Search

Accurate extraction of multiplet parameters, such as J-couplings and chemical shifts, play a vital role in small molecule analysis using nuclear magnetic resonance (NMR) spectroscopy. These parameters provide essential quantitative information about molecular structures, interatomic interactions, and chemical environments, enabling precise characterization of small organic compounds. This poster presents an innovative omputational approach that utilizes state-of-the-art deep learning techniques, specifically detection transformers, to automate and optimize the extraction of multiplet parameters from 1D NMR spectra of small molecules. By incorporating these advanced computational methods, experimenters can achieve improved efficiency, accuracy, and speed in analyzing and characterizing small organic compounds using NMR spectroscopy.

RNA, widely recognized as an information-carrier molecule, is capable of catalyzing essential biological processes through ribozymes. Despite their ubiquity, specific functions in a biological context and phenotypes based on the ribozymes' activity are often unknown. Here, we present the discovery of a subgroup of minimal HDV-like ribozymes, which reside 3' to viral tRNAs and appear to cleave the 3'-trailers of viral premature tRNA transcripts. This proposed tRNA-processing function is unprecedented for any ribozymes, thus, we designate this subgroup as theta ribozymes. Most theta ribozymes were identified in Caudoviricetes bacteriophages, the main constituent (>90%) of the mammalian gut virome. Intriguingly, our findings further suggest the involvement of theta ribozymes in the transition of certain bacteriophages between distinct genetic codes, thus possibly contributing to the phage lysis trigger. Our discovery expands the limited repertoire of biological functions attributed to HDV-like ribozymes and provides insights into the fascinating world of RNA catalysis.

Trillions of microorganisms, collectively known as the microbiome, inhabit our bodies with the gut microbiome being of particular interest in biomedical research. Bacteriophages, the dominant virome constituents, can utilize suppressor tRNAs to switch to alternative genetic codes (e.g., the UAG stop-codon is reassigned to glutamine) while infecting hosts with the standard bacterial code. However, what triggers this switch and how the bacteriophage manipulates its host is poorly understood. Here, we report the discovery of a subgroup of minimal hepatitis delta virus (HDV)-like ribozymes - theta ribozymes - potentially involved in the code switch leading to the expression of recoded lysis and structural phage genes. We demonstrate their HDV-like self-scission behavior in vitro and find them in an unreported context often located with their cleavage site adjacent to tRNAs, indicating a role in viral tRNA maturation and/or regulation. Every fifth associated tRNA is a suppressor tRNA, further strengthening our hypothesis. The vast abundance of tRNA-associated theta ribozymes - we provide 1753 unique examples - highlights the importance of small ribozymes as an alternative to large enzymes that usually process tRNA 3'-ends. Our discovery expands the short list of biological functions of small HDV-like ribozymes and introduces a previously unknown player likely involved in the code switch of certain recoded gut bacteriophages.

Holliday 4-way junctions are key to important biological DNA processes (insertion, recombination, and repair) and are dynamic structures that adopt either open or closed conformations, the open conformation being the biologically active form. Tetracationic metallo-supramolecular pillarplexes display aryl faces about a cylindrical core, an ideal structure to interact with open DNA junction cavities. Combining experimental studies and MD simulations, we show that an Au pillarplex can bind DNA 4-way (Holliday) junctions in their open form, a binding mode not accessed by synthetic agents before. Pillarplexes can bind 3-way junctions too, but their large size leads them to open up and expand that junction, disrupting the base pairing, which manifests in an increased hydrodynamic size and lower junction thermal stability. At high loading, they rearrange both 4-way and 3-way junctions into Y-shaped forks to increase the available junction-like binding sites. Isostructural Ag pillarplexes show similar DNA junction binding behavior but lower solution stability. This pillarplex binding contrasts with (but complements) that of metallo-supramolecular cylinders, which prefer 3-way junctions and can rearrange 4-way junctions into 3-way junction structures. The pillarplexes' ability to bind open 4-way junctions creates exciting possibilities to modulate and switch such structures in biology, as well as in synthetic nucleic acid nanostructures. In human cells, the pillarplexes do reach the nucleus, with antiproliferative activity at levels similar to those of cisplatin. The findings provide a new roadmap for targeting higher-order junction structures using a metallo-supramolecular approach, as well as expanding the toolbox available to design bioactive junction binders into organometallic chemistry.

We introduce a deep learning-based deconvolution approach for 1H NMR spectra, developed by leveraging concepts from the field of physics informed-learning, intelligent labeling, and tailored high dynamic range (HDR) spectral preprocessing. Since automation and faster workflows are major concerns in NMR spectroscopy, the algorithm handles uncorrected spectra without strict assumptions on phase and baseline correction as well as line shape. Due to the lack of high quality and consistently labeled experimental spectra in quantities needed to train modern deep learning models, we relied on synthetic spectra creation. Moreover, instead of training with synthetic spectra consisting of single lines, we created synthetic multiples that further supported a realistic deconvolution. We achieved super-human performance on corrected and uncorrected synthetic spectra. Finally, and most importantly, the results on synthetic data translate well to experimental spectra despite the covariate shift. Thus, this tool is a promising candidate for automated expert-level deconvolution of experimental HDR 1H NMR spectra.

Riboswitches are structural elements of mRNA involved in the regulation of gene expression by responding to specific cellular metabolites. To fulfil their regulatory function, riboswitches prefold into an active state, the so-called binding competent form, that guarantees metabolite binding and allows a consecutive refolding of the RNA. Here, we describe the folding pathway to the binding competent form as well as the ligand free structure of the moaA riboswitch of E. coli. This RNA proposedly responds to the molybdenum cofactor (Moco), a highly oxygen-sensitive metabolite, essential in the carbon and sulfur cycles of eukaryotes. K+- and Mg2+-dependent footprinting assays and spectroscopic investigations show a high degree of structure formation of this RNA already at very low ion-concentrations. Mg2+ facilitates additionally a general compaction of the riboswitch towards its proposed active structure. We show that this fold agrees with the earlier suggested secondary structure which included also a long-range tetraloop/tetraloop-receptor like interaction. Metal ion cleavage assays revealed specific Mg2+-binding pockets within the moaA riboswitch. These Mg2+ binding pockets are good indicators for the potential Moco binding site, since in riboswitches, Mg2+ was shown to be necessary to bind phosphate-carrying metabolites. The importance of the phosphate and of other functional groups of Moco is highlighted by binding assays with tetrahydrobiopterin, the reduced and oxygen-sensitive core moiety of Moco. We demonstrate that the general molecular shape of pterin by its own is insufficient for the recognition by the riboswitch.

Single-molecule microscopy is often used to observe and characterize the conformational dynamics of nucleic acids (NA). Due to the large variety of NA structures and the challenges specific to single-molecule observation techniques, the data recorded in such experiments must be processed via multiple statistical treatments to finally yield a reliable mechanistic view of the NA dynamics. In this chapter, we propose a comprehensive protocol to analyze single-molecule trajectories in the scope of single-molecule Förster resonance energy transfer (FRET) microscopy. The suggested protocol yields the conformational states common to all molecules in the investigated sample, together with the associated conformational transition kinetics. The given model resolves states that are indistinguishable by their observed FRET signals and is estimated with 95% confidence using error calculations on FRET states and transition rate constants. In the end, a step-by-step user guide is given to reproduce the protocol with the Multifunctional Analysis Software to Handle single-molecule FRET data (MASH-FRET).

Fast and efficient site-specific labeling of long RNAs is one of the main bottlenecks limiting distance measurements by means of Förster resonance energy transfer (FRET) or electron paramagnetic resonance (EPR) spectroscopy. Here, we present an optimized protocol for dual end-labeling with different fluorophores at the same time meeting the restrictions of highly labile and degradation-sensitive RNAs. We describe in detail the dual-labeling of a catalytically active wild-type group II intron as a typical representative of long functional RNAs. The modular procedure chemically activates the 5′-phosphate and the 3′-ribose for bioconjugation with a pair of fluorophores, as shown herein, or with spin labels. The mild reaction conditions preserve the structural and functional integrity of the biomacromolecule and results in covalent, dual-labeled RNA in its pre-catalytic state in yields suitable for both ensemble and single-molecule FRET experiments.

This chapter focuses on the interaction of metal ions mainly with RNA. It is an update of the previous Chapter 3.21 in the 2nd Edition of Comprehensive Inorganic Chemistry II (2013) but focusing solely on RNA. Metal ions are key to folding, structure, and function of any nucleic acid. These interactions are generally of a weak and highly dynamic nature as they concern mostly K+ and Mg2+ in living organisms. Aside from the large excess of loosely bound ions for charge compensation, a network of inner-sphere and outer-sphere interactions holds more specifically bound ions in place. Hence, metal ion binding to larger RNAs is rather complicated and has many facets. After a few general considerations on the basic properties of metal ions and the potential coordination sites on the RNA, the thermodynamics of metal ion binding to RNA and known metal ion binding motifs in RNA are described. This is followed by today's knowledge on the role of metal ions in folding, dynamics, sensing, and/or catalysis of riboswitches and ribozymes, respectively, is summarized.

Single-molecule FRET (smFRET) is a versatile technique to study the dynamics and function of biomolecules since it makes nanoscale movements detectable as fluorescence signals. The powerful ability to infer quantitative kinetic information from smFRET data is, however, complicated by experimental limitations. Diverse analysis tools have been developed to overcome these hurdles but a systematic comparison is lacking. Here, we report the results of a blind benchmark study assessing eleven analysis tools used to infer kinetic rate constants from smFRET trajectories. We test them against simulated and experimental data containing the most prominent difficulties encountered in analyzing smFRET experiments: different noise levels, varied model complexity, non-equilibrium dynamics, and kinetic heterogeneity. Our results highlight the current strengths and limitations in inferring kinetic information from smFRET trajectories. In addition, we formulate concrete recommendations and identify key targets for future developments, aimed to advance our understanding of biomolecular dynamics through quantitative experiment-derived models.

Single-molecule FRET (smFRET) is a versatile technique to study the dynamics and function of biomolecules since it makes nanoscale movements detectable as fluorescence signals. The powerful ability to infer quantitative kinetic information from smFRET data is, however, complicated by experimental limitations. Diverse analysis tools have been developed to overcome these hurdles but a systematic comparison is lacking. Here, we report the results of a blind benchmark study assessing eleven analysis tools used to infer kinetic rate constants from smFRET trajectories. We test them against simulated and experimental data containing the most prominent difficulties encountered in analyzing smFRET experiments: different noise levels, varied model complexity, non-equilibrium dynamics, and kinetic heterogeneity. Our results highlight the current strengths and limitations in inferring kinetic information from smFRET trajectories. In addition, we formulate concrete recommendations and identify key targets for future developments, aimed to advance our understanding of biomolecular dynamics through quantitative experiment-derived models.