Our Research is funded by:     Der Wissenschaftsfonds.         European Community


The activities in the laboratory are centered around RNA. Our goal is to understand how RNA adopts its structure to achieve its function and how RNA interacts with other molecules like metal ions, proteins and antibiotics. Our model molecules are the self-splicing group I intron from phage T4 (the td ribozyme), in vitro selected RNA aptamers, small non-coding RNAs and small synthetic hairpins. Lately, we became interested in loop-loop interactions as a basic system to study RNA-RNA interactions.

Group I intron splicing  | Folding of the td intron in vivo  | Proteins with in vivo RNA-chaperone activity |
Loop-loop interactions and small non coding RNAs  | RNA/Antibiotic Interactions |
Aminoglycosides as inhibitors of ribozymes  | In vitro Selection of Antibiotic Binding Aptamers |
StreptoTag: a novel method for the isolation of RNA-binding proteins  |

   Group I intron splicing          

EMBO cover Many group I introns are able to self-splice in vitro in the absence of proteins. To be able to do so, the pre-RNA has to fold into a specific 3D structure with the help of magnesium ions. The splicing process of group I introns is initiated by the guanosine cofactor which binds to the core of the intron. To study the splicing reaction of the td intron in more detail, we developed a trans-cleaving ribozyme system. We compared in vitro activities of mutant ribozymes with in vivo splicing activities and found that the main problem of the td intron is to define its 5 splice-site.

Most catalytic RNAs depend on divalent metal ions for folding and catalysis. For a thorough structure-function analysis it is therefore essential to identify their metal ion binding sites. Several divalent metal ions (Ca2+, Sr2+ and Pb2+) do not promote splicing, but instead induce cleavage at a single site in the conserved group I intron core at elevated pH, generating products with 5 OH and 3 phosphate ends. Based on the core cleavage position and on the previously proposed coordination sites for Mg2+ we propose a structural location for two metal ions surrounding the splice site in the Michel-Westhof 3D model of the group I intron core (Streicher et al., 1996). The proposed location was strengthened by a first mutational analysis which supported the suggested interaction between one of the metal ions and the bulged residue in P7. We also used metal-hydroxide mediated RNA cleavage and Fenton chemistry to probe for metal ion binding sites in the group I intron RNA (Berens et al., 1998) and we evaluated uranyl-photocleavage as a probe for monitoring ion binding and flexibility in RNAs (Wittberger et al., 2000).

   Folding of the td intron in vivo          

Although many group I introns self-splice in vitro, it is generally assumed that in vivo splicing is facilitated by proteins. A specific in vivo splicing factor for the self-splicing td intron from the T4 phage thymidylate synthase gene has not yet been found, and attempts to isolate trans-active splicing defective mutants in E. coli failed.

We are studying the folding pathway of the td intron in vivo. We observed, that in vivo, splicing is dependent on translation. Splicing of the td intron was uncoupled from translation by introducing stop codons in the upstream exon. This resulted in splicing deficiency. Over expression of a UGA suppressor tRNA partially rescued splicing, suggesting that this in vitro self-splicing intron requires translation for splicing in vivo. Inhibition of translation by the antibiotics chloramphenicol and spectinomycin also resulted in splicing deficiency. Ribosomal protein S12, a protein with RNA chaperone activity, and CYT-18, a protein which stabilizes the 3D-structure of group I introns, efficiently rescued the stop codon mutants. We identified a region in the upstream exon, which interferes with splicing. Point mutations in this region partially alleviate the effect of a nonsense codon. We infer from these results that the ribosome acts as an RNA chaperone to facilitate proper folding of the intron (Semrad & Schroeder, 1998 and Waldsich et al., 1998).

Katharina's cartoon

   Proteins with in vivo RNA-chaperone activity          

RNAs have fundamental folding problems. These arise from their tendency to form alternative base pairings kinetically trapping the molecules in inactive conformations. This problem can be solved with the aid of RNA-binding proteins. Two classes of such proteins, based on their binding specificities, have been identified:

We are studying the activity of proteins with RNA chaperones, in vivo. Many proteins have been found which exert strand annealing and strand exchange activities and which accelerate the folding of ribozymes in vitro. Here some examples for proteins which help RNA to fold: the ribosomal protein S12 is involved in the fidelity of the decoding process; the A1 hnRNP binds to pre-RNAs in eukaryotic cells before their assembly with the spliceosome; the Ncp7 protein of HIV is a component of the nucleocapsid of the virus; some cold shock proteins of Bacillus subtilis facilitate initiation of translation.

We identified an interaction between the upstream exon of the td gene with nine bases at the very 3' end of the intron. This nine base pair interaction represents a folding trap, which has to be overcome for correct intron folding. Proteins with RNA chaperone activity are able to resolve or impede the formation of this trap in vivo. We are thus able to assay RNA chaperone activity in vivo, by co-expressing the td gene containing a stop codon and a putative RNA chaperone. We defined the chaperone activity as that one which rescues the folding trap but not cis-acting intron mutants with defect structure (Clodi et al., 1999).

   Loop-loop interactions and small non coding RNAs          

RNA secondary structure prediction is based on thermodynamic parameters of small RNA duplexes. To extend the structure prediction to three-dimensional structures, we are currently systematically assessing thermodynamic parameters of loop-loop interactions in bimolecular systems. Loop-loop interactions, also termed kissing loops, play a fundamental role in the process of RNA annealing. The formation of RNA duplexes in the cell is difficult, due to the tendency of RNA not to be single stranded. To achieve efficient RNA annealing, the process is often initiated from complementary hairpin loops. RNA annealing is part of many processes like HIV dimerization and of anti-sense RNA strategies in the prokaryotic translational control. Small non-coding RNAs in E.coli, also termed riboregulators, anneal to their target sequence in a yet unknown mechanism. We are interested in the proteins involved in this process.

   RNA/Antibiotic Interactions          

Antibiotics are produced by microorganisms as products of secondary metabolism. The versatility of secondary metabolites is extremely large and we must assume that we know only a small fraction of all of these compounds. Most RNA-binding antibiotics target the ribosomal RNA. Some of these drugs, which had long been thought to act specifically on the ribosome, interfere with other functional RNAs as well. Among these alternative antibiotic target sites are the self-splicing group I introns, the hammerhead ribozyme derived from the Avocado Sunblotch Viroid, the ribozyme derived from the human Hepatitis Delta Virus, RNAse P and the HIV RNA. For reviews see Wallis & Schroeder, 1997, Schroeder et al., 2000, or Walter et al., 1999.

   Aminoglycosides as inhibitors of ribozymes          

Neomycin in the intron core The aminoglycoside antibiotic neomycin B induces misreading of the genetic code during translation and inhibits several ribozymes. We analysed the mode of binding and inhibition of the self-splicing T4 phage derived thymidylate synthase (td) group I intron (von Ahsen et al.,1991). Footprinting experiments identified two major neomycin B binding sites: one in the internal loop between the P4 and P5 stems and the other in the catalytic core surrounding the G-binding site (von Ahsen & Noller, 1993). Mutational analyses defined the latter as the inhibitory site. Splicing inhibition is strongly dependent on pH and Mg2+ concentration suggesting electrostatic interactions and interference with divalent metal ions. Fe-OH radicals (Berens et al., 1998) were used to monitor changes in the surroundings of the metal ions. Neomycin B protected several positions in the catalytic core from Fe-OH radical cleavages suggesting that metal ions are displaced in the presence of the antibiotic. Mutation of the bulged nucleotide in the P7 stem, a position which is strongly protected by neomycin B from Fe-OH radical cleavage and which has been proposed to contact an essential metal ion, renders splicing resistant to neomycin B. These results allowed the docking of neomycin to the core of the group I intron (Hoch et al., 1998).

   In vitro Selection of Antibiotic Binding Aptamers          

Neomycin aptamer (NMR) For a better understanding of the mode of action and binding of antibiotics to RNA, we set out to in vitro select small antibiotic-binding aptamers. So far, we have isolated and partially characterized aptamers, which bind neomycin B (Wallis et al., 1995), viomycin (Wallis et al., 1997), streptomycin (Wallace & Schroeder, 1998), and tetracycline (Berens et al., 2001). Although the starting selection pool was totally random and the isolated molecules had no sequence homologies to the natural target sites, each selection revealed basic features of the binding mode of each drug. The aminoglycoside neomycin B binds to RNA stem-loops, whose stems contain widened major grooves and the peptide viomycin has a strong specificity for RNA pseudoknots.
To see the NMR structure of a neomycin B aptamer complexed with the antibiotic, go to Jiang at al. 1999.

   StreptoTag: a novel method for the isolation of RNA-binding proteins          

The streptomycin-binding aptamer used for the development of a protein isolation and purification method. We developed a fast and simple one-step affinity-purification method for the isolation of specific RNA-binding proteins. An in vitro-transcribed hybrid RNA consisting of an aptamer sequence with high binding specificity to the antibiotic streptomycin and a putative protein-binding RNA sequence is incubated with crude extract. After complex formation, the sample is applied to an affinity column containing streptomycin immobilized to Sepharose. The binding of the in vitro-assembled RNA-protein complex to streptomycin-Sepharose is mediated by the aptamer RNA and the specifically bound proteins are recovered from the affinity matrix by elution with the antibiotic (Bachler et al., 1999).

Streptotag method