Estudo bioquímico e estrutural de uma lectina da alga vermelha Alsidium seaforthii expressa em sistema heterólogo

Abstract

Marine environments, due to their high biodiversity, stand out as promising sources of bioactive natural products, whose unique structures and chemical properties offer great potential for the development of new drugs. Algae are a rich source of compounds with relevant biotechnological and pharmacological properties, including lectins: proteins and glycoproteins capable of specific, reversible, and non-catalytic binding to carbohydrates. Heterologous expression enables the production and purification of lectins, overcoming limitations of natural sources such as seasonality, low yield, and the presence of isoforms. The aim of this work was to heterologously produce the lectin from the red marine alga Alsidium seaforthii, characterize it biochemically, and predict its three-dimensional structure in silico. The isolectin 3 of Alsidium seaforthii (BSL3) was expressed using the pET32a-TEV vector and obtained from the soluble fraction of Escherichia coli BL21(DE3) bacterial cultures. The protein (rBSL3) was purified by nickel-affinity chromatography, and its molecular mass was estimated at 30 kDa by 15% SDS-PAGE. The final purification yield was approximately 1.35 mg per 2.5 liters of induced culture. The hemagglutination assay showed that rBSL3 is capable of agglutinating trypsin-treated rabbit erythrocytes, similarly to the wild-type form. rBSL3 demonstrated affinity for porcine thyroglobulin, a glycoprotein rich in α1,6-core-fucosylated oligosaccharides, a glycan marker associated with tumor cells. Primary structure determination by mass spectrometry confirmed identity with the native lectin sequence. The dissociation constant of the lectin–ligand complex was determined by isothermal titration calorimetry as 53 μM. The hydrodynamic size of the lectin, 5.89 nm, was determined by dynamic light scattering. Secondary structure prediction by circular dichroism revealed that rBSL3 is composed of 18.5% α-helices, 29.3% β-sheets, 22.8% β-turns, and 28.5% random coils. In silico structural prediction using AlphaFold2, based on machine learning, indicated a secondary structure predominantly composed of β-sheets, with 98.51% of amino acid residues located in favored regions of the Ramachandran plot. Therefore, considering the characterization of rBSL3 in this study alongside the properties of its wild-type form, the results highlight its potential as a tool for cancer diagnostics and studies of cytotoxicity against tumor cells.