Since the outbreak of the coronavirus disease 2019 (COVID-19) pandemic, it has posed a severe impact on health, community and the global economy [1]. According to data released by the World Health Organization (WHO), as of August 25, 2025, over 7099,716 deaths from COVID-19 have been reported globally [2]. This crisis also sparked concerns and discussions about the sudden outbreak of unknown X-diseases [3]. During the pandemic, vaccination could help effectively reduce the severe disease rate and mortality rate [4]. Currently, multiple vaccine platforms such as inactivated vaccines [5], attenuated vaccines, subunit vaccines, viral vector vaccines, and mRNA vaccines have been successfully developed and were widely used for disease prevention [[6], [7], [8], [9]]. Among them, inactivated vaccines, attenuated vaccines, and subunit vaccines offer increased safety and immunogenicity [10], with inactivated vaccines being the preferred choice for newly emerging infectious diseases [9].

Large-scale vaccine manufacturing is typically a lengthy challenging and expensive process [11], especially downstream process. Downstream processing always involved many steps to collect the product and remove complex impurities simultaneously [12]. These steps include several typical unit operations such as ultracentrifugation, tangential flow filtration, and chromatography [13]. Among them, affinity chromatography has been regarded as the most powerful tool for the capture of the targets, regardless of the scenarios for the production of monoclonal antibodies, viral vectors or vaccines [[14], [15], [16], [17]]. Since 2020, several affinity chromatographic techniques have been investigated for the purification of the COVID-19 virus and its antigens [[18], [19], [20], [21]]. Moreover, several COVID-19 diagnostic tests and conjugates for the therapeutics have also been developed [22,23]. In affinity chromatography, receptor recognition of viral proteins is an underlying fundamental for the design and development of these techniques [24]. The spike (S) protein has a native trimeric structure on the surface of the COVID-19 virus and is involved in virus entry into host cells [25]. It is initiated by binding with the natural receptor, human angiotensin-converting enzyme 2 (ACE2) on human cells through the receptor-binding domain (RBD) on the S protein [26]. By means of various libraries generated from combinatorial chemistry [27,28], phage display [29], mRNA display [23] and other entity libraries, several affinity candidate molecules have been identified targeting the RBD of the S protein for the application of the purification and rapid diagnostic tests [22,30,31]. However, entity libraries always have several inherent limitations including limited library size, codon and sequence bias, long time cycles, and expensive selection [32]. The emergence and application of computer-assisted virtual screening technology have significantly reduced the time and economic costs [33]. For example, Yang et al. screened five inhibitory molecules with high binding energies targeting the RBD region via molecular docking and molecular dynamics (MD) simulations [34]. Recently, molecular docking and simulation have been applied successfully in the screening of peptide ligands for vaccine purification that target the RBD on the S protein of the COVID-19 virus and influenza virus hemagglutinin [21,35].

The affibody is a small-molecule protein scaffold consisting of 58 amino acids, derived from the Z domain of Staphylococcus protein A [36]. The scaffold has three antiparallel α-helices [37]. By substituting 13 replaceable residues on the first and second helices of the affibody, it was possible to endow the affibody with the ability to bind all proteins [38]. Currently, affibodies are employed as affinity ligands for the purification of blood factors and antibodies [39]. Recently, Shi and Song reported the development of affibody ligands targeting the ACE2 binding site in the RBD by mutating residues in the Z-domain of protein A and conducting virtual screening [40]. They successfully obtained three mutant ligands with significantly higher affinities than polypeptide ligands did, thereby expanding the affibody applications.

The COVID-19 virus is characterized by rapid mutation, with four variants, namely Alpha, Beta, Gamma, and Delta, emerging during the initial four-month transmission [41]. The Omicron variant, discovered in South Africa, even has a substantial number of mutations, with up to 15 mutations within the RBD region on S protein [42], and most of the mutations are concentrated at the ACE2-binding site. This not only results in the variant having greater immune escape capabilities and increased transmissibility [43,44] but also points to a great indeterminacy for those affinity ligands targeting the ACE2-binding site on the RBD in the purification of COVID-19 vaccines. Therefore, the discovery of affinity ligands suitable for the purification of various COVID-19 vaccines is particularly crucial.

Based on the previous works [30,45], we reported a strategy for the development of affibody ligands targeting the conserved region on the RBD for the purification of COVID-19 omicron vaccine as shown in Fig. 1. In this work, affinity modeling was established by analyzing the binding complex of the 3–2A2–4 nanobody with the RBD, and a virtual affibody library was generated for ligand screening via molecular docking and MD simulation. The candidates were further evaluated by affinity measurements via microscale thermophoresis (MST) [46]. The candidate was subsequently coupled onto a Sepharose Fast Flow (SepFF) gel and the resulting affinity gel was applied for the purification of the COVID-19 omicron vaccine.