Periodic Reporting for period 1 - ECONOMICS (In cellular conformations of microRNA-34a via magnetic resonance)
Reporting period: 2023-01-01 to 2026-04-30
Summary of the context and overall objectives of the project
microRNAs (miRNA) are short 20-22 nucleotides RNA that regulates cellular biochemistry and their regulation in turn is crucial to guard against diseases including various cancers. Since their discovery in 1990s and four Nobel prizes later, a mechanistic understanding of miRNA action is beginning to be unravelled. A prevalent mechanism suggests that miRNA guides the human Argonaute protein to a location on messenger RNA (mRNA) and typically halts protein production. Conventionally, miRNA locates it target mRNA but finding complementary sequence to its initial 2-7 nucleotides, known as the “seed” region. For the past five years many research including those from the host laboratory, have shown that regions beyond the seed plays a crucial role in miRNA-based mRNA regulation. Crystallography and Cryo-electron microscopy have provided crucial information on the organisation of miRNA within Argonaute, however the regions beyond the seed remains poorly defined due to its inherent dynamics. NMR spectroscopy can bridge this knowledge gap which leads to the first objective of this project; to elucidate the structure and dynamics of the entire miRNA in Argonaute both in presence and absence of a known mRNA target focussing on miRNA-34a (miR-34a).
Among the various disease relevant miRNA, miR-34a have been implicated in 50% of cancers and is being considered as potential cancer therapeutics. A miR-34a mimic was tested against last-stage hepatocellular carcinoma patients where partial reemission of the disease showed promise for future development. This clinical trial was ultimately halted due to high mortality which is attributed to the ability of miR-34a mimic targeting multiple mRNA, known as off-targeting phenomenon. The unmodified miR-34a is known to have ~100 validated targets which poses a challenge to design safe, precise and effective miRNA-based therapeutics. The first objective will aid in addressing this challenge where rules of miRNA binding to mRNA can be expanded to include miRNA regions beyond the seed. It has been shown by the host laboratory that in miR-34a, a nucleotide at position 8 can dynamically become part of the seed albeit with 1% population. Stabilising this low-populated state increased its efficacy in regulating mRNA encoding the protein sirtuin-1, a key regulator of the p53 oncoprotein. However, this observation was performed in controlled in vitro environment hence, the second objective of the project is to develop methods to study the structure and dynamics of miRNA in cellular environment.
Overall, the two objectives will provide the much-needed atomic information on structure and dynamics of miRNA beyond the seed region and methods to detect and study them in cellular environment with a potential to design effective miRNA-based therapeutics.
Among the various disease relevant miRNA, miR-34a have been implicated in 50% of cancers and is being considered as potential cancer therapeutics. A miR-34a mimic was tested against last-stage hepatocellular carcinoma patients where partial reemission of the disease showed promise for future development. This clinical trial was ultimately halted due to high mortality which is attributed to the ability of miR-34a mimic targeting multiple mRNA, known as off-targeting phenomenon. The unmodified miR-34a is known to have ~100 validated targets which poses a challenge to design safe, precise and effective miRNA-based therapeutics. The first objective will aid in addressing this challenge where rules of miRNA binding to mRNA can be expanded to include miRNA regions beyond the seed. It has been shown by the host laboratory that in miR-34a, a nucleotide at position 8 can dynamically become part of the seed albeit with 1% population. Stabilising this low-populated state increased its efficacy in regulating mRNA encoding the protein sirtuin-1, a key regulator of the p53 oncoprotein. However, this observation was performed in controlled in vitro environment hence, the second objective of the project is to develop methods to study the structure and dynamics of miRNA in cellular environment.
Overall, the two objectives will provide the much-needed atomic information on structure and dynamics of miRNA beyond the seed region and methods to detect and study them in cellular environment with a potential to design effective miRNA-based therapeutics.
Work performed from the beginning of the project to the end of the period covered by the report and main results achieved so far
Objective 1: Structural and dynamic characterisation of miR-34a within Argonaute
A central goal of this project was to elucidate the structure and dynamics of miR-34a within Argonaute in both the binary (miRNA-loaded) and ternary (miRNA + target mRNA) complexes, systems that pose extraordinary challenges due to their size (104 kDa and 110 kDa, respectively) and the inherent dynamics of miRNA regions beyond the seed. To overcome these barriers, dynamic nuclear polarisation (DNP) magic angle spinning (MAS) NMR was employed, achieving a signal enhancement of approximately 200-fold. This provided, for the first time, detailed insight into the conformational properties of the full miR-34a guide strand within Argonaute, an information inaccessible to crystallography and cryo-EM due to the poor local resolution of the dynamic non-seed regions. Conformational ensembles trapped in frozen states were characterised, exhibiting distinct distributions of C2′- versus C3′-endo sugar pucker, demonstrating that DNP-enhanced solid-state NMR can report on room-temperature conformational heterogeneity under cryogenic conditions. This study provides the first direct observation of differential sugar puckering preferences among individual nucleotides across all regions of miR-34a, including the dynamic central region, closing a structural gap that crystal structures have consistently failed to address. Together, these findings begin to expand the rules governing miRNA–mRNA recognition beyond the seed, directly supporting the aims of Objective 1.
Critically, the presence of a 21-nt SIRT1 target mRNA was found to alter the conformational preferences of miR-34a and reduce its non-helicity by approximately 10%. This has direct implications for the off-targeting problem that led to the clinical failure of the miR-34a mimic: if different mRNA targets induce distinct conformational states of miR-34a within Argonaute, structural characterisation of these states offers a basis for rational design of conformation-selective miRNA therapeutics. These observations open avenues for systematic study of miR-34a conformation in the presence of multiple target sequences, with the ultimate goal of building a mechanistic framework for miRNA-based target discrimination and regulation.
Objective 2: Development of methods to study miRNA structure and dynamics in the cellular environment
The second objective of this project was to develop experimental methods capable of extending structural and dynamic studies of miRNA from controlled in vitro settings into the cellular environment. Two methodological advances were achieved in this direction.
First, through modifications to the acquisition and analysis of existing control experiments and the use of biological replicates, the selective longitudinal relaxation time (selT1) of nucleic acids inside living cells was determined. The measured selT₁ was used to optimise the recovery delay of the 1D-SOFAST experiment, yielding maximum in-cell sensitivity exceeding that of the jump-return experiment and, based on in vitro benchmarks, exceeding that of the un-optimised SOFAST experiment. This approach is applicable to any in-cell nucleic acid or protein system and provides a practical route toward more demanding in-cell NMR experiments, including determination of base pair opening rates, molecular tumbling times, and the physico-chemical effects of the cellular environment on nucleic acid structure. These measurements are directly relevant to validating whether the low-populated dynamic states of miR-34a identified in vitro persist inside cells.
Second, a strategy for targeted DNP MAS NMR of RNA was established using AsymPol-NCS-SDSL. Conjugation of this isothiocyanate derivative of AsymPol to amino-modified RNA was achieved under mild conditions, and the approach is broadly adaptable to alternative polarising agents or labelling chemistries. Using spin-labelled mSirt1 mRNA fragments hybridised with 13C,15N-cytidine-labelled miR-34a, DNP enhancements of up to 27-fold were obtained. Systematic optimisation showed that matrix deuteration markedly improved intramolecular localisation of the polarisation enhancement, and that Cu2+ could serve as an alternative paramagnetic dopant. The relatively low effective biradical concentration employed (20–40 μM, compared with ~10 mM in conventional cross-effect DNP) is particularly relevant for future in-cell applications, where minimising perturbation of cellular components is essential. This method enables targeted DNP studies of nucleic acids without requiring pre-existing metal ion binding sites, providing a broadly applicable platform for the structural investigation of miRNA–protein complexes in conditions approaching the cellular context.
Methodological advances supporting both objectives
A detailed theoretical and experimental treatment of ¹H R1ρ relaxation dispersion (RD) was developed, addressing the influence of dipolar-coupled protons (¹Hdip) on exchange-sensitive sites. It was demonstrated that for ¹Hdip distances greater than 3 Å cross-relaxation contributions to R1ρ are negligible, validating the use of standard exchange models for fitting RD data and substantially broadening the range of nucleotide sites that can be probed for dynamics. Applied to A2 DNA, the method revealed a previously unrecognised second excited state (ES2) within the Watson–Crick–Franklin to Hoogsteen (WCF–HG) exchange pathway. Molecular dynamics and metadynamics simulations indicate that ES2 likely corresponds to an intermediate stabilised by a hydrogen bond between the A16 amino group and T9 O2. Stabilisation of ES2 in the presence of Actinomycin D further implicates this transient state in anticancer drug binding. The ability to characterise such low-populated conformational states through integrated experimental and computational approaches directly enhances the methodological toolkit available to study miRNA dynamics, both in vitro (Objective 1) and, with ongoing development, within the cellular environment (Objective 2).
A central goal of this project was to elucidate the structure and dynamics of miR-34a within Argonaute in both the binary (miRNA-loaded) and ternary (miRNA + target mRNA) complexes, systems that pose extraordinary challenges due to their size (104 kDa and 110 kDa, respectively) and the inherent dynamics of miRNA regions beyond the seed. To overcome these barriers, dynamic nuclear polarisation (DNP) magic angle spinning (MAS) NMR was employed, achieving a signal enhancement of approximately 200-fold. This provided, for the first time, detailed insight into the conformational properties of the full miR-34a guide strand within Argonaute, an information inaccessible to crystallography and cryo-EM due to the poor local resolution of the dynamic non-seed regions. Conformational ensembles trapped in frozen states were characterised, exhibiting distinct distributions of C2′- versus C3′-endo sugar pucker, demonstrating that DNP-enhanced solid-state NMR can report on room-temperature conformational heterogeneity under cryogenic conditions. This study provides the first direct observation of differential sugar puckering preferences among individual nucleotides across all regions of miR-34a, including the dynamic central region, closing a structural gap that crystal structures have consistently failed to address. Together, these findings begin to expand the rules governing miRNA–mRNA recognition beyond the seed, directly supporting the aims of Objective 1.
Critically, the presence of a 21-nt SIRT1 target mRNA was found to alter the conformational preferences of miR-34a and reduce its non-helicity by approximately 10%. This has direct implications for the off-targeting problem that led to the clinical failure of the miR-34a mimic: if different mRNA targets induce distinct conformational states of miR-34a within Argonaute, structural characterisation of these states offers a basis for rational design of conformation-selective miRNA therapeutics. These observations open avenues for systematic study of miR-34a conformation in the presence of multiple target sequences, with the ultimate goal of building a mechanistic framework for miRNA-based target discrimination and regulation.
Objective 2: Development of methods to study miRNA structure and dynamics in the cellular environment
The second objective of this project was to develop experimental methods capable of extending structural and dynamic studies of miRNA from controlled in vitro settings into the cellular environment. Two methodological advances were achieved in this direction.
First, through modifications to the acquisition and analysis of existing control experiments and the use of biological replicates, the selective longitudinal relaxation time (selT1) of nucleic acids inside living cells was determined. The measured selT₁ was used to optimise the recovery delay of the 1D-SOFAST experiment, yielding maximum in-cell sensitivity exceeding that of the jump-return experiment and, based on in vitro benchmarks, exceeding that of the un-optimised SOFAST experiment. This approach is applicable to any in-cell nucleic acid or protein system and provides a practical route toward more demanding in-cell NMR experiments, including determination of base pair opening rates, molecular tumbling times, and the physico-chemical effects of the cellular environment on nucleic acid structure. These measurements are directly relevant to validating whether the low-populated dynamic states of miR-34a identified in vitro persist inside cells.
Second, a strategy for targeted DNP MAS NMR of RNA was established using AsymPol-NCS-SDSL. Conjugation of this isothiocyanate derivative of AsymPol to amino-modified RNA was achieved under mild conditions, and the approach is broadly adaptable to alternative polarising agents or labelling chemistries. Using spin-labelled mSirt1 mRNA fragments hybridised with 13C,15N-cytidine-labelled miR-34a, DNP enhancements of up to 27-fold were obtained. Systematic optimisation showed that matrix deuteration markedly improved intramolecular localisation of the polarisation enhancement, and that Cu2+ could serve as an alternative paramagnetic dopant. The relatively low effective biradical concentration employed (20–40 μM, compared with ~10 mM in conventional cross-effect DNP) is particularly relevant for future in-cell applications, where minimising perturbation of cellular components is essential. This method enables targeted DNP studies of nucleic acids without requiring pre-existing metal ion binding sites, providing a broadly applicable platform for the structural investigation of miRNA–protein complexes in conditions approaching the cellular context.
Methodological advances supporting both objectives
A detailed theoretical and experimental treatment of ¹H R1ρ relaxation dispersion (RD) was developed, addressing the influence of dipolar-coupled protons (¹Hdip) on exchange-sensitive sites. It was demonstrated that for ¹Hdip distances greater than 3 Å cross-relaxation contributions to R1ρ are negligible, validating the use of standard exchange models for fitting RD data and substantially broadening the range of nucleotide sites that can be probed for dynamics. Applied to A2 DNA, the method revealed a previously unrecognised second excited state (ES2) within the Watson–Crick–Franklin to Hoogsteen (WCF–HG) exchange pathway. Molecular dynamics and metadynamics simulations indicate that ES2 likely corresponds to an intermediate stabilised by a hydrogen bond between the A16 amino group and T9 O2. Stabilisation of ES2 in the presence of Actinomycin D further implicates this transient state in anticancer drug binding. The ability to characterise such low-populated conformational states through integrated experimental and computational approaches directly enhances the methodological toolkit available to study miRNA dynamics, both in vitro (Objective 1) and, with ongoing development, within the cellular environment (Objective 2).
Progress beyond the state of the art and expected potential impact (including the socio-economic impact and the wider societal implications of the project so far)
This project has delivered several results that advance the field significantly beyond the current state of the art in structural biology of miRNA–protein complexes and NMR methodology.
The structural characterisation of miR-34a within hAgo2 at atomic resolution represents a breakthrough that crystallography and cryo-EM have been unable to achieve due to the dynamic nature of the non-seed regions. By establishing DNP MAS NMR as a viable tool for complexes exceeding 100 kDa, this work opens a new experimental window on a class of biologically critical assemblies that has remained structurally intractable. The first nucleotide-resolution map of sugar puckering preferences across all regions of a guide miRNA within RISC, including the dynamic central region, constitutes direct structural evidence for conformational heterogeneity that has until now only been inferred indirectly. The finding that target mRNA binding modulates miR-34a conformation within hAgo2 provides a mechanistic basis for target discrimination beyond the seed, a question of fundamental importance for understanding how one miRNA can regulate hundreds of distinct targets with varying efficacy.
On the methodological side, the project has produced two broadly applicable advances. The AsymPol-NCS-SDSL spin-labelling strategy for targeted DNP of RNA removes a longstanding barrier to the application of DNP to nucleic acid systems, which lack natural metal ion binding sites exploitable for paramagnetic enhancement. The low biradical concentrations required make the method particularly promising for in-cell DNP applications. In parallel, the optimised in-cell SOFAST NMR protocol and the theoretical framework validating ¹H R1ρ relaxation dispersion at an expanded set of nucleotide sites collectively lower the experimental barriers to studying nucleic acid dynamics inside living cells. These methodological outputs are not specific to miR-34a or Argonaute; they are immediately transferable to the structural study of other non-coding RNAs, RNA–protein complexes, and DNA systems of biomedical relevance. Further development toward in-cell DNP and routine in-cell observation of low-populated RNA states would benefit from dedicated instrumentation access, isotope labelling infrastructure, and interdisciplinary collaboration with cell biology and pharmacology groups.
Collectively, these results position NMR-based structural biology as an indispensable complement to cryo-EM and crystallography in the study of dynamic RNA–protein complexes, with a clear pathway toward impact in the rational design of next-generation miRNA therapeutics.
The structural characterisation of miR-34a within hAgo2 at atomic resolution represents a breakthrough that crystallography and cryo-EM have been unable to achieve due to the dynamic nature of the non-seed regions. By establishing DNP MAS NMR as a viable tool for complexes exceeding 100 kDa, this work opens a new experimental window on a class of biologically critical assemblies that has remained structurally intractable. The first nucleotide-resolution map of sugar puckering preferences across all regions of a guide miRNA within RISC, including the dynamic central region, constitutes direct structural evidence for conformational heterogeneity that has until now only been inferred indirectly. The finding that target mRNA binding modulates miR-34a conformation within hAgo2 provides a mechanistic basis for target discrimination beyond the seed, a question of fundamental importance for understanding how one miRNA can regulate hundreds of distinct targets with varying efficacy.
On the methodological side, the project has produced two broadly applicable advances. The AsymPol-NCS-SDSL spin-labelling strategy for targeted DNP of RNA removes a longstanding barrier to the application of DNP to nucleic acid systems, which lack natural metal ion binding sites exploitable for paramagnetic enhancement. The low biradical concentrations required make the method particularly promising for in-cell DNP applications. In parallel, the optimised in-cell SOFAST NMR protocol and the theoretical framework validating ¹H R1ρ relaxation dispersion at an expanded set of nucleotide sites collectively lower the experimental barriers to studying nucleic acid dynamics inside living cells. These methodological outputs are not specific to miR-34a or Argonaute; they are immediately transferable to the structural study of other non-coding RNAs, RNA–protein complexes, and DNA systems of biomedical relevance. Further development toward in-cell DNP and routine in-cell observation of low-populated RNA states would benefit from dedicated instrumentation access, isotope labelling infrastructure, and interdisciplinary collaboration with cell biology and pharmacology groups.
Collectively, these results position NMR-based structural biology as an indispensable complement to cryo-EM and crystallography in the study of dynamic RNA–protein complexes, with a clear pathway toward impact in the rational design of next-generation miRNA therapeutics.