Deciphering lung adenocarcinoma prognosis and immunotherapy response through an AI-driven stemness-related gene signature.

Lung adenocarcinoma (LUAD) is a leading cause of cancer-related deaths, and improving prognostic accuracy is vital for personalised treatment approaches, especially in the context of immunotherapy. In this study, we constructed an artificial intelligence (AI)-driven stemness-related gene signature (SRS) that deciphered LUAD prognosis and immunotherapy response. CytoTRACE analysis of single-cell RNA sequencing data identified genes associated with stemness in LUAD epithelial cells. An AI network integrating traditional regression, machine learning, and deep learning algorithms constructed the SRS based on genes associated with stemness. Subsequently, we conducted a comprehensive exploration of the connection between SRS and both intrinsic and extrinsic immune environments using multi-omics data. Experimental validation through siRNA knockdown in LUAD cell lines, followed by assessments of proliferation, migration, and invasion, confirmed the functional role of CKS1B, a top SRS gene. The SRS demonstrated high precision in predicting LUAD prognosis and likelihood of benefiting from immunotherapy. High-risk groups classified by the SRS exhibited decreased immunogenicity and reduced immune cell infiltration, indicating challenges for immunotherapy. Conversely, in vitro experiments revealed CKS1B knockdown significantly impaired aggressive cancer phenotypes like proliferation, migration, and invasion of LUAD cells, highlighting its pivotal role. These results underscore a close association between stemness and tumour immunity, offering predictive insights into the immune landscape and immunotherapy responses in LUAD. The newly established SRS holds promise as a valuable tool for selecting LUAD populations likely to benefit from future clinical stratification efforts.

A novel artificial intelligence network to assess the prognosis of gastrointestinal cancer to immunotherapy based on genetic mutation features.

Background: Immune checkpoint inhibitors (ICIs) have revolutionized gastrointestinal cancer treatment, yet the absence of reliable biomarkers hampers precise patient response prediction. Methods: We developed and validated a genomic mutation signature (GMS) employing a novel artificial intelligence network to forecast the prognosis of gastrointestinal cancer patients undergoing ICIs therapy. Subsequently, we explored the underlying immune landscapes across different subtypes using multiomics data. Finally, UMI-77 was pinpointed through the analysis of drug sensitization data from the Genomics of Drug Sensitivity in Cancer (GDSC) database. The sensitivity of UMI-77 to the AGS and MKN45 cell lines was evaluated using the cell counting kit-8 (CCK8) assay and the plate clone formation assay. Results: Using the artificial intelligence network, we developed the GMS that independently predicts the prognosis of gastrointestinal cancer patients. The GMS demonstrated consistent performance across three public cohorts and exhibited high sensitivity and specificity for 6, 12, and 24-month overall survival (OS) in receiver operating characteristic (ROC) curve analysis. It outperformed conventional clinical and molecular features. Low-risk samples showed a higher presence of cytolytic immune cells and enhanced immunogenic potential compared to high-risk samples. Additionally, we identified the small molecule compound UMI-77. The half-maximal inhibitory concentration (IC50) of UMI-77 was inversely related to the GMS. Notably, the AGS cell line, classified as high-risk, displayed greater sensitivity to UMI-77, whereas the MKN45 cell line, classified as low-risk, showed less sensitivity. Conclusion: The GMS developed here can reliably predict survival benefit for gastrointestinal cancer patients on ICIs therapy.

Single-cell sequencing reveals novel proliferative cell type: a key player in renal cell carcinoma prognosis and therapeutic response.

Renal cell carcinoma (RCC) is characterized by a variety of subtypes, each defined by unique genetic and morphological features. This study utilizes single-cell RNA sequencing to explore the molecular heterogeneity of RCC. A highly proliferative cell subset, termed as "Prol," was discovered within RCC tumors, and its increased presence was linked to poorer patient outcomes. An artificial intelligence network, encompassing traditional regression, machine learning, and deep learning algorithms, was employed to develop a Prol signature capable of predicting prognosis. The signature demonstrated superior performance in predicting RCC prognosis compared to other signatures and exhibited pan-cancer prognostic capabilities. RCC patients with high Prol signature scores exhibited resistance to targeted therapies and immunotherapies. Furthermore, the key gene CEP55 from the Prol signature was validated by both proteinomics and quantitative real time polymerase chain reaction. Our findings may provide new insights into the molecular and cellular mechanisms of RCC and facilitate the development of novel biomarkers and therapeutic targets.

Systemic immune-inflammation index predicts survival in patients with resected lung invasive mucinous adenocarcinoma.

BACKGROUND: The prognostic-related factors of lung invasive mucinous adenocarcinoma(IMA) are unclear because of its rarity. Various inflammation-based biomarkers were reported to predict the survival of malignant diseases. This study aims to explore the prognostic significance of the systemic immune-inflammation index(SII), which is calculated using absolute platelet, neutrophil, and lymphocyte counts, among patients with invasive mucinous adenocarcinoma. METHODS: From January 2015 to December 2019, 106 patients were identified as having IMA accepted radical resection and enrolled in the retrospective study. We analyzed the overall survival and disease-free survival using the Kaplan-Meier method and log-rank test. Receiver operating characteristic curve was used to find the optimal SII cut-off values for survival. A Cox regression model was carried out for multivariable analyses. RESULTS: The study cohort included 53 men and 53 women, with a mean age of 60 years (range 29 to 78 years, median 61 years). The median SII measured before surgery was 378.47 (range: 79.87-1701.97). ROC analyses revealed that the optimal cut-off values of SII was 379.43 for predicting both OS and DFS. An elevated SII (≥379.43) was observed in 52 patients (49.1 %), and was associated with younger age (P = 0.02), advanced T staging (P = 0.042), lymph node metastasis (P = 0.018) and pneumonic-type IMA (P = 0.018). Multivariable analysis showed that SII and pneumonic-type IMA were independent prognostic predictors of OS and DFS in radically resected IMA patients (P < 0.05). CONCLUSION: High SII is correlated with worse outcome and can be a novel prognostic biomarker for IMA patients accepted radical surgery.

MOICS, a novel classier deciphering immune heterogeneity and aid precise management of clear cell renal cell carcinoma at multiomics level.

Recent studies have indicated that the tumor immune microenvironment plays a pivotal role in the initiation and progression of clear cell renal cell carcinoma (ccRCC). However, the characteristics and heterogeneity of tumor immunity in ccRCC, particularly at the multiomics level, remain poorly understood. We analyzed immune multiomics datasets to perform a consensus cluster analysis and validate the clustering results across multiple internal and external ccRCC datasets; and identified two distinctive immune phenotypes of ccRCC, which we named multiomics immune-based cancer subtype 1 (MOICS1) and subtype 2 (MOICS2). The former, MOICS1, is characterized by an immune-hot phenotype with poor clinical outcomes, marked by significant proliferation of CD4+ and CD8+ T cells, fibroblasts, and high levels of immune inhibitory signatures; the latter, MOICS2, exhibits an immune-cold phenotype with favorable clinical characteristics, characterized by robust immune activity and high infiltration of endothelial cells and immune stimulatory signatures. Besides, a significant negative correlation between immune infiltration and angiogenesis were identified. We further explored the mechanisms underlying these differences, revealing that negatively regulated endopeptidase activity, activated cornification, and neutrophil degranulation may promote an immune-deficient phenotype, whereas enhanced monocyte recruitment could ameliorate this deficiency. Additionally, significant differences were observed in the genomic landscapes between the subtypes: MOICS1 exhibited mutations in TTN, BAP1, SETD2, MTOR, MUC16, CSMD3, and AKAP9, while MOICS2 was characterized by notable alterations in the TGF-ß pathway. Overall, our work demonstrates that multi-immune omics remodeling analysis enhances the understanding of the immune heterogeneity in ccRCC and supports precise patient management.

Identification of peripheral blood immune infiltration signatures and construction of monocyte-associated signatures in ovarian cancer and Alzheimer's disease using single-cell sequencing.

Background: Ovarian cancer (OC) is a common tumor of the female reproductive system, while Alzheimer's disease (AD) is a prevalent neurodegenerative disease that primarily affects cognitive function in the elderly. Monocytes are immune cells in the blood that can enter tissues and transform into macrophages, thus participating in immune and inflammatory responses. Overall, monocytes may play an important role in Alzheimer's disease and ovarian cancer. Methods: The CIBERSORT algorithm results indicate a potential crucial role of monocytes/macrophages in OC and AD. To identify monocyte marker genes, single-cell RNA-seq data of peripheral blood mononuclear cells (PBMCs) from OC and AD patients were analyzed. Enrichment analysis of various cell subpopulations was performed using the "irGSEA" R package. The estimation of cell cycle was conducted with the "tricycle" R package, and intercellular communication networks were analyzed using "CellChat". For 134 monocyte-associated genes (MRGs), bulk RNA-seq data from two diseased tissues were obtained. Cox regression analysis was employed to develop risk models, categorizing patients into high-risk (HR) and low-risk (LR) groups. The model's accuracy was validated using an external GEO cohort. The different risk groups were evaluated in terms of immune cell infiltration, mutational status, signaling pathways, immune checkpoint expression, and immunotherapy. To identify characteristic MRGs in AD, two machine learning algorithms, namely random forest and support vector machine (SVM), were utilized. Results: Based on Cox regression analysis, a risk model consisting of seven genes was developed in OC, indicating a better prognosis for patients in the LR group. The LR group had a higher tumor mutation burden, immune cell infiltration abundance, and immune checkpoint expression. The results of the TIDE algorithm and the IMvigor210 cohort showed that the LR group was more likely to benefit from immunotherapy. Finally, ZFP36L1 and AP1S2 were identified as characteristic MRGs affecting OC and AD progression. Conclusion: The risk profile containing seven genes identified in this study may help further guide clinical management and targeted therapy for OC. ZFP36L1 and AP1S2 may serve as biomarkers and new therapeutic targets for patients with OC and AD.

Crosstalk of disulfidptosis-related subtypes, establishment of a prognostic signature and immune infiltration characteristics in bladder cancer based on a machine learning survival framework.

Background: Bladder cancer (BLCA) is the most common malignancy of the urinary tract. On the other hand, disulfidptosis, a mechanism of disulfide stress-induced cell death, is closely associated with tumorigenesis and progression. Here, we investigated the impact of disulfidptosis-related genes (DRGs) on the prognosis of BLCA, identified various DRG clusters, and developed a risk model to assess patient prognosis, immunological profile, and treatment response. Methods: The expression and mutational characteristics of four DRGs were first analyzed in bulk RNA-Seq and single-cell RNA sequencing data, IHC staining identified the role of DRGs in BLCA progression, and two DRG clusters were identified by consensus clustering. Using the differentially expressed genes (DEGs) from these two clusters, we transformed ten machine learning algorithms into more than 80 combinations and finally selected the best algorithm to construct a disulfidptosis-related prognostic signature (DRPS). We based this selection on the mean C-index of three BLCA cohorts. Furthermore, we explored the differences in clinical characteristics, mutational landscape, immune cell infiltration, and predicted efficacy of immunotherapy between high and low-risk groups. To visually depict the clinical value of DRPS, we employed nomograms. Additionally, we verified whether DRPS predicts response to immunotherapy in BLCA patients by utilizing the Tumour Immune Dysfunction and Rejection (TIDE) and IMvigor 210 cohorts. Results: In the integrated cohort, we identified several DRG clusters and DRG gene clusters that differed significantly in overall survival (OS) and tumor microenvironment. After the integration of clinicopathological features, DRPS showed robust predictive power. Based on the median risk score associated with disulfidptosis, BLCA patients were divided into low-risk (LR) and high-risk (HR) groups, with patients in the LR group having a better prognosis, a higher tumor mutational load and being more sensitive to immunotherapy and chemotherapy. Conclusion: Our study, therefore, provides a valuable tool to further guide clinical management and tailor the treatment of BLCA patients, offering new insights into individualized treatment.

Single-cell RNA sequencing identifies a novel proliferation cell type affecting clinical outcome of pancreatic ductal adenocarcinoma.

Background: Pancreatic ductal adenocarcinoma (PDAC) is an extremely deadly neoplasm, with only a 5-year survival rate of around 9%. The tumor and its microenvironment are highly heterogeneous, and it is still unknown which cell types influence patient outcomes. Methods: We used single-cell RNA sequencing (scRNA-seq) and spatial transcriptome (ST) to identify differences in cell types. We then applied the scRNA-seq data to decompose the cell types in bulk RNA sequencing (bulk RNA-seq) data from the Cancer Genome Atlas (TCGA) cohort. We employed unbiased machine learning integration algorithms to develop a prognosis signature based on cell type makers. Lastly, we verified the differential expression of the key gene LY6D using immunohistochemistry and qRT-PCR. Results: In this study, we identified a novel cell type with high proliferative capacity, Prol, enriched with cell cycle and mitosis genes. We observed that the proportion of Prol cells was significantly increased in PDAC, and Prol cells were associated with reduced overall survival (OS) and progression-free survival (PFS). Additionally, the marker genes of Prol cell type, identified from scRNA-seq data, were upregulated and associated with poor prognosis in the bulk RNA-seq data. We further confirmed that mutant KRAS and TP53 were associated with an increased abundance of Prol cells and that these cells were associated with an immunosuppressive and cold tumor microenvironment in PDAC. ST determined the spatial location of Prol cells. Additionally, patients with a lower proportion of Prol cells in PDAC may benefit more from immunotherapy and gemcitabine treatment. Furthermore, we employed unbiased machine learning integration algorithms to develop a Prol signature that can precisely quantify the abundance of Prol cells and accurately predict prognosis. Finally, we confirmed that the LY6D protein and mRNA expression were markedly higher in pancreatic cancer than in normal pancreatic tissue. Conclusions: In summary, by integrating bulk RNA-seq and scRNA-seq, we identified a novel proliferative cell type, Prol, which influences the OS and PFS of PDAC patients.

A novel T-cell exhaustion-related feature can accurately predict the prognosis of OC patients.

The phenomenon of T Cell exhaustion (TEX) entails a progressive deterioration in the functionality of T cells within the immune system during prolonged conflicts with chronic infections or tumors. In the context of ovarian cancer immunotherapy, the development, and outcome of treatment are closely linked to T-cell exhaustion. Hence, gaining an in-depth understanding of the features of TEX within the immune microenvironment of ovarian cancer is of paramount importance for the management of OC patients. To this end, we leveraged single-cell RNA data from OC to perform clustering and identify T-cell marker genes utilizing the Unified Modal Approximation and Projection (UMAP) approach. Through GSVA and WGCNA in bulk RNA-seq data, we identified 185 TEX-related genes (TEXRGs). Subsequently, we transformed ten machine learning algorithms into 80 combinations and selected the most optimal one to construct TEX-related prognostic features (TEXRPS) based on the mean C-index of the three OC cohorts. In addition, we explored the disparities in clinicopathological features, mutational status, immune cell infiltration, and immunotherapy efficacy between the high-risk (HR) and low-risk (LR) groups. Upon the integration of clinicopathological features, TEXRPS displayed robust predictive power. Notably, patients in the LR group exhibited a superior prognosis, higher tumor mutational load (TMB), greater immune cell infiltration abundance, and enhanced sensitivity to immunotherapy. Lastly, we verified the differential expression of the model gene CD44 using qRT-PCR. In conclusion, our study offers a valuable tool to guide clinical management and targeted therapy of OC.

Single-cell RNA sequencing highlights the role of PVR/PVRL2 in the immunosuppressive tumour microenvironment in hepatocellular carcinoma.

Introduction: The conflict between cancer cells and the host immune system shapes the immune tumour microenvironment (TME) in hepatocellular carcinoma (HCC). A deep understanding of the heterogeneity and intercellular communication network in the TME of HCC will provide promising strategies to orchestrate the immune system to target and eradicate cancers. Methods: Here, we performed single-cell RNA sequencing (scRNA-seq) and computational analysis of 35786 unselected single cells from 3 human HCC tumour and 3 matched adjacent samples to elucidate the heterogeneity and intercellular communication network of the TME. The specific lysis of HCC cell lines was examined in vitro using cytotoxicity assays. Granzyme B concentration in supernatants of cytotoxicity assays was measured by ELISA. Results: We found that VCAN+ tumour-associated macrophages (TAMs) might undergo M2-like polarization and differentiate in the tumour region. Regulatory dendritic cells (DCs) exhibited immune regulatory and tolerogenic phenotypes in the TME. Furthermore, we observed intensive potential intercellular crosstalk among C1QC+ TAMs, regulatory DCs, regulator T (Treg) cells, and exhausted CD8+ T cells that fostered an immunosuppressive niche in the HCC TME. Moreover, we identified that the TIGIT-PVR/PVRL2 axis provides a prominent coinhibitory signal in the immunosuppressive TME. In vitro, antibody blockade of PVR or PVRL2 on HCC cell lines or TIGIT blockade on immune cells increased immune cell-mediated lysis of tumour cell. This enhanced immune response is paralleled by the increased secretion of Granzyme B by immune cells. Discussion: Collectively, our study revealed the functional state, clinical significance, and intercellular communication of immunosuppressive cells in HCC at single-cell resolution. Moreover, PVR/PVRL2, interact with TIGIT act as prominent coinhibitory signals and might represent a promising, efficacious immunotherapy strategy in HCC.

A novel immunogenomic signature to predict prognosis and reveal immune infiltration characteristics in pancreatic ductal adenocarcinoma.

Background: The immune response in the tumor microenvironment (TME) plays a crucial role in cancer progression and recurrence. We aimed to develop an immune-related gene (IRG) signature to improve prognostic predictive power and reveal the immune infiltration characteristics of pancreatic ductal adenocarcinoma (PDAC). Methods: The Cancer Genome Atlas (TCGA) PDAC was used to construct a prognostic model as a training cohort. The International Cancer Genome Consortium (ICGC) and the Gene Expression Omnibus (GEO) databases were set as validation datasets. Prognostic genes were screened by using univariate Cox regression. Then, a novel optimal prognostic model was developed by using least absolute shrinkage and selection operator (LASSO) Cox regression. Cell type identification by estimating the relative subsets of RNA transcripts (CIBERSORT) and estimation of stromal and immune cells in malignant tumors using expression data (ESTIMATE) algorithms were used to characterize tumor immune infiltrating patterns. The tumor immune dysfunction and exclusion (TIDE) algorithm was used to predict immunotherapy responsiveness. Results: A prognostic signature based on five IRGs (MET, ERAP2, IL20RB, EREG, and SHC2) was constructed in TCGA-PDAC and comprehensively validated in ICGC and GEO cohorts. Multivariate Cox regression analysis demonstrated that this signature had an independent prognostic value. The area under the curve (AUC) values of the receiver operating characteristic (ROC) curve at 1, 3, and 5 years of survival were 0.724, 0.702, and 0.776, respectively. We further demonstrated that our signature has better prognostic performance than recently published ones and is superior to traditional clinical factors such as grade and tumor node metastasis classification (TNM) stage in predicting survival. Moreover, we found higher abundance of CD8+ T cells and lower M2-like macrophages in the low-risk group of TCGA-PDAC, and predicted a higher proportion of immunotherapeutic responders in the low-risk group. Conclusions: We constructed an optimal prognostic model which had independent prognostic value and was comprehensively validated in external PDAC databases. Additionally, this five-genes signature could predict immune infiltration characteristics. Moreover, the signature helped stratify PDAC patients who might be more responsive to immunotherapy.

Non­small cell lung cancer carrying PBRM1 mutation suggests an immunologically cold phenotype leading to immunotherapy failure even with high TMB.

High tumor mutation load (TMB-H, or TMB ≥ 10) has been approved by the U.S. FDA as a biomarker for pembrolizumab treatment of solid tumors, including non­small cell lung cancer (NSCLC). Patients with cancer who have immunotherapy-resistant gene mutations cannot achieve clinical benefits even in TMB-H. In this study, we aimed to identify gene mutations associated with immunotherapy resistance and further informed mechanisms in NSCLC. A combined cohort of 350 immune checkpoint blockade-treated patients from Memorial Sloan Kettering Cancer Center (MSKCC) was used to identify genes whose mutations could negatively influence immunotherapy efficacy. An external NSCLC cohort for which profession-free survival (PFS) data were available was used for independent validation. CIBERSORT algorithms were used to characterize tumor immune infiltrating patterns. Immunogenomic features were analysed in the TCGA NSCLC cohort. We observed that PBRM1 mutations independently and negatively influence immunotherapy efficacy. Survival analysis showed that the overall survival (OS) and PFS of patients with PBRM1 mutations (MT) were significantly shorter than the wild type (WT). Moreover, compared with PBRM1-WT/TMB-H group, OS was worse in the PBRM1-MT/TMB-H group. Notably, in patients with TMB-H/PBRM1-MT, it was equal to that in the low-TMB group. The CIBERSORT algorithm further confirmed that the immune infiltration abundance of CD8+ T cells and activated CD4+ memory T was significantly lower in the MT group. Immunogenomic differences were observed in terms of immune signatures, T-cell receptor repertoire, and immune-related genes between WT and MT groups. Nevertheless, we noticed an inverse relationship, given that MT tumors had a higher TMB than the WT group in MSKCC and TCGA cohort. In conclusion, our study revealed that NSCLC with PBRM1 mutation might be an immunologically cold phenotype and exhibited immunotherapy resistance. NSCLC with PBRM1 mutation might be misclassified as immunoresponsive based on TMB.