Ultraconserved cDNA segments in the human transcriptome exhibit resistance to folding and implicate function in translation and alternative splicing

Abstract

Ultraconservation, defined as perfect human-to-rodent sequence identity at least 200-bp long, is a strong indicator of evolutionary and functional importance and has been explored extensively at the genome level. However, it has not been investigated at the transcript level, where such extreme conservation might highlight loci with important post-transcriptional regulatory roles. We present 96 ultraconserved cDNA segments (UCSs), stretches of human mature mRNAs that match identically with orthologous regions in the mouse and rat genomes. UCSs can span multiple exons, a feature we leverage here to elucidate the role of ultraconservation in post-transcriptional regulation. UCS sites are implicated in functions at essentially every post-transcriptional stage: pre-mRNA splicing and degradation through alternative splicing and nonsense-mediated decay (AS-NMD), mature mRNA silencing by miRNA, fast mRNA decay rate and translational repression by upstream AUGs. We also found UCSs to exhibit resistance to formation of RNA secondary structure. These multiple layers of regulation underscore the importance of the UCS-containing genes as key global RNA processing regulators, including members of the serine/arginine-rich protein and heterogeneous nuclear ribonucleoprotein (hnRNP) families of essential splicing regulators. The discovery of UCSs shed new light on the multifaceted, fine-tuned and tight post-transcriptional regulation of gene families as conserved through the majority of the mammalian lineage.

Figure 1.

A cross-exon UCS, not previously found in UCE, exhibits extreme conservation across vertebrate genomes and overlaps the start codon of an alternatively spliced isoform.

Figure 2.

Expression patterns of cassette exons. Cassette exons containing cross-exon UCSs tend to be more ubiquitously expressed. This means UCS cassette exons are often spliced together with their neighboring UCS exons, forming one functional unit.

Figure 3.

Mechanism through which stop codon-containing exons are alternatively spliced and degraded by means of NMD. This regulatory mechanism is prevalent in splicing regulators such as the SR and hnRNP protein genes. Figured inspired by (12).

Figure 4.

SR protein gene SRP54 contains two UCSs, one of which is the same as UCE uc.28; the other is a cross-exon UCS KDN02407. Two alternative splicing events are associated with KDN02407, while none are found in uc.28. However, because of the upstream position of uc.28 from the alternative promoter, it may play role in regulating alternative transcription initiation instead.

Figure 5.

Examples of 5′-UTR diversity created through cassette exons, bleeding exons and retained introns. (A) 5′-UTR diversity (two light blue isoforms) is created through selective inclusion and exclusion of two ultraconserved cassette exons. The corresponding start codons are also contained within the UCS. (B) 5′-UTR diversity (two light blue isoforms) is created through retained introns and a bleeding exon. The start codon in one isoform becomes uAUG of the other. These exemplify how uAUG and alternative splicing might regulate translation efficiency.

Figure 6.

Gibbs free folding energy comparison shows significant fold resistance of ultra- and highly conserved elements as compared with random cDNA sequences. The gray horizontal line marks the mean of the fold energy of random sequences. miRNA coding sequences form significantly more stable structures while miRNA target sequences exhibit resistance to folding similar to conserved elements and suggesting an open structure for regulatory functions.

Figure 7.

UCS-containing mRNAs degrade at faster rates and have shorter half-life than average mRNAs. Note that decay rate is presented in log scale.