A cuproptosis-related long non-coding RNA risk coefficient model to predict outcomes and therapeutic responses in papillary renal cell carcinoma

doi:10.21203/rs.3.rs-2488229/v1

Download PDF

Article

A cuproptosis-related long non-coding RNA risk coefficient model to predict outcomes and therapeutic responses in papillary renal cell carcinoma

https://doi.org/10.21203/rs.3.rs-2488229/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background

Papillary renal cell carcinoma (pRCC) is the second most common subtype of renal cell carcinoma, second only to clear cell renal cell carcinoma (ccRCC). Compared with ccRCC, metastatic pRCC has worse outcomes. Copper is an essential micronutrient; its accumulation is related to cancer proliferation, growth, angiogenesis, and metastasis. The abnormal expression of long non-coding RNA (lncRNA) plays an essential role in developing pRCC. Our study aimed to predict outcomes in pRCC patients using a prognostic risk coefficient model for cuproptosis-related lncRNAs.

Methods

We obtained data from pRCC patients and cuproptosis-related genes from The Cancer Genome Atlas and the literature. Cuproptosis-related lncRNAs were obtained using co-expression analysis based on R language software. The prognostic risk model of cuproptosis-related pRCC was established using univariate and multivariate Cox regressions and LASSO regression analysis, and the model was validated. The high- and low-risk groups were divided by the median value of the risk coefficient. Finally, we performed correlation analysis and potential drug identification for high- and low-risk subgroups.

Results

Seven cuproptosis-related lncRNAs were included (AC019080.5, AC092807.3, AC107464.2, AL5941845.1, GCC2-AS1, NINJ2-AS1, and ZNF710-AS1). Riskscore and tumor stage were independent prognostic factors in pRCC. Gene ontology and Kyoto Encyclopedia of Genes and Genomes pathway functional enrichment analysis showed that differentially expressed genes in pRCC cuproptosis-related lncRNAs were associated with the pathways of collagen-containing extracellular matrix formation and intracellular and extracellular structure formation. Correlation analysis of immune function showed that type II interferon responses, immune checkpoints, co-inhibition of T cells, cytolysis, co-inhibition of antigen-presenting cells, and human leukocyte antigen were significantly different between high- and low-risk groups. Higher Tumor Immune Dysfunction and Exclusion scores were observed in the low-risk group in the immune escape association analysis, illustrating the suboptimal effect of immunotherapy in low-risk patients. We identified 5-fluorouracil, epothilone B, gemcitabine, paclitaxel, pazopanib, and sunitinib as potentially more beneficial in the high-risk group; bortezomib, erlotinib, and sorafenib were more beneficial in the low-risk group.

Conclusion

We identified seven cuproptosis-related lncRNAs associated with pRCC outcomes and constructed a prognostic model that provides a basis for basic research and pRCC precision treatment.

Biological sciences/Cancer/Urological cancer

Biological sciences/Computational biology and bioinformatics/Predictive medicine

Biological sciences/Computational biology and bioinformatics/Virtual drug screening

TCGA database

papillary renal cell carcinoma

cuproptosis-related lncRNAs

risk coefficient model

prognosis

Globally, the top ten widespread malignancies include renal cell carcinoma (RCC). In 2022, the case of RCC in the US reached approximately 79,000 which accounted for 5% of all new cancer cases and 13,920 new deaths each year [1]. An estimated 77,410 new cases and 46,345 new deaths were recorded in China in 2022 [2]. Localized RCC is generally treated by surgery; however, up to 30% of patients have metastatic disease at initial diagnosis and report poor outcomes. The five-year survival is about 13.9%. An additional 30% of patients who have had their tumors removed develop recurrence [3]. Patients with metastatic RCC are usually resistant to chemoradiotherapy. Although targeted therapy has been used in clinical practice, about 75% of renal cancer are clear cell renal cell carcinoma (ccRCC), the most studied subtype in the field of biomarkers and targeted therapy; therefore, there are fewer treatments options for non-ccRCC patients [4]. The second most prevalent subtype of RCC, papillary renal cell carcinoma (pRCC), accounts for 15–20% of all RCC cases and is second only to ccRCC [5]. The outcomes of localized pRCC are better than that of ccRCC; however, a study reported worse outcomes for metastatic pRCC [6].

The class of RNA molecules that do not code for proteins and have a length of more than 200 bp are known as long non-coding RNAs (lncRNAs). Initially, it was believed to be a nonfunctional by-product of RNA polymerase II transcription and was later shown to interact with proteins, DNA, and other types of RNA to regulate gene expression. These lncRNAs in the nucleus are involved in the regulation of histone modifications [7], DNA methylation [8], and chromatin remodeling [9]. The lncRNAs located in the cytoplasm are involved in protein-related roles in the cytoplasm [10] and regulation of mRNA metabolism [11, 12]. They also regulate the initiation and development of tumors and immune tolerance [13–15]. Any abnormalities in the expression of some lncRNAs influence the development and outcomes of pRCC; for example, competitive endogenous RNAs on the MiR-140-3p-GPRIN1 axis, and lncRNAs upstream of them are associated with pRCC and may influence outcomes and therapeutic efficacy [16]. Chen et al. found that RP11-63A11.1 inhibited pRCC and may be a potential immune target [17]. Chen and his colleagues also found four immune-related lncRNAs (LIFR-AS1, AC099850.3, AL031710.1, and AC015922.3) associated with pRCC outcomes that could serve as potential therapeutic targets [18].

Cell death is a normal phenomenon that has become a research hot spot. There are several cell death pathways such as apoptosis, pyroptosis, necrosis, and ferroptosis. Copper is an essential micronutrient typically maintained at very low levels in mammalian cells. When copper ions' intracellular concentrations exceed the homeostasis threshold, copper can be cytotoxic [19]. Tsvetkov et al. proposed the term cuproptosis, a mechanism of controlled cell death distinct from the known cell death mechanisms. Analysis of mechanisms and disease models revealed that cuproptosis results in cell death by triggering the toxic stress of protein. The process starts with the tricarboxylic acid cycle’s fatty acylation components binding to copper. This results in aggregated lipoacylated proteins and the reduction in iron-sulfur cluster proteins eventually causing cell death [19]. FDX1 abundance and lipoacylated proteins highly correlate with several human tumors. The potential of copper ionophore as a potential cancer therapy is indicated by the sensitivity of cell lines with high levels of lipoacylated proteins to cuproptosis. Many studies showed that cuproptosis and lncRNA are associated with cancer development and outcomes [20–23]. However, little research has been done on the link between cuproptosis and lncRNAs in pRCC. Hence, this research dealt with investigating the link between cuproptosis and lncRNAs in pRCC using bioinformatics. A prognostic model of pRCC patients was developed that will contribute to future basic research and clinical practice.

Obtaining transcriptome and clinical data of pRCC

The RNA transcriptome of pRCC and clinical data were accessed at The Cancer Genome Atlas (TCGA) (https://cancergenome.nih.gov/) using the GDC Data Transfer Tool [24]. A total of 291 tumors and 32 normal tissue samples were included. The expression of the transcriptome in each sample was integrated, and lncRNAs were extracted using the Strawberry-Perl language.

Co-expression analysis of cuproptosis-linked genes and lncRNAs in pRCC

The literature yielded 19 cuproptosis-linked genes [19, 25], and their levels of expression in samples were extracted through the limma package in the R language software [26]. Normal tissue sample data were deleted, and cuproptosis-linked lncRNAs were determined using the Pearson correlation test based on a co-expression strategy, with filtering criteria for correlation coefficient > 0.4 and P < 0.001 [27]. The visualization of the co-expression analysis data was conducted through the ggplot2 and ggalluvial packages in R language software.

Development and validation of the prognostic risk model

Published clinical prognostic data of 289 pRCC patients were retrieved from TCGA. The model was built, and data visualization was conducted with the aid of the R software packages survival, caret, glmnet, and survminer. The patients were randomized to obtain data on 145 individuals in the training group to develop the risk model, and 144 individuals in the testing group were utilized to validate the risk model. The chi-square test was employed for the comparison of the clinical trait data between the training and testing groups. P-values above 0.05 implied that no remarkable difference (variation) existed between the groups. A univariate Cox analysis of clinical data and cuproptosis-linked lncRNAs with P < 0.05 was conducted in the training group to obtain the univariate cuproptosis-linked significant lncRNAs [28]. The least absolute shrinkage and selection operator (LASSO) regression model was developed for these significant lncRNAs [29]. The number of lncRNAs that needed to be identified was obtained by cross-validation, and a multivariate Cox analysis was conducted to obtain the optimal model and output the model formula (Riskscore=\(\sum lncRNA coef\left(n\right)\ast expression\left(n\right)\)). The model was employed for the prediction of the sample risk score of the training group. As per the median risk score serving as a cutoff value, the classification of the patients into groups of high- and low-risk patients was carried out. To validate the model’s formula in the testing group's samples, the survival variation between the two risk groups was assessed. The packages limma, reshape2, tidyverse, and ggplot2 of the R software were used to extract lncRNAs and the level of expression of cuproptosis-linked genes in risk models. Correlation analysis was carried out, and the correlation heatmap was drawn. The training and testing groups' high- and low-risk patients were subjected to survival analysis, and any variation in the results was determined through Kaplan-Meier plots. According to the median risk output of training and testing groups, risk curves were constructed using R language software. Subgroup analysis was conducted for each clinical characteristic of the risk model, with P < 0.05 considered a differential clinical feature.

Independent prognostic analysis of risk coefficient model and construction of receiver operating characteristic (ROC) curves

Univariate and multivariate independent prognostic analyses of clinical characteristics and ROC curves for constructing this risk coefficient model were performed using R language software's survival, survminer, and timeROC packages [30]. The clinical data from the model was circulated, and the ROC curves were plotted at one, three, and five years for the selected patients with pRCC to determine the value of the model's predictions by calculating the area under the curve (AUC). The bootcv simulation was utilized to calculate the c-index values of every clinical feature in the model using the aforementioned packages of the R language.

Principal component analysis (PCA) of pRCC data

PCA was performed on the downloaded whole lncRNAs transcriptome, as well as genes and lncRNAs linked to cuproptosis in pRCC patients using the limma and scatterplot3d packages. The data were visualized using PCA plots.

Differential analysis and functional pathway enrichment analysis of cuproptosis-linked lncRNAs

The limma package was utilized for differential analysis of the cuproptosis-linked lncRNAs between high- and low-risk groups in the model. Afterward, the genes depicting remarkable variations were selected with the following filtering criteria: |log (FC) |>1 and FDR<0.05. Following obtaining the differential genes, gene ontology (GO) enrichment analysis of the genes was carried out in terms of biological process, molecular function, and cellular component, and Kyoto Encylopedia of Genes and Genomes (KEGG) pathway enrichment analysis using R language software, applying the human genome annotation package at org.hs.eg.db, with filtering criteria of P < 0.05, and the data were visualized.

Immune-related functional analysis

The limma, GSVA, GSEAbase, heatmap, and reshape packages were employed to examine the variations in immune-related functions between the two risk groups. The known immune function-related gene set files in TISIDB were retrieved to quantify immune infiltration using ssGSEA analysis to visualize the data and draw heatmaps.

Correlation analysis between mutation frequency and burden

Publicly available nucleotide mutation data of pRCC samples were accessed from TCGA, and the data were processed. Tumor mutation burden (TMB) was calculated using Perl language scripts for each sample. The frequency of mutations in the risk groups’ tumor samples was assessed with the aid of the maftools package in R language. The subsequent visualization of 15 genes with the highest mutation frequency resulted in waterfall plots of high- and low-risk groups. Finally, the cutoff value of the cancer mutational burden was set for the purpose of generating survival curves using the survival and survmiver packages.

Tumor immune escape correlation analysis

To analyze the sensitivity of predicted tumors to immunotherapy, a correlation analysis of tumor immune escape among the patients in the risk groups was conducted. This analysis was based on the Tumor Immune Dysfunction and Exclusion (TIDE) (http://tide.dfci.harvard.edu) [31]. The resulting expression input files of the patients with pRCC were uploaded previously, and TIDE scores were obtained for the individual tumor samples. The patient immune escape data and risk coefficient scores were analyzed through the packages limma and ggpubr to determine if there was any variation between the two risk groups with P < 0.05 serving as the significance threshold. Finally, the resulting data were visualized to obtain violin and box plots.

Identification of potential beneficial drugs for patients with pRCC

To predict the variation in drug sensitivity between the two pRCC risk groups and identify their potential drugs, the half-maximal inhibitory concentration (IC₅₀) of drugs was utilized as an indicator to measure drug sensitivity. The analysis was conducted through the limma, ggpubr, pRRophetic, and ggplot2 packages, and the data was visualized by drawing box graphs.

Identification of cuproptosis-linked lncRNAs in pRCC

The flow chart of the model establishment and correlation analysis depicts the process in detail (Fig. 1); 16876 lncRNAs-related data of pRCC were acquired from TCGA. Using Pearson correlation analysis, 1418 cuproptosis-linked lncRNAs were identified with co-expression correlation and coefficient > 0.4 and P < 0.001 as identification criteria. R language software was applied to visualize the data, and Sankey diagrams were drawn (Fig. 2A). Correlation heatmaps were constructed for the lncRNAs involved in the model establishment and the 19 cuproptosis-linked lncRNAs (Fig. 2B).

Establishment of a risk coefficient model for pRCC

Data from 289 pRCC patients were retrieved from TCGA, followed by assigning 145 of the patients randomly to a training group to build the risk coefficient model. Univariate Cox regression analysis of the aforementioned training group-related 1418 cuproptosis-linked lncRNAs depicted that 19 of these lncRNAs were remarkably linked to the prognosis of pRCC patients (Fig. 3A). LASSO regression reduces the weight of features that do not contain useful information to reduce overfitting and enhance the stability of the prognostic model. Hence, for multivariate Cox regression analysis, LASSO regression analysis was utilized for the identification of 14 outcomes-related lncRNAs of cuproptosis (Figs. 3B and 3C). Finally, seven cuproptosis-linked lncRNAs were obtained to construct a model utilized for determining pRCC patients’ prognoses.

The association of the risk coefficient score with the high- and low-risk pRCC patients was assessed with the aid of the R language software (Fig. 4A). The association between the survival of the patient and the risk coefficient score was obtained over time (Fig. 4B). The correlation heatmap of the expression level of seven prognosis-linked cuproptosis lncRNAs was obtained in the training group samples (Fig. 4C). As per the survival status of patients in the high-risk group, the Kaplan-Meier curve was generated (Fig. 4D). In contrast with the group with high risk, the ten-year survival rate of patients in the group with low risk was considerably greater (P < 0.001).

Validation of the cuproptosis-linked lncRNAs-based risk coefficient model in pRCC

The risk coefficient model formula was placed into the testing group (Fig. 4E–4H) and the whole sample (Fig. 5A–5D) for verification of the predictive value of the training group. Characteristics such as whether the patient was older than 65, had tumor grade I-II and III-IV, and had stages T1-2 and T3-4 were assessed. Overall survival (OS) of patients with lower risk was greater than that of patients with higher risk (Fig. 6).

PCA of the risk coefficient model

PCA verifies the ability of lncRNAs participating in the model development to distinguish patients in high- and low-risk groups. PCA analysis was conducted using the R language, and data visualization was performed. PCA analysis of all sample genes (Fig. 7A), cuproptosis-linked genes (Fig. 7B), cuproptosis-linked lncRNAs (Fig. 7C), and associated lncRNAs that constructed the risk prediction model was performed (Fig. 7D).

GO and KEGG pathway function enrichment analysis of differentially expressed genes (DEGs)

R language software was utilized for functional enrichment analysis. GO functional enrichment analysis revealed a close relationship between cuproptosis-linked lncRNAs and collagen-containing extracellular matrix formation in pRCC, as visualized in the bubble and circle data (Figs. 8A and 8C). The KEGG pathway functional enrichment analysis indicated that the pathway was associated with the formation of intracellular and extracellular structures, as visualized in the bubble and circle data (Figs. 8B and 8D).

Correlation analysis of immune function, escape, and TMB

The correlation analysis of immune function identified the associated variation in immune function between high-risk groups, and a heatmap was generated (Fig. 9A). The heatmap implied that the immune functions of type II interferon responses, immune checkpoints, co-inhibition of T cells, cytolytic activity, co-inhibition of antigen-presenting cells, and human leukocyte antigen varied remarkably in high-risk groups (P < 0.05). R language software, examining the 15 genes with the highest mutation frequency, determined greater tumor mutation frequency in the group with higher risk than that in the group with lower risk-. The data were visualized, and a waterfall diagram was obtained (Figs. 9B and 9C). A violin plot of the immune escape correlation analysis was generated (Fig. 9D). The plot showed that in contrast with the high-risk group, the TIDE score was greater in the low-risk group (P<0.001), indicating a more remarkable potential of such lower-risk patients for immune escape. The evidence suggested that patients with lower risk would obtain less benefit from immunotherapy. Ultimately, a survival analysis of TMB yielded a Kaplan-Meier plot, which found that patients with a greater mutation burden had a higher survival rate in comparison with patients with a low mutation burden (P < 0.05) (Fig. 9E). However, in the survival analysis of the high- and low-risk subgroups, the patients with low risk had the highest survival rate with a low mutation burden (P < 0.001) (Fig. 9F).

Independent prognostic analysis of clinically relevant characteristics in the prognostic model

Using Cox analysis of the clinical characteristics, such as sex, age, tumor stage, and grade of the prognostic model, the features that function as prognostic factors were obtained. Univariate and multivariate Cox regression analyses were conducted, and it was revealed that tumor grade (hazard risk 2.127, P = 0.004) and risk coefficient score (hazard risk 1.002, P = 0.008) could independently predict patient prognosis (Figs. 10A and 10B). The one-, three-, and five-year ROC curves of pRCC patients were constructed (Fig. 10A), and the one-year AUC (0.98) maximum was calculated (Fig. 10C). The AUCs for three and five years were 0.966 and 0.857, respectively. ROC curves for clinically relevant indicators and risk coefficient scores for comparison were also generated (Fig. 10D). The risk coefficient score (AUC = 0.988) and tumor grade (AUC = 0.991) of one-year outcomes were the strongest predictors.

Identification of potential drugs

Potential drugs for patients with pRCC were identified using R language and its software packages. The 50% inhibitory concentration (IC₅₀) was used to determine the sensitivity in the high- and low-risk groups. The drugs such as 5-fluorouracil (Fig. 11A), epothilone B (Fig. 11C), gemcitabine (Fig. 11E), paclitaxel (Fig. 11F), pazopanib (Fig. 11G), and sunitinib (Fig. 11I) were determined as potentially more beneficial in patients with high risk. Bortezomib (Fig. 11B), erlotinib (Fig. 11D), and sorafenib (Fig. 11H) were beneficial in patients with low risk.

Depending on the pathology, renal cell carcinoma is classified into clear cell carcinoma, papillary cell carcinoma, and chromophobe cell carcinoma. Renal cell carcinoma’s second highly prevalent subtype is pRCC. The standard traditional surgical method is limited to patients with localized pRCC with a five-year OS of 78–79% [6]. Treating patients with initial metastasis, postoperative recurrence, and drug and radiochemotherapy tolerance remains challenging. Therefore, research on cancer biomarkers and outcome prediction of pRCC has been a research hotspot recently. Jing et al. found that methylation of five genes (CCNB2, CDKN1C, CTSH, E2F2, and ERMP1) was related to the prognosis of pRCC patients and could become potential targets of methylation-mediated immunotherapy [32]. Dang et al. identified 15 ferroptosis-linked lncRNAs associated with the prognosis of pRCC patients through TCGA data mining to generate a prognosis-linked risk model. Using correlation analysis, CD80, IDO1, and LAG3 were determined to be the therapeutic targets of immune checkpoint inhibitors [33]. Yu Liu et al. found four immune-related lncRNAs related to the prognosis of pRCC patients (AP001,267.3, AC026,471.3, SNHG16, and ADAMTS9-AS1) and established a model utilized for prognostic prediction. Subgroup analysis showed that SNHG16 expression significantly increased in female patients with higher T stage [34]. Yang et al. analyzed the data from pRCC patients using R language software and found two molecular markers (AC024022.1 and GAS6-AS1) that were beneficial and three molecular markers (AC087379.2, AL352984.1, and AL499627.1) that were related to poor outcomes [35]. A co-expression analysis by Wu et al. depicted that lncRNA expression was linked to ferroptosis, in which the expression levels of AL033397.2, AC099850.3, AC090197.1, LINC00462, B3GALT1-AS1, and CASC19 were elevated considerably in the high-risk group, suggesting that these markers were related to progression and might be cancer-promoting factors [36].

Copper is an essential trace element. The steady state of copper is crucial to various physiological processes. Its redox characteristics make it beneficial or harmful to cells because the imbalance of copper ion utilization in cells can induce oxidative stress and cytotoxicity [37]. Cuproptosis depends on the accumulation of copper, which differs from other cell death modes such as pyroptosis, apoptosis, necroptosis, and ferroptosis. Tsvetkov et al. found that the toxicity of copper was highly related to the activity of mitochondria by using a copper ion carrier that allowed small molecules to transport copper ions into cells. Compared with cells that provide energy by anaerobic glycolysis, cells that rely on mitochondrial activity to provide energy for aerobic respiration are sensitive to this copper carrier. The participation of critical components of the tricarboxylic acid cycle causes the accumulation of copper in cells, eventually resulting in cell death [19]. Regarding the relationship between disease status and copper, studies showed that copper content in tumor cells was higher compared with normal tissues, and copper accumulation was linked to the proliferation of tumors, along with their growth, metastasis, and angiogenesis [38–40].

Cuproptosis-linked genes perform an essential role in tumor occurrence and progression. Researchers conducted in-depth research on bladder cancer [41], clear cell renal cell carcinoma [42], hepatocellular carcinoma [43], lung adenocarcinoma [44], and other diseases. However, until now, no research has been conducted to analyze cuproptosis-linked genes in patients with pRCC. This research used R language software to analyze the co-expression of the transcriptome of lncRNA of 289 pRCC patients obtained from TCGA and known cuproptosis-linked genes. Univariate and multivariate Cox and LASSO regression analyses were conducted to identify seven cuproptosis-linked lncRNAs (AC019080.5, AC092807.3, AC107464.2, AL5941845.1, GCC2-AS1, NINJ2-AS1, and ZNF710-AS1) to develop the risk prognostic model for pRCC patients. These lncRNAs in pRCC patients increased significantly.

The lncRNA GCC2-AS1 gene is on chromosome 2q12.3 and may act as a reverse transcriptional regulator of GCC2. Thus, promoting the oligomerization of the ALK enzyme domain leads to the activation of the ALK enzyme. GCC2-ALK is a novel targeted fusion product. Patients with non-small cell lung cancer were sensitive to ALK inhibitors [45, 46]. AC019,080.5 was reported in establishing an immune-related lncRNA-based prognostic model for endometrial cancer patients, but it has not been studied yet [47]. The remaining five cuproptosis-linked lncRNAs in this study have not been reported previously; they might become pRCC tumor markers and potential drug targets.

Using the training group, a formula for a prognostic model was derived. High- and low-risk groups were classified as per the median risk value and were subjected to the testing group for verification. In contrast with the high-risk group, the group with low risk depicted greater OS for various clinical characteristics. The link between each pRCC patient’s risk score and the clinical features, including age, sex, and tumor grade was assessed. Multivariate Cox regression analysis depicted the ability of tumor grade and risk score to independently predict the prognosis of patients. ROC curves of clinically relevant indicators were generated, along with the subsequent comparison of the ROC curves of one-year risk coefficient scores. The tumor stage and one-year outcome risk coefficient score depicted the potential to independently predict the prognosis.

PCA was conducted to depict that the distinction between high-risk and low-risk groups was evident. To analyze the model, the functional enrichment of DEGs was assessed in pRCC patients. These cuproptosis-linked lncRNAs were associated with forming a collagen-containing extracellular matrix and forming internal and external structures. The correlation analysis of immune function showed significant differences among the high-risk groups in type II interferon responses, immune checkpoints, co-inhibition of T cells, cytolytic activity, co-inhibition of antigen-presenting cells, and human leukocyte antigen. The level of expression of these immune-related functions in high-risk groups was low, indicating low mobilization of immune function in such groups. The correlation analysis of tumor mutation and immune escape depicted increased tumor mutation frequency in the group with high risk than in the group with low risk.

The TIDE score, utilized for evaluating the effectiveness of immunotherapy, was employed to assess the immunotherapeutic response of patients in the two risk groups [48]. The group with low risk depicted an increased TIDE score, indicating the significant potential of patients in this group for immune escape, whereas the patients with increased risk depicted a better response to immunotherapy.

The tumor mutation load-related survival analysis depicted longer OS duration of pRCC patients with greater tumor mutation load. In most cancers, higher TMB indicates better responses to immune checkpoint inhibitors. TMB can be utilized as a cancer biomarker [49]. However, the survival analysis of high-risk subgroups depicted an increased survival rate of low-risk patients with a low mutation load.

To explore the potential beneficial drugs for patients with metastatic pRCC, medications were identified using R language and IC₅₀ as drug sensitivity indicators. 5-fluorouracil, epothilone B, gemcitabine, paclitaxel, pazopanib, and sunitinib were identified as potentially more beneficial in the high-risk group. Bortezomib, erlotinib, and sorafenib were more beneficial in the low-risk group. These findings have not been reported previously for pRCC patients.

This research has some limitations. The data on pRCC patients were retrieved from TCGA. Most patient data were from European and American countries and were retrospective studies; therefore, there may have been selection bias. Additionally, it is essential to use external clinical data to confirm the robustness of the risk coefficient model. The focus of our follow-up work is to collect external clinical data of patients and expand the sample size to validate the reliability of this prognostic risk model.

TCGA and R language were employed to identify seven cuproptosis-linked lncRNAs related to the prognosis of pRCC individuals through Cox regression analysis. A prognostic risk coefficient scoring model was established, a correlation analysis was conducted on the biomolecules, and the potential benefit of medications was predicted. The findings of this research establish a basis and direction for the future basic research of pRCC and clinical precision medicine.

Author contributions

C. Tang wrote the main manuscript text and analyzed the data; G. Qu performed experiments and analyzed data; G. Qu and Y. Xu proposed the idea; G. Yang, W. Huang, and D. Chen collected data. All the authors reviewed the manuscript and discussed the results and edited the manuscript.

Data availability statement

The datasets presented in this study can be found in online repositories. The RNA transcriptome of pRCC and clinical data were accessed at TCGA (https://cancergenome.nih.gov/). All data generated or analyzed during this study are included in this article.

Competing Interests Statement

The authors declare that they have no competing interests.

Siegel, R.L., Miller, K.D., Fuchs, H.E. & Jemal, A. Cancer statistics, 2022. CA: a cancer journal for clinicians .72, 7-33 (2022).
Xia, C., et al. Cancer statistics in China and United States, 2022: profiles, trends, and determinants. Chinese medical journal .135, 584-590 (2022).
Rini, B.I., Campbell, S.C. & Escudier, B. Renal cell carcinoma. Lancet (London, England) .373, 1119-1132 (2009).
Vera-Badillo, F.E., et al. Systemic therapy for non-clear cell renal cell carcinomas: a systematic review and meta-analysis. European urology .67, 740-749 (2015).
Angori, S., Lobo, J. & Moch, H. Papillary renal cell carcinoma: current and controversial issues. Current opinion in urology .32, 344-351 (2022).
Steffens, S., et al. Incidence and long-term prognosis of papillary compared to clear cell renal cell carcinoma--a multicentre study. European journal of cancer (Oxford, England : 1990). 48, 2347-2352 (2012).
Li, L., et al. Targeted disruption of Hotair leads to homeotic transformation and gene derepression. Cell reports. 5, 3-12 (2013).
Venkatraman, A., et al. Maternal imprinting at the H19-Igf2 locus maintains adult haematopoietic stem cell quiescence. Nature .500, 345-349 (2013).
Chen, D., et al. ANRIL inhibits p15(INK4b) through the TGFβ1 signaling pathway in human esophageal squamous cell carcinoma. Cellular immunology. 289, 91-96 (2014).
Chu, C., et al. Systematic discovery of Xist RNA binding proteins. Cell .161, 404-416 (2015).
Cooper, D.R., et al. Long Non-Coding RNA NEAT1 Associates with SRp40 to Temporally Regulate PPARγ2 Splicing during Adipogenesis in 3T3-L1 Cells. Genes .5, 1050-1063 (2014).
Cai, R., et al. Adiponectin AS lncRNA inhibits adipogenesis by transferring from nucleus to cytoplasm and attenuating Adiponectin mRNA translation. Biochimica et biophysica acta. Molecular and cell biology of lipids .1863, 420-432 (2018).
St Laurent, G., Wahlestedt, C. & Kapranov, P. The Landscape of long noncoding RNA classification. Trends in genetics : TIG .31, 239-251 (2015).
Goodall, G.J. & Wickramasinghe, V.O. RNA in cancer. Nature reviews. Cancer .21, 22-36 (2021).
Statello, L., Guo, C.J., Chen, L.L. & Huarte, M. Gene regulation by long non-coding RNAs and its biological functions. Nature reviews. Molecular cell biology .22, 96-118 (2021).
Zhou, Q., et al. A novel lncRNA-miRNA-mRNA competing endogenous RNA regulatory network in lung adenocarcinoma and kidney renal papillary cell carcinoma. Thoracic cancer .12, 2526-2536 (2021).
Chen, B., Dong, D., Yao, Q., Zou, Y. & Hu, W. A novel prognostic cancer-related lncRNA signature in papillary renal cell carcinoma. Cancer cell international .21, 545 (2021).
Chen, S.H., et al. An immune-related lncRNA prognostic model in papillary renal cell carcinoma: A lncRNA expression analysis. Genomics .113, 531-540 (2021).
Tsvetkov, P., et al. Copper induces cell death by targeting lipoylated TCA cycle proteins. Science (New York, N.Y.) .375, 1254-1261 (2022).
Li, X., et al. Multi-omics analysis reveals prognostic and therapeutic value of cuproptosis-related lncRNAs in oral squamous cell carcinoma. Frontiers in genetics .13, 984911 (2022).
Ding, L., et al. Identification of cuproptosis-related subtypes, cuproptosis-related gene prognostic index in hepatocellular carcinoma. Frontiers in immunology .13, 989156 (2022).
Wang, Y., et al. In Silico Identification and Validation of Cuproptosis-Related LncRNA Signature as a Novel Prognostic Model and Immune Function Analysis in Colon Adenocarcinoma. Current oncology (Toronto, Ont.) .29, 6573-6593 (2022).
Pan, S., et al. Identification of cuproptosis-related subtypes in lung adenocarcinoma and its potential significance. Frontiers in pharmacology .13, 934722 (2022).
Tomczak, K., Czerwińska, P. & Wiznerowicz, M. The Cancer Genome Atlas (TCGA): an immeasurable source of knowledge. Contemporary oncology (Poznan, Poland) .19, A68-77 (2015).
Xu, S., et al. Cuproptosis-Associated lncRNA Establishes New Prognostic Profile and Predicts Immunotherapy Response in Clear Cell Renal Cell Carcinoma. Frontiers in genetics .13, 938259 (2022).
Ritchie, M.E., et al. limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic acids research .43, e47 (2015).
Pripp, A.H. [Pearson's or Spearman's correlation coefficients]. Tidsskrift for den Norske laegeforening : tidsskrift for praktisk medicin, ny raekke .138(2018).
Zhao, Z., et al. Necroptosis-Related lncRNAs: Predicting Prognosis and the Distinction between the Cold and Hot Tumors in Gastric Cancer. Journal of oncology .2021, 6718443 (2021).
Tibshirani, R. The lasso method for variable selection in the Cox model. Statistics in medicine .16, 385-395 (1997).
Hoo, Z.H., Candlish, J. & Teare, D. What is an ROC curve? Emergency medicine journal : EMJ .34, 357-359 (2017).
Brinkman, E.K. & van Steensel, B. Rapid Quantitative Evaluation of CRISPR Genome Editing by TIDE and TIDER. Methods in molecular biology (Clifton, N.J.). 1961, 29-44 (2019).
Jing, X., et al. A five-gene methylation signature predicts overall survival of patients with clear cell renal cell carcinoma. Journal of clinical laboratory analysis .35, e24031 (2021).
Dang, R., et al. A Novel Ferroptosis-Related lncRNA Signature for Prognosis Prediction in Patients with Papillary Renal Cell Carcinoma. International journal of general medicine .15, 207-222 (2022).
Liu, Y., et al. Bioinformatics profiling integrating a four immune-related long non-coding RNAs signature as a prognostic model for papillary renal cell carcinoma. Aging .12, 15359-15373 (2020).
Yang, F., et al. Long non-coding RNAs as prognostic biomarkers in papillary renal cell carcinoma. Oncology letters .18, 3691-3697 (2019).
Wu, Z., Huang, X., Cai, M. & Huang, P. Potential biomarkers for predicting the overall survival outcome of kidney renal papillary cell carcinoma: an analysis of ferroptosis-related LNCRNAs. BMC urology. 22, 152 (2022).
Ge, E.J., et al. Connecting copper and cancer: from transition metal signalling to metalloplasia. Nature reviews. Cancer .22, 102-113 (2022).
Michniewicz, F., et al. Copper: An Intracellular Achilles' Heel Allowing the Targeting of Epigenetics, Kinase Pathways, and Cell Metabolism in Cancer Therapeutics. ChemMedChem .16, 2315-2329 (2021).
Aubert, L., et al. Copper bioavailability is a KRAS-specific vulnerability in colorectal cancer. Nature communications .11, 3701 (2020).
Saleh, S.A.K., Adly, H.M., Abdelkhaliq, A.A. & Nassir, A.M. Serum Levels of Selenium, Zinc, Copper, Manganese, and Iron in Prostate Cancer Patients. Current urology .14, 44-49 (2020).
Song, Q., Zhou, R., Shu, F. & Fu, W. Cuproptosis scoring system to predict the clinical outcome and immune response in bladder cancer. Frontiers in immunology .13, 958368 (2022).
Bian, Z., Fan, R. & Xie, L. A Novel Cuproptosis-Related Prognostic Gene Signature and Validation of Differential Expression in Clear Cell Renal Cell Carcinoma. Genes .13(2022).
Zhang, Z., et al. Cuproptosis-Related Risk Score Predicts Prognosis and Characterizes the Tumor Microenvironment in Hepatocellular Carcinoma. Frontiers in immunology. 13, 925618 (2022).
Hu, Q., Wang, R., Ma, H., Zhang, Z. & Xue, Q. Cuproptosis predicts the risk and clinical outcomes of lung adenocarcinoma. Frontiers in oncology .12, 922332 (2022).
Jiang, J., et al. GCC2-ALK as a targetable fusion in lung adenocarcinoma and its enduring clinical responses to ALK inhibitors. Lung cancer (Amsterdam, Netherlands) .115, 5-11 (2018).
Vendrell, J.A., et al. Detection of known and novel ALK fusion transcripts in lung cancer patients using next-generation sequencing approaches. Scientific reports 7, 12510 (2017).
Wang, Z., et al. An Immune-Related Long Noncoding RNA Signature as a Prognostic Biomarker for Human Endometrial Cancer. Journal of oncology .2021, 9972454 (2021).
Jiang, P., et al. Signatures of T cell dysfunction and exclusion predict cancer immunotherapy response. Nature medicine .24, 1550-1558 (2018).
Samstein, R.M., et al. Tumor mutational load predicts survival after immunotherapy across multiple cancer types. Nature genetics .51, 202-206 (2019).

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

A cuproptosis-related long non-coding RNA risk coefficient model to predict outcomes and therapeutic responses in papillary renal cell carcinoma

Status:

Version 1

Abstract

Background

Methods

Results

Conclusion

Figures

Introduction

Materials And Methods

Obtaining transcriptome and clinical data of pRCC

Co-expression analysis of cuproptosis-linked genes and lncRNAs in pRCC

Development and validation of the prognostic risk model

Independent prognostic analysis of risk coefficient model and construction of receiver operating characteristic (ROC) curves

Principal component analysis (PCA) of pRCC data

Differential analysis and functional pathway enrichment analysis of cuproptosis-linked lncRNAs

Immune-related functional analysis

Correlation analysis between mutation frequency and burden

Tumor immune escape correlation analysis

Identification of potential beneficial drugs for patients with pRCC

Results

Identification of cuproptosis-linked lncRNAs in pRCC

Establishment of a risk coefficient model for pRCC

Validation of the cuproptosis-linked lncRNAs-based risk coefficient model in pRCC

PCA of the risk coefficient model

GO and KEGG pathway function enrichment analysis of differentially expressed genes (DEGs)

Correlation analysis of immune function, escape, and TMB

Independent prognostic analysis of clinically relevant characteristics in the prognostic model

Identification of potential drugs

Discussion

Declarations

References

Additional Declarations

Status:

Version 1