Translation initiation with 70S ribosomes : a single molecule study

Until now, three initiation pathways for ribosomal protein synthesis have been described: the 30S-binding mechanism of canonical mRNAs, which assumes that the first step is binding of the small subunit to the start codon of the mRNA with the help of initiation factors, the rare initiation of leaderless mRNAs by full 70S ribosomes, and the 70S-scanning mode through which the second cistron of a polycistronic mRNA is translated. According to the standard model from classical textbook, translation of canonical mRNAs occurs exclusively via the 30S-binding mechanism, which requires dissociation of subunits. In contrast to this, we demonstrated here using an in vitro transcription-translation system that canonical mRNAs can be also initiated by full 70S ribosomes, without dissociation. The standard and non-standard initiation events were quantified with single molecule experiments, since only by tracking one particle at a time one can identify and count parallel mechanistic initiation pathways.To perform single molecule experiments, the sample preparation had first to be optimized by establishing the protocols for reassociation of ribosomes, biotinylation and labeling. The reassociation procedure involves purification of tightly-coupled 70S from E. coli, dissociation into subunits and reassociation to form 70S monosomes, and it leads to a highly pure sample, in which the endogenous bacterial mRNAs and factors are discarded. Concerning the biotinylation, two approaches are presented: on the one hand, in vivo biotinylation of the uL4 ribosomal protein, and on the other hand, binding of a biotinylated oligonucleotide to an RNA loop engineered on the small subunit. The first approach lead to a 1/4 decrease in activity of the modified ribosomes, in contrast to the second method, which did not affect the activity of the ribosomes. In addition to this, ribosomes were fluorescently labeled for single molecule experiments, using two strategies: unspecific labeling of accessible lysines with NHS dyes and labeling via oligonucleotide binding on engineered loops of rRNA. In both cases, the samples were characterized in regard to the labeling ratio, presence of aggregates and free dye, and activity. The first method allowed multilabeling of ribosomes, but a 1/3 decrease in activity was observed, while the second approach yielded site-specifically single labeled ribosomes with an activity comparable to the wild type. The second step was improving the stalling properties of the DNA used for GFPem synthesis. A poor stalling efficiency would release the GFPem from the ribosomal complexes, so it would not be detected on a wide field. Moreover, the ribosomal subunits would dissociate, meaning that the 70S which initiated translation would not be identified. Previous studies used the secretion monitor (SecM) arrest sequence to stall ribosome nascent chains. However, the SecM showed unsatisfactory stalling characteristics in our experiments, considering that most of the translated protein was released from the ribosome. Therefore, a new SecM arrest sequence was designed (SecMstr), that was proven to have enhanced stalling properties, achieving an almost absolute stalling of the ribosomes (>95%). The good stalling properties allowed also the quantification of the active fraction of ribosomes in an in vitro transcription-translation kit and the average number of productive cycles the ribosomes can perform in these conditions. Moreover, the techniques for single molecule experiments were optimized. In order to limit the bleaching and blinking of the fluorophores, two oxygen scavenging systems were studied: the protocatechuic acid/protocatechuate-3,4-dioxygenase (PCA/PCD) system in combination with Trolox, and the Glucose oxydase/catalase (Gox) system. The first one considerably improved the detection of Cy5, both in Two-Colour Coincidence Detection (TCCD) and wide field measurements. For specific tethering of biotinylated ribosomes on modified glass slides, a protocol for surface coating using a mixture of PEG and bioPEG has been developed, focusing on obtaining a low fluorescence background, good blocking properties and maintaining the activity of the immobilized ribosomes. After achieving an optimized system, the study on translation initiation was performed, using site-specifically single labeled reassociated ribosomes, and the canonical mRNA containing the SecMstr in order to stall the produced GFPem on the ribosomes. The initiation modes were quantified with TCCD experiments, demonstrating that from the total successful translation events, 30% occurred by initiation with 70S, and 70% occurred via the 30S-binding mode. This is the first study that investigates and quantifies initiation of canonical mRNAs by full 70S ribosomes, without dissociation into subunits. Furthermore, the optimized system presented here can be used in wide field measurements to get insights into the kinetics of the processes involved in this non-standard initiation mechanism. Moreover, optimization of the stalling properties of the SecM arrest sequence in combination with single molecule studies offers a broad spectrum of further studies on transcription, translation and co-translational protein folding