Decryption of sequence, structure, and functional features of SINE repeat elements in SINEUP non-coding RNA-mediated post-transcriptional gene regulation

Abstract

RNA structure folding largely influences RNA regulation by providing flexibility and functional diversity. In silico and in vitro analyses are limited in their ability to capture the intricate relationships between dynamic RNA structure and RNA functional diversity present in the cell. Here, we investigate sequence, structure and functional features of mouse and human SINE-transcribed retrotransposons embedded in SINEUPs long non-coding RNAs, which positively regulate target gene expression post-transcriptionally. In-cell secondary structure probing reveals that functional SINEs-derived RNAs contain conserved short structure motifs essential for SINEUP-induced translation enhancement. We show that SINE RNA structure dynamically changes between the nucleus and cytoplasm and is associated with compartment-specific binding to RBP and related functions. Moreover, RNA-RNA interaction analysis shows that the SINE-derived RNAs interact directly with ribosomal RNAs, suggesting a mechanism of translation regulation. We further predict the architecture of 18 SINE RNAs in three dimensions guided by experimental secondary structure data. Overall, we demonstrate that the conservation of short key features involved in interactions with RBPs and ribosomal RNA drives the convergent function of evolutionarily distant SINE-transcribed RNAs.

Fig. 1. SINEB2 RNAs from different mouse antisense lncRNAs can act as SINEUP effector domain and upregulate GFP protein expression in HEK293T/17.

a Schematic representation of the sense/antisense EGFP system used to validate the effect of different SINEB2 RNAs on translation. BD, binding domain; ED, effector domain. b SINEUP construct design to verify independent and combined effects of two SINEB2 elements from the same antisense lncRNA. c Western blot-based EGFP band intensities normalized to ACTB expression as a measurement of SINEUP-induced protein fold change observed in the HEK293T/17 cells. Empty vector, Direct AS Uchl1 SINEB2, and ΔAS Uchl1 SINEB2 (SINEB2 deletion) were used as controls. N = 3 biologically independent experiments. d Western blot images with anti-GFP and anti-ACTB antibodies. AS, antisense; Direct, direct orientation of SINEB2; Δ, deletion; a and b, two different SINEB2 elements from the same antisense lncRNA; ab and ba, the two SINEB2 elements in original and reversed order, respectively; AS Uchl1 SINEB2–RS, antisense Uchl1 SINEB2 construct without any restriction enzyme sites around the SINEB2 elements. e GFP protein fold changes in HEK293T cells after co-transfection with sense EGFP and miniSINEUP. Western blot images and corresponding GFP band intensities normalized to ACTB expression level. Empty vector = negative control, and antisense Uchl1 SINEB2 containing miniSINEUP (AS Uchl1) = positive control. Sample labels indicate names of antisense lncRNAs from which the SINEB2 sequences were isolated, which represent four sub-families (based on RepeatMasker annotation): B3, B3A, Mm2, and Mm1t. SINEB2 elements are from the B3 subfamily unless specifically stated. AS Uchl1 + AS Uxt-b = combination of AS Uchl1 SINEB2 and the second SINEB2 of AS Uxt; 3× indicates three repeats of SINEB2; and SINEB2/B3 consensus is the B3 subfamily consensus sequence taken from the RepBase database. N = 5 biologically independent experiments. In (c) and (e), the data are presented as mean values ± SEM, data points for biological replicates are depicted by blue dots and significant P values calculated by two-tailed Student’s t-test in comparison to empty vector are noted on the charts. Source data are provided as a Source Data file.

Fig. 2. In-cell SINE structures and conserved motifs in functionally active SINEUPs.

a Antisense (AS) Uchl1 SINEB2 RNA secondary structure based on icSHAPE data (right) is compared with the one previously identified using chemical footprinting. In icSHAPE-based structure, squares highlight the common sequences and structure motifs in in-cell structures of SINE, with their corresponding motif number. AGG (blue) and UGG (red) motifs are shown. Location of SINEUP motifs in icSHAPE-derived structures of functionally active SINE RNAs from (b) antisense Txnip SINEB2, (c) antisense Uxt SINEB2-b and (d) human SINE element FRAM. e icSHAPE-derived RNA secondary structures and local SINEUP motifs (colors as in Fig. 2a) of functionally dormant wild-type antisense Gadd45α SINEB2-a and its deletion mutant antisense Gadd45α SINEB2-a [Δ27–36] with gained SL1. f Effect of antisense Gadd45α SINEB2-a deletion mutants on GFP protein expression as measured by Western blot. An empty vector was used as the negative control. Four mutants with random deletion, as listed in the Figure, were compared with the full-length construct (Full). N = 3 biologically independent experiments, data are mean values ± standard deviation, P value by two-tailed Student’s t-test. AS antisense, Δ deletion, SL1 stem-loop 1. Source data for (f) are provided as a Source Data file.

Fig. 3. SINEUP structure dynamics and interaction with HNRNPK in the nucleus and cytoplasm.

a Normalized icSHAPE reactivity score profiles of mouse antisense (AS) Uchl1 SINEB2 RNA in the nucleus and cytoplasm and b related 2D-structure models created from icSHAPE data. SINEUP motifs from M1 to M7 are indicated by dashed squares. c Normalized icSHAPE reactivity score profiles of human SINE RNA FRAM in the nucleus and cytoplasm and d their respective 2D-structure models. Different structural regions between the compartments are shaded in blue boxes. e seCLIP data for HNRNPK binding to SINEUP-GFP RNA (of AS Uchl1 SINEB2) in the nuclear and cytoplasmic fractions of HEK293T cells. IP, immunoprecipitation against HNRNPK; input, not immunoprecipitated, contains all RNA–protein complexes in the given fraction. Normalized peak regions representing significant enrichment of HNRNPK binding site in IP compared to input are indicated as horizontal bars. Identified HNRNPK binding sites are marked by a red line on 2D-structure models of AS Uchl1 SINEB2 in (b).

Fig. 4. Mouse SINEB2 RNA interacts with human rRNA in cell.

Antisense (AS) Uchl1 SINEB2 interacting regions (Int) in human a 18S rRNA (Int1) and b 28S rRNA (Int2 and 3) identified by PARIS2. Different interaction sites are color-coded, and rRNA-SINEB2 pairs are shown in the corresponding color. Possible base-pairing based on sequence complementarity of interacting regions is shown. c SINEB2 regions interacting with 18S and 28S rRNA regions marked in (a) and (b) on icSHAPE data-driven SINEB2 structure from whole-cell lysate. SINEUP motifs from M1 to M7 are indicated by dashed squares.

Fig. 5. In-cell interaction of functional human SINE element FRAM with rRNA.

a Regions of 18S rRNA found to pair with the SINEUP FRAM region in PARIS2 data. b Corresponding interacting regions (Int) to (a) are highlighted on FRAM RNA. c Description of FRAM-18S rRNA pairs and interaction sites marked in (a) and (b). d 28S rRNA regions identified as FRAM interaction sites by PARIS2 and e their corresponding binding regions on FRAM. f Details of FRAM-28S rRNA interactions highlighted in (d) and (e). Different interaction sites (Int) are represented by different colors. Possible base-pairing based on sequence complementarity of RNA–RNA pairs is shown in (c) and (f). In (b) and (e) the icSHAPE-guided FRAM structure from whole-cell lysate is shown. Matching regions to SINEUP motifs from M1 to M7 are indicated by dashed squares.

Fig. 6. Machine-translated 3D RNA structure models based on in vivo SINE RNA 2D structures.

3D structure models for a–d mouse antisense (AS) Uchl1 SINEB2 RNA, in the nucleus (a, b) and cytoplasm (c, d) and for FRAM RNA, in the nucleus (e, f) and cytoplasm (g, h). SINEUP structure motifs M1–M7 are shown. Red frames show enlarged views of the SL1 region. b, d, f, h show the molecular surface view of the 3D models.