Letter to editor regarding “piR-36249 and DHX36 together inhibit testicular cancer cells progression by upregulating OAS2”

Several reports describing PIWI-interacting RNAs (piRNAs) in human cancer cells or in the bloodstream are affected by the presence of false positives in piRNA databases. A recent report suggested that piR-36249 regulates testicular cancer progression by engaging with DHX36 to regulate OAS2. However, piR-36249 is a tRNA-Cys 5’ half capable of forming intermolecular G-quadruplexes. It is therefore expected that DHX36, a helicase with high affinity for DNA and RNA G-quadruplexes, was pulled down using piR-36249 mimicking probes. The suggestion of using piR-36249 as a therapeutic target for testicular cancer is therefore questionable, due to the consequences that tRNA inhibition could have on healthy cells.

Several reports describing PIWI-interacting RNAs (piRNAs) in human cancer cells or in the bloodstream are affected by the presence of false positives in piRNA databases.A recent report suggested that piR-36249 regulates testicular cancer progression by engaging with DHX36 to regulate OAS2.However, piR-36249 is a tRNA-Cys 5' half capable of forming intermolecular G-quadruplexes.It is therefore expected that DHX36, a helicase with high affinity for DNA and RNA G-quadruplexes, was pulled down using piR-36249 mimicking probes.The suggestion of using piR-36249 as a therapeutic target for testicular cancer is therefore questionable, due to the consequences that tRNA inhibition could have on healthy cells.

I have read the manuscript by Wang et al. recently published in
Noncoding RNA Research, suggesting that the piR-36249/DHX36/OAS2 axis regulates testicular cancer progression [1].While the paper includes several interesting experiments, I am afraid it is affected by the PIWI-interacting RNA (piRNA) annotation problem that is widespread in the mammalian somatic piRNA literature [2,3].
piR-36249 (DQ598183) is a tRNA-Cys 5' half.The same genomic region in chromosome 17 that is shown in Wang et al. [1] as the loci producing piR-36249 is annotated in the UCSC Genome Browser as tRNA-Cys, anticodon GCA (see HUGO or RepeatMasker tracks).Consistently, the sequence of piR-36249 is identical to the first 30 nucleotides of tRNA-Cys-GCA-2 based on the genomic tRNA database (http://gtrnadb.ucsc.edu/).
Like tRNA-Cys 5′ halves, piR-36249/DQ598183 starts with 5 consecutive guanines.This 5′ terminal oligoguanine (5′ TOG) motif induces the tetramerization of tRNA-Cys 5′ halves, due to the formation of intermolecular G-quadruplex structures [4,5].Tetrameric tRNA-Cys 5′ halves regulate translation by interacting with eIF4G [6,7].The differences observed by Wang et al. in cell viability, proliferation and migration when using piR-36249 mimics vs. control oligonucleotides or inhibitors can therefore be explained by the lack of the 5′ TOG motif in the control sequences.Additionally, while the biotinylated piR-36249 probe used for pull-down experiments contained the 5' TOG motif, the control probe did not.It is therefore not surprising that DHX36 was experimentally identified as a piR-36249 interactor [1], because DHX36 binds G-quadruplexes with extremely high affinity [8].
In their conclusion, Wang et al. state that their data supports piR-36249 as a novel therapeutic target in testicular cancer [1].However, targeting a tRNA gene would probably be toxic.
This article [1] illustrates why avoiding mis-classifying noncoding RNA fragments as piRNAs is of paramount importance.While somatic piRNA expression is well documented in arthropods, mollusks, and birds [9][10][11], piRNA expression in human cancer and the presence of piRNAs in the bloodstream remains to be demonstrated (despite a conundrum of papers suggesting so).
We have introduced the term "miscellaneous piRNAs" (m-piRNAs) to contemplate the possibility of certain noncoding RNA fragments being associated with PIWI proteins under certain conditions [3].However, no direct evidence linking piR-36249 and PIWI proteins was provided by Wang et al. so this piRNA should be considered a wrongly annotated tRNA-derived fragment rather than a canonical piRNA or an m-piRNA.
While results in Wang et al. are still valid, the interpretation of these results and the likelihood of using piR-36249 as a therapeutic target should be revisited.

Declaration of competing interest
The author has no competing interests regarding this submission.