The lack of a universal influenza vaccine is a global health problem. Interest is now focused on structurally conserved protein domains capable of eliciting protection against a broad range of influenza virus strains. The long alpha helix (LAH) is an attractive vaccine component since it is one of the most conserved influenza hemagglutinin (HA) stalk regions. For an improved immune response, the LAH domain from H3N2 strain has been incorporated into virus-like particles (VLPs) derived from hepatitis B virus core protein (HBc) using recently developed tandem core technology. Fermentation conditions for recombinant HBc-LAH were established in yeast Pichia pastoris and a rapid and efficient purification method for chimeric VLPs was developed to match the requirements for industrial scale-up. Purified VLPs induced strong antibody responses against both group 1 and group 2 HA proteins in mice. Our results indicate that the tandem core technology is a useful tool for incorporation of highly hydrophobic LAH domain into HBc VLPs. Chimeric VLPs can be successfully produced in bioreactor using yeast expression system. Immunologic data indicate that HBc VLPs carrying the LAH antigen represent a promising universal influenza vaccine component.

The core proteins (HBc) of the hepatitis B virus (HBV) genotypes A, B, C, D, E, F, and G were cloned and expressed in Escherichia coli (E. coli), and HBc-formed virus-like particles (VLPs) were purified with ammonium sulfate precipitation, gel filtration, and ion exchange chromatography (IEX). The best VLP yield was found for the HBc of the HBV genotypes D and G. For the HBc of the HBV genotypes D, F, and G, the possibility of dissociation and reassociation maintaining the native HBc structure was demonstrated. Single-stranded (ss) and double-stranded (ds) ribonucleic acid (RNA) was successfully packed into HBc VLPs for the HBV genotypes D and G.

Influenza, an acute, highly contagious respiratory disease, remains a significant threat to public health. More effective vaccination strategies aimed at inducing broad cross-protection not only against seasonal influenza variants, but also zoonotic and emerging pandemic influenza strains are urgently needed. A number of conserved protein targets to elicit such cross-protective immunity have been under investigation, with long alpha-helix (LAH) from hemagglutinin stalk and ectodomain of matrix protein 2 ion channel (M2e) being the most studied ones. Recently, we have reported the three-dimensional structure and some practical applications of LAH expressed in Escherichia coli system (referred to as tri-stalk protein). In the present study, we investigated the immunogenicity and efficacy of a panel of broadly protective influenza vaccine prototypes based on both influenza tri-stalk and triple M2e (3M2e) antigens integrated into phage AP205 virus-like particles (VLPs). While VLPs containing the 3M2e alone induced protection against standard homologous and heterologous virus challenge in mice, only the combination of both conserved influenza antigens into a single VLP fully protected mice from a high-dose homologous H1N1 influenza infection. We propose that a combination of genetic fusion and chemical coupling techniques to expose two different foreign influenza antigens on a single particle is a perspective approach for generation of a broadly-effective vaccine candidate that could protect against the constantly emerging influenza virus strains.

Plant-based meat alternatives, exemplified by Impossible Foods’ Impossible Burger, offer a sustainable, ethical substitute for traditional meat, closely mimicking the taste and appearance of meat by utilizing soy leghemoglobin (LegH), a 16 kDa holoprotein found in soy plants structurally similar to heme in animal meat. Cultivation medium plays an important role in bioprocess development; however, medium development or optimization can be labor intensive, and thus the use of previously reported media can be enticing. In this study, we explored the expression of recombinant LegH in Pichia pastoris in various reported cultivation media (BSM, BMGY, FM22, D’Anjou, BSM/2, and RDM) and using different feeding approaches (µ-stat and mixed feed with sorbitol). Our findings indicate that optimization techniques tailored to the specific process did not increase LegH yields, highlighting the need to investigate strain-specific strategies. We also utilized the collected process data to create and train a novel artificial neural network-based soft sensor for estimating cell biomass, relying solely on standard bioreactor measurements (such as stirrer speed, dissolved oxygen, O₂ enrichment, base feed, glycerol feed, methanol feed, and reactor volume). This soft sensor proved to be robust and exhibited a strong correlation (3.72% WCW) with experimental data.