Structural and thermodynamic analysis of siRNA

To predict the secondary structures of the single stranded siRNA candidates, the MaxExpect tool from RNAstructure web server (Version 6.0.1, Mathews Lab, University of Rochester Medical Center)³² was employed. It is a tool for predicting RNA or DNA secondary structures which are optimized to include base pairs with the highest probability of being correct. For this study, the 23-nucleotide long siRNA guide strand was given as an input while maintaining the other parameters to default values. MaxExpect then provides the maximum expected accuracy structure, which is defined as the structure with the highest sum of correctly predicted base pairing probabilities. This approach has been shown to more accurately reflect experimentally determined RNA secondary structures than traditional minimum free energy-based methods⁵¹.

The DuplexFold tool from the RNAstructure web server³² was utilized to predict the hybridization of two separate RNA strands and to evaluate the energetic stability of their interaction. This tool computes the most thermodynamically stable structures by evaluating all possible intermolecular base pairs between the two nucleotide strands while disallowing intramolecular base pairing, thereby determining the configuration with the minimum free energy (ΔG). Designed siRNA guide strands were aligned to their complementary target regions within the mRNA sequence using DuplexFold to model RNA-RNA duplex formation. Predictions were performed under default conditions at 37 °C with standard ionic strength (1 M NaCl), and the output included the predicted intermolecular base-pairing structure and the associated minimum free energy of the duplex.

The secondary structure formed by each of the single stranded siRNA candidates was analysed using Mfold tool³³ from the Unified Nucleic Acid Folding (UNAFold) Web Server⁵². The tool calculates the most thermodynamically stable structures possible for the given nucleotide sequence, mainly through minimization of free energy using dynamic programming algorithms and established thermodynamic data. The nucleotide sequence of each siRNA guide strand was input in linear single-stranded RNA format, as appropriate for siRNA structure prediction. With default parameters such as the fixed folding temperature of 37 °C and 1 M NaCl ionic concentration, the associated free energy of folding upon secondary structure formation was predicted and used to assess the stability of each siRNA.

The DI-Nucleic Acid hybridization and melting prediction (DINAmelt) tool³⁴from the UNAFold Web Server⁵⁰was utilized to predict the melting behaviour of the hybridized siRNA and mRNA strand. Nearest-neighbor thermodynamic models, which consider the stability of each base pair with respect to its adjacent pairs, are used for this prediction. It simulates the thermal denaturation process and provides equilibrium melting profiles as a function of temperature along with melting temperature values such as Tm (Conc) and Tm (Cp), which are crucial parameters for evaluating the effectiveness of siRNAs. The single nucleotide sequence of each of the siRNA candidates and their corresponding target mRNA was given as an input. Using the energy minimization model and default parameters the melting temperature values were calculated for each duplex to help evaluate the most efficient siRNAs.

The OligoWalk module available through the RNAstructure web server³⁶ was used to predict the silencing efficacy of the designed siRNA guide strands. This tool allows for the evaluation of each siRNA’s ability to bind effectively to its target site within the full-length GPR10 mRNA sequence. Each guide strand was submitted individually, and the analysis was performed using default parameters. OligoWalk predicts siRNA efficacy by considering factors such as duplex stability, target site accessibility, and the overall energy profile of binding. For each siRNA, the tool generates an efficacy score expressed as a percentage, which reflects the likelihood of successful gene silencing. These percentage scores were used to rank the siRNA candidates and identify the most effective guide strands for further analysis.

The GC content of each siRNA guide strand was calculated using the Oligonucleotide Properties Calculator provided by Bio-Synthesis Inc³⁵. The nucleotide sequences were entered in single-stranded RNA format, and the GC percentage was computed automatically by the tool to assess duplex stability and ensure optimal base pairing characteristics. The thermodynamic stability of each siRNA guide strand was also estimated using the same tool³⁵. The Gibbs free energy (ΔG) values, indicating the spontaneity of siRNA folding, were calculated under standard conditions set to 1 M NaCl, 25 °C, and pH 7. These conditions, which represent the default parameters of the Oligonucleotide Properties Calculator, may differ from those used in the molecular dynamics (MD) simulations. However, the ΔG values (expressed in kcal/mol) were used solely for the relative comparison of thermodynamic stability among siRNA candidates during the selection process and were not intended to replicate physiological environments.