CAP BINDING PROPERTIES OF POLY(A)-SPECIFIC RIBONUCLEASE (PARN)

Per Nilsson and Niklas Henriksson

The two processes, polyadenylation and deadenylation, regulate the maintenance and length of the poly(A) tail, making the poly(A) tail a highly dynamic structure. Recently, several studies have shown that the poly(A) tail plays important roles during translational initiation and mRNA decay. Poly(A) specific ribonuclease (PARN) is one of the major deadenylases. A unique property of PARN is that it does not only interact with the 3’end located poly(A) tail but it also binds the 5’ end located cap structure of mRNA. The cap structure stimulates deadenylation by PARN and makes the enzyme more processive. We have used fluorescence spectroscopy to identify amino acids involved in cap binding of PARN. We found that two tryptophans were crucial for cap binding. A recently determined NMR structure of PARN(437-529) containing the two tryptophans important for cap binding, revealed that this domain of PARN is folded into a RRM. We found that amino acids important for cap-binding in PARN were located behind the surface of the RRM. We draw the conclusion that cap binding and RNA binding takes place at different positions within the RRM. Further we investigated the RNA binding properties of PARN. PARN(443-560) shows the same substrate length requirement and poly(A)-specificity as full length PARN, implying that the RRM domain harbors these properties. A recent crystal structure of PARN has revealed a dimeric structure of PARN (Wu et al 2006 EMBO J). Our research has at present two objectives based on structural data. 1) To completely identify the cap-binding site, where we will continue the mutational analysis and identify amino acids involved in cap-binding by using fluorescence spectroscopy, 2) To characterise the surfaces of interaction of the different domains in the holo enzyme using in vitro deadenylation assays, gel filtration and RNA binding studies.

HOW POLY(A)-SPECIFIC RIBONUCLEASE (PARN) RECOGNIZES POLY(A)

Niklas Henriksson and Per Nilsson

Poly(A)-specific ribonuclease (PARN) is a 3’ exoribonuclease and one of the few known mammalian nucleases that degrades poly(A) with high specificity. In this project, we are investigating the reasons behind the adenosine specificity of this enzyme. PARN degrades poly(A) during three distinct reaction phases, a fast phase generates A4, a second slower phase occurs between A4-A2 and a third very slow phase from A2-A1. Improved Km for longer substrates is the main reason for this difference in catalytic efficiency. By using a large repertoire of trinucleotide substrates we have investigated how PARN recognizes the different bases in the active site. This analysis revealed that PARN harbors at least three specific binding slots in its active site that together provide specificity for adenosine recognition. We propose that that these three binding slots recognize the 3’ located leaving nucleotide, the penultimate nucleotide and the nucleotide located 5’ of the penultimate nucleotide, respectively. Of the three binding slots the one that interacts with the penultimate nucleotide provides the highest specificity towards recognition of adenosine residues in the active site of PARN. Further we investigate the RNA binding properties of PARN. Gel shift analysis shows that a 20 nt’s long poly(A) molecule binds to PARN with high affinity and that this binding is highly poly(A)-specific. PARN(443-560), which includes a classical RNA-regognition motif (RRM) shows the same substrate length requirement and poly(A)-specificity as full length PARN, implying that the RRM domain harbors these properties. PARN(170-245) where an R3H-domain is located, can also by itself bind RNA. Taken together, our results show that both RNA-binding properties of the RRM and R3H-domains in addition to base recognition in the active site contributes to PARN poly(A)-specificity.

FUNCTIONAL CHARACTERIZATION OF U1-LIKE SNRNPs

Lei Liu Conze, Jens Schuster, Pontus Larsson and Leif Kirsebom

We have detected a surprising heterogeneity among human spliceosomal U1 small nuclear RNA (snRNA). Most interestingly, we have identified three U1 snRNA variants that lack complementarity to the canonical 5’ splice site (5’SS) GU dinucleotide and display features of functional spliceosomal snRNAs, i.e. tri-methylated cap structures, association with Sm proteins and presence in nuclear RNA-protein complexes. However, it is necessary to extend the biochemical characterization beyond these initial observations. To investigate the composition of the U1-like snRNP complexes, two different techniques are used: i) MS2 affinity selection ii) Streptavidin-agarose affinity selection. The first approach utilizes an in vitro transcript of the U1-like snRNA fused to a MS2-hairpin (MS2 coat protein binding site). Complexes that assemble onto this RNA in a nuclear extract can be affinity purified using a fusion protein (MS2-coat protein fused to Maltose-binding-Protein). The second approach utilizes a biotinlyated oligonucleotide complementary to the snRNA for affinity purification. Combined, these experiments will give an insight into the composition of the U1-like snRNPs and allow identification of specific components. To address the functionality of U1-like snRNAs in vivo, a transient transfection assay system is applied. Through our collaboration with Drs. Schaal and Schindler (Germany), a reporter construct with a non-canonical 5’SS is used. Focusing initially on U1A7 snRNA, we investigate whether it can be utilized in vivo to interact with the non-canonical 5’SS and give rise to spliced products. The results from such assays will provide a catalogue of functional snRNA genes. The same assay system can also be used to experimentally verify if bioinformatically identified potential target pre-mRNAs are regulated or dependent on a specific snRNA variant.

EVOLUTIONARY ANALYSIS OF THE SPLICEOSOMAL RNA GENES

Pontus Larsson, Leif A. Kirsebom, David H. Ardell

The human genome is known to encode a large number of genes for spliceosomal small nuclear RNAs (snRNAs). The majority of these genes are believed to be pseudogenes lacking any function. However, we have recently shown that a number of genes encoding variant U1 snRNAs are apparently functional but lack complementarity to the canonical 5’ splice site (5’SS) GU dinucleotide. This unanticipated heterogeneity among spliceosomal snRNAs might reflect a strategy to accommodate a broad variety of 5’SS sequences while assuring correct splicing and could contribute to the complexity of vertebrates by expanding the coding capacity of the genome. Interestingly, evolutionary analyses suggest that the variant U1 snRNA genes have appeared recently in evolution and evolve rapidly from bona fide U1 snRNA genes. The revelation that the variant U1 snRNA loci are conserved within the primate lineage but significantly less conserved between primates and other vertebrate and invertebrate species indicates a link between snRNA heterogeneity and speciation. This project aims at investigating the evolutionary relationships of spliceosomal snRNA gene loci between primate species and other vertebrates and invertebrates. The degree of conservation and the pattern of substitution between orthologous loci will help us to assay the variant snRNA genes for evolutionary constraints and distinguish functional genes from pseudogenes. In conclusion, we hope to provide a broad overview of the spliceosomal snRNA genes in the human genome and their fate during the course of evolution.

RNA DEGRADATION IN Drosophila melanogaster

Susanne Stier

The aim of this project is to study the eukaryotic mRNA degradation machineries in Drosophila melanogaster. Degradation of the mRNA is one important mechanism of gene regulation. Much effort has been placed on detailed studies of isolated components of the mRNA degradation machinery, especially in yeast. In this project we aim to dissect various degradation pathways in Drosophila melanogaster, using techniques such as microarray analysis, RNA interference and various biochemical assays already established in our laboratory. We are currently setting up microarray experiments where we want to find differentially expressed genes in cells in the absence or the presence of key players in the deadenylation machineries. Furthermore, we are planning to perform genetic screens for more candidate factors involved in the degradation machinery using RNAi strategies. Factors believed to be involved in the mRNA degradation pathways will be knocked down one at a time or in combinations in insect cell cultures. Each extract will be assayed for various RNA degradation capabilities. The relationship between components can then be studied by searching for combinations of cell extracts that can rescue inactivated reactions. Components of interest will be further studied in vivo in transgenic flies.

SMALL LIGANDS AND RNA

Shiying Wu, Ema Kikovska, Nils-Egil Mikkelsen and Anders Virtanen

The recognition that RNA molecules can adopt very complex three-dimensional structures has made it a potential drug target. Aminoglycosides interact with various functionally important RNA molecules and inhibit their function. Available data suggest that binding of aminoglycosides to RNA results in displacement of divalent metal ions that are essential for function. We have demonstrated that the function of the essential and ubiquitous endoribonuclease RNase P is inhibited by various aminoglycosides. Our data also suggest that, in keeping with other systems, aminoglycosides bind to RNase P RNA and thereby displace Mg2+ ions that are important for function. In many cases Mg2+ and Pb2+ compete for the same binding site. Under appropriate conditions Pb2+ binding to RNA results in cleavage of the phosphodiester backbone at specific positions. We have taken advantage of this and demonstrated that aminoglycosides (and other inhibitors of RNA function) also interferes with Pb2+ induced cleavage of RNA. To understand the interaction between RNase P RNA and various inhibitors in more detail we are currently using small model RNA molecules representing functionally important domains of RNase P RNA and other RNA molecules to probe binding of inhibitors using various biochemical protocols and X-ray crystallography. For example, we have recently obtained yeast tRNAPhe tobramycin co-crystals that diffracts to 2.1Å. To conclude, our data clearly show that binding of an inhibitor to an RNA indeed can result in displacement of a divalent metal ion in vitro and we are in the process to study whether some of these inhibitors inhibits bacterial growth.

This research is financially supported by NordForsk, The Swedish Research Council and The Swedish Research Council to Uppsala RNA Research Center in the form of Linnéstöd.