Long noncoding RNAs with snoRNA ends

2012-10-26 00:01:46

Molecular Cell; 26 Oct 2012;48(2):219-30; DOI:10.1016/j.molcel.2012.07.033

Qing-Fei Yin, Li Yang, Yang Zhang, Jian-Feng Xiang, Yue-Wei Wu, Gordon G. Carmichael, Ling-Ling Chen


We describe the discovery of sno-lncRNAs, a class of nuclear-enriched intron-derived long noncoding RNAs (lncRNAs) that are processed on both ends by the snoRNA machinery. During exonucleolytic trimming, the sequences between the snoRNAs are not degraded, leading to the accumulation of lncRNAs flanked by snoRNA sequences but lacking 5' caps and 3' poly(A) tails. Such RNAs are widely expressed in cells and tissues and can be produced by either box C/D or box H/ACA snoRNAs. Importantly, the genomic region encoding one abundant class of sno-lncRNAs (15q11-q13) is specifically deleted in Prader-Willi Syndrome (PWS). The PWS region sno-lncRNAs do not colocalize with nucleoli or Cajal bodies, but rather accumulate near their sites of synthesis. These sno-lncRNAs associate strongly with Fox family splicing regulators and alter patterns of splicing. These results thus implicate a previously unannotated class of lncRNAs in the molecular pathogenesis of PWS.


Small nucleolar RNAs (snoRNAs) are a family of conserved nuclear RNAs (about 70-200 nt) that are usually concentrated in Cajal bodies or nucleoli where they either function in the modification of snRNAs or rRNA, or participate in the processing of rRNA during ribosome subunit maturation. There are several hundred known cellular snoRNAs and the great majority of these are encoded in the introns of protein-coding genes. SnoRNAs are processed from excised and debranched introns by exonucleolytic trimming and carry out their functions in complex with specific protein components, forming ribonucleoprotein complexes (snoRNPs). There are two main classes of snoRNAs: box C/D snoRNAs and box H/ACA snoRNAs, both of which serve as guide RNAs complementary to specific target sequences. Box C/D snoRNAs guide 2'-O-ribose methylation and box H/ACA snoRNAs guide pseudouridine modifications. Box C/D snoRNAs contain four conserved sequence elements: box C (RUGAUGA) and box D (CUGA) near the 5' and 3' termini, respectively, and an internal copy of each box, called C' and D'.

Some snoRNAs expressed from an imprinted region have been implicated in an important human disease, Prader-Willi Syndrome (PWS). While most autosomal genes are expressed biallelically, genomic imprinting causes some to be expressed only from either the paternal or maternal chromosome. The 15q11-q13 region is imprinted, leading to expression of the SNURF-SNRPN gene and downstream noncoding region from the paternal chromosome, while the UBE3A gene is expressed biallelically in most cells, but only from the maternal chromosome in neurons. The long neuron-specific paternal transcript contains at least 148 exons (which along with their introns appear to have arisen by duplications followed by sequence divergence) and spans 470 kb. All paternal transcripts downstream of the SNRPN gene are noncoding and have been considered primarily as precursors for small RNAs. Since the minimal paternal deletion region associated with PWS (108 kb) removes only SNORD109A, the SNORD116 cluster of 29 similar snoRNAs and IPW, the most current published model is that SNORD116 deficiency is the primary cause of disease. The function of SNORD116s is unknown. While most snoRNAs exhibit complementarity to rRNA or snRNA targets, it is noteworthy that even after numerous attempts, no noncoding RNA, pre-mRNA or mRNA targets for any of the SNORD116 snoRNAs from 15q11-q13 have been confirmed. These snoRNAs show minimal complementarity to rRNA or other RNAs so far and show greater homology to one another than to other snoRNAs. A few potential targets have been predicted bioinformatically, but none has been validated. The human and mouse SNORD116s are similar but not identical (even in putative targeting regions), but their deficiency in the mouse recapitulates some of the features of PWS. This has led to some confusion in the field as to what the precise molecular cause of PWS is.

Increasing lines of evidence have demonstrated that long noncoding RNAs (lncRNAs) are involved in gene regulation and human diseases. The majority of lncRNAs are transcribed from intergenic regions, promoters and enhancers. Few, however, have been shown to derive from introns. Such molecules would be expected to lack both 5' cap structures and 3' poly(A) tails. We report here the discovery of a class of lncRNAs whose ends correspond to positions of intronic snoRNAs and which we have named snoRNA-related lncRNAs (sno-lncRNAs). We characterize the sno-lncRNAs from the PWS critical region of chr15 and demonstrate that at least some of these associate strongly with the splicing factor Fox2 and can alter splicing patterns in cells. Thus, these lncRNAs are implicated in the pathogenesis of an important human disease.

To Read Full Access, Click Here


Intron-derived sno-lncRNAs are processed on both ends by the snoRNA machinery | Sno-lncRNAs are functionally distinct from classical snoRNAs | The genomic region encoding the most abundant sno-lncRNAs is associated with PWS | PWS region sno-lncRNAs interact with Fox2 and alter patterns of splicing