Development and validation of ferroptosis-related lncRNAs prognosis signatures in kidney renal clear cell carcinoma

Abstract

Background: Ferroptosis is a recently recognised new type of cell death which may be a potential target for cancer therapy. In the present study, we aimed to screen ferroptosis-related differentially expressed long non-coding RNAs as biomarkers to predict the outcome of kidney renal clear cell carcinoma.

Methods: RNAseq count data and corresponding clinical information were obtained from the Cancer Genome Atlas database. Lists of ferroptosis-related genes and long non-coding RNAs were obtained from the FerrDb and GENCODE databases, respectively. The candidate prognostic signatures were screened by Cox regression analyses and least absolute shrinkage and selection operator analyses.

Results: Three ferroptosis-related long non-coding RNAs (DUXAP8, LINC02609, and LUCAT1) were significantly correlated with the overall survival of kidney renal clear cell carcinoma independently. Kidney renal clear cell carcinoma patients with high-risk values displayed worse OS. Meanwhile, the expression of these three ferroptosis-related long non-coding RNAs and their risk scores were significantly correlated with clinicopathological features. Principal component analyses showed that patients with kidney renal clear cell carcinoma have differential risk values were well distinguished by the three ferroptosis-related long non-coding RNAs.

Conclusions: The present study suggests that the risk assessment model constructed by these three ferroptosis-related long non-coding RNAs could accurately predict the outcome of kidney renal clear cell carcinoma. We also provide a novel perspective for cancer prognosis screening.

Keywords: Ferroptosis; KIRC; Prognosis signatures; lncRNA.

Fig. 1

Differential expression analyses. Differential expression analyses of KIRC (a DEGs. b FR-DEG. c DELs). d Univariate Cox regression and K–M analyses illustrated nine FR-DELs associated with prognosis. g Multivariate Cox regression independently illustrated three FR-DELs associated with prognosis. d–f K–M plots of those 3 FR-DELs [DUXAP8 (d), LINC02609 (e), and LUCAT1(f)]

Fig. 2

Development and validation of prognosis lncRNAs signature. Risk value and survival status (a), expression (b), K–M curve (c), ROC curve (d) of the prognostic signature in the training group. Risk value and survival status (e), expression (f), K–M curve (g), and ROC curve (h) of the prognostic signature in the validation group. Risk value and survival status (i), expression (j), K–M curve (k), and ROC curve (l) of the prognostic signature in the entire group. *p  < 0.05, **p  < 0.01, ***p  < 0.001

Fig. 3

Independent prognostic factors of overall survival. a–c Univariate (blue) and multivariate (red) Cox regression of prognostic factors in the training, validation, and entire groups, respectively. ROC curve plot of risk model for all patients (d), patients at 3 (e), 5 (f), and 10-year (g) in the entire group

Fig. 4

Correlation analyses with clinicopathological features. Correlation of risk value with clinicopathological features [age (a), sex (b), pathologic T (c), pathologic N (d), pathologic M (e), and pathologic stage (f)]. Correlation of expression of ferroptosis-related lncRNAs with clinicopathological features [age (a), sex (b), pathologic T (c), pathologic N (d), pathologic M (e), and pathologic stage (f)]. *p  < 0.05, **p  < 0.01, ***p  < 0.001

Fig. 5

Principal component analyses. Principal component analysis plots displayed the distribution of patients with KIRC with high and low risk values based on 62 FR-DEGs (a), 361 DELs (b), 46 FR-DEGs (c), 251 FR-DELs (d), 9 FR-DELs (e), and 3 FR-DELs (f). Blue means low risk. Red means high risk

Fig. 6

Functional enrichment analyses. Significantly enriched GO term (top 10). BP biological process (a) CC cellular component (b). MF molecular function (c). d Significantly enriched KEGG pathway (top 10)