabolism participates in the progression of the tumor via various mechanisms. However, glutamine-metabolism-associated signatures and its significance in prostate cancer are still unclear. In this current study, we identified five genes associated with glutamine metabolism by univariate and Lasso regression analysis and constructed a model to predict the biochemical recurrence free survival (BCRFS) of PCa. Further validation of the prognostic risk model demonstrated a good efficacy in predicting the BCRFS in PCa patients. Interestingly, based on the CIBERSORTx, ssGSEA and ESTIMATE algorithms predictions, we noticed a distinct immune cell infiltration and immune pathway pattern in the prediction of the two risk groups stratified by the risk model. Drug sensitivity prediction revealed that patients in the high-risk group were more suitable for chemotherapy. Last but not least, glutamine deprivation significantly inhibited cell growth in GLUL or ASNS knock down prostate cancer cell lines. Therefore, we proposed a novel prognostic model by using glutamine metabolism genes for PCa patients and identified potential mechanism of PCa progression through glutamine-related tumor microenvironment remodeling.
BibTeX
@article{wang2023five,
  author = {Hai Wang and Yuxiao Chen and Wei Zhao and Haolin Liu and Hongtao Tu and Zhongyou Xia and Rui Wang and Jinze Tang and Chuang Zhu and Rui Li and Xiaodong Liu and Peng Gu},
  title = {{A Five Glutamine-Associated Signature Predicts Prognosis of Prostate Cancer and Links Glutamine Metabolism with Tumor Microenvironment}},
  journal = {Journal of Clinical Medicine},
  year = {2023},
  doi = {10.3390/jcm12062243},
}
Figure