Exploring the selective constraint on the sizes of insertions and deletions in 5' untranslated regions in mammals

A low-coverage 3' RNA-seq to detect homeolog expression in polyploid wheat

Although allopolyploid species are common among natural and crop species, it is not easy to distinguish duplicated genes, known as homeologs, during their genomic analysis. Yet, cost-efficient RNA sequencing (RNA-seq) is to be developed for large-scale transcriptomic studies such as time-series analysis and genome-wide association studies in allopolyploids. In this study, we employed a 3' RNA-seq utilizing 3' untranslated regions (UTRs) containing frequent mutations among homeologous genes, compared to coding sequence. Among the 3' RNA-seq protocols, we examined a low-cost method Lasy-Seq using an allohexaploid bread wheat, Triticum aestivum. HISAT2 showed the best performance for 3' RNA-seq with the least mapping errors and quick computational time. The number of detected homeologs was further improved by extending 1 kb of the 3' UTR annotation. Differentially expressed genes in response to mild cold treatment detected by the 3' RNA-seq were verified with high-coverage conventional RNA-seq, although the latter detected more differentially expressed genes. Finally, downsampling showed that even a 2 million sequencing depth can still detect more than half of expressed homeologs identifiable by the conventional 32 million reads. These data demonstrate that this low-cost 3' RNA-seq facilitates large-scale transcriptomic studies of allohexaploid wheat and indicate the potential application to other allopolyploid species.

The non-random patterns of genetic variation induced by asymmetric somatic hybridization in wheat

BACKGROUND

Asymmetric somatic hybridization is an efficient crop breeding approach by introducing several exogenous chromatin fragments, which leads to genomic shock and therefore induces genome-wide genetic variation. However, the fundamental question concerning the genetic variation such as whether it occurs randomly and suffers from selection pressure remains unknown.

RESULTS

Here, we explored this issue by comparing expressed sequence tags of a common wheat cultivar and its asymmetric somatic hybrid line. Both nucleotide substitutions and indels (insertions and deletions) had lower frequencies in coding sequences than in un-translated regions. The frequencies of nucleotide substitutions and indels were both comparable between chromosomes with and without introgressed fragments. Nucleotide substitutions distributed unevenly and were preferential to indel-flanking sequences, and the frequency of nucleotide substitutions at 5'-flanking sequences of indels was obviously higher in chromosomes with introgressed fragments than in those without exogenous fragment. Nucleotide substitutions and indels both had various frequencies among seven groups of allelic chromosomes, and the frequencies of nucleotide substitutions were strongly negatively correlative to those of indels. Among three sets of genomes, the frequencies of nucleotide substitutions and indels were both heterogeneous, and the frequencies of nucleotide substitutions exhibited drastically positive correlation to those of indels.

CONCLUSIONS

Our work demonstrates that the genetic variation induced by asymmetric somatic hybridization is attributed to both whole genomic shock and local chromosomal shock, which is a predetermined and non-random genetic event being closely associated with selection pressure. Asymmetric somatic hybrids provide a worthwhile model to further investigate the nature of genomic shock induced genetic variation.

Using high-throughput sequencing transcriptome data for INDEL detection: challenges for cancer drug discovery

INTRODUCTION

A cancer cell is a mosaic of genomic and epigenomic alterations. Distinct cancer molecular signatures can be observed depending on tumor type or patient genetic background. One type of genomic alteration is the insertion and/or deletion (INDEL) of nucleotides in the DNA sequence, which may vary in length, and may change the encoded protein or modify protein domains. INDELs are associated to a large number of diseases and their detection is done based on low-throughput techniques. However, high-throughput sequencing has also started to be used for detection of novel disease-causing INDELs. This search may identify novel drug targets.

AREAS COVERED

This review presents examples of using high-throughput sequencing (DNA-Seq and RNA-Seq) to investigate the incidence of INDELs in coding regions of human genes. Some of these examples successfully utilized RNA-Seq to identify INDELs associated to diseases. In addition, other studies have described small INDELs related to chemo-resistance or poor outcome of patients, while structural variants were associated with a better clinical outcome.

EXPERT OPINION

On average, there is twice as much RNA-Seq data available at the most used repositories for such data compared to DNA-Seq. Therefore, using RNA-Seq data is a promising strategy for studying cancer samples with unknown mechanisms of drug resistance, aiming at the discovery of proteins with potential as novel drug targets.

Are all of the human exons alternatively spliced?

Alternative mRNA splicing (AS) is a major mechanism for increasing regulatory complexity. A key concept in AS is the distinction between alternatively and constitutively spliced exons (ASEs and CSEs, respectively). ASEs and CSEs have been reported to be differentially regulated, and to have distinct biological properties. However, the recent flood of RNA-sequencing data has obscured the boundary between ASEs and CSEs. Researchers are beginning to question whether ‘authentic CSEs’ do exist, and whether the ASE/CSE distinction is biologically invalid. Here, I examine the influences of increasing transcriptome data on the human ASE/CSE classification and our past understanding of the properties of these two types of exons. Interestingly, although the percentage of human ASEs has increased dramatically in recent years, the overall distinction between ASEs and CSEs remain valid. For example, CSEs are longer, evolve more slowly, and less frequently correspond to intrinsically disordered protein regions than ASEs. In addition, only a relatively small number of human genes have their transcripts composed entirely of ASEs despite the large amount of high-throughput transcriptome information. Therefore, the ‘backbone’ concept of AS, in which CSEs constitute the invariant part and ASEs the flexible part of the transcript, appears to be generally true despite the increasing percentage of ASEs in the human exome.

Alternative RNA structure-coupled gene regulations in tumorigenesis

Alternative RNA structures (ARSs), or alternative transcript isoforms, are critical for regulating cellular phenotypes in humans. In addition to generating functionally diverse protein isoforms from a single gene, ARS can alter the sequence contents of 5'/3' untranslated regions (UTRs) and intronic regions, thus also affecting the regulatory effects of these regions. ARS may introduce premature stop codon(s) into a transcript, and render the transcript susceptible to nonsense-mediated decay, which in turn can influence the overall gene expression level. Meanwhile, ARS can regulate the presence/absence of upstream open reading frames and microRNA targeting sites in 5'UTRs and 3'UTRs, respectively, thus affecting translational efficiencies and protein expression levels. Furthermore, since ARS may alter exon-intron structures, it can influence the biogenesis of intronic microRNAs and indirectly affect the expression of the target genes of these microRNAs. The connections between ARS and multiple regulatory mechanisms underline the importance of ARS in determining cell fate. Accumulating evidence indicates that ARS-coupled regulations play important roles in tumorigenesis. Here I will review our current knowledge in this field, and discuss potential future directions.

Emerging evidence for functional peptides encoded by short open reading frames

Short open reading frames (sORFs) are a common feature of all genomes, but their coding potential has mostly been disregarded, partly because of the difficulty in determining whether these sequences are translated. Recent innovations in computing, proteomics and high-throughput analyses of translation start sites have begun to address this challenge and have identified hundreds of putative coding sORFs. The translation of some of these has been confirmed, although the contribution of their peptide products to cellular functions remains largely unknown. This Review examines this hitherto overlooked component of the proteome and considers potential roles for sORF-encoded peptides.

uPEPperoni: an online tool for upstream open reading frame location and analysis of transcript conservation

BACKGROUND

Several small open reading frames located within the 5' untranslated regions of mRNAs have recently been shown to be translated. In humans, about 50% of mRNAs contain at least one upstream open reading frame representing a large resource of coding potential. We propose that some upstream open reading frames encode peptides that are functional and contribute to proteome complexity in humans and other organisms. We use the term uPEPs to describe peptides encoded by upstream open reading frames.

RESULTS

We have developed an online tool, termed uPEPperoni, to facilitate the identification of putative bioactive peptides. uPEPperoni detects conserved upstream open reading frames in eukaryotic transcripts by comparing query nucleotide sequences against mRNA sequences within the NCBI RefSeq database. The algorithm first locates the main coding sequence and then searches for open reading frames 5' to the main start codon which are subsequently analysed for conservation. uPEPperoni also determines the substitution frequency for both the upstream open reading frames and the main coding sequence. In addition, the uPEPperoni tool produces sequence identity heatmaps which allow rapid visual inspection of conserved regions in paired mRNAs.

CONCLUSIONS

uPEPperoni features user-nominated settings including, nucleotide match/mismatch, gap penalties, Ka/Ks ratios and output mode. The heatmap output shows levels of identity between any two sequences and provides easy recognition of conserved regions. Furthermore, this web tool allows comparison of evolutionary pressures acting on the upstream open reading frame against other regions of the mRNA. Additionally, the heatmap web applet can also be used to visualise the degree of conservation in any pair of sequences. uPEPperoni is freely available on an interactive web server at http://upep-scmb.biosci.uq.edu.au.

Identification and complete genome analysis of a virus variant or putative new foveavirus associated with apple green crinkle disease

A virus identified as "apple green crinkle associated virus" (AGCaV) was isolated from Aurora Golden Gala apple showing severe symptoms of green crinkle disease. Evidence was obtained of a potential causal relationship to the disease. The viral genome consists of 9266 nucleotides, excluding the poly(A) tail at the 3'-terminus. It has a genome organization similar to that of members of the species Apple stem pitting virus (ASPV), the type species of the genus Foveavirus, family Betaflexiviridae. ORF1 of AGCaV encodes a replicase-complex polyprotein with a molecular mass of 247 kDa; the proteins of ORFs 2, 3, and 4 (TGB proteins) are estimated to be 25.1 kDa, 12.8 kDa, and 7.4 kDa, respectively; and ORF5 encodes the CP, with an estimated molecular mass of 43.3 kDa. Interestingly, AGCaV utilizes different stop codons for ORF1, ORF3, and ORF5 compared to the ASPV type isolate PA66, and between the two viruses, six distinct indel events were observed within ORF5. AGCaV has four non-coding regions (NCRs), including a 5'-NCR (60 nt), a 3'-NCR (134 nt), and two intergenic (IG) NCRs: IG-NCR1 (69 nt) and IG-NCR2 (91 nt). A conserved stable hairpin structure was identified in the variable 5'-NCR of members of the genus Foveavirus. AGCaV may be a variant or strain of ASPV with unique biological properties, but there is evidence that it may be a distinct putative foveavirus.

Selective constraint on the upstream open reading frames that overlap with coding sequences in animals

Upstream open reading frames (uORFs) are translational regulatory elements located in 5′ untranslated regions. They can significantly repress the translation of the downstream coding sequences (CDS), and participate in the spatio-temporal regulations of protein translation. Notwithstanding this biological significance, the selective constraint on uORFs remains underexplored. Particularly, the uORFs that partially overlap with CDS with a different reading frame (overlapping uORFs, or “VuORFs”) may lead to strong translational inhibition or N-terminal truncation of the peptides encoded by the affected CDS. By analyzing VuORF-containing transcripts (designated as “VuORF transcripts”) in human, mouse, and fruit fly, we demonstrate that VuORFs are in general slightly deleterious - the proportion of genes that encode at least one VuORF transcript is significantly smaller than expected in all of the three examined species. In addition, this proportion is significantly smaller in fruit fly than in mammals, indicating a higher efficiency of removing VuORFs in the former species because of its larger effective population size. Furthermore, the deleterious effect of a VuORF depends on the sequence context of its start codon (VuAUG). VuORFs with an optimal VuAUG context are more strongly disfavored than those with a suboptimal context in all of the three examined species. And the propensity to remove optimal-context VuAUGs is stronger in fruit fly than in mammals. Intriguingly, however, the currently observable optimal-context VuAUGs (but not suboptimal-context VuAUGs) are more conserved than expected. These observations suggest that the regulatory functions of VuORFs may have been gained fortuitously in organisms with a small effective population size because the slightly deleterious effect of these elements can be better tolerated in these organisms, thus allowing opportunities for the development of novel biological functions. Nevertheless, once the functions of VuORFs were established, they became subject to negative selection.

The genomic features that affect the lengths of 5' untranslated regions in multicellular eukaryotes

Background

The lengths of 5’UTRs of multicellular eukaryotes have been suggested to be subject to stochastic changes, with upstream start codons (uAUGs) as the major constraint to suppress 5’UTR elongation. However, this stochastic model cannot fully explain the variations in 5’UTR length. We hypothesize that the selection pressure on a combination of genomic features is also important for 5’UTR evolution. The ignorance of these features may have limited the explanatory power of the stochastic model. Furthermore, different selective constraints between vertebrates and invertebrates may lead to differences in the determinants of 5’UTR length, which have not been systematically analyzed.

Methods

Here we use a multiple linear regression model to delineate the correlation between 5’UTR length and the combination of a series of genomic features (G+C content, observed-to-expected (OE) ratios of uAUGs, upstream stop codons (uSTOPs), methylation-related CG/UG dinucleotides, and mRNA-destabilizing UU/UA dinucleotides) in six vertebrates (human, mouse, rat, chicken, African clawed frog, and zebrafish) and four invertebrates (fruit fly, mosquito, sea squirt, and nematode). The relative contributions of each feature to the variation of 5’UTR length were also evaluated.

Results

We found that 14%~33% of the 5’UTR length variations can be explained by a linear combination of the analyzed genomic features. The most important genomic features are the OE ratios of uSTOPs and G+C content. The surprisingly large weightings of uSTOPs highlight the importance of selection on upstream open reading frames (which include both uAUGs and uSTOPs), rather than on uAUGs per se. Furthermore, G+C content is the most important determinants for most invertebrates, but for vertebrates its effect is second to uSTOPs. We also found that shorter 5’UTRs are affected more by the stochastic process, whereas longer 5’UTRs are affected more by selection pressure on genomic features.

Conclusions

Our results suggest that upstream open reading frames may be the real target of selection, rather than uAUGs. We also show that the selective constraints on genomic features of 5’UTRs differ between vertebrates and invertebrates, and between longer and shorter 5’UTRs. A more comprehensive model that takes these findings into consideration is needed to better explain 5’UTR length evolution.