Yaw Sing Tan

Protein–RNA docking is a valuable tool for predicting the structures of protein–RNA complexes, which allow us to understand the structural basis for gene expression and regulation, thus facilitating drug development. Despite the development of several protein–RNA docking programs, the field remains relatively underdeveloped compared to protein–protein docking, and a systematic comparison of these programs in terms of accuracy and efficiency is still lacking. Recent advances in deep learning-based structure prediction, such as AlphaFold 3, offer a promising alternative for modeling protein–RNA complexes. Here, we have compiled a consolidated benchmark data set of 235 protein–RNA complexes (freely available at https://github.com/tanys-group/protein-rna-docking-benchmark), which were curated from PDB structures deposited up to July 2024, to assess the performance of five template-free docking web servers and the AlphaFold Server. Among the docking web servers, HDOCK performed the best, achieving success rates of 31.1% and 44.7% within the top 1 and top 5 predictions, respectively, as assessed by CAPRI (Critical Assessment of PRedicted Interactions) metrics. Although AlphaFold 3 outperformed all the docking web servers with an overall success rate of 87.0% in its top 5 predictions, it failed in nine cases where docking approaches succeeded and showed a markedly lower success rate of 40% for protein–RNA complexes outside its training set, comparable to that of HDOCK (35%). Our study provides valuable insights into the strengths and limitations of current protein–RNA docking servers and AlphaFold 3, offering practical guidance for selecting the appropriate tool for protein–RNA complex structure prediction. These results also suggest that hybrid approaches combining physics-based and machine learning methods hold significant promise for achieving higher prediction accuracy.

HNF4A and HNF1A encode transcription factors that are important for the development and function of the pancreas and liver. Mutations in both genes have been directly linked to Maturity Onset Diabetes of the Young (MODY) and type 2 diabetes (T2D) risk. To better define the pleiotropic gene regulatory roles of HNF4A and HNF1A, we generated a comprehensive genome-wide map of their binding targets in pancreatic and hepatic cells using ChIP-Seq. HNF4A was found to bind and regulate known (ACY3, HAAO, HNF1A, MAP3K11) and previously unidentified (ABCD3, CDKN2AIP, USH1C, VIL1) loci in a tissue-dependent manner. Functional follow-up highlighted a potential role for HAAO and USH1C as regulators of beta cell function. Unlike the loss-of-function HNF4A/MODY1 variant I271fs, the T2D-associated HNF4A variant (rs1800961) was found to activate AKAP1, GAD2 and HOPX gene expression, potentially due to changes in DNA-binding affinity. We also found HNF1A to bind to and regulate GPR39 expression in beta cells. Overall, our studies provide a rich resource for uncovering downstream molecular targets of HNF4A and HNF1A that may contribute to beta cell or hepatic cell (dys)function, and set up a framework for gene discovery and functional validation.

The Non-POU domain-containing octamer-binding protein (NONO) is a nucleic acid binding protein with diverse functions that has been identified as a potential cancer target in cell biology studies. Little is known about structural motifs that mediate binding to NONO apart from its ability to form homodimers, as well as heterodimers and oligomers with related homologues. We report a stapling approach to macrocyclise helical peptides derived from the insulin-like growth factor binding protein (IGFBP-3) that NONO interacts with, and also from the dimerisation domain of NONO itself. Using a range of chemistries including Pd-catalysed cross-coupling, cysteine arylation, and cysteine alkylation, we successfully improved the helicity and observed modest peptide binding to the NONO dimer, although binding could not be saturated at micromolar concentrations. Unexpectedly, we observed cell permeability and preferential nuclear localisation of various dye-labelled peptides in live confocal microscopy, indicating the potential for developing peptide-based tools to study NONO in a cellular context.

Regulation of global transcription output is important for normal development and disease, but little is known about the mechanisms involved. DNA topoisomerase I (TOP1) is an enzyme well-known for its role in relieving DNA supercoils for enabling transcription. Here, we report a non-enzymatic function of TOP1 that downregulates RNA synthesis. This function is dependent on specific DNA-interacting residues located on a conserved protein surface. A loss-of-function knock-in mutation on this surface, R548Q, is sufficient to cause hypertranscription and alter differentiation outcomes in mouse embryonic stem cells (mESCs). Hypertranscription in mESCs is accompanied by reduced TOP1 chromatin binding and change in genomic supercoiling. Notably, the mutation does not impact TOP1 enzymatic activity; rather, it diminishes TOP1-DNA binding and formation of compact protein-DNA structures. Thus, TOP1 exhibits opposing influences on transcription through distinct activities which are likely to be coordinated. This highlights TOP1 as a safeguard of appropriate total transcription levels in cells.

Biomarkers have the potential to be utilized in disease diagnosis, prediction and monitoring. The cancer cell type is a leading candidate for next-generation biomarkers. Although traditional digital biomolecular sensor (DBS) technology has shown to be effective in assessing cell-based interactions, low cell-population detection of cancer cell types is extremely challenging. Here, we controlled the electrical signature of a two-dimensional (2D) nanomaterial, tungsten disulfide (WS₂), by utilizing a combination of the Phage-integrated Polymer and the Nanosheet (PPN), viz., the integration of the M13-conjugated polyethylene glycol (PEG) and the WS₂, through shape-complementarity phenomena, and developed a sensor system, i.e., the Phage-based DBS (P-DBS), for the specific, rapid, sensitive detection of clinically-relevant MCF-7 cells. The P-DBS attains a detection limit of 12 cells per μL, as well as a contrast of 1.25 between the MCF-10A sample signal and the MCF-7 sample signal. A reading length of 200 μs was further achieved, along with a relative cell viability of ∼100% for both MCF-7 and MCF-10A cells and with the PNN. Atomistic simulations reveal the structural origin of the shape complementarity-facilitated decrease in the output impedance of the P-DBS. The combination of previously unreported exotic sensing materials and digital sensor design represents an approach to unlocking the ultra-sensitive detection of cancer cell types and provides a promising avenue for early cancer diagnosis, staging and monitoring.

Nucleic acid aptamers as antibody alternatives bind specifically to target molecules. These aptamers are generated by isolating candidates from libraries with random sequence fragments, through an evolutionary engineering system. We recently reported a high-affinity DNA aptamer generation method that introduces unnatural bases (UBs) as a fifth letter into the library, by genetic alphabet expansion. By incorporating hydrophobic UBs, the affinities of DNA aptamers to target proteins are increased over 100-fold, as compared with those of conventional aptamers with only the natural four letters. However, there is still plenty of room for improvement of the methods for routinely generating high-affinity UB-containing DNA (UB-DNA) aptamers. The success probabilities of the high-affinity aptamer generation depend on the existence of the aptamer candidate sequences in the initial library. We estimated the success probabilities by analysing several UB-DNA aptamers that we generated, as examples. In addition, we investigated the possible improvement of conventional aptamer affinities by introducing one UB at specific positions. Our data revealed that UB-DNA aptamers adopt specific tertiary structures, in which many bases including UBs interact with target proteins for high affinity, suggesting the importance of the UB-DNA library design. This article is part of the theme issue ‘Reactivity and mechanism in chemical and synthetic biology’.

Promising results in clinical studies have been demonstrated by the utilization of electrothermal agents (ETAs) in cancer therapy. However, a difficulty arises from the balance between facilitating the degradation of ETAs, and at the same time, increasing the electrothermal performance/stability required for highly efficient treatment. In this study, we controlled the thermal signature of the MoS₂ by harnessing MoS₂ nanostructures with M13 phage (MNM) via the structural assembling (hydrophobic interaction) phenomena and developed a combined PANC-1 cancer cell–MNM alternating current (AC)-stimulus framework for cancer cell ablation and electrothermal therapy. A percentage decrease in the cell viability of ~23% was achieved, as well as a degradation time of 2 weeks; a stimulus length of 100 μs was also achieved. Molecular dynamics (MD) simulations revealed the assembling kinetics in integrated M13 phage–cancer cell protein systems and the structural origin of the hydrophobic interaction-enabled increase in thermal conduction. This study not only introduced an ‘ideal’ agent that avoided the limitations of ETAs but also provided a proof-of-concept application of MoS₂-based materials in efficacious cancer therapy.

Biosensors are of vital significance for healthcare by supporting the management of infectious diseases for preventing pandemics and the diagnosis of life-threatening conditions such as cancer. However, the advancement of the field can be limited by low sensing accuracy. Here, we altered the bioelectrical signatures of the cells using carbon nanotubes (CNTs) via structural loosening effects. Using an alternating current (AC) pulse under light irradiation, we developed a photo-assisted AC pulse sensor based on CNTs to differentiate between healthy breast epithelial cells (MCF-10A) and luminal breast cancer cells (MCF-7) within a heterogeneous cell population. We observed a previously undemonstrated increase in current contrast for MCF-7 cells with CNTs compared to MCF-10A cells with CNTs under light exposure. Moreover, we obtained a detection limit of ∼1.5 × 103 cells below a baseline of ∼1 × 104 cells for existing electrical-based sensors for an adherent, heterogeneous cell population. All-atom molecular dynamics (MD) simulations reveal that interactions between the embedded CNT and cancer cell membranes result in a less rigid lipid bilayer structure, which can facilitate CNT translocation for enhancing current. This as-yet unconsidered cancer cell-specific method based on the unique optoelectrical properties of CNTs represents a strategy for unlocking the detection of a small population of cancer cells and provides a promising route for the early diagnosis, monitoring, and staging of cancer.

A few-volt nanosecond alternating-current stimulation method with an increased number of pulses exploiting orientational-disordering processes.

The identification and characterization of binding sites is a critical component of structure-based drug design (SBDD). Probe-based/cosolvent molecular dynamics (MD) methods that allow for protein flexibility have been developed to predict ligand binding sites. However, cryptic pockets that appear only upon ligand binding and occluded binding sites with no access to the solvent pose significant challenges to these methods. Here, we report the development of accelerated ligand-mapping MD (aLMMD), which combines accelerated MD with LMMD, for the detection of these challenging binding sites. The method was validated on five proteins with what we term “recalcitrant” cryptic pockets, which are deeply buried pockets that require extensive movement of the protein backbone to expose, and three proteins with occluded binding sites. In all the cases, aLMMD was able to detect and sample the binding sites. Our results suggest that aLMMD could be used as a general approach for the detection of such elusive binding sites in protein targets, thus providing valuable information for SBDD.