GC content around splice sites affects splicing through pre-mRNA secondary structures
© Zhang et al; licensee BioMed Central Ltd. 2011
Received: 19 August 2010
Accepted: 31 January 2011
Published: 31 January 2011
Alternative splicing increases protein diversity by generating multiple transcript isoforms from a single gene through different combinations of exons or through different selections of splice sites. It has been reported that RNA secondary structures are involved in alternative splicing. Here we perform a genomic study of RNA secondary structures around splice sites in humans (Homo sapiens), mice (Mus musculus), fruit flies (Drosophila melanogaster), and nematodes (Caenorhabditis elegans) to further investigate this phenomenon.
We observe that GC content around splice sites is closely associated with the splice site usage in multiple species. RNA secondary structure is the possible explanation, because the structural stability difference among alternative splice sites, constitutive splice sites, and skipped splice sites can be explained by the GC content difference. Alternative splice sites tend to be GC-enriched and exhibit more stable RNA secondary structures in all of the considered species. In humans and mice, splice sites of first exons and long exons tend to be GC-enriched and hence form more stable structures, indicating the special role of RNA secondary structures in promoter proximal splicing events and the splicing of long exons. In addition, GC-enriched exon-intron junctions tend to be overrepresented in tissue-specific alternative splice sites, indicating the functional consequence of the GC effect. Compared with regions far from splice sites and decoy splice sites, real splice sites are GC-enriched. We also found that the GC-content effect is much stronger than the nucleotide-order effect to form stable secondary structures.
All of these results indicate that GC content is related to splice site usage and it may mediate the splicing process through RNA secondary structures.
Pre-mRNA splicing in eukaryotes removes introns and joins exons together. It is catalyzed by the splicesome that is a large ribonucleoprotein complex with several hundred proteins and five small nuclear RNAs [1, 2]. The recognition of splice sites requires multiple RNA binding proteins to bind to various splicing signals in pre-mRNAs. Genes can choose different sets of splice sites to produce multiple transcript isoforms, which further increases the complexity of splicing regulation. In eukaryotes, besides some short consensus sequence elements around the 5' splice site (5'ss), the 3' splice site (3'ss), the branch point, and the polypyrimidine tract, the splicing process needs other splicing regulatory elements (SREs) such as splicing enhancers or silencers [3–6]. In addition, pre-mRNA secondary structures also play an important role in splicing regulation .
It has been reported that local RNA secondary structures affect splice site selection through experimental observations from individual genes [8–12]. With the growing amount of genomic data and tools available, more genome-wide studies were carried out to support the hypothesis that pre-mRNA secondary structures are involved in the splicing process. For example, Patterson et al.  reported that the splice site prediction can be improved by adding the localized pre-mRNA secondary structure information to the conventional sequence-based approaches. Hiller et al.  found that some experimentally verified splicing enhancers and silencers near splice sites are significantly enriched in the single-stranded regions of the local secondary structures. Conserved secondary structures in Drosophila genomes were identified and they may modulate splicing regulation through long distance interactions . Shepard et al.  discovered that stable and conserved pre-mRNA secondary structures around splice sites may promote alternative splicing to a large extent. All of these results indicate that the secondary structure of pre-mRNA is part of the mRNA splicing code [6, 17, 18].
In this work, we first confirmed that for internal exons, structures around alternative splice sites are significantly more stable than those around constitutive and skipped splice sites in multiple species. More importantly, we found that these splice sites have distinct GC content. The GC content differences can remarkably explain these stability differences because GC content is positively associated with structural stability and sites with thermodynamic advantages tend to be GC enriched. We also found that splice site of the first exon in humans and mice tends to be more stable, no matter whether it is an alternative splice site or a constitutive splice site, because the promoter regions are generally GC enriched . In addition, splice sites of long exons tend to be GC-enriched and hence more stable in the structural level. Tissue-specific alternative splice sites in humans are also GC-enriched, which indicates the functional consequence of the GC effect. We further show that regions around splice sites are GC enriched as compared to regions far away or decoy splice site regions, which suggests a selection pressure on splice site regions to form stable secondary structures. By contrast, the nucleotide-order effect to the structural stability around splice sites is insignificant. All these support that GC may be an important factor in splicing through forming stable secondary structures.
Difference of pre-mRNA secondary structures between alternative, constitutive, and skipped splice sites
Difference of GC content between alternative, constitutive, and skipped splice sites and its association with the stability of pre-mRNA secondary structure
Since structural stability was significantly associated with the GC percentage in junction sequences, it is possible that long-range GC enrichment patterns, instead of the local GC variation, would result in the genome-wide thermodynamic advantages in alternatively spliced sites. For example, if alternative splice sites were more frequently selected in high GC isochors, the junction sequences therein would be significantly biased to more stable base pairings compared with the constitutive and skipped splice sites. To clarify these two factors, pairwise free energy comparisons were performed with strict distance control to ensure similar GC background but allowing local GC variations. Across the human genome, we only selected the alternative splice sites with at least one constitutive splice site within a distance of 3000 bp. The alternative splice sites and the nearby constitutive splice sites shared similar GC background, but had local variations. Pairwise energy comparison of these two groups still suggested slight yet significant enrichment of more stable structures in alternatively spliced sites (-45.54 vs. -44.71 kcals/mol, paired Wilcoxon test P = 0.001 for the donor sites, -43.95 vs. -43.46 kcals/mol, P = 0.03 for the acceptor sites). Similar results were also observed in the comparison between alternative and skipped splice sites (data not shown), further confirming the contribution of local GC variations to stabilize mRNA secondary structures.
Splice sites of the first exons or long exons are GC enriched and hence more stable in humans and mice
It is well known that human promoter regions are enriched in GC . We suspect that splice sites near transcript start sites may form more stable structures compared with those in internal regions. To test our hypothesis, a list of constitutively and alternatively spliced donor sites of the first exons was generated and structures were predicted by the RNAfold program (see methods). Note that for studies in other places of this paper, only splice sites of internal exons were considered. As expected, both alternative (-52.87 vs. -41.28 kcals/mol) and constitutive splice sites (-55.56 vs. -38.43 kcals/mol) near the transcription start sites preferentially formed more stable structures than the middle ones (Wilcoxon tests, P < 2.2 × 10-16). However, in contrast with the splice sites of the middle exons, for the first exons, alternative splice sites displayed less stable structures than constitutive splice sites (-52.87 vs. -55.56 kcals/mol, Wilcoxon test, P = 4.9 × 10-4). As expected, the GC content of these alternative splices was lower than that of the constitutive splice sites (0.62 vs. 0.64, Wilcoxon test, P = 7.0 × 10-6). The results indicate that the intervention of RNA secondary structures in splicing may vary upon regions and it depends on GC content.
Shepard et al.  found that long exons tended to have more stable structures around splice sites. Similarly, we observed a significant bias in long exons (length > 200 bp) toward more stable structures in humans (Wilcoxon tests, P < 2.2 × 10-16 for the donor sites and P = 5.3 × 10-8 for acceptor sites). More importantly, this difference can also be explained by the GC difference. Specifically, the GC content of the splice sites around long exons was higher compared with that around short exons (length ≤ 200 bp) (Wilcoxon tests, P < 2.2 × 10-16 for both the donor and acceptor sites). Hence, these results suggest that pre-mRNA secondary structures may play different roles in the splicing of exons with different lengths, depending on the GC content.
Similar results for the first exons and long exons were obtained for mice. However, in nematodes or fruit flies, since promoter regions were not GC-enriched , the splice sites near the promoter regions did not exhibit more stable structures. Besides, no obvious bias toward more stable structures has been observed in long exons of nematodes or flies.
Tissue-specific alternative splice sites are GC enriched
Selection on GC content around splice sites
GC effect is more dominant than the nucleotide-order effect
In this paper, we studied GC content around splice sites in terms of its effect on the splice site usage and the stability of pre-mRNA secondary structures in multiple species. For middle exons, alternative splice sites were more enriched with GC than constitutive and skipped splice sites, and hence exhibited stronger potential to form stable secondary structures (as shown in Figures 1 and 2). More importantly, we showed that the GC content was the major factor to account for the structural stability discrepancy. Given the same GC content, the predicted free energy is similar no matter whether it is alternative splice sites or other splice sites (Figure 4). We also notice that although the structural topology predicted from RNAfold has already been proved to demonstrate significantly high correlations with the experimental results in yeast , it is still possible that the estimated secondary structure may be different from that in vivo for the species we tested. In addition, it is still unclear whether RNA tertiary structure is also a common mechanism to regulate splicing . Advanced experimental technologies are in need to better predict RNA structures.
Several possible regulation mechanisms can be proposed for our genome-wide structural stability observations. First, mRNA secondary structures may mediate the splicing process via affecting the motif recognition rate to facilitate or prevent the binding of splicing regulators. It has been reported that splicing regulators have different RNA structural topology preferences . The stable stem regions (tend to be GC-enriched) in alternative splice sites might mask some key motifs and thus repress the site recognition. Second, long exons exhibit pronounced GC enrichment and hence more stable structures, indicating that RNA secondary structures may be actively involved in the splicing event via long distance mediations to bring the distal signals together [30, 31]. Third, the GC-enriched sequences around the first donor sites indicate that the role of stable RNA secondary structures in splicing may vary across regions. It has been reported that splicing near the promoter region enhances transcription . Our discovered stable secondary structures around the promoter regions may contribute to the interaction between splicing and transcription via serving as stable binding platforms of the transcription/splicing regulators. Furthermore, increasing evidence shows that in the beginning of the transcription process, the polymerase II may fall into some paused status, or even go forward and backward from time to time. Theoretical computations demonstrate that the stable hairpins in the nascent RNA is beneficial for the backtracking of polymerase II , which explain our discovery about the more stable structures of the first donor sites.
It is well known that the structure prediction software, such as RNAfold and Mfold, executes the pre-mRNA secondary structure calculation under simplified conditions, probably resulting in inappropriate minimum free energy predictions. Besides, the folding of nascent mRNA sequences may change frequently in different biological environments . We therefore performed studies on functional alternative splice sites (i.e., tissue-specific alternative splice sites). These sites also tended to be GC-enriched, further suggesting the functional consequence of GC content in splicing.
Sequences around splice sites were more GC enriched compared to either the positions far away from the splice sites or nearby decoy splice sites, indicating the selection pressure on splice site regions to form stable structures. We also investigated whether additional factors exist to affect the structural stability. Permutation analysis reveals only limited nucleotide-order effect in the native sequence to keep a favorable context with larger thermodynamic advantages (see Figure 7). Thus, the stability variation introduced by GC was more dominant that that caused by the nucleotide order.
In order to check whether the regulation role of GC content in splicing is a universal phenomenon in multiple species, we extended our work on humans to nematodes, fruit files, and mice. The results of energy vs. GC were further summarized in a table (Additional file 11). In spite of the lower quality of the splicing event lists due to the incompleteness of the gene annotation compared to humans, we also observed the enrichment of stable structures in alternatively spliced sites as well as slightly different yet still statistically significant correlations between the GC percentage and the free energy in all these species. Thus, the involvement of GC content in the splicing regulation process might have been in existence from ancient times.
All together, our results show that GC content around splice sites may play an important role in splicing regulation by forming stable secondary structures. Through the selection of GC enriched sequences, exons with alternative splice sites can maintain stable pre-mRNA structures to promote alternative splicing. This GC effect is more dominant than the nucleotide-order effect. On the other hand, constitutive exons and cassette exons are not enriched with stable structures. It indicates that that the pre-mRNA structure is part of, but not the whole of, the splicing code. We expect to investigate the biological significance in details when related experimental data become available in the future.
Splice site assembly
Splice site positions of different human exons were obtained from the UCSC Genome Browser (alternative splicing event track, version hg18). For alternative splice sites, we required a distance of at least 8 bp between two 3' splice sites of the same exon, and at least 5 bp between two 5' splice sites of the same exon. A total of 4,128 internal alternative 5'ss, 5,899 internal alternative 3'ss, and 44,337 skipped splice sites (21,473 5'ss and 22,864 3'ss) for cassette exons were extracted. We generated our own list of internal constitutive exons using the UCSC transcript annotation information with two requirements: (1) a constitutive exon should appear in all isoforms (with at least three exons) of the gene without overlapping with any other exon and preserve exactly the same starting and ending positions; (2) the gene should contain at least four different isoforms. Thus, 30,992 internal constitutive splicing sites (15,496 5' ss and 15,496 3'ss) were selected in total. Note that for exons with alternative splice sites, all of the splice sites were used for the structure and GC analysis.
For the donor sites of the first exons study, we generated the list of first alternative and constitutive sites by examining transcript isoforms. We focused on transcripts with at least three exons. Only the first exons which share the same transcription starting sites in all the transcript isoforms of the same genes were considered. Constitutive first donor sites were defined as the ones appearing at the 5'ss of the initial exons and keeping exactly the same position in all transcript isoforms (at least three isoforms). If the donor site of the first did not appear in all of the transcript isoforms, it was counted as an alternative first donor site. According to these criteria, 405 alternative first donor sites and 1,761 constitutive first donor sites were generated in our work.
Refseq gene annotations (version WS190) were downloaded from the UCSC genome browser http://genome.ucsc.edu/ in the analysis of nematodes. For the fruit fly and mouse studies, gene annotation lists were downloaded from http://genome.ewha.ac.kr/ECgene/, in which only transcripts with high quality annotations (confidence high group) were selected. Alternative splice sites and skipped splice sites were assembled from the annotation lists. Similar criteria as in humans were used to generate the constitutive exons lists.
Pre-mRNA secondary structure prediction
NCBI build 36 for the mouse and human, BDGP Release 4 for the fruit fly, and WS190 for the nematode genome sequences were used for the structure prediction. For each splice site, 70 nucleotides were added both up and downstream of the intron-exon junction to form a 141 nt window, and then the RNAfold program was used to predict the minimum free energy. The default settings of RNAfold were used for the energy prediction in humans and mice. However, for nematodes and fruit flies, the prediction temperature was set to 25 and 24 degrees centigrade respectively to adjust for the different temperatures for growth.
Tissue-specific alternative splice sites and decoy splice sites
Tissue-specific alternative spliced sites were obtained from the microarray study where the expression of alternative splicing events in 48 different human cells was profiled . For each alternative donor or acceptor sites, if the isoform proportionality change was greater than 10 percent and at the same time the corresponding P value was less than 0.3, we claimed that they were tissue-specific alternative splice sites. The non-tissue specific alternative splice sites were those with the proportionality change less than 5 percent or the P value were larger than 0.3.
As a control group of real splice sites, decoy splice sites that share similar consensus sequence information but seldom experience the splicing events were selected. Firstly, all the AG/GU dinucleotides in the flanking introns within 300-150 nt to the real splice site were extracted as candidates for decoy splice sites, and MaxEntScan code from http://genes.mit.edu/burgelab/maxent/Xmaxentscan_scoreseq.html was used to compute the splice sites score for each candidate. A higher score indicates a larger similarity to the consensus sequence of their training data. The candidates with a higher score within 300-150 nt to the real splice sites were chosen as decoy splice sites. For the splice sites with multiple decoy sites nearby, only the nearest one was selected as the control to the real junction.
Two random permutations were used to generate the control data to evaluate the nucleotide-order effect on secondary structures. In the permutation that kept only the first order nucleotide frequencies, each 141 nt sequence was randomly permuted and the pre-mRNA secondary structure was predicted for the permuted sequence. We did the permutation ten times. The results were all similar (data not shown). It is well known that the dinucleotide frequency affects the predicted stability significantly by the RNAfold program . Thus for a fair comparison, the ushuffle program at http://digital.cs.usu.edu/~mjiang/ushuffle/ was used to generate the permutations that fixed the dinucleotide frequencies.
All of the statistical analyses including the Wilcoxon tests, the pairwise Wilcoxon tests, and regression analyses were performed using the R software.
- GC content:
5' splice site
3' splice site
This research was supported in part by the National Institutes of Health [P50 HG 002790] and a start-up fund from USC.
- Jurica MS, Moore MJ: Pre-mRNA splicing: awash in a sea of proteins. Mol Cell. 2003, 12 (1): 5-14. 10.1016/S1097-2765(03)00270-3.PubMedView ArticleGoogle Scholar
- Zhou Z, Licklider LJ, Gygi SP, Reed R: Comprehensive proteomic analysis of the human spliceosome. Nature. 2002, 419 (6903): 182-185. 10.1038/nature01031.PubMedView ArticleGoogle Scholar
- Cartegni L, Chew SL, Krainer AR: Listening to silence and understanding nonsense: exonic mutations that affect splicing. Nat Rev Genet. 2002, 3 (4): 285-298. 10.1038/nrg775.PubMedView ArticleGoogle Scholar
- Faustino NA, Cooper TA: Pre-mRNA splicing and human disease. Genes Dev. 2003, 17 (4): 419-437. 10.1101/gad.1048803.PubMedView ArticleGoogle Scholar
- Stadler MB, Shomron N, Yeo GW, Schneider A, Xiao X, Burge CB: Inference of splicing regulatory activities by sequence neighborhood analysis. PLoS Genet. 2006, 2 (11): e191-10.1371/journal.pgen.0020191.PubMed CentralPubMedView ArticleGoogle Scholar
- Wang Z, Burge CB: Splicing regulation: from a parts list of regulatory elements to an integrated splicing code. RNA. 2008, 14 (5): 802-813. 10.1261/rna.876308.PubMed CentralPubMedView ArticleGoogle Scholar
- Buratti E, Baralle FE: Influence of RNA secondary structure on the pre-mRNA splicing process. Mol Cell Biol. 2004, 24 (24): 10505-10514. 10.1128/MCB.24.24.10505-10514.2004.PubMed CentralPubMedView ArticleGoogle Scholar
- Clouet d'Orval B, d'Aubenton Carafa Y, Sirand-Pugnet P, Gallego M, Brody E, Marie J: RNA secondary structure repression of a muscle-specific exon in HeLa cell nuclear extracts. Science. 1991, 252 (5014): 1823-1828.PubMedView ArticleGoogle Scholar
- Eperon LP, Graham IR, Griffiths AD, Eperon IC: Effects of RNA secondary structure on alternative splicing of pre-mRNA: is folding limited to a region behind the transcribing RNA polymerase?. Cell. 1988, 54 (3): 393-401. 10.1016/0092-8674(88)90202-4.PubMedView ArticleGoogle Scholar
- Graveley BR: Mutually exclusive splicing of the insect Dscam pre-mRNA directed by competing intronic RNA secondary structures. Cell. 2005, 123 (1): 65-73. 10.1016/j.cell.2005.07.028.PubMed CentralPubMedView ArticleGoogle Scholar
- Jacquenet S, Ropers D, Bilodeau PS, Damier L, Mougin A, Stoltzfus CM, Branlant C: Conserved stem-loop structures in the HIV-1 RNA region containing the A3 3' splice site and its cis-regulatory element: possible involvement in RNA splicing. Nucleic Acids Res. 2001, 29 (2): 464-478. 10.1093/nar/29.2.464.PubMed CentralPubMedView ArticleGoogle Scholar
- Loeb DD, Mack AA, Tian R: A secondary structure that contains the 5' and 3' splice sites suppresses splicing of duck hepatitis B virus pregenomic RNA. J Virol. 2002, 76 (20): 10195-10202. 10.1128/JVI.76.20.10195-10202.2002.PubMed CentralPubMedView ArticleGoogle Scholar
- Patterson DJ, Yasuhara K, Ruzzo WL: Pre-mRNA secondary structure prediction aids splice site prediction. Pac Symp Biocomput. 2002, 223-234.Google Scholar
- Hiller M, Zhang Z, Backofen R, Stamm S: Pre-mRNA secondary structures influence exon recognition. PLoS Genet. 2007, 3 (11): e204-10.1371/journal.pgen.0030204.PubMed CentralPubMedView ArticleGoogle Scholar
- Raker VA, Mironov AA, Gelfand MS, Pervouchine DD: Modulation of alternative splicing by long-range RNA structures in Drosophila. Nucleic Acids Res. 2009, 37 (14): 4533-4544. 10.1093/nar/gkp407.PubMed CentralPubMedView ArticleGoogle Scholar
- Shepard PJ, Hertel KJ: Conserved RNA secondary structures promote alternative splicing. RNA. 2008, 14 (8): 1463-1469. 10.1261/rna.1069408.PubMed CentralPubMedView ArticleGoogle Scholar
- Warf MB, Berglund JA: Role of RNA structure in regulating pre-mRNA splicing. Trends Biochem Sci. 2009, 35 (3): 169-178. 10.1016/j.tibs.2009.10.004.PubMed CentralPubMedView ArticleGoogle Scholar
- Barash Y, Calarco JA, Gao W, Pan Q, Wang X, Shai O, Blencowe BJ, Frey BJ: Deciphering the splicing code. Nature. 2010, 465 (7294): 53-59. 10.1038/nature09000.PubMedView ArticleGoogle Scholar
- Kalari KR, Casavant M, Bair TB, Keen HL, Comeron JM, Casavant TL, Scheetz TE: First exons and introns--a survey of GC content and gene structure in the human genome. In Silico Biol. 2006, 6 (3): 237-242.PubMedGoogle Scholar
- Schroeder R, Grossberger R, Pichler A, Waldsich C: RNA folding in vivo. Curr Opin Struct Biol. 2002, 12 (3): 296-300. 10.1016/S0959-440X(02)00325-1.PubMedView ArticleGoogle Scholar
- Bompfunewerer AF, Backofen R, Bernhart SH, Hertel J, Hofacker IL, Stadler PF, Will S: Variations on RNA folding and alignment: lessons from Benasque. J Math Biol. 2008, 56 (1-2): 129-144. 10.1007/s00285-007-0107-5.PubMedView ArticleGoogle Scholar
- Hofacker IL, Stadler PF: Memory efficient folding algorithms for circular RNA secondary structures. Bioinformatics. 2006, 22 (10): 1172-1176. 10.1093/bioinformatics/btl023.PubMedView ArticleGoogle Scholar
- Rach EA, Yuan HY, Majoros WH, Tomancak P, Ohler U: Motif composition, conservation and condition-specificity of single and alternative transcription start sites in the Drosophila genome. Genome Biol. 2009, 10 (7): R73-10.1186/gb-2009-10-7-r73.PubMed CentralPubMedView ArticleGoogle Scholar
- Castle JC, Zhang C, Shah JK, Kulkarni AV, Kalsotra A, Cooper TA, Johnson JM: Expression of 24,426 human alternative splicing events and predicted cis regulation in 48 tissues and cell lines. Nat Genet. 2008, 40 (12): 1416-1425. 10.1038/ng.264.PubMed CentralPubMedView ArticleGoogle Scholar
- Yeo G, Burge CB: Maximum entropy modeling of short sequence motifs with applications to RNA splicing signals. J Comput Biol. 2004, 11 (2-3): 377-394. 10.1089/1066527041410418.PubMedView ArticleGoogle Scholar
- Seffens W, Digby D: mRNAs have greater negative folding free energies than shuffled or codon choice randomized sequences. Nucleic Acids Res. 1999, 27 (7): 1578-1584. 10.1093/nar/27.7.1578.PubMed CentralPubMedView ArticleGoogle Scholar
- Workman C, Krogh A: No evidence that mRNAs have lower folding free energies than random sequences with the same dinucleotide distribution. Nucleic Acids Res. 1999, 27 (24): 4816-4822. 10.1093/nar/27.24.4816.PubMed CentralPubMedView ArticleGoogle Scholar
- Kertesz M, Wan Y, Mazor E, Rinn JL, Nutter RC, Chang HY, Segal E: Genome-wide measurement of RNA secondary structure in yeast. Nature. 2010, 467 (7311): 103-107. 10.1038/nature09322.PubMedView ArticleGoogle Scholar
- Li X, Quon G, Lipshitz HD, Morris Q: Predicting in vivo binding sites of RNA-binding proteins using mRNA secondary structure. RNA. 2010, 16 (6): 1096-1107. 10.1261/rna.2017210.PubMed CentralPubMedView ArticleGoogle Scholar
- Thompson-Jager S, Domdey H: Yeast pre-mRNA splicing requires a minimum distance between the 5' splice site and the internal branch acceptor site. Mol Cell Biol. 1987, 7 (11): 4010-4016.PubMed CentralPubMedView ArticleGoogle Scholar
- Rogic S, Montpetit B, Hoos HH, Mackworth AK, Ouellette BF, Hieter P: Correlation between the secondary structure of pre-mRNA introns and the efficiency of splicing in Saccharomyces cerevisiae. BMC Genomics. 2008, 9: 355-10.1186/1471-2164-9-355.PubMed CentralPubMedView ArticleGoogle Scholar
- Furger A, O'Sullivan JM, Binnie A, Lee BA, Proudfoot NJ: Promoter proximal splice sites enhance transcription. Genes Dev. 2002, 16 (21): 2792-2799. 10.1101/gad.983602.PubMed CentralPubMedView ArticleGoogle Scholar
- Klopper AV, Bois JS, Grill SW: Influence of secondary structure on recovery from pauses during early stages of RNA transcription. Phys Rev E Stat Nonlin Soft Matter Phys. 2010, 81 (3 Pt 1): 030904-10.1103/PhysRevE.81.030904.PubMedView ArticleGoogle Scholar
- Mahen EM, Watson PY, Cottrell JW, Fedor MJ: mRNA secondary structures fold sequentially but exchange rapidly in vivo. PLoS Biol. 2010, 8 (2): e1000307-10.1371/journal.pbio.1000307.PubMed CentralPubMedView ArticleGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.