Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
, 20 (7), e47789

Dynamic Expression of tRNA-derived Small RNAs Define Cellular States

Affiliations

Dynamic Expression of tRNA-derived Small RNAs Define Cellular States

Srikar Krishna et al. EMBO Rep.

Abstract

Transfer RNA (tRNA)-derived small RNAs (tsRNAs) have recently emerged as important regulators of protein translation and shown to have diverse biological functions. However, the underlying cellular and molecular mechanisms of tsRNA function in the context of dynamic cell-state transitions remain unclear. Expression analysis of tsRNAs in distinct heterologous cell and tissue models of stem vs. differentiated states revealed a differentiation-dependent enrichment of 5'-tsRNAs. We report the identification of a set of 5'-tsRNAs that is upregulated in differentiating mouse embryonic stem cells (mESCs). Notably, interactome studies with differentially enriched 5'-tsRNAs revealed a switch in their association with "effector" RNPs and "target" mRNAs in different cell states. We demonstrate that specific 5'-tsRNAs can preferentially interact with the RNA-binding protein, Igf2bp1, in the RA-induced differentiated state. This association influences the transcript stability and thereby translation of the pluripotency-promoting factor, c-Myc, thus providing a mechanistic basis for how 5'-tsRNAs can modulate stem cell states in mESCs. Together our study highlights the role of 5'-tsRNAs in defining distinct cell states.

Keywords: Igf2bp1; c-Myc; stem cell differentiation; tRNA-derived small RNAs (tsRNAs); translation.

Conflict of interest statement

The authors declare that they have no conflict of interest.

Figures

Figure EV1
Figure EV1. Profiling and characterisation of small RNAs and tsRNAs in heterologous cell and tissue models of stem vs. differentiated states

Expression of pluripotency genes in mESCs grown in stem or differentiating conditions (n = 2, error bars represent SD, and significance was calculated using Student's t‐test).

Size distribution of the sequence reads from small RNA libraries that mapped to the genome.

Size distribution of the small RNA reads that map to the tRNA.

Correlation of tRNA gene copy number and tsRNA abundance. tRNA gene copy numbers for every anticodon are represented as percentages of the total tRNA copies on the genome.

Comparison of small RNA population changes in RA‐induced differentiating cells treated with T4 PNK.

tsRNA profile of bulge stem cell (CD34+α6+) or basal cell (CD34α6+) populations isolated from adult murine skin (see Appendix Fig S1).

Images of HMECs (Panels I‐IV) engineered to represent stages of oncogenic transformation: I: parental HMECs immortalized with hTERT; II: immortalized HMECs with tamoxifen‐inducible oncogenic HRAS (hTERT/sT/HRASG12V:ER/EV); III: HMEC as described in II with additional p53 knockdown (hTERT/sT/HRASG12V:ER/shP53); HMECs I‐III were treated with 130 nM tamoxifen (5 days); and IV: HMEC II treated with tamoxifen (10 days).

Quantification of 30–35 nt RNA levels in HMECs I–IV (n = 3, error bars represent SD, and significance was calculated using Student's t‐test).

tsRNA profiles of stem (CD44+CD24) vs. differentiated (CD24+) populations sorted from breast cancer cell lines: MDA‐MB‐231 (H) and HS578T (I) (see Appendix Figs S2 and S3).

Heat map of tsRNA reads mapped to parent tRNA residues across various samples sequenced in this study. Majority of the tsRNAs identified were processed specifically from the 5′‐half of the mature tRNA.

Data information: *P value < 0.05, **P value < 0.01; ***P value < 0.001; n.s, non‐significant.
Figure 1
Figure 1. Small RNA profiling of RA‐induced differentiating mESCs

Distribution of 25–35 mer reads mapping to the mouse genome.

Proportion of 30–35 nt genomic reads matching mouse tRNA sequences. The sequencing was done in duplicates. Error bars represent SEM, and significance was calculated using unpaired t‐test.

Pie charts depicting the different small RNA populations sequenced in the three stem cell states.

Per base sequence coverage of 30–35 nt reads across parental tRNA sequence. Anticodon positions are demarcated by discontinuous lines (blue = LIF‐treated, gray = Wnt3 + LIF‐treated, red = RA‐treated mESCs).

Per base coverage of 30–35 nt reads across the 468 individual parental tRNAs plotted as heat maps.

Positional heat map of reads mapped to tRNAGlnCTG, tRNAGluTTC, tRNAValAAC, and tRNAGlyGCC residues. The yellow box indicates the anticodon region.

Figure 2
Figure 2. Characterization of 5‐tsRNAs during mESC differentiation

Pie chart depicting the various species of 5′‐tsRNAs expressed in the three different states of stem cells tested.

Northern blot of candidate tRNAs and tsRNAs tsGlnCTG, tsGlyGCC, tsGluTTC, tsValCAC, and tsLysTTT. The black arrows represent the tRNAs, and the red arrows mark the tsRNAs.

Northern blot of 5′‐tsRNAs: tsGlyGCC, tsGlnCTG, tsGluTTC, tsLysTTT, and tsValCAC at various time points of RA‐induced mESCs differentiation showing dynamic expression of these 5′‐tsRNAs.

Relative quantification of the tsRNA bands from northern hybridization tested across two biological replicates. The solid lines represent the expression of tsRNAs, and the dashed lines represent tRNAs.

Figure EV2
Figure EV2. Northern blot and qPCR validation of tsRNA expression in LIF and RA treated mESCs

qPCR validation of tsRNAs between LIF and RA at 48 h of differentiation (n = 2, error bars represent SD, and significance was calculated using Student's t‐test).

Gel showing the tsRNA band amplified in the qPCRs.

Replicate of northern blots of candidate tRNA and 5′‐tsRNAs across time points of RA differentiation. Black arrows indicate tRNA species, and the red arrows indicate tsRNA species.

Northern blot of parent tRNA across various time points of RA‐induced differentiation.

Figure EV3
Figure EV3. Functional interrogation of tsRNAs in the modulation of stem and differentiated states in mESCs

qPCR quantitation of tsGlnCTG, transfected in LIF conditions (n = 3, error bar represent SEM, and significance was calculated using unpaired t‐test).

Alkaline phosphatase assays of tsRNA mimic overexpression in LIF condition (n = 3, error bar represent SD, and significance was calculated using unpaired t‐test).

Relative expression of stemness marker in tsRNA transfected in LIF conditions (n = 2, error bars represent SEM, and significance was calculated using unpaired t‐test).

Volcano plot showing transcripts upregulated (in green) and downregulated (in pink) in LIF samples transfected with tsRNA mimics compared with mock controls.

Northern blot of tsGlnCTG in RA‐ vs. ASO‐transfected RA cells. F Components of various stemness pathways upregulated (in green) in ASO‐treated RA conditions.

Data information: *P value < 0.05, ***P value < 0.001, ****P value < 0.0001.
Figure 3
Figure 3. Effect of 5′‐tsRNA knockdown on RA‐induced differentiation of mESCs

Effect of ASO‐mediated inhibition of 5′‐tsRNAs on alkaline phosphatase activity in RA‐treated mESCs. (n = 3), error bars represent SEM, and significance was calculated using unpaired t‐test.

Relative expression of stemness markers in RA‐induced differentiation mESCs blocked for 5′‐tsRNA function with ASOs. (n = 2), error bars represent SEM, and significance was calculated using unpaired t‐test.

Volcano plot of the transcriptome depicting the overexpressed transcripts (in green) and downregulated transcripts (in pink) in ASO‐treated differentiating cells.

KEGG pathway analysis of the transcripts upregulated in a RA‐mediated differentiating mESCs blocked for tsRNA function.

qPCR validation of transcripts involved in the stem cell signaling pathway that are upregulated upon 5′‐tsRNA knockdowns using ASOs. (n = 2), error bars represent SEM, and significance was calculated using unpaired t‐test.

Data information: *P value < 0.05; **P value < 0.01; ***P value < 0.001; ****P value < 0.0001; n.s., not significant.
Figure 4
Figure 4. Identification and functional characterization of tsRNA interactome in mESCs cultured with LIF or RA

Venn diagram depicting the common interacting proteins between different tsRNAs in RA‐treated cells.

Scatter plot showing Log2 fold‐change enrichment of protein interactome between LIF‐ and RA‐treated mESCs. All peptides identified by LC‐MS/MS had < 1% FDR (Fig EV4E). LC‐MS/MS was conducted twice (R 2 = 0.85 between LIF duplicates and R 2 = 0.82 between RA‐treated mESC duplicates).

Scatter plot showing Log2 fold‐change enrichment of mRNAs associated with tsGlnCTG over Scramble (Scr1) in LIF‐ vs. RA‐treated mESCs.

Bar plots and representative gene lists depicting association of tsGlnCTG with “pluripotency‐associated” (D) and “differentiation‐responsive” (D′) genes in LIF‐ or RA‐treated mESCs.

Figure EV4
Figure EV4. Characterization of mechanism and molecular function of candidate tsRNAs in pluripotent vs. differentiating mESCs

GO annotation of 109 common tsRNA‐interacting proteins.

Effects of candidate 5′‐tsRNAs: tsGlnCTG, tsGlyGCC, tsMetCAT, and tsIleGAT on uncapped luciferase mRNA in an in vitro translation system (n = 2–3, error bars represent SEM, and significance was calculated using unpaired t‐test).

Effects of tsGlnCTG on capped GFP mRNA in an in vitro translation system (n = 2–3, error bars represent SD, and significance was calculated using unpaired t‐test).

Polysome profiles of different stem vs. differentiating states tested in this study. Gray areas indicate the fractions collected for sequencing.

tsRNA distribution among the polysome fractions from indicated cell states.

Workflow to identify proteomic and transcriptomic interactions of tsGlnCTG in pluripotency vs. differentiating mESCs.

False discovery rate (FDR) of peptides detected in LC‐MS/MS of biotinylated tsGlnCTG pulldown.

Western blot showing differential interaction of Igf2bp1, Ybx1, and Rpl10 with tsGlnCTG between LIF and RA conditions.

Overlap between transcripts that are associated with tsRNAs in RA condition and the transcripts that are affected upon ASO transfection in RA conditions.

Schematic of the experimental setup to identify protein‐mediated and complementarity‐based tsRNA target interaction.

qPCR analysis showing the enrichment of transcripts pulled down by tsGlnCTG in lysates devoid of proteins compared with normal lysate (n = 2, error bars represent SEM, and significance was calculated using unpaired t‐test).

Top 10 tsGlnCTG‐associated proteins in indicated cell states.

Data information: *P value < 0.05, **P value < 0.01; ***P value < 0.001; n.s, non‐significant.
Figure EV5
Figure EV5. Functional characterization of tsGlnCTG in the regulation of IGF2BP1 and c‐Myc

Merged triplicate Western immunoblots of Igf2bp1 and β‐actin in mESCs grown under LIF or RA conditions (top) and densitometric quantitation of the same (below) (n = 3, error bars represent SD, and significance was calculated using Student's t‐test).

Validation of siRNA‐mediated knockdowns of Igf2bp1 in RA‐treated mESCs (n = 3, error bars represent SD, and significance was calculated using Student's t‐test).

Expression of pluripotency genes in Igf2bp1 knockdown cells (n = 3, error bars represent SD, and significance was calculated using Student's t‐test).

Small RNAs identified (18–35 nt) from input and Igf2bp1 pulldown in RA‐treated mESCs.

qPCR validation of the tsGlnCTG enrichment upon Igf2bp1 pulldown in RA‐treated mESCs.

Binding curve showing the interaction of tsGlnCTG and Igf2bp1 (Kd = 33 nM). The intensities of the Igf2bp1 band were normalized to band intensity of the total Igf2bp1 protein used for pulldown.

Biological duplicate for c‐Myc quantitation from Igf2bp1 pulldown between LIF vs. RA conditions (n = 2, for duplicate check Fig 5B).

Data information: *P value < 0.05, **P value < 0.01; ***P value < 0.001; n.s, non‐significant.
Figure 5
Figure 5. 5′‐tsRNA‐mediated regulation of c‐Myc through Igf2bp1 interaction.

In vitro binding analysis of tsGlnCTG to Igf2bp1 in the presence and absence of antisense oligo (ASO) against tsGlnCTG. ASOs effectively disrupt the binding of tsGlnCTG to Igf2bp1. (n = 2) error bars represent SEM.

Quantification of the Igf2bp1‐bound c‐Myc mRNA between LIF‐ vs. RA‐treated mESCs (n = 2; see Fig EV5F for duplicate data).

Association of c‐Myc transcript in different translating pools in RA‐treated mESCs as compared to LIF condition. (n = 2), error bars represent SD, and significance was calculated by one‐tailed unpaired t‐test.

Relative enrichment of c‐Myc mRNA in translating (80S and polysome) and non‐translating (mRNPs) pools (fractionated from polysome profiling) between ASO‐treated and mock‐treated RA‐induced differentiating mESCs. (n = 2), error bars represent SD, and significance was calculated using one‐tailed unpaired t‐test.

Relative levels of c‐Myc mRNA in ASO‐treated and mock‐treated RA‐induced differentiating mESCs compared with LIF‐treated mESCs. (n = 3), error bars represent SD, and significance was calculated using one‐tailed unpaired t‐test.

Epistatic analysis of 5′‐tsRNAs and IGF2BP1 in regulating Myc transcriptional reporter activity (n = 3), error bars represent SD, and significance was calculated using unpaired t‐test.

Schematic representing tsRNA based c‐Myc transcript regulation.

Data information: *P value < 0.05, **P value < 0.01; ***P value < 0.001.
Figure 6
Figure 6. Proposed mechanistic model for tsRNA cellular and molecular functions
The dynamic expression of 5′‐tsRNAs plays a crucial role in modulating stem cell differentiation. During differentiation, 5′‐tsRNAs regulate the translation and/or the stability of several transcripts through its interaction with ribosomes, RNA‐binding proteins (ribonucleoproteins), such as IGF2BP1, or through direct sequence complementarity.

Similar articles

See all similar articles

Cited by 2 PubMed Central articles

References

    1. Ivey KN, Srivastava D (2010) MicroRNAs as regulators of differentiation and cell fate decisions. Cell Stem Cell 7: 36–41 - PubMed
    1. Kumar P, Anaya J, Mudunuri SB, Dutta A (2014) Meta‐analysis of tRNA derived RNA fragments reveals that they are evolutionarily conserved and associate with AGO proteins to recognize specific RNA targets. BMC Biol 12: 78 - PMC - PubMed
    1. Keam S, Hutvagner G (2015) tRNA‐derived fragments (tRFs): emerging new roles for an ancient RNA in the regulation of gene expression. Life 5: 1638–1651 - PMC - PubMed
    1. Cole C, Sobala A, Lu C, Thatcher SR, Bowman A, Brown JWS, Green PJ, Barton GJ, Hutvagner G (2009) Filtering of deep sequencing data reveals the existence of abundant Dicer‐dependent small RNAs derived from tRNAs. RNA 15: 2147–2160 - PMC - PubMed
    1. Fu H, Feng J, Liu Q, Sun F, Tie Y, Zhu J, Xing R, Sun Z, Zheng X (2009) Stress induces tRNA cleavage by angiogenin in mammalian cells. FEBS Lett 583: 437–442 - PubMed

Publication types

Associated data

LinkOut - more resources

Feedback