RESA identifies mRNA-regulatory sequences at high resolution

Abstract

Gene expression is extensively regulated at the levels of mRNA stability, localization and translation. However, decoding functional RNA-regulatory features remains a limitation to understanding post-transcriptional regulation in vivo. Here, we developed RNA-element selection assay (RESA), a method that selects RNA elements on the basis of their activity in vivo and uses high-throughput sequencing to provide a quantitative measurement of their regulatory functions at near-nucleotide resolution. We implemented RESA to identify sequence elements modulating mRNA stability during zebrafish embryogenesis. RESA provides a sensitive and quantitative measure of microRNA activity in vivo and also identifies novel regulatory sequences. To uncover specific sequence requirements within regulatory elements, we developed a bisulfite-mediated nucleotide-conversion strategy for large-scale mutational analysis (RESA–bisulfite). Finally, we used the versatile RESA platform to map candidate protein–RNA interactions in vivo (RESA–CLIP).

Access options

Rent or Buy article

Get time limited or full article access on ReadCube.

from$8.99

All prices are NET prices.

Figure 1: RESA is a high-throughput method to systematically map mRNA-regulatory sequences in vivo.
Figure 2: RESA identifies sequences that promote mRNA destabilization.
Figure 3: RESA-B implements high-throughput mutational analysis of regulatory sequences.
Figure 4: The RESA framework can be adapted to map RNA–protein interactions and other RNA biology in vivo.

Accession codes

Primary accessions

Sequence Read Archive

References

  1. 1

    Reinhart, B.J. et al. The 21-nucleotide let-7 RNA regulates developmental timing in Caenorhabditis elegans. Nature 403, 901–906 (2000).

    CAS  Article  Google Scholar 

  2. 2

    Meola, N., Gennarino, V.A. & Banfi, S. microRNAs and genetic diseases. PathoGenetics 2, 7 (2009).

    Article  Google Scholar 

  3. 3

    Giraldez, A.J. et al. MicroRNAs regulate brain morphogenesis in zebrafish. Science 308, 833–838 (2005).

    CAS  Article  Google Scholar 

  4. 4

    Lai, E.C. & Posakony, J.W. The Bearded box, a novel 3′ UTR sequence motif, mediates negative post-transcriptional regulation of Bearded and Enhancer of split Complex gene expression. Development 124, 4847–4856 (1997).

    CAS  PubMed  Google Scholar 

  5. 5

    Treisman, R. Transient accumulation of c-fos RNA following serum stimulation requires a conserved 5′ element and c-fos 3′ sequences. Cell 42, 889–902 (1985).

    CAS  Article  Google Scholar 

  6. 6

    Piqué, M., López, J.M., Foissac, S., Guigó, R. & Méndez, R. A combinatorial code for CPE-mediated translational control. Cell 132, 434–448 (2008).

    Article  Google Scholar 

  7. 7

    Mowry, K.L. & Melton, D.A. Vegetal messenger RNA localization directed by a 340-nt RNA sequence element in Xenopus oocytes. Science 255, 991–994 (1992).

    CAS  Article  Google Scholar 

  8. 8

    Zubiaga, A.M., Belasco, J.G. & Greenberg, M.E. The nonamer UUAUUUAUU is the key AU-rich sequence motif that mediates mRNA degradation. Mol. Cell. Biol. 15, 2219–2230 (1995).

    CAS  Article  Google Scholar 

  9. 9

    Giraldez, A.J. et al. Zebrafish MiR-430 promotes deadenylation and clearance of maternal mRNAs. Science 312, 75–79 (2006).

    CAS  Article  Google Scholar 

  10. 10

    Lim, L.P. et al. Microarray analysis shows that some microRNAs downregulate large numbers of target mRNAs. Nature 433, 769–773 (2005).

    CAS  Article  Google Scholar 

  11. 11

    Nielsen, C.B. et al. Determinants of targeting by endogenous and exogenous microRNAs and siRNAs. RNA 13, 1894–1910 (2007).

    CAS  Article  Google Scholar 

  12. 12

    Lai, E.C. Micro RNAs are complementary to 3′ UTR sequence motifs that mediate negative post-transcriptional regulation. Nat. Genet. 30, 363–364 (2002).

    CAS  Article  Google Scholar 

  13. 13

    Grimson, A. et al. MicroRNA targeting specificity in mammals: determinants beyond seed pairing. Mol. Cell 27, 91–105 (2007).

    CAS  Article  Google Scholar 

  14. 14

    Bartel, D.P. MicroRNAs: target recognition and regulatory functions. Cell 136, 215–233 (2009).

    CAS  Article  Google Scholar 

  15. 15

    Chi, S.W., Zang, J.B., Mele, A. & Darnell, R.B. Argonaute HITS-CLIP decodes microRNA–mRNA interaction maps. Nature 460, 479–486 (2009).

    CAS  Article  Google Scholar 

  16. 16

    Hafner, M. et al. Transcriptome-wide identification of RNA-binding protein and microRNA target sites by PAR-CLIP. Cell 141, 129–141 (2010).

    CAS  Article  Google Scholar 

  17. 17

    Geisberg, J.V., Moqtaderi, Z., Fan, X., Ozsolak, F. & Struhl, K. Global analysis of mRNA isoform half-lives reveals stabilizing and destabilizing elements in yeast. Cell 156, 812–824 (2014).

    CAS  Article  Google Scholar 

  18. 18

    Oikonomou, P., Goodarzi, H. & Tavazoie, S. Systematic identification of regulatory elements in conserved 3′ UTRs of human transcripts. Cell Rep. 7, 281–292 (2014).

    CAS  Article  Google Scholar 

  19. 19

    Rosenberg, A.B., Patwardhan, R.P., Shendure, J. & Seelig, G. Learning the sequence determinants of alternative splicing from millions of random sequences. Cell 163, 698–711 (2015).

    CAS  Article  Google Scholar 

  20. 20

    Weingarten-Gabbay, S. et al. Comparative genetics: systematic discovery of cap-independent translation sequences in human and viral genomes. Science. 351, aad4939 (2016).

    Article  Google Scholar 

  21. 21

    Zhao, W. et al. Massively parallel functional annotation of 3′ untranslated regions. Nat. Biotechnol. 32, 387–391 (2014).

    CAS  Article  Google Scholar 

  22. 22

    Wissink, E.M., Fogarty, E.A. & Grimson, A. High-throughput discovery of post-transcriptional cis-regulatory elements. BMC Genomics 17, 177 (2016).

    Article  Google Scholar 

  23. 23

    Wolter, J.M., Kotagama, K., Babb, C.S. & Mangone, M. Detection of miRNA targets in high-throughput using the 3′LIFE assay. J. Vis. Exp. e52647 (2015).

  24. 24

    Yartseva, V. & Giraldez, A.J. The maternal-to-zygotic transition during vertebrate development: a model for reprogramming. Curr. Top. Dev. Biol. 113, 191–232 (2015).

    CAS  Article  Google Scholar 

  25. 25

    Lee, M.T. et al. Nanog, Pou5f1 and SoxB1 activate zygotic gene expression during the maternal-to-zygotic transition. Nature 503, 360–364 (2013).

    CAS  Article  Google Scholar 

  26. 26

    Bazzini, A.A., Lee, M.T. & Giraldez, A.J. Ribosome profiling shows that miR-430 reduces translation before causing mRNA decay in zebrafish. Science 336, 233–237 (2012).

    CAS  Article  Google Scholar 

  27. 27

    Staton, A.A., Knaut, H. & Giraldez, A.J. Reply to: “On the robustness of germ cell migration and microRNA-mediated regulation of chemokine signaling”. Nat. Genet. 45, 1266–1267 (2013).

    CAS  Article  Google Scholar 

  28. 28

    Brennecke, J., Stark, A., Russell, R.B. & Cohen, S.M. Principles of microRNA-target recognition. PLoS Biol. 3, e85 (2005).

    Article  Google Scholar 

  29. 29

    Vella, M.C., Choi, E.Y., Lin, S.Y., Reinert, K. & Slack, F.J. The C. elegans microRNA let-7 binds to imperfect let-7 complementary sites from the lin-41 3′UTR. Genes Dev. 18, 132–137 (2004).

    CAS  Article  Google Scholar 

  30. 30

    Moreno-Mateos, M.A. et al. CRISPRscan: designing highly efficient sgRNAs for CRISPR-Cas9 targeting in vivo. Nat. Methods 12, 982–988 (2015).

    CAS  Article  Google Scholar 

  31. 31

    Tuerk, C. & Gold, L. Systematic evolution of ligands by exponential enrichment: RNA ligands to bacteriophage T4 DNA polymerase. Science 249, 505–510 (1990).

    CAS  Article  Google Scholar 

  32. 32

    Ellington, A.D. & Szostak, J.W. In vitro selection of RNA molecules that bind specific ligands. Nature 346, 818–822 (1990).

    CAS  Article  Google Scholar 

  33. 33

    Blackwell, T.K. & Weintraub, H. Differences and similarities in DNA-binding preferences of MyoD and E2A protein complexes revealed by binding site selection. Science 250, 1104–1110 (1990).

    CAS  Article  Google Scholar 

  34. 34

    Martin, K.C. & Ephrussi, A. mRNA localization: gene expression in the spatial dimension. Cell 136, 719–730 (2009).

    CAS  Article  Google Scholar 

  35. 35

    Rabani, M., Kertesz, M. & Segal, E. Computational prediction of RNA structural motifs involved in post-transcriptional regulatory processes. Methods Mol. Biol. 714, 467–479 (2011).

    CAS  Article  Google Scholar 

  36. 36

    Bazzini, A.A. et al. Codon identity regulates mRNA stability and translation efficiency during the maternal-to-zygotic transition. EMBO J. 35, 2087–2103 (2016).

    CAS  Article  Google Scholar 

  37. 37

    Mishima, Y. & Tomari, Y. Codon usage and 3′ UTR length determine maternal mRNA stability in zebrafish. Mol. Cell 61, 874–885 (2016).

    CAS  Article  Google Scholar 

  38. 38

    Presnyak, V. et al. Codon optimality is a major determinant of mRNA stability. Cell 160, 1111–1124 (2015).

    CAS  Article  Google Scholar 

  39. 39

    Davis, C.A., Monnier, J.M. & Nick, H.S. A coding region determinant of instability regulates levels of manganese superoxide dismutase mRNA. J. Biol. Chem. 276, 37317–37326 (2001).

    CAS  Article  Google Scholar 

  40. 40

    Wellington, C.L., Greenberg, M.E. & Belasco, J.G. The destabilizing elements in the coding region of c-fos mRNA are recognized as RNA. Mol. Cell. Biol. 13, 5034–5042 (1993).

    CAS  Article  Google Scholar 

  41. 41

    Rozen, S. & Skaletsky, H. Primer3 on the WWW for general users and for biologist programmers. Methods Mol. Biol. 132, 365–386 (2000).

    CAS  Google Scholar 

  42. 42

    Jiang, H., Lei, R., Ding, S.W. & Zhu, S. Skewer: a fast and accurate adapter trimmer for next-generation sequencing paired-end reads. BMC Bioinformatics 15, 182 (2014).

    Article  Google Scholar 

  43. 43

    Trapnell, C., Pachter, L. & Salzberg, S.L. TopHat: discovering splice junctions with RNA-Seq. Bioinformatics 25, 1105–1111 (2009).

    CAS  Article  Google Scholar 

  44. 44

    Li, H. et al. The Sequence Alignment/Map format and SAMtools. Bioinformatics 25, 2078–2079 (2009).

    Article  Google Scholar 

  45. 45

    Bailey, T.L. et al. MEME SUITE: tools for motif discovery and searching. Nucleic Acids Res. 37, W202–W208 (2009).

    CAS  Article  Google Scholar 

  46. 46

    Thisse, C. & Thisse, B. High-resolution in situ hybridization to whole-mount zebrafish embryos. Nat. Protoc. 3, 59–69 (2008).

    CAS  Article  Google Scholar 

Download references

Acknowledgements

We thank K. Bilguvar, S. Mane, C. Castaldi and I. Tikhonova for sequencing support; H. Codore for technical assistance; A. Bazzini, P. Oikonomou and S. Tavazoie for discussions; and all the members of the Giraldez laboratory for intellectual and technical support. This research was supported by the US National Institutes of Health (R01 HD074078, GM103789, GM102251, GM101108 and GM081602), the Pew Scholars Program in the Biomedical Sciences, the March of Dimes (1-FY12-230), the Yale Scholars Program, the HHMI Faculty Scholars Program, Whitman fellowship funds provided by E.E. Just, Lucy B. Lemann, and Evelyn and Melvin Spiegel, and the H. Keffer Hartline and Edward F. MacNichol, Jr. Fellowship Fund of the Marine Biological Laboratory (Woods Hole, Massachusetts, USA) to A.J.G.; the Swiss National Science Foundation (grant P2GEP3_148600) to C.E.V.; NIH Fellowship F32 HD061194 to C.M.T.; NIH Fellowship F32 HD071697 and start-up funds from the University of Pittsburgh to M.T.L.; and NIH Training Grants T32 GM007223 and T32 HD007149, an Edward L. Tatum Fellowship (Yale University) and the Yale MRSP to V.Y.

Author information

Affiliations

Authors

Contributions

V.Y., C.M.T. and A.J.G. designed and conceived the project. V.Y. generated the RESA UTR, RESA-B and RESA–CLIP libraries, and performed the validation experiments. V.Y. and M.T.L. developed the RESA and RESA-B analysis. M.T.L. and C.E.V. performed the computational analyses. All authors interpreted and analyzed the data. V.Y., M.T.L. and A.J.G. wrote the manuscript with input from the other authors.

Corresponding authors

Correspondence to Miler T Lee or Antonio J Giraldez.

Ethics declarations

Competing interests

The authors declare no competing financial interests.

Integrated supplementary information

Supplementary Figure 1 RESA libraries are complex and reproducible.

(a) Adaptor ligation to DNA fragments results in incorporation of both strands into the mRNA reporter library. Reverse complement variants can serve as internal controls in RESA experiments. (b) Proportion of unique inserts in each sample (defined as having the same aligned start and end position), segregated by read orientation. R1 = biological replicate 1, etc. (c) Distribution of sequence fragment length in library input sample. (d) Distribution of sequence fragment length recovered after injection into zebrafish. (e) Histogram of median sequencing coverage for successfully amplified 3’UTRs (N = 434). 97% of assayed 3’UTRs have coverage ≥ 20X and 87% have coverage ≥ 100X. (f) Histogram of nucleotides proportion overlapped by unique read start sites across the UTRs (N = 434). For 83% of 3’UTRs, greater than 50% of positions are unique read start sites. (g) Randomly selected 2000 reads derived from early stage (yellow) and late stage (blue) samples overlapping miR-430 target locus txnrd1. Below, positional coverage for each condition separated by replicate. miR-430 seed sites are marked with dashed lines. (h) Sequencing read coverage over non-overlapping 50-nt windows across the 3’UTRs comparing early stage replicates and (i) late stage replicates. Pearson’s correlation between early and late stage replicates is R = 0.97 and R=0.98, respectively.

Supplementary Figure 2 miR-430-dependent regulation demonstrates the efficacy of the RESA platform.

(a) miR-430 target seed definitions where Watson-Crick pairing occurs between positions 2-8 of the mature miR-430 sequence and complementary target sequence. An A in the target site opposite position 1 is also favorable. (b) Boxplot showing RESA coverage fold change late vs early stage centered on miR-430 target 8-mers (AGCACUUA), 7-mers (AGCACUU or GCACUUA), 6-mers (GCACUU), offset 6-mers (AGCACU), sequences encoding the non target seed CCACUU, and reverse complement coverage over the 7-mers and 8-mers. Depletion at late stage for the miRNA targets is significant (comparing all targets to non targets, p = 3.8E-39, Kruskal-Wallis Test; comparing sense 7-mers and 8-mers to their reverse complement, p=7.1E-40, Wilcoxon Signed-Rank Test). (c) Similar to (b), comparing RESA coverage fold change wild type late stage vs 430LNA treated, which lacks miR-430 activity. Depletion at wild type late stage for the miRNA targets is significant (comparing all targets to non targets, p = 3.8E-44, Kruskal-Wallis Test; comparing sense 7-mers and 8-mers to their reverse complement, p=9.4E-44, Wilcoxon Signed-Rank Test). (d) Meta profiles of coverage ratios (wild type late stage vs 430LNA) centered on all the miR-430 target seed in the study.

Supplementary Figure 3 FDR for regulated-region identification.

(a) Full depth sequencing profile for the acat1 locus with miR-430 target seed. Below profiles generated using randomly sampled reads from early and late stage embryos at varying depth (500 rounds per target depth). False-positive valley is highlighted in 4X coverage condition. (b) Distribution of false-positive valley magnitude in 500 simulations of varied read coverage for the acat1 locus. (c) Determination of a magnitude threshold for identifying regulated regions to reduce false discovery. For each simulation in (b), the nth percentile of the distribution of false positive magnitudes is plotted, for n = 0.05 and 0.1; this implies that 95% or 90% of false positive valleys have shallower magnitude (less negative) than this value, respectively. Thus, at 100X coverage, a threshold of -1.33-fold (-0.41 on a log2 scale) falls between a false-discovery rate (FDR) of 0.05 and 0.1. This parameter becomes the maximum (least negative) log2 coverage ratio allowed to confidently define a regulated region.

Supplementary Figure 4 Regulatory-element width measurement with the RESA platform.

(a) Histogram of RESA destabilized region (valley) for miR-430 targets seeds. 97% of mir-430 target seeds overlap valleys that are ≥ 65 nt wide where width is measured at 25% of valley minimum magnitude. (b) Model of the relationship between regulatory element size and RESA valley widths. For reporter fragments (black bars) of a constant length distribution, longer regulatory elements (blue) will be fully overlapped by fewer fragments than elements with short sequence requirements, resulting in correspondingly narrower valleys. (c) Average coverage depletion over miR-430 target seeds (GCACUU) comparing short reporter inserts (56-80nt, pink), long fragments (105-129nt, green), and all fragments (56-129, blue). (d) The three profiles in (c) separated, showing nucleotide width as measured at 50% or 10% of the valley minimum (maximum depletion point).

Supplementary Figure 5 RESA approach to distinguish mRNA-regulatory mechanisms.

(a-c) Plots showing the dissection of mechanisms regulating mRNA stability. Destabilized regions that depend on miR-430 activity are revealed by comparing wild type late stage RESA coverage to 430LNA coverage, in which miR-430 is inhibited (a), while regions that do not depend on miR-430 activity are revealed by comparing 430LNA coverage to early stage (b). In each comparison, the horizontal line indicates the median coverage ratio of the regions that are not regulated. For a subset (red squares), the region is composed of adjacent miR-430-dependent and miR-430-independent regulatory elements, as they show regulation in both comparisons. (c) An example of a region (tcea2) with adjacent miR-430 dependent and independent mechanisms that are revealed using the 430LNA condition. (d) Plot comparing destabilization magnitudes at late stage when compared against early stage (x axis) or 430LNA (y axis). miR-430 dependent regions (blue) are correlated in these conditions. (e) Plot comparing destabilization magnitudes when comparing wild-type (x axis) or 430LNA (y axis) late stage to early stage. Regions that do not depend on miR-430 (orange) are correlated in these conditions.

Supplementary Figure 6 RESA identifies differentially regulated mRNA regions.

(a) High confidence (see Online Methods) miR-430 dependent (blue) regions, miR-430 independent regions (orange), and regions with adjacent mechanisms (red). Source gene name and relative position within the respective 3’UTR is marked for referenced regions (see Supplementary Data 3). (b) Log2 coverage profiles showing examples of a miR-430 dependent destabilized region (top), a non-regulated region (second), a non-miR-430 dependent destabilized region (third), and adjacent destabilized and stabilized regions (fourth). Source gene and relative position in the UTR amplicon are shown below each. (c) Histogram of regulatory element density in 3’UTRs. (d) (left) Diagram of the predicted destabilization element in the ptgfrn 3’UTR (dashed box) and the CRISPR-mediated deletion (red) in MZptgfrnΔ222Δ42. (right) RESA coverage profile over the destabilized region.

Supplementary Figure 7 RESA identifies regulated regions with high precision.

(a) Coverage ratio plots over the gdpd5a UTR separated by three replicate late stage samples when compared against the pooled early stage each. Responsive regions are marked with errow. (b) RESA destabilized regions from the pooled analysis are also identified when single replicate late-stage samples are used individually. 86% of the regions are found individually by at least 2 separate replicates. (c) Plot comparing log2 coverage ratio using 2 different single replicate late-stage samples, showing high agreement for the RESA destabilized regions from the pooled analysis (green). Additional destabilized regions are predicted in one but not the other replicate (purple); however, these are only weakly predicted. (d) High confidence (see Online Methods) stabilized regions. Wild-type sequence coverage enrichment over time indicates stability-promoting regions, since stable regions accumulate over time with respect to the overall mRNA population. (e) Dot plot of miR-430 dependent destabilized regions containing miR-430 target seeds (within 50 nts of the region minimum, for the late versus early stage comparison ratio). 82% of miR-430 seed-containing regions are located within 10 nts of the region minimum, demonstrating the high predictive power of RESA valley shape. Only regions in the interior of amplicons (i.e., not truncated by edge effects) were considered.

Supplementary Figure 8 Compensatory pairing and mismatch seeds are detected in miR-430-dependent regions.

(a) Proportion of RESA predicted miR-430 dependent destabilized regions that contains a canonical miR-430 target seed match is 70%. Of the remaining regions (non target seed), the majority (30 out of 61) contains a 1-nt mismatch seed, as illustrated below (b) Scheme illustrating 1-nt mismatch and compensatory pairing for miR-430 target region. (c) The “non target seed” regions are significantly enriched for containing mismatch seeds, as well as base-pairing complementarity to the 3’ end of mature miR-430 (d), compared to the non-miR-430 dependent RESA destabilized regions or a set of background, unregulated regions (see Online Methods). (e) Target seed match has a significant effect (P = 3.2e-9, Kruskal-Wallis Test) on destabilization strength, with non-seed match containing regions weaker than offset 6-mer containing regions.

Supplementary Figure 9 RESA–bisulfite mapping strategy.

(a) RESA-Bisulfite used bisulfite treatment to induce C to U or G to A mutations within the reporter library pool. Partially mutated reporters are injected in vivo to determine the effect of mutations (red, purple) on the destabilization of wild type (WT, blue) regions. (b) Mapping partially converted reads to the genome first requires building a reference genome where all C bases are converted to T (top) and one where all G bases are converted to A (bottom). Reads are expected to encode either C to T conversions or G to A conversions (relative to the sense orientation of the read pair), but not both. (c) Because fragments are only partially converted, all reads are subjected to in silico conversion to completion, by changing all C bases to T (top) or separately all G bases to A (bottom). This is to ensure that they align to the converted genomes from (b). These two sets of reads will be treated separately moving forward. Base identity is with respect to the orientation of the fragment, which is defined by read 1 of the read pair; thus, conversions on read 2 are reverse complements due to the design of Illumina paired-end reads (i.e., if read 1 is subjected to C to T conversion, read 2 is subjected to G to A). For a given read pair, it is unknown a priori whether the fragment originally encodes C to T or G to A conversions, so both in silico conversions are always applied, with the expectation that only one will yield an alignable read pair. (d) Mapping is done four ways, with each set of converted reads aligned to each of the two converted genomes; only one of the four pairings should produce a successful alignment per read pair. If a gene is in the sense orientation with respect to the reference genome sequence, then successful alignment will occur sense to sense (1 or 4). If a gene is in the antisense orientation with respect to the genome, successful alignment will occur sense to antisense (2 or 3). (e) Once read pairs are mapped, the reads are reverted to their original, partially converted sequences. This is easily implemented by encoding the original sequence as part of the read ID during the in silico conversion phase.

Supplementary Figure 10 RESA–bisulfite reveals nucleotide positions associated with differential regulation.

(a, b) Comparison of the reference base identities of each nucleotide position to the bases in the aligned reads induces a conversion rate, which follows an approximately normal distribution centered at 0.67-0.69. Only positions with at least 10 overlapping reads were used to calculate this distribution. (c) Histograms of the positions of bases where significant depletion of the wild-type allele was observed, with respect to RESA-predicted destabilized regions. Distance is measured from the destabilized region minimum to the nearest depleted wild-type base, up to 50 nts. (d) Summary of positions with nucleotide biases (depletion of wild-type or converted base) at late stage. Positions with biases observed as linked dibases or tribases are distinguished from positions with independent effects. (e) Pooled base conversion adjacent to miR-430 offset 6-mer target sites enhances destabilization. Y-axis shows the magnitude of change in the proportion of converted bases observed (U, red; A, purple) from early to late stage. X-axis is oriented sense to the target. P value from a G test of independence is shown. (f) miR-430 target seeds with A or U at position 9 (outside of the canonical microRNA seed region) are significantly more destabilized than those with C or G, across target types. 8-mers and 7-mers (pairing positions 2-8) are pooled. (g) GFP reporter analysis comparing the stability of endogenous loci containing offset 6-mer miR-430 sites with either G (pam) or A (itpr3) at position 9, versus mutant variants where the G is mutated to A (pam) or A to G (itpr3). GFP mRNA level is normalized to control dsRed mRNA and assayed by qRT-PCR. In both cases reporters with A in position 9 show significantly stronger destabilization: substituting the endogenous G with an A in the pam sequence increased destabilization 35%, while conversely substituting the endogenous A with a G in itpr3 attenuated destabilization 29% (Welch’s t-test; bars show mean +/- SEM).

Supplementary Figure 11 RESA applications and CLIP method comparison.

(a) Diagram of proposed examples of RESA applications in diverse biological contexts. (b) Overview of RESA-CLIP strategy. Tagged Ago2 is co-injected with the RESA library, injected into zebrafish embryos, crosslinked, immunoprecipitated, and sequenced to identify Ago2-bound regions. (c) Comparison of RESA-CLIP to other methods to interrogate protein-RNA interactions.Supplementary References[47] Licatalosi, D. D. et al. HITS-CLIP yields genome-wide insights into brain alternative RNA processing. Nature 456, 464–470 (2008). [48] Hafner, M. et al. Transcriptome wide identification of RNA binding protein and microRNA target sites by PAR-CLIP. Cell 141, 129–141 (2010). [49] Huppertz, I. et al. iCLIP: Protein-RNA interactions at nucleotide resolution. Methods 65, 274–287 (2014). [50] Sugimoto, Y. et al. hiCLIP reveals the in vivo atlas of mRNA secondary structures recognized by Staufen 1. Nature 519, 491–494 (2015). [51] Flynn, R. A. et al. Dissecting noncoding and pathogen RNA – protein interactomes. RNA 21, 135–143 (2014). [52] Zarnegar, B. J. et al. irCLIP platform for efficient characterization of protein – RNA interactions. Nat. Methods 13, (2016). [53] Van Nostrand, E. L. et al. Robust transcriptome-wide discovery of RNA-binding protein binding sites with enhanced CLIP (eCLIP). Nat. Methods 1–9 (2016).

Supplementary Figure 12 RESA–CLIP shows specific Ago2 association with miR-430-dependent destabilized regions.

(a) (Top) Plot of reads-per-million normalized coverage over the gdpd5a 3’UTR showing RESA-CLIP levels (shaded blue) compared to input RNA levels (orange line). (Bottom) Corresponding plot of log2 ratio CLIP over input. Only the region overlapping the miR-430 target seed has enrichment of CLIP signal. (b) Proportion of CLIP peaks overlapping RESA-predicted miR-430 dependent regions, at minimum CLIP read threshold 100X (top) or 50X (bottom). (c) Proportion of RESA destabilized regions classified as miR-430 dependent (blue) and non miR-430 dependent (orange) overlapped by CLIP peaks, as minimum read threshold is varied. A higher proportion of miR-430 dependent regions are overlapped throughout the threshold range. (d) Motif analysis in Ago2 CLIP peaks shows significant enrichment of the miR-430 target seed sequence. (e) CLIP peaks that do not overlap RESA destabilized regions are enriched in partial sequence matches to mature miR-430, correlated with the strength of the CLIP signal (P = 2.0e-5, Chi-squared test comparing presence/absence of mismatch seeds; P = 2.4e-8, Wilcoxon rank sum test comparing miR-430 base-pairing score, see Online Methods). This suggests that RESA-CLIP is capturing RISC interactions as it samples for matches to the miR-430 sequence. (f) Boxplot showing RESA late versus early stage coverage ratio of miR-430 dependent destabilized regions, grouped according to CLIP sequencing read coverage. Regions with greater CLIP coverage have significantly stronger destabilization (P = 6.1e-8, Wilcoxon rank sum test). (g) The proportion of miR-430 dependent RESA destabilized regions overlapped by CLIP peaks (≥50X coverage) is correlated with the strength of the miR-430–target seed interaction. However, even miR-430 dependent destabilized regions lacking a target seed match are significantly enriched for CLIP peaks compared to non-miR-430 dependent destabilized regions (P = 3.6e-5, Chi-squared test).

Supplementary information

Supplementary Text and Figures

Supplementary Figures 1–12 and Supplementary Tables 1–3 (PDF 2537 kb)

Supplementary Data 1

List of primer pairs used to amplify the RESA library. (TXT 47 kb)

Supplementary Data 2

BED file of the UTR regions covered by the RESA library. (TXT 37 kb)

Supplementary Data 3

Regulatory regions identified by RESA. (TXT 164 kb)

Supplementary Data 4

Positions with significant RESA-bisulfite induced allele biases. (TXT 246 kb)

Supplementary Data 5

Ago2 bound regions identified by RESA-CLIP at ≥ 50X coverage. (TXT 9 kb)

Source data

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Yartseva, V., Takacs, C., Vejnar, C. et al. RESA identifies mRNA-regulatory sequences at high resolution. Nat Methods 14, 201–207 (2017). https://doi.org/10.1038/nmeth.4121

Download citation

Further reading