Prader Willi Syndrome (PWS) is a congenital disease that results from the loss of paternal expression of
maternally imprinted genes located on chromosome 15q11-13. It manifests itself as excessive appetite
and life threatening obesity as well as mental and growth retardation, symptoms that are due to a
hormonal imbalance. The only therapeutic approach that is so far available, administration of growth
hormone, addresses only one phenotypic aspect of the syndrome and is therefore insufficient. The
molecular mechanism behind PWS is unknown. The affected genetic locus comprises several protein
coding genes along with multi-copy small nucleolar RNAs (snoRNAs). Recent studies of individual
patients demonstrated that the snoRNAs of this locus play a crucial role in the development of PWS. The
mechanism is unclear, because in contrast to the majority of snoRNAs that guide modification of
ribosomal RNAs through base complementarity, the snoRNAs expressed from the Prader-Willi locus are
not complementary to rRNAs or other typical snoRNA targets. Rather, the HBII-52 snoRNA bears a
striking 18nt-long complementarity to the serotonin receptor 2C mRNA, which, given that both the
snoRNA and the putative target have brain-specific expression, strongly suggests that loss of HBII-52
results in an aberrant expression of serotonin receptor. Studies in patients with PWS indicate however that
the loss of HBII-52 expression is not the major cause of PWS. The finding that snoRNAs can be further
processed in various smaller RNAs, including miRNAs suggests new directions of research but also
complicates the search for targets of the “orphan snoRNAs” encoded on chromosome 15q11-13 because
it becomes unclear which subsequence of the snoRNA recognizes the target(s).
Understanding the molecular function of the snoRNAs implicated in PWS remains a crucial bottleneck in
developing appropriate therapeutics. Here we propose to tackle this problem taking advantage of
recent advances in techniques like Crosslinking and Immunoprecipitation (CLIP). In a few previous
studies, we showed that CLIP (especially its photoreactive ribonucleotide-enhanced variant called PARCLIP)
performed on Argonaute proteins identifies both guide microRNAs and their mRNA targets. We
therefore generated preliminary PAR-CLIP data on core proteins of snoRNP in human embryonic kidney
cells (HEK 293) to demonstrate that the technique enables us to recover both snoRNAs and their targets.
We now propose to make further use of this method to identify targets of the brain specific-snoRNAs that
are encoded in the PWS locus, all of which are C/D box snoRNAs that associate with the core proteins
NOP56, NOP58 and Fibrillarin. We will crosslink fine microsections of mouse brain, perform
CLIP/PAR-CLIP on various snoRNA core proteins and sequence the protein-bound RNA fragments.
Through analysis of the high throughput sequencing data with advanced bioinformatics tools that were
previously developed by our group, we will identify antisense regions corresponding to the PWS
snoRNAs. Putative targets will be subjected to functional validation such as primer extension assays to
identify sites of snoRNA-guided ribose sugar modifications. The major advantage of the CLIP approach
relative to most others employed so far is that it does not make strong assumptions about the nature of the
possible targets. We will therefore be able to identify targets other than the conventional rRNA or
snoRNAs, in which case we will have to adapt the functional assays to the nature of the identified targets.
This approach may not work when the snoRNAs act through smaller, processed, functional variants
called snoRNA-derived RNAs (sdRNAs) that may associate with completely different proteins (e.g.
hnRNPs and Argonautes). We will pursue this alternative first performing paired-end sequencing of small
(up to 200 nucleotides) RNAs derived from mouse brain, which will allow us to identify sdRNAs derived
from the PWS locus. These sdRNAs will be chemically synthesized and biotin tagged to facilitate pull
down of associated proteins, which will then be identified with mass spectrometry. These proteins can
then be targeted by CLIP. Alternatively, synthetic sdRNAs will be psoralen modified to facilitate
sdRNA:target RNA crosslinking and facilitate target identification. Appropriate downstream functional
assays will be employed to confirm the regulatory role of sdRNAs. Thus, through novel, unbiased,
genome-wide approaches, our project aims to uncover the molecular underpinnings of PWS, with the
hope that these will pave the way to new therapeutic approaches.