Integrated analysis of bulk and single cell datasets with experimental validation of cancer stemness function in thyroid cancer prognosis and immunotherapy

doi:10.21203/rs.3.rs-5314905/v1

Download PDF

Research Article

Integrated analysis of bulk and single cell datasets with experimental validation of cancer stemness function in thyroid cancer prognosis and immunotherapy

https://doi.org/10.21203/rs.3.rs-5314905/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Cancer stem cells (CSCs) are vital in tumor development, immune therapy resistance, and thyroid carcinoma (THCA) advancement. Nonetheless, the exact mechanisms are still unclear.

In this study, 28 mRNAs linked to cancer stemness were identified through an analysis of 13 publicly available THCA transcriptomic datasets alongside a CRISPR dataset for thyroid cancer cell lines. Importantly, we found a negative association between cancer stemness and the effectiveness of immunotherapy. By applying multiple machine learning techniques, a tumor stemness cell (TSC) model was both developed and validated, encompassing genes related to cancer stemness. This model effectively predicted patient prognosis and immunotherapy results for Cytotoxic T-Lymphocyte Associated Protein 4 (CTLA4) immune checkpoint inhibitors (ICIs) treatment across various cancer types. Additionally, chromosomal amplifications in regions 1q and 8p were identified as intrinsic contributors to the onset of thyroid cancer. Specifically, based on single cell dataset, the amplification of CKS1B on chromosome 1q was revealed to advance the progression of thyroid cancer cells by boosting their proliferation. Moreover, experiments uncovered that the over-expression of CKS1B also significantly promoted the proliferation and invasion abilities of THCA, which maybe a potential therapeutic target for THCA.

In conclusion, our research illuminates the correlation between cancer stem cell characteristics and the effectiveness of immunotherapy, offering a new framework for forecasting the outlook and response to immunotherapy in THCA patients on an individual basis.

Cancer stemness

Immunotherapy

Thyroid cancer

Single cell

CKS1B

Thyroid cancer, originating in the thyroid gland cells, stands as a common form of cancer and is the most frequently occurring malignant tumor within the endocrine system, showing an average annual growth rate of 4.5%[1]. Even though there have been advancements in diagnostic techniques and therapeutic measures, the capacity to swiftly diagnose and reliably forecast patient outcomes continues to be lacking[2]. Targeting inhibitory interactions between immune checkpoint receptors and their ligands, such as the programmed death 1 (PD-1) pathway, has revolutionized the treatment of advanced cancers with immune checkpoint inhibitors. The integration of immunotherapy has transformed the approach to cancer care and enhanced patient survival rates. However, due to the genetic diversity of tumors, current biomarkers are not adequate for accurately categorizing patients for optimal treatment outcomes[3]. Some patients may not respond as favorably and could experience negative effects, underscoring the importance of identifying robust markers and optimizing biomarker combinations. Precisely identifying patients who will benefit from immunotherapy could expand the pool of eligible candidates for this treatment, potentially lessening the impact of chemotherapy side effects in individuals with thyroid cancer.

CSCs exhibit a pronounced capacity for self-renewal, display considerable heterogeneity along with metastatic potential, and contribute significantly to tumor initiation, progression, and dissemination[4, 5]. The diversity of cellular components and the immune-suppressive milieu within the tumor serve as distinct yet interrelated factors driving tumor progression and fostering treatment resistance[6]. Prior investigations have demonstrated extensive negative associations between cancer stem cell traits and the immune system's response to cancer therapy. This phenomenon was noted even though tumors with high stemness exhibited elevated mutation rates, cancer-specific antigen expression, and intratumor heterogeneity across 21 different solid cancer types[7]. Additionally, Zhang Z et.al. discovered a negative association between tumor stemness and outcomes of immune checkpoint inhibitor therapy in various cancer groups. They formulated a robust model to investigate the link between cancer stemness and responsiveness to immunotherapy[8]. Nonetheless, the connection between tumor stemness and the success of ICI therapy in thyroid cancer has been generally overlooked.

In this research, we effectively employed the robust mRNA-based stemness index (mRNAsi)[9] to pinpoint 28 mRNAs associated with cancer stemness. Our findings indicated an inverse relationship between cancer stemness and immunotherapy across various transcriptome, single-cell, and CRISPR cell line datasets. Following this, we developed and validated a TCS scoring model aimed at predicting the prognosis and immunotherapy responses in patients with thyroid cancer. Moreover, we uncovered that the amplification of CKS1B, a pivotal gene linked to cancer stemness, can drive the proliferation of thyroid cancer cells. This gene may serve as an effective biomarker for predicting the cellular origins of thyroid malignant cells, as evidenced by cell line experiments.

Overall, our study provides new understandings into the relationship between immunotherapy and cancer stem cells in thyroid cancer. Additionally, it presents a strong model for classifying THCA patients according to their immune response and survival rates.

Acquisition thyroid cancer datasets

The dataset concerning thyroid cancer with survival data from The Cancer Genome Atlas (TCGA) was acquired from the UCSC Xena database[10]. In addition, 12 GEO cohorts were studied to focus on thyroid cancer (GSE104005, GSE196264, GSE29265, GSE32662, GSE35570, GSE53157, GSE5364, GSE60542, GSE65074, GSE65144, GSE76039, and GSE82208) to examine mRNAs associated with cancer stemness. Furthermore, the gene expression profiles of a single-cell dataset for thyroid cancer with accession number GSE184362[11] were collected from the GEO database to explore mRNAs linked to intrinsic stemness in cancer cells. This dataset comprised 7 primary tumors, 6 para-tumors, 8 metastatic LNs, and 2 subcutaneous metastatic sites of thyroid cancer tissues.

Collection pan-cancer anti-CTLA4 treated Immunotherapy-associated datasets

To authenticate the precision of the model in forecasting the response to CTLA4 immunotherapy, we collected diverse datasets from cohorts treated with anti-CTLA4. These included the cohort by Riaz N[12] (GSE91061: melanoma patients administered with anti-CTLA4 and PD-1) and the cohort by Necchi[13] (IMvigor210: individuals with advanced or metastatic urothelial carcinoma treated with Atezolizumab) obtained via the "IMvigor210CoreBiologies" R package. We amassed gene expression and clinical data from these cohorts.

Collection of single cell datasets for ICI-treated SKCM cohort

To explore the relationship between the stemness properties of cancer cells and their response to immunotherapy, a cohort of melanoma patients treated with ICIs was analyzed. This analysis incorporated single-cell RNA sequencing data and ICI response information, accessible from GEO under accession number GSE115978[14].

Identifying essential genes for thyroid carcinoma

Thyroid carcinoma cells underwent CRISPR screening via the DepMap portal (https://depmap.org/portal/download). Utilizing the CERES algorithm, dependency scores for around 17,000 genes were computed. Genes deemed vital displayed a CERES score below − 1 in 75% of the 21 THCA cell lines[15].

Development and validation of tumor stemness cell prediction model

The construction of a new pipeline was undertaken to create a forecast model for cancer stem cells, illustrated in Fig. 1. Firstly, the calculation of mRNAsi was conducted on 12 GEO datasets as well as the TCGA-THCA dataset. Subsequently, the connection between overall mRNA and mRNAsi was evaluated, and mRNAs showing noteworthy positive correlations in a minimum of seven out of thirteen sets (Cor > 0.2 and P < 0.05) were recognized as mRNAsi-related mRNAs, leading to the identification of 135 mRNAsi-linked mRNAs.

As a result of the scarcity of clinical survival data on thyroid cancer, the TCGA-THCA dataset was split into two halves, Cohort1 and Cohort2, while all samples were merged into Cohort3. Afterwards, the effectiveness of the stemness model for cancer was created and assessed. A range of machine learning methods including random forest (RSF), elastic net (Enet), gradient boosting (GBM), survival-SVM, ridge regression, Stepcox, plsRcox, CoxBoost, lasso, and SuperPC were employed to construct the stemness model for cancer.

Prediction of immunotherapy outcomes using TIDE webserver

To assess the efficacy of PD-1/CTLA4 immunotherapy, we first calculated scores for tumor immune dysfunction and exclusion (TIDE) by manipulating expression data from individuals with thyroid cancer. The resulting array of expression patterns was then examined on the TIDE database platform (http://tide.dfci.harvard.edu/) to assess patient outcomes. Following this, we employed the submap algorithm on the GenePattern portal to contrast response probabilities among low- and high-TSC categories.

Cell lines

Cell lines for human thyroid cancer (KHM-5M, TPC-1) were sourced from the Cell Bank of the Committee for Conservation of Typical Cultures at the Chinese Academy of Sciences. These specific cell strains were grown in Dulbecco's Modified Eagle Medium (DMEM) acquired from HyClone, located in Logan, UT, USA, and supplemented with 10% fetal bovine serum. Additionally, the medium contained 100 IU/mL of penicillin and streptomycin from Gibco based in New York, USA.

Knockout and overexpression in ATC cell lines

Lentiviral vectors for CKS1B overexpression were provided by Genechem (Shanghai, China). The cells were stably transfected with CKS1B overexpression lentiviruses and the respective control plasmids to induce puromycin resistance. According to the manufacturer's instructions, stable transfectants were selected using 2 µg/mL puromycin for 2–3 days to establish stable CKS1B overexpressing cell lines. The siRNA targeting CSK1B was synthesized by Biotend Co., Ltd. Cells were transfected with 50 nM of siRNA for 24 hours utilizing the jetPRIME transfection kit from Polyplus-transfection (Strasbourg, France).

Western blotting

The procedure for Western blotting was as follows: Cells in culture were rinsed with ice-cold PBS, and total cellular protein lysates were extracted at 4°C using RIPA lysis buffer (Beyotime, Shanghai, China) containing 1% protease inhibitor cocktail (MedChemExpress, New Jersey, USA). After centrifugation at 12,000 g for 20 minutes at 4°C, the resulting supernatant was gathered and combined with loading buffer. Subsequently, the samples were separated through 10% SDS-PAGE and transferred onto a PVDF membrane. The membrane was then obstructed using 5% skim milk for 2 hours at room temperature, followed by an overnight incubation at 4°C with primary antibodies. Following a wash with Tris Buffered Saline, the membrane underwent incubation with secondary antibodies. Detection was conducted using enhanced chemiluminescence reagents (Beyotime, Shanghai, China). The subsequent antibodies were employed for the western blot analysis: CKS1B (ab72639 Abcam, Waltham, Massachusetts, USA), GAPDH (60004-1-Ig proteintech, Wuhan, China). The dilution ratios for the antibodies used were as specified.

Assessment of cell proliferation and migration abilities

In order to assess the capacity for clone formation, a total of 1000 to 2000 cells were introduced into individual wells of a 6-well plate and incubated for approximately 7 to 10 days. Upon reaching a size of over 50 cells, the clones were treated with 0.2% crystal violet solution for half an hour. Subsequent to a thorough rinsing with PBS thrice, the clones underwent imaging and enumeration.

Cells were distributed evenly in a 6-well plate. When they achieved confluence, a 10-µL pipette tip was used to scratch a straight line along a ruler. After rinsing with PBS, the cells were cultured in a medium without serum. Incubation was done at 37°C with 5% CO2, and images were taken at both 0 and 24 hours.

To evaluate the migration capacity of cells, 3–5×10^4 cells were resuspended in 200 µL of culture medium and positioned in the upper chamber of Transwell plates (BD Biosciences, Bedford, MA, USA). Concurrently, 600 µL of culture medium containing 10% FBS was introduced into the lower chamber. Following an incubation period of 24 hours at 37°C, the cells were fixed using 4% paraformaldehyde for a duration of 30 minutes and subsequently stained with 0.25% crystal violet for another 30 minutes. The cells in the upper chamber were then carefully removed, and those that had migrated to the bottom side of the membrane were imaged and counted.

Statistical Analysis

Various clinical features between the high and low TSC score groups were compared using the Wilcoxon test. Analyzing the diversity in the response to immunotherapy among the groups with low and high TSC score was carried out using the Chisq test. Calculating Pearson's correlation coefficient served to explore the linkage between mRNA and mRNAsi. Investigations into the relationship among TSC, CKS1B, and survival were conducted using Kaplan-Meier survival analysis; the log-rank test determined the significance of observed differences. Time-dependent receiver operating characteristic (ROC) curves were utilized to assess both the prognostic and immunotherapy advantages of TSC, employing the 'pROC' R package. Based on the optimal threshold identified by the 'survminer' R package, patients were categorized accordingly. Statistical significance was defined by a significance level of P or adjP < 0.05.

Cancer stemness is negatively associated with immunotherapy benefits

An extensive analysis was initially conducted to investigate the connection between the cancer stemness and the effectiveness of immunotherapy. Thirteen THCA transcriptome datasets were extracted from the GEO and TCGA databases. The goal was to identify mRNAs linked to cancer stemness, and the mRNAsi value was calculated for each patient (Fig. 1A). By utilizing the Pearson correlation coefficient, mRNAs that showed a significant positive correlation with mRNAsi in most cohorts (more than 50%; 7 out of 13; Cor > 0.2 & P < 0.05) were pinpointed. A total of 135 mRNAs were then singled out as being related to cancer stemness, signifying the stemness level within the tumor cells. Additionally, a comprehensive analysis of CRISPR-based loss-of-function screens was carried out on a global scale using the DepMap database to identify key candidate genes associated with THCA malignancy. This analysis revealed a total of 647 genes deemed crucial for the survival of 28 THCA cell lines (CERE score < -1 in 75% of THCA malignant cells). Among these, 26 mRNAs were selected based on the overlap of THCA mRNAsi-associated mRNAs with those highlighted by CRISP (Fig. 1C).

Consistent with above results, the majority of the identified genes displayed significant overexpression in tumor tissues compared to adjacent tissues, providing evidence of their role in tumor progression (Fig. 1D). Furthermore, our KEGG enrichment analysis revealed that these genes are primarily involved in biological processes such as the cell cycle and DNA repair, further supporting their association with the stemness of tumor stem cells (Fig. 1E). All selected genes were chosen for further investigation. To assess the impact of cancer stemness-related genes on the effectiveness of immunotherapy, we analyzed a previously published single-cell dataset of ICI-treated melanoma patients (SKCM) to investigate the relationship between cancer stemness and ICI responses. After excluding samples lacking malignant cells, we focused on 23 patients, including 10 non-responders (NR) and 13 treatment-naïve (TN) individuals. While ideally comparing cancer stemness levels between responders and non-responders would have been optimal, due to data constraints, we compared stemness levels between the NR and TN groups. Using AUCell, we observed significantly higher levels of stemness in malignant cells from the NR subgroup (P < 0.001, Fig. 1H), indicating an inverse correlation between cancer stemness and immune checkpoint inhibitor outcomes.

Taken together, our results indicated that these 26 mRNAs above selected can accurately characterized cancer stemness, and the stemness of cancer is negatively correlated with immunotherapy benefits.

Establishment of cancer stemness score based on THCA RNA-seq datset

To further develop a prediction model for immunotherapy response based on cancer stemness associated mRNAs, various machine learning algorithms were utilized. The scores for each sample in three cohorts were calculated using these predictors. The performance was assessed by determining the average C-index for each algorithm, revealing that most predictors had a high average C-index (Fig. 2A). Among these models, the random forest (RSF) algorithm exhibited the highest precision (average C-index = 0.993, Fig. 2A) and was termed as the TSC score. Patients were then grouped based on the optimal threshold determined by the 'survminer' R package, showing that those with higher TSC scores had significantly reduced overall survival (OS) compared to those with lower TSC scores (Fig. 2B). Additionally, our analysis indicated that TSC was a reliable predictor, as evidenced by the area under the curve (AUC) values. The AUC curve demonstrated excellent predictive performance, with an average AUC > 0.98 across the three cohorts examined (Fig. 2C). To enhance the clinical applicability of our model, we investigated the potential independence of TSC from other clinical features. Multiple cox analysis revealed that TSC could be used as an independent prognostic factor for predicting THCA prognosis (Fig. 2D). The advancement in next-generation sequencing and large-scale data mining technologies has enabled a comprehensive exploration and improvement of gene expression-derived markers for prognosis prediction. To thoroughly evaluate the effectiveness of TSC compared to alternative markers, we systematically reviewed previously published markers from the last decade, analyzing a total of 79 models in this study. Notably, the reliability of the TSC in predicting survival outcomes exceeded that of all other models across eight different THCA cohorts, achieving an average AUC > 0.98 in the mentioned cohorts (Fig. 2E).

To sum up, these results suggested that TSC model is a reliable indicator for predicting the prognosis of THCA and it is significantly better than other models in terms of accuracy.

Tumor stemness cell score can server as a promising predictor for predicting CTLA4 immunotherapy outcomes

Given the negative correlation between cancer stemness and immunotherapy response, we investigated whether TSC could be utilized as a predictive factor for immunotherapy outcomes in patients with THCA. Employing the submap algorithm on the GenePattern platform, we sought to assess the likelihood of immunotherapy response in THCA patients categorized as high- or low-TSC. Interestingly, our analysis revealed that individuals in the low TSC category exhibited a notable response to CTLA4 immunotherapy (P = 0.072, Bonferroni corrected P < 0.01, see Fig. 3A), underscoring the predictive utility of the TSC model in determining the effectiveness of CTLA4 immunotherapy.

To further validate the predictive capability of TSC in assessing the therapeutic efficacy of CTLA4 ICIs, we compiled diverse datasets related to CTLA4 immune checkpoint inhibitors. Our results consistently showed that patients with lower TSC scores had significantly better overall survival rates following CTLA4 immunotherapy compared to those with higher TSC scores (Fig. 3B and 3C). This suggests that higher TSC scores may impede the benefits of CTLA4 immunotherapy. Patients with higher TSC scores had suboptimal responses to the treatment, with only 48% showing a positive response (Fig. 3D). Notably, our analysis indicated that TSC could accurately predict the response to CTLA4 ICI immunotherapy, as reflected by the high area under the curve (AUC) values. The AUC curve demonstrated excellent predictive performance, with 1-year AUC = 0.996, 2-year AUC = 0.989, and 3-year AUC = 0.989 in the IMvigor210 dataset; and 1-year AUC = 0.900, 2-year AUC = 0.975, and 3-year AUC = 0.854 in the SKCM dataset (Fig. 3E and 3F).

To summarize, the TSC model serves as a dependable predictor for projecting the outlook of THCA patients and acts as an efficient biomarker for anticipating patient response to immunotherapy.

The amplification of CKS1B is associated with the proliferation of malignant cells in THCA

To investigate the clonal architecture and cell origins of thyroid malignant cells, we conducted single-cell RNA profiling of thyroid carcinoma. Cells expressing fewer than 200 genes and those with more than 20% expression of mitochondrial genes were excluded. After quality control, the filtered cells were categorized into eight main cell types based on established biomarkers. Figure 4A and 4B display various cell populations, including B cells (MS4A1 and CD79A), Endothelial cells (PLVAP and VWF), Epithelial cells (KRT7 and EPCAM), Fibroblasts (COL1A1 and ACTA2), Myeloid cells (LYZ and CD68), Plasma cells (JCHAIN and MZB1), T & NK cells (CD3D and IL7R), as well as plasmacytoid dendritic cells (IRF7 and LILRA4).

Subsequently, the inferCNV algorithm was employed to evaluate copy number variations (CNV) and clonality in thyroid malignant cells derived from epithelial cells (ECs). Out of the 29499 ECs analyzed from THCA tissues, 21052 exhibited high CNV scores, indicating malignancy (Fig. 4C). Particularly, amplifications in chromosomal 8p and 1q were identified as key driving variations in thyroid cancer, with CKS1B amplification in chromosomal 1q being associated with stemness-related genes (Fig. 4D). Prior studies have demonstrated that CKS1B enhances cell proliferation and invasion by activating STAT3/AKT signaling in papillary thyroid carcinoma[16]. Consistently, CKS1B expression was significantly upregulated in tumors compared to adjacent para-cancer tissues (Fig. 4E). Moreover, in agreement with our bioinformatic results, the protein levels of CKS1B were notably elevated in LIHC tumor tissues compared to normal liver samples (Fig. 4F). Furthermore, we observed that higher CKS1B expression was specifically correlated with the advancement of tumor stage and size (Fig. 4G).

In summary, the amplification of CKS1B can drive the progression of thyroid malignant cells and it may serve as an efficient biomarker in foreseeing cellular origins of THCA malignant cells.

CKS1B enhances thyroid malignant cells’ proliferation, migration, and invasion

To investigate the impact of CKS1B on thyroid cancer cells, we categorized malignant tumor cells with an expression value above 1 as CKS1B + malignant tumor cells, while those with an expression value of 0 were classified as CKS1B- malignant tumor cells based on CKS1B expression levels. Our HALLMARK enrichment analysis revealed that CKS1B is capable of activating cell cycle-related pathways such as the G2M checkpoint and E2F targets (Figure S1A). To further validate the oncogenic role of CKS1B in thyroid cancer, we generated CKS1B knockdown and overexpressing models in KHM-5M and TPC-1 cell lines. Their efficacy was confirmed through western blot analyses at the protein level (Fig. 5A, B). Notably, CKS1B knockdown led to a marked inhibition of clonogenic capacity in KHM-5M cells (Fig. 5C), suggesting a role for CKS1B in promoting KHM-5M cell growth potential. Meanwhile, CKS1B overexpression significantly enhanced the clonogenic ability of TPC-1 cells (Fig. 5D), confirming the crucial role of CKS1B in promoting proliferation potential in thyroid cancer. Furthermore, in wound healing and Transwell assays, CKS1B knockdown significantly impaired the migratory capacity of KHM-5M cells (Fig. 5E), whereas CKS1B overexpression markedly enhanced the migratory ability of TPC-1 cells (Fig. 5F), indicating an important role for CKS1B in promoting the metastatic potential of thyroid cancer.

Immunotherapy, particularly immune checkpoint inhibitors, is now widely accepted as a second-line treatment for advanced cancers and is increasingly being approved for first-line use. Tumor heterogeneity and stemness are key factors in immune evasion and response to immunotherapy. Although the correlation between cancer stemness and anti-tumor immunity has been extensively researched, there is currently no direct evidence linking tumor stemness to ICI response in thyroid cancer. Immunotherapy for thyroid cancer faces various challenges, with the main hurdle being the identification of suitable candidates for ICI treatment.

In this research study, we employed an established stemness index[9] to detect 28 cancer stemness-linked mRNAs in 13 thyroid cancer datasets. These gene sets were identified to participate in various important pathways including cell cycle, DNA replication, nucleotide excision repair, and the p53 signaling route. These pathways are crucial for regulating cancer proliferation[17–19]. Furthermore, our investigation revealed an inverse relationship between stemness and the outcomes of immune checkpoint inhibitors, supported by data from an ICI-treated single-cell cohort of SKCM[14]. Moreover, several studies have indicated that patients with heightened cancer stemness experience deteriorated prognosis and heightened resistance to immunotherapy[8, 20]. Subsequently, to assess the prognostic effect of tumor stemness genes in thyroid cancer patients, we utilized diverse machine learning algorithms to establish a prognostic model for survival and response to immunotherapy. The performance of this model was authenticated in three datasets, with the Random Survival Forest model being identified as the favored TSC model due to its improved reliability and precision. Kaplan-Meier analysis indicated that patients with high-TSC scores in THCA exhibited significantly worse survival outcomes compared to patients with low-TSC scores. Furthermore, the AUC curve illustrated exceptional predictive capability, with an average AUC exceeding 0.98 across the three datasets. Additionally, our TSC score was validated as an autonomous predictor in multivariable Cox regression analysis. The TSC model showcased superior predictive performance for survival in contrast to conventional clinical and pathological characteristics, surpassing seventy-nine previously proposed models.

The predictive value of cancer stemness for immunotherapy in various cancer types has been well-established, but its potential role in thyroid cancer remains largely unexplored[8]. To evaluate the ability of the cancer stemness index to predict immunotherapy response, we employed TIDE and Submap webserver tools for the analysis of immunotherapy response in thyroid cancer[21]. Our results highlight the efficacy of the TSC model in predicting the response to CTLA4 immune checkpoint inhibitors specifically in thyroid cancer. These findings suggest that the TSC model may have broad utility in predicting CTLA4 ICI treatment responses across different cancer types. Subsequent comprehensive analysis revealed that TSC demonstrated high accuracy in predicting CTLA4 ICI responses in various datasets using bulk RNA-Seq data, with an average AUC exceeding 0.9.

The presence of stemness is positively associated with increased intratumoral heterogeneity, potentially providing a protective mechanism for antigenic clones against elimination by the immune system[7]. Copy number variation (CNV) in cancerous cells plays a crucial role in shaping tumor heterogeneity. Initial identification of inherent CNVs in malignant thyroid cells at the single-cell level indicated amplification of chromosomal 8p and 1q. Subsequent analysis linking mRNAsi and CNV profiles unveiled CKS1B as an endogenous mRNA linked to amplification-driven cancer stemness. CKS1B, a gene responsible for coding proteins, interacts with the catalytic subunit of cyclin-dependent kinases, facilitating proliferation, migration, invasion, and angiogenesis in retinoblastoma cells via MEK/ERK signaling[22]. Previous research has also demonstrated CKS1B's role as a predictor for immunotherapy response in pancreatic cancer[23]. In line with these findings, our study confirms a notable upregulation of CKS1B in thyroid tumor tissue, correlating with tumor advancement. On a mechanistic level, overexpression of CKS1B triggers the G2M checkpoint and E2F target-related pathways, bolstering proliferation, invasion, and metastasis of thyroid cancer cells, presenting CKS1B as a potential therapeutic target for THCA.

To summarize, our study has revealed a significant adverse correlation between cancer stemness and immunotherapy in thyroid cancer. Notably, we have developed a strong cancer stemness signature that can assess prognosis and potential benefits of immunotherapy, providing a useful tool for improving decision-making and monitoring strategies for individual patients with THCA. While we have shown the impressive predictive precision of the TSC in evaluating thyroid cancer prognosis and immunotherapy benefits, it is crucial to recognize some limitations in our research. The TSC model's ability to forecast immunotherapy outcomes for thyroid cancer is dependent on predictions from the submap and TIDE algorithm, and further validation of the TSC model using real thyroid cancer CTLA4 immunotherapy cohorts is required. Moreover, conducting studies with larger cohort data is vital to validate the performance of the TSC model in thyroid cancer.

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Funding

Not applicable.

Availability of data and materials

Sources of data for this research comprise public databases like TCGA, GEO, DepMap, and dbGaP. The manuscript and supplementary materials contain the data generated in this study. Additional data supporting this study's conclusions can be obtained from the corresponding author upon request.

Author contributions

Z.F., Y.Z. designed this project. Y.X. performed the bioinformatics analysis. Z.F. wrote the manuscript and supervised the project. Y.D., X.S., Q.Z., J.Z, Q.C., and Y.Z.. performed the data review and modified manuscript. All authors read and approved the manuscript.

Competing interests

The authors declare no competing interests.

Acknowledgments

Not applicable.

Siegel, R.L., K.D. Miller, and A. Jemal, Cancer statistics, 2015. CA: a cancer journal for clinicians, 2015. 65(1).
Miranda-Filho, A., et al., Thyroid cancer incidence trends by histology in 25 countries: a population-based study. The lancet Diabetes & endocrinology, 2021. 9(4): p. 225-234.
Hegde, P.S. and D.S. Chen, Top 10 challenges in cancer immunotherapy. Immunity, 2020. 52(1): p. 17-35.
Miao, Y., et al., Adaptive immune resistance emerges from tumor-initiating stem cells. Cell, 2019. 177(5): p. 1172-1186. e14.
Tang, C., et al., GART Functions as a Novel Methyltransferase in the RUVBL1/β‐Catenin Signaling Pathway to Promote Tumor Stemness in Colorectal Cancer. Advanced Science, 2023. 10(25): p. 2301264.
Bayik, D. and J.D. Lathia, Cancer stem cell–immune cell crosstalk in tumour progression. Nature Reviews Cancer, 2021. 21(8): p. 526-536.
Miranda, A., et al., Cancer stemness, intratumoral heterogeneity, and immune response across cancers. Proceedings of the National Academy of Sciences, 2019. 116(18): p. 9020-9029.
Zhang, Z., et al., Integrated analysis of single-cell and bulk RNA sequencing data reveals a pan-cancer stemness signature predicting immunotherapy response. Genome Medicine, 2022. 14(1): p. 45.
Lian, H., et al., Integrative analysis of gene expression and DNA methylation through one‐class logistic regression machine learning identifies stemness features in medulloblastoma. Molecular oncology, 2019. 13(10): p. 2227-2245.
Goldman, M.J., et al., Visualizing and interpreting cancer genomics data via the Xena platform. Nature biotechnology, 2020. 38(6): p. 675-678.
Pu, W., et al., Single-cell transcriptomic analysis of the tumor ecosystems underlying initiation and progression of papillary thyroid carcinoma. Nature Communications, 2021. 12(1): p. 6058.
Riaz, N., et al., Tumor and microenvironment evolution during immunotherapy with nivolumab. Cell, 2017. 171(4): p. 934-949. e16.
Necchi, A., et al., Atezolizumab in platinum-treated locally advanced or metastatic urothelial carcinoma: post-progression outcomes from the phase II IMvigor210 study. Annals of Oncology, 2017. 28(12): p. 3044-3050.
Jerby-Arnon, L., et al., A cancer cell program promotes T cell exclusion and resistance to checkpoint blockade. Cell, 2018. 175(4): p. 984-997. e24.
Meyers, R.M., et al., Computational correction of copy number effect improves specificity of CRISPR–Cas9 essentiality screens in cancer cells. Nature genetics, 2017. 49(12): p. 1779-1784.
Wang, H., et al., CKS1B promotes cell proliferation and invasion by activating STAT3/PD‐L1 and phosphorylation of Akt signaling in papillary thyroid carcinoma. Journal of Clinical Laboratory Analysis, 2021. 35(1): p. e23565.
Abad, E., D. Graifer, and A. Lyakhovich, DNA damage response and resistance of cancer stem cells. Cancer Letters, 2020. 474: p. 106-117.
Caglar, H.O. and C. Biray Avci, Alterations of cell cycle genes in cancer: unmasking the role of cancer stem cells. Molecular biology reports, 2020. 47(4): p. 3065-3076.
Sengupta, S., et al., Transition of amyloid/mutant p53 from tumor suppressor to an oncogene and therapeutic approaches to ameliorate metastasis and cancer stemness. Cancer Cell International, 2022. 22(1): p. 416.
Shi, X., et al., Cancer stemness associated with prognosis and the efficacy of immunotherapy in adrenocortical carcinoma. Frontiers in oncology, 2021. 11: p. 651622.
Jiang, P., et al., Signatures of T cell dysfunction and exclusion predict cancer immunotherapy response. Nature medicine, 2018. 24(10): p. 1550-1558.
Tang, Y., et al., CKS1B as a potential target for prognostic assessment and intervention in pancreatic cancer and its role in abnormal proliferation and cellular phenotype through mediation of cell cycle signaling pathways. Saudi Medical Journal, 2024. 45(2): p. 128.
Li, L., et al., Identification of CKS1B as a prognostic indicator and a predictive marker for immunotherapy in pancreatic cancer. Frontiers in immunology, 2022. 13: p. 1052768.

No competing interests reported.

FigureS1A.tif
Figure S1. CKS1B activates cell cycle-related pathways. A. GSEA enrichment of HALLMARK pathways between CKS1B- and CKS1B+ malignant cells.

Download PDF

Version 1

posted

You are reading this latest preprint version

Integrated analysis of bulk and single cell datasets with experimental validation of cancer stemness function in thyroid cancer prognosis and immunotherapy

Status:

Version 1

Abstract

Figures

Introduction

Methods

Acquisition thyroid cancer datasets

Collection pan-cancer anti-CTLA4 treated Immunotherapy-associated datasets

Collection of single cell datasets for ICI-treated SKCM cohort

Identifying essential genes for thyroid carcinoma

Development and validation of tumor stemness cell prediction model

Prediction of immunotherapy outcomes using TIDE webserver

Cell lines

Knockout and overexpression in ATC cell lines

Western blotting

Assessment of cell proliferation and migration abilities

Statistical Analysis

Results

Cancer stemness is negatively associated with immunotherapy benefits

Establishment of cancer stemness score based on THCA RNA-seq datset

Tumor stemness cell score can server as a promising predictor for predicting CTLA4 immunotherapy outcomes

The amplification of CKS1B is associated with the proliferation of malignant cells in THCA

CKS1B enhances thyroid malignant cells’ proliferation, migration, and invasion

Discussion

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1