Identification of RPL5 and RPL10 as novel diagnostic biomarkers of Atypical teratoid/rhabdoid tumors

Rhabdoid tumors (RTs) are aggressive tumors that occur most frequently in children under 2 years old, which often invade kidney (KRTs) and Center Nervous System, named Atypical teratoid/rhabdoid tumors (AT/RTs). RTs often progress fast and lead to a high lethality. RTs have a low incidence, we can hardly accumulate enough samples to elicit the diagnosis. More importantly, histologically, RTs present a host of neural, epithelial, mesenchymal, or ependymal patterns, which makes them rather variable and difficult to diagnose. Molecularly, RTs are diagnosed mainly on the lack of SMARCB1/INI1 protein expression, which, on the one hand, accounts for 75% of RTs, on the other hand, loss of expression of SMARCB1 is not exclusive to RTs. So, there is a need to find more accurate diagnose markers of RTs.

In this study, we analyzed 109 samples including AT/RT, KRT and corresponding normal samples downloaded form NCBI GEO database. First, we identified the differentially expressed lncRNAs and PCGs in AT/RT, KRT and corresponding normal samples. Second, we evaluated the co-expression relationship between lncRNA and PCG, and defined four types of the dysregulated PCG-lncRNA pairs. Third, we compared the differentially expressed genes, the dysregulated PCG-lncRNA pairs and commonly known cancer genes, we get potential diagnostic markers. Then, the potential diagnostic markers were subjected to Receiver operating characteristic (ROC) analysis to assess the diagnostic accuracy. Importantly, differential expression of the marker genes in different tumors was shown to distinguish AT/RT and KRT from other pediatric tumors specifically.

We compared the expression profiles between 47 AT/RTs, 31 KRTs, 8 normal brain samples, and 23 normal kidney samples. After applying a stringent set of criteria on the gene expression profiles, we identified 3667 PCGs and 81 lncRNAs differentially expressed in AT/RT, 3809 PCGs and 34 lncRNAs differentially expressed in KRT tissues. Next, we compared the three sets(AT/RT versus control brain samples, KRT versus control kidney samples, and AT/RT versus KRT) of differentially expressed lncRNAs and PCGs, 491 PCGs and 2 lncRNAs appeared in all three sets. We examined the correlation of the expression levels of these genes in the ‘three-set overlap’ group and identified four types of dysregulated lncRNAs and PCGs. By compared these genes to the well-known cancer driver genes, 19 PCGs were selected as potential candidates of diagnostic markers. Filtered with the number of the corresponding co-expressed lncRNA (namely “degree”), eight PCGs with more than five lncRNAs in the ‘three-set overlap’ group were selected as candidate diagnostic markers. Among them, RPL5 and RPL10 exhibited high sensitivity and specificity in diagnosis of AT/RT and KRT. However, when these two genes were used to distinguish AT/RT and KRT from other pediatric tumors, only AT/RT can be distinguished from medulloblastoma.

Our study mined existing GEO datasets for novel diagnostic markers associated with Rhabdoid tumors, and identified RPL5 and RPL10 as potential diagnostic markers for AT/RT. These two biomarkers may be used as supplementary biomarkers to canonical diagnostic tools such as biopsy and immunohistochemistry.