Browsing by Author "Nurasyikin Hamzah, Ph.D"

Ovarian cancer (OC) poses a significant risk as it usually remains asymptomatic until advanced stages, resulting in delayed diagnosis and lowering the chances of survival. This fifth most common cancer among women globally lacks efficient screening approaches for early stages OC (stage I and II), increasing the threat of OC progression and mortality. In this study, the in silico method was applied to design DNA aptamer hairpins for the detection of human epididymis protein 4 (HE4), an OC biomarker. The in silico work outcome was supported by an in vitro assay. The work began with the HE4 protein modelling using AlphaFold, I-TASSER, and Robetta protein structure prediction servers. Subsequently, molecular dynamics (MD) simulation was conducted on each predicted model for 100 ns using the OPLS force field, for structure refinement. The tertiary structure quality was validated by PROCHECK and ERRAT, showing the refined model from AlphaFold, RF1, was the highest-quality HE4 tertiary conformation. All amino acids were located in the favoured regions of Ramachandran plot with ERRAT overall quality score of 97.701. Next, this HE4 structure was docked using AutoDock Vina with four HE4 aptamer candidates (A1, A2, A3, and A4). The HE4-A4 binding energy was -6.0 kcal/mol, and the complex formed 24 hydrogen bonds (H-bonds); 5 identified at the aptamer hairpin loop region. A4 was chosen as the most suitable candidate to be utilised in the designing of the DNA hairpin, as it exhibited good affinity with highest number of H-bonds at the hairpin loop. To initiate the in silico design of new hairpins, the 25-mer A4 aptamer was truncated at the hairpin region, 5’-CGCAAG-3’ and the stem was extended, forming the 5’-GCGCAAGC-3’ sequence. The loop nucleotides, -GCAA-, were substitutionally mutated, producing 256 sequences. These 256 hairpins were docked with HE4 individually, and H16, H101, and H256 have shown good binding affinities with binding energies ranging between -10.6 and -11.6 kcal/mol. Consequently, 100 ns MD simulation using CHARMM27 force field was applied to the HE4-H16, HE4-H101, and HE4-H256 complexes. Based on the RMSD, radius of gyration, number of H-bonds, H-bond occupancy, and the overall total energy, H256 was deduced as the most promising DNA hairpin against HE4 marker with great affinity and stability throughout the simulation. This H256 hairpin was synthesised and its binding with HE4 was analysed via in vitro electrophoretic mobility shift assay (EMSA). Based on the DNA band intensities, the designed H256 (3.27 %) bound four times better to HE4 than the A4 aptamer (0.84 %). Finally, a preliminary study for future diagnostic potential was carried out by conjugating gold nanoparticle (GNP) with H256. The FTIR and Raman spectra confirmed the presence of amide group, formed by the successful conjugation of the carboxylated GNP with the aminated-H256. The GNP solution changed from red to purple-red, indicating the size increment after conjugation, that was confirmed by particle size analyser. In conclusion, H256 is a promising DNA hairpin in HE4 screening and is recommended for future development of a fully functional OC diagnostic kit, suitable to be used in routine screening for all women, with or without symptoms. This potentially improves the detection among early stages (I and II) patients, enhancing patients’ outcome.

Antimicrobial resistance (AMR) poses a critical global health threat, driving increased mortality, morbidity, and healthcare costs. The ability of microbes to mutate and develop resistance against existing antibiotics necessitates the discovery of new antibacterial agents. β-Carboline, a naturally abundant heterocyclic alkaloid, exhibits diverse pharmacological activities, including antibacterial effects. However, limitations in its drug-likeness restrict its potential as a novel antibiotic. This study aimed to design, synthesise, and evaluate β-carboline derivatives in-vitro targeting bacterial DNA gyrase, a key enzyme involved in bacterial replication. Initially, 51 β-carboline derivatives were virtually screened against DNA gyrase (PDB ID: 3G75) using protein’s bound ligand B48 as a reference, followed by ADMET prediction. Most β-carboline derivatives exhibited improved binding energies (-5.8 to -8.3 kcal/mol) with derivative (3.1o) (-6.1 kcal/mol), surpassing B48 (-5.8 kcal/mol) while the β-carboline scaffold (1.1) showed comparable affinity (-5.8 kcal/mol). Derivative (3.1o) interacted with the binding site of the target protein by forming two strong conventional hydrogen bonds with two important residues, Ser55 and Asp81. ADMET analysis showed that all β-carboline derivatives fulfil Lipinski’s Rule of Five with good permeability, although most of the derivatives exhibited low sp3 fraction, indicating poor solubility due high unsaturation. Among them, derivative (3.1o) was further designed using Fragment-Based Drug Design (FBDD) using growing strategy targeting hydrogen atom from each site, R1 and R2 where about 23550 and 67 compounds were generated, respectively. Virtual screening prioritised compound (4.2a) for synthesis as it exhibited improved binding affinity (-8.1 kcal/mol) compared to B48 (-5.8 kcal/mol). The new substituent at the modification site (R2) can accommodate into the binding site of the protein forming few strong conventional hydrogen bonds with critical residues, Ser55 and Asp81. On top of that, it also strictly adhered with Lipinski’s Rule of Five with high GI absorption underscoring its favourable drug-likeness and potential for oral bioavailability However, due to persistent synthetic challenges, an alternative route using Pictet-Spengler condensation with different aromatic aldehydes were proposed and successfully synthesised five tetrahydro-β-carboline derivatives (3.14a–e) with low yields (17.45 – 29.28%). The chemical structures of synthesised compounds were verified using Nuclear Magnetic Resonance (NMR) and Fourier Transform Infrared (FTIR). These derivatives were tested in-vitro for antibacterial activity against Gram-positive (S. aureus, B. subtilis) and Gram-negative bacteria (E. coli, P. aeruginosa). Of these, only compound (3.14a) showed modest activity at 50 µg/mL against Gram-negative strains, where others exhibited a weak inhibition at 100 µg/mL. Molecular docking supported these findings, with compound (3.14a) showing the highest binding affinity (-7.2 kcal/mol), comparable to streptomycin (-7.3 kcal/mol) and novobiocin (-7.5 kcal/mol). This research highlights the need for further optimisation and comprehensive virtual screening of β-carboline derivatives to establish a detailed structure-activity relationship (SAR).

Hyaluronidase (Hyal) and lipoxygenase (LOX) enzymes produced inflammatory inducers in the human body, making them suitable target for the design of anti-inflammatory agents. Xanthorrhizol (XNT) is a sesquiterpenoid isolated from Curcuma xanthorrhiza that possesses anti-inflammatory properties and holds promise as therapeutic candidate. The potential of XNT as anti-inflammatory agent can be further explored through modification of its structure. The use of computational tools can optimise the structure of XNT by precisely targeting Hyal and LOX and enhance its anti-inflammatory action. The objective of this study was to virtually screen and design potential XNT derivatives as new Hyal-1 and LOX-3 inhibitors from in-house database and in silico fragment-based drug design (FBDD) approaches, respectively. Then, the promising derivatives from both approaches were synthesised and characterised, followed by biological evaluation through in vitro enzyme inhibitory assays. The most active derivative against the enzyme in in vitro studies was subjected to further analysis using molecular dynamics (MD) simulation. Thirty XNT derivatives obtained from in-house database were virtually screened against Hyal-1 enzyme via molecular docking. Benzyl xanthorrhizyl ether (76) exhibited stronger binding energy (-8.0 kcal/mol) against Hyal-1 enzyme compared to XNT (-6.8 kcal/mol) and proposed to be synthesised as Hyal-1 inhibitor. Additionally, the in silico FBDD generated five XNT derivatives as potential Hyal-1 inhibitors, specifically (2-pyridinyl)methyl xanthorrhizyl ether (102), 4-nitrobenzyl xanthorrhizyl ether (103), 3-trifluoromethylbenzyl xanthorrhizyl ether (104), (2-tetrahydro-2H-pyranyl)methyl xanthorrhizyl ether (105) and (2-tetrahydrofuranyl)methyl xanthorrhizyl ether (106). Among the six proposed inhibitors for Hyal-1 enzyme, five derivatives (76, 102-105) were successfully synthesised and characterised using spectroscopic analyses. Meanwhile, the synthesis of derivative (106) yielded an unexpected product, characterised as propyl xanthorrhizyloxyacetate (117). The in vitro analysis against Hyal-1 enzyme showed that derivative (102) was the most active against Hyal-1 enzyme with IC50 value of 44.54 µg/mL compared to XNT (IC50 = 203.56 µg/mL). The MD simulation revealed that derivative (102) displayed stable interaction in the binding site of Hyal-1 enzyme throughout 100 ns simulation. Derivative (76) was identified as potential LOX-3 inhibitor as it exerts stronger binding energy (-8.6 kcal/mol) compared to XNT (-7.7 kcal/mol). The design of XNT derivatives through in silico FBDD approach generated four XNT derivatives as LOX-3 inhibitors, however, their synthesis were unsuccessful. Derivative (76) was synthesised, and tested via in vitro LOX-3 inhibitory assay, where it displayed weaker activity (IC50 = 619.92 µg/mL) compared to XNT (IC50 = 142.34 µg/mL). These findings suggest that XNT derivatives, particularly (2-pyridinyl)methyl xanthorrhizyl ether (102) showed promising anti-inflammatory activities through inhibition of Hyal-1 enzyme.