Search

Cryo-electron microscopy (cryo-EM) is an indispensable tool for structural biologists, allowing the reconstruction of biomolecules in their native, hydrated state to nanometre, or sub-nanometre, resolution. Unfortunately in situ cryo-EM characterisation is currently limited by our ability to interpret micrographs and tomograms, as a low signal-to-noise ratio means different structures are usually difficult to distinguish. To overcome this limitation, microscopists often use cryo-EM tags: electron-dense, nano-scale objects with specific targeting properties. This thesis presents two related projects involving the use of nucleic acid nanostructures as cryo-EM tags. In Project 1, I introduce functionalised DNA-origami nanostructures as tags. I begin with the cryo-EM characterisation of various nanostructures in cell lysate to assess their imaging properties in a biological context. I find that solid lattice-based nanostructures provide high contrast in this environment. By cryo-EM, the outline of nanostructures is often difficult to visualise, whereas the internal arrangement of helices results in a distinctive `stripy' effect. As a result, honeycomb-lattice origami nanostructures with different shapes can be difficult to distinguish, whereas nanostructures of different lattice types are easily identifiable. Based on this preliminary work, I present a signpost origami nanostructures for tagging, comprising a high-contrast, asymmetric `sign' for identification, and a six-helix bundle `post' for targeting. To functionalise nanostructures for targeting, I have characterised an RNA aptamer which binds green fluorescent protein (GFP) and quantified its binding affinity to GFP analogues. The aptamer is attached to a signpost nanostructure by hybridisation and in situ binding is demonstrated with a murine leukaemia virus test system. High-contrast, high-efficiency labelling of viral glycoproteins is observed. Electroporation of polyamine-stabilised nanostructures into E. coli is also shown for intracellular labelling applications. In Project 2, I present methods for designing 3D RNA nanostructures for co-transcriptional folding. This provides the groundwork for the development of genetically-encoded cryo-EM tags. I firstly introduce a grid-like architecture, where an RNA nanostructure consists of a set of interacting units arranged on a square lattice. Each unit is a branched hairpin structure, with units connected by a path of single-stranded linkers and through kissing interactions between complementary hairpin arms. For this design, six different nanostructures are synthesised. Supermolecular assembly of nanostructures into finite super-structures is demonstrated, as is the incorporation of an aptamer sequence for GFP binding. Finally, I investigate an origami-like route toward RNA nanostructure design, by introducing two new assembly motifs for incorporation into an already-established 2D RNA origami architecture in an effort to build in 3D. My first origami-like design uses paranemic crossovers formed between bulge loops situated on adjacent sheets. I show that these nanostructures do not fold into the intended 3D structure and suggest explanations for this observation. In my second origami-like design, I introduce the combined bulge-loop/kissing-loop motif. Two-sheet nanostructures, held together through these motifs, have been designed. Although a more promising prospect than the previous design, the yield of well-folded structures is low. Based on these observations, the grid-like architecture will provide a more promising launching pad for the development of genetic tags.

© Published under licence by IOP Publishing Ltd. A system has been developed to provide flexible, efficient and robust processing of radiotherapy planning and treatment data collected in the VoxTox project, which investigates differences between planned and delivered dose, and dose-toxicity correlations. This paper outlines the system requirements and implementation, highlighting the use made of software tools and computing models developed for experiments at the Large Hadron Collider. Experience with VoxTox data processing is summarised.

Nuclear waste repositories are being installed in deep excavated rock formations in some places in Europe to isolate and store radioactive waste. In France, the Callovo-Oxfordian formation (COx) is a potential candidate for a nuclear waste repository. The redox reactivity (kinetics and thermodynamic redox potential) of COx clay rock samples are already under study using microscopic, spectrometric and wet analysis techniques. In order to cross and overcome certain limits by improvement in the knowledge, specific electrodes should be constructed and devoted to the deepening of the electrochemical behaviour of the COx system in different situations. Iron is one of the most common redox species in soils and sedimentary rocks. Iron-bearing phyllosilicates play key roles in various biogeochemical processes. The complexity of the physical and chemical (along with structural) changes involving their structural iron makes the studies of its redox properties challenging. Most of the recent reported efforts were focused on probing Fe redox on finely powdered clay (and often micas) particles, and have been hampered by inadequate interactions between particles and electrodes. Moreover, such experiments usually involve redox probe ions, thus adding supplementary difficulties in the determination of structural iron redox parameters such as redox potential (Eh) and kinetics. The present study aims at qualitatively investigating the above mentioned phenomena on minerals like iron-bearing micas. In the current work, we present initial insights regarding efforts to build a direct electrical interface between solid electrodes and conveniently shaped macroscopic mica crystals in order to investigate the redox properties of structural iron in dry and aqueous environments, in the presence of representative perturbations. A classical three electrode system has been used for voltammetric measurements. Platinum plate (1cm²) was the counter electrode. Potentials have been measured against either silver-silver chloride electrode (Ag-AgCl/3M KCl) or Saturated Calomel Electrode (SCE/KClsat) as reference electrodes. Open Circuit Potential (OCP) measurements and Cyclic Voltammetry (CV) were realised as well as Long term OCP measurements, along with pH and other parameters, are also measured. Most of the experiments discussed here have been conducted in unbuffered NaCl or KCl 0.1 M solutions, at 25°C. Anoxic conditions were maintained by first bubbling N2 and then maintaining a gas layer in the head space of the reactor. Long term measurements showed that in these conditions, pH stays at a value around 8.9 and is stable for several weeks. Experiments realised in buffered solution, at pH 7.5, using 1 mM Piperazine-1,4-bis(2-ethanesulfonic acid) (PIPES) and 2 mM NaOH did not show any observable change. In some experiments, 5 mM ferricyanide (Fe(CN)63-) have also been used as a redox probe. High resistivities previously have been reported for this type of material, ranging from 1010 to 1015 Ω.cm at room temperature. Given the current flows detection limit of the apparatus used for measuring (~nA), the mm-thick pieces used in first experiments should normally behave like insulating screens on Ag (or maybe Cu-Ag) electrodes, e.g. no current could be measured across the sample with low iron content. OCP measurement consists in measuring the electrode rest potential against a reference electrode, in absence of observable current. It provides a first indication on the type of material present on the support electrode; normally no potential should be measured in absence of an electrically conductive path. Unfortunately, first records showed that OCP can be measured immediately after the immersion of all the first electrodes prepared (Figure 1, left) even on muscovite and lepidolite, clearly indicating the presence of water in the different samples. However, signals obtained for lepidolite and muscovite are less stable and can only be conveniently recorded in absence of external perturbation around the electrode such as vibration or even air movement, whereas signals obtained for the different biotite samples are very stable. Changing parameters in the electrolyte such as pH (figure 1, right) removal of O2, or change in Cl- does not seems to affect the measured OCP of this type of electrode as it stay stable for days. Several week-longs monitoring on ten biotite electrodes, dipped in unbuffered NaCl 0.1 M, showed that the OCP slowly decreases and stays stable, at values between 0.1 and 0.14 mV/NHE (data not shown), despite occasional pH oscillation between 4 and 9. Results of these experiments show that several aspects of the designs of such electrodes have to be improved for further experiments. From these observations, we can't preclude that water might participate in the electrical contact between the support electrode and the solution, but we can still hypothesize that it could be due to interlayer traces of water, offering a poorly electrically conductive path in lepidolite and muscovite, whereas another mechanism might participate in the biotite sample. First, lowering of the water content in the mineral must be realised, using smaller crystals, longer drying and vacuuming time, and using more penetrating and impregnating resins. Platinum plates and evaporated platinum or carbon are envisioned for the inert support electrode. Using thinned slices should reduce electrical resistivity of the whole bulk structure, allowing a more convenient observation of the coupled electrons transfers that might occur between separated crystal surfaces. Hence, more suitable cutting techniques, such as wire sawing, should be planned to obtain thinner slices. More mica samples must also be selected, with higher and lower iron contents than the one used in the present experiments. This first set of experiments in interfacing solid electrodes and conveniently shaped mica crystals offers hope that this will be a valuable technique for probing structural iron. .

Understanding longitudinal trends in the processing of carbon in rivers represents a much conceptualised, but infrequently tested, issue in aquatic ecology. In this study, we conducted concurrent longitudinal examinations of three very different rivers in eastern Australia to determine whether general principles in river functioning exist across broad geographic and hydrologic scales. Specifically, we examined trends in ambient basic water chemistry, nutrient concentrations, dissolved organic carbon (DOC), extracellular enzymes and food web structure and functioning and conducted bioassays to examine the degree to which DOC and nutrients limit heterotrophic bacterial respiration. These parameters revealed striking similarities across all sites. For metazoan communities, stable isotope analysis showed that algal carbon was the dominant basal resource utilised by consumers in all three rivers, suggesting that in-stream primary producers strongly underpin trophic pathways regardless of the position within a catchment or catchment condition. Analyses of extracellular enzymes revealed that microbial communities are actively utilising DOC at all sites. In fact, heterotrophic microbial respiration was strongly limited by DOC at all sites, with nutrient additions resulting in only relatively minor increases in respiration. Ultimately, this study demonstrates that DOC and algal carbon are critically important drivers of ecosystem processes in Australian riverine ecosystems. Furthermore, across all of our sites and rivers, ambient nutrient concentrations did not influence carbon processing. The consistent longitudinal trends in river function identified in this study provide useful insights for catchment managers and modellers with respect to identifiying key principles that underpin ecosystem functioning in Australian rivers. Copyright © 2009 John Wiley & Sons, Ltd. Copyright © 2009 John Wiley & Sons, Ltd.

International audience

National audience