A gene signature linked to fibroblast differentiation for prognostic prediction of mesothelioma

Our plain-English summary of A gene signature linked to fibroblast differentiation for prognostic prediction of mesothelioma, by J. Liu, Y. Lu, Y. Liu, et al (Cell & Bioscience, 2024).

The study

• Authors: J. Liu, Y. Lu, Y. Liu, W. Zhang, S. Xian, S. Wang, Z. Zheng, R. Lin, M. Jin, M. Zhang, W. Qian, J. Tang, B. Lu, Y. Yang, Z. Liu, M. Qu, H. Ma, X. Wu, Z. Chang, J. Zhang, Y. Zhang
• Journal: Cell & Bioscience
• Published: 2024;14(1):33
• Cited by: 3 articles, as counted by NIH iCite
• Read it: PubMed · Publisher · Free full text

Abstract

Background: Malignant mesothelioma is a type of infrequent tumor that is substantially related to asbestos exposure and has a terrible prognosis. We tried to produce a fibroblast differentiation-related gene set for creating a novel classification and prognostic prediction model of MESO.

Method: Three databases, including NCBI-GEO, TCGA, and MET-500, separately provide single-cell RNA sequencing data, bulk RNA sequencing profiles of MESO, and RNA sequencing information on bone metastatic tumors. Dimensionality reduction and clustering analysis were leveraged to acquire fibroblast subtypes in the MESO microenvironment. The fibroblast differentiation-related genes (FDGs), which were associated with survival and subsequently utilized to generate the MESO categorization and prognostic prediction model, were selected in combination with pseudotime analysis and survival information from the TCGA database. Then, regulatory network was constructed for each MESO subtype, and candidate inhibitors were predicted. Clinical specimens were collected for further validation.

Result: A total of six fibroblast subtypes, three differentiation states, and 39 FDGs were identified. Based on the expression level of FDGs, three MESO subtypes were distinguished in the fibroblast differentiation-based classification (FDBC). In the multivariate prognostic prediction model, the risk score that was dependent on the expression level of several important FDGs, was verified to be an independently effective prognostic factor and worked well in internal cohorts. Finally, we predicted 24 potential drugs for the treatment of MESO. Moreover, immunohistochemical staining and statistical analysis provided further validation.

Conclusion: Fibroblast differentiation-related genes (FDGs), especially those in low-differentiation states, might participate in the proliferation and invasion of MESO. Hopefully, the raised clinical subtyping of MESO would provide references for clinical practitioners.

© 2024. The Author(s).

In plain English

This was a computer-based (bioinformatics) study of gene activity in malignant mesothelioma, with a small check on hospital tissue samples. It did not study asbestos exposure. The authors used public single-cell data from three mesothelioma samples and grouped 1969 fibroblasts (connective-tissue cells) into six subtypes and three differentiation (maturity) states. They combined this with survival data from the TCGA database and selected 39 genes linked to both fibroblast differentiation and survival. They then used these genes to sort TCGA patients into three groups. The group tied to highly differentiated fibroblasts had the best survival, and the group tied to lowly differentiated fibroblasts had the worst. A six-gene risk score had an accuracy measure (AUC) of 0.916 in the whole cohort, but it was checked only within that same cohort. The authors also used a computer method to predict 24 possible drugs. In tumors from 11 patients at a Shanghai hospital, the three groups differed in differentiation, immune infiltration and metastasis (p = 0.02 for each). The authors concluded that their model had sound prognostic performance independent of clinical measures. They stated that most results lack laboratory verification and need external validation.

We wrote this plain-English summary from the full article with the help of AI, and a second AI model checked each sentence against the article. It is not the authors' wording; the abstract and the article are their own account.

What the study did and found

• Design: Bioinformatics (computer-based) analysis of public single-cell and bulk RNA sequencing data, with immunohistochemical staining and a small retrospective statistical check on hospital tumor samples
• Who was studied: Single-cell sequencing data from three mesothelioma samples in the public GEO database (GSE201925); bulk RNA sequencing and survival data on mesothelioma patients from the TCGA database (number not stated in the text); bone-metastatic tumor data from MET500. The validation set was 11 tumor samples from 11 patients diagnosed with malignant mesothelioma between January 1, 2020 and December 31, 2022 at the First Affiliated Hospital of Naval Medical University, Shanghai, China. The article uses 'malignant mesothelioma'/'MESO' and does not separate pleural from peritoneal cases in its analyses.
• Exposure: No exposure was studied. Asbestos is mentioned only as background on the cause of mesothelioma; no fiber types, jobs or doses were analyzed. The study examined gene expression in tumors.
• Main results:
  • In the three single-cell samples, 1969 fibroblasts were identified and divided into six fibroblast subtypes.
  • 39 genes were linked both to fibroblast differentiation and to survival in TCGA data (called fibroblast differentiation-related genes, FDGs).
  • Among TCGA patients sorted into three groups by these genes, the highly differentiated fibroblast-related group (HDFM) had better survival than the other two groups, and the lowly differentiated group (LDFM) had the worst.
  • The six-gene risk score model had AUC 0.916 in the whole TCGA cohort, 0.911 in the training cohort and 0.914 in the test cohort.
  • The risk score was an independent prognostic factor in analyses that included age, gender, stage and bone metastasis (p less than 0.001; hazard ratio 2828.449 univariate and 3201.584 multivariate; no confidence intervals given in the text).
  • A computer drug-sensitivity prediction identified 24 inhibitors to which the LDFM group was predicted to be more sensitive than the HDFM group.
  • In the 11 hospital tumor samples, the three subtypes differed significantly in differentiation (p = 0.02); immune infiltration and metastasis also differed (p = 0.02 each).
  • In the hospital samples, the authors interpreted HDFM as indicating good differentiation, high immune infiltration and no metastasis, and LDFM the opposite.
• The authors' conclusion: The authors concluded that their fibroblast differentiation-based classification and prognostic model worked independently of clinical parameters and had sound prognostic performance. They also suggested that these genes, especially those in low-differentiation states, might take part in mesothelioma proliferation and invasion.
• Limitations the authors state:
  • Most of the work relied on bioinformatics tools, and results lack experimental verification.
  • The prognostic model was validated only in internal cohorts; external validation is needed.
  • The predicted drug effects need cellular and animal experiments and clinical trials to confirm.
  • The number of clinical specimens for validation was limited because mesothelioma is rare.

Topics

Medical terms in this study, explained in our Lexicon: Fibroblast, Mesothelioma, Malignant Mesothelioma, Asbestos Exposure, Prognosis, Immunohistochemistry.

Related studies

• Identification of prognostic and bone metastasis​-related alternative splicing signatures in mesothelioma — R. Huang, Z. Zheng, et al. (2021)
• Mesothelioma survival prediction based on a six-gene transcriptomic signature — K. Behrouzfar, S.E. Mutsaers, et al. (2024)
• Towards a potential pan-cancer prognostic signature for gene expression based on probesets and ensemble machine learning — D. Chicco, A. Alameer, et al. (2022)
• Development of unenhanced CT-based imaging signature for BAP1 mutation status prediction in malignant pleural mesothelioma: Consideration of 2D and 3D segmentation — X.J. Xie, S.Y. Liu, et al. (2021)
• MicroRNA signature of malignant mesothelioma with potential diagnostic and prognostic implications — S. Busacca, S. Germano, et al. (2010)
• Osteopontin modulates inflammation, mucin production, and gene expression signatures after inhalation of asbestos in a murine model of fibrosis — T. Sabo-Attwood, M.E. Ramos-Nino, et al. (2011)

Browse return to medical research.