Transfer RNA beyond translation.

Figure 1. Main categories of tRFs and tiRNAs. Adapted from Xie et al (2020). Please notice there is no unified naming, and the different categories may appear under slightly different names in the literature.

tiRNAs have a length of ~30–40 nucleotides and are produced by cleaving across the anticodon loop under stress conditions and viral infections². tiRNAs are classified in two groups according to whether the sequence contains the original 5' or 3' end (Figure 1).

Notably, full-length tRNAs, tRFs, and tiRNAs have been consistently detected in extracellular vesicles as well as in the cell-free RNA fraction of biofluids, indicating potential roles in intercellular communication and utility as minimally invasive biomarkers⁷.

Known or proposed tRNA functions

Translational regulation and stress response

Some tiRNAs inhibit global translation initiation by binding to eIF4G/A, promoting stress granule assembly. For example, angiogenin generated 5′ tiRNAs can displace eIF4F complexes, reducing cap dependent initiation under stress^8,9. How broad is this mechanism across cell and tissue types is not yet understood. It is also unclear whether all stress conditions can generate tiRNAs, and what distinguishes functional species from background cleavage.

RNA interference

Several studies have reported that certain tRFs are loaded into Argonaute (AGO) complexes and can act like microRNAs, modulating target mRNA stability. The degree to which tRFs mimic miRNA targeting mechanism versus novel modes remains unclear.

For example, it has been shown displacement of YBX1 by tRFs suppresses oncogenic transcripts in breast cancer³. In human HEK293T cells, upregulating tRFs leads to the widespread repression of target genes. Rhizobial tRF-3 use host AGO1 to repress negative symbiotic regulators in the host. tRFs are also shown to hinder the replication of endogenous retroviruses by targeting its primer-binding site triggering immunity¹⁰.

Stem cells, development, fertility

Emerging work in mammalian stem cells and sperm maturation indicates that tRFs influence cell fate decisions and intergenerational information transfer¹¹. The challenge remains to define what specific tRF species are necessary or sufficient in developmental fate decisions. Also, it is unclear how these fragments integrate within known epigenetic and signaling networks.

Link with tRNA modifications and network specificity

A major open frontier is how tRNA modifications (e.g. methylations, pseudouridine) influence tRF generation, stability, or function. Some fragments resist detection due to modification induced reverse transcription blockage. The impact of modifications on cleavage and how do these layers regulate fragment repertoires in physiology or disease remains largely uncharted.

Sequencing tRNAs and tRFs

Small RNA sequencing NEXTFLEX Small RNA Sequencing Kit V4 Discover  is the standard method to profile RNAs with length <200 nt, including tRNA, tRFs and tiRNAs. It enables high throughput discovery and quantification of species present in a sample at single-base resolution, although there a few challenges to consider.

For example, many tRFs and tiRNAs (e.g. angiogenin generated), bear 5' OH or 2',3' cyclic phosphate ends which hinder standard adapter ligation. This can be overcome by treating the RNA upstream of library prep (e.g. T4 PNK or phosphatase). Protocols are readily available upon request.

Another challenge is the presence of internal modifications such as m¹A, m₃C, m¹G etc. They can block reverse transcriptase, causing drop outs and bias. Methods like PANDORA seq (which enzymatically demethylates before RT) have revealed previously under detected tRFs and tiRNAs¹².

Finally, accurate alignment of tRNA-derived reads remains a significant bioinformatic challenge. Mature tRNAs often share high sequence similarity, sometimes differing by only one or two nucleotides. As a result, standard small RNA-seq alignment pipelines frequently struggle with multimapping, misassignment, or low confidence in read origin, particularly when dealing with short or modified fragments. To address these limitations, specialized tools such as tDRmapper, MINTmap, SPORTS1.0, and tsRNAsearch have been developed. These pipelines integrate curated tRNA annotation databases, consider both precursor and mature tRNA isoacceptors, and implement logic to account for CCA addition, mismatches due to modifications, and fragment ambiguity.