Research achievements during my PhD.

CPEB2 subfamily 3'UTRs

Background. Developmental biology in higher organisms critically depends on RNA mediated gene expression regulation. As a consequence, the length and complexity of 3' untranslated regions (3'UTRs) of mRNAs is dramatically expanded in mammals. The aim of my Ph.D. was to make sense of the different signatures present in the 3’UTRs, in particular miRNA targets and highly conserved elements.

Research. I showed that CPEB2, CPEB3 and CPEB4 transcripts are regulated by a shared ancestral miRNA signature¹. Moreover, I showed that the miRNA binding sites preceded the generation of highly conserved elements in their 3'UTRs (Figure 1).

Figure 1. The 3’UTRs of CPEB2, CPEB3 and CPEB4 share an ancestral miRNA signature that preceded the generation of highly conserved elements. a, Evolution of the CPEB2 subfamily consisting of two duplications preceding vertebrate speciation. The 3’UTR of the ancestral and current CPEBs for different vertebrates is depicted. The binding sites for different miRNA are indicated. b, Alignment of segments of human CPEB2, CPEB3 and CPEB4 3’UTRs. c, Alignment of the same segment of CPEB2 3’UTRs now between different vertebrates. The binding sites for mir-26 and mir-92 are shown.

3'UTR annotation

The explosion of genomic and transcriptomic data created a bottleneck in the gene annotation pipelines at the level of data analysis. While working on the CPEBs, I noticed that the long 3'UTR isoforms of these and other genes were mis-annotated. Importantly, these isoforms were highly conserved across vertebrates and abundantly expressed in the brain (Figure 2). I then moved on to develop different approaches to annotate 3'UTRs in vertebrates. I found a few hundred mis-annotated genes in mice and humans and a few thousand in other vertebrates². I also showed that long 3'UTRs tend to be misclassified as long non-coding RNAs.

Figure 2. Identification of 3’ ends for transcripts encoding different K⁺ channels. a, Genomic region encoding the 3’UTRs of Kcnq3 (left panel) and Kcnb1 (right panel). The annotated transcripts are shown on top (Ensembl) together with mapping ESTs, the conservation score for the genomic region (PhyloP) and RNASeq reads from brain (B), testis (T) and heart (H). The proposed 3’ ends are indicated with colored arrowheads. b, Northern blots of brain (B), testis (T) and heart (H) using probes for Kcnq3 (left) and Kcnb1 (right) are shown. The bands corresponding to the proposed new 3’ ends are indicated by the arrowheads.

Perspective. In the study of the CPEB transcripts, I showed that miRNA binding sites embedded in highly conserved stretches of 3’UTRs are functional and more interestingly, they also preceded the formation of these highly conserved elements. Also, the annotations that we proposed for conserved 3’UTRs were later validated by others³ and with time incorporated to the commonly used databases.

1.        Morgan, M., Iaconcig, A. & Muro, A. F. CPEB2, CPEB3 and CPEB4 are coordinately regulated by miRNAs recognizing conserved binding sites in paralog positions of their 3’-UTRs. Nucleic Acids Res. 38, 7698–7710 (2010).

2.        Morgan, M., Iaconcig, A. & Muro, A. F. Identification of 3’ gene ends using transcriptional and genomic conservation across vertebrates. BMC Genomics 13, 708 (2012).

3.        Miura, P. et al. Widespread and extensive lengthening of 3 ′ UTRs in the mammalian brain Widespread and extensive lengthening of 3 9 UTRs in the mammalian brain. 812–825 (2013). doi:10.1101/gr.146886.112