Introduction

The Coronavirus disease-2019 (COVID-19), caused by the severe acute respiratory syndrome Coronavirus 2 (SARS-CoV-2) (Harrison et al. 2020), belongs to the order Nidovirales (Gulyaeva and Gorbalenya 2021), which includes viruses with a single-stranded, positive-sense RNA with a large genome ranging in size from 26 to 32 kb (Gorbalenya et al. 2006, 2020). SARS-CoV-2 has a genome of ~ 30 kb. (Malone et al. 2022). This large genome is annotated to possess 14 open reading frames (ORFs) encoding 9890 amino acids, making 27 proteins (Gorbalenya et al. 2006; Wu et al. 2020). SARS-CoV-2 comprises four main structural proteins, including envelope, membrane, nucleocapsid, and spike proteins (Huang et al. 2020). The fragment of spike protein, from amino acid 319 to 541, is known as the receptor-binding domain (RBD) (Shin et al. 2021), which plays a central role in the disease process by binding to the angiotensin-converting enzyme-2 (ACE2) receptor, allowing the virus to enter the host cell and initiate infection (Li et al. 2003).

According to the World Health Organization (WHO), 775,678,432 confirmed cases have been infected with COVID-19, resulting in 7,052,472 deaths (WHO Coronavirus (COVID-19) Dashboard, 2024). Alongside the mortality caused by COVID-19, the long-term effects of this disease and the enduring effects of vaccines are still under discussion and investigation (Byambasuren et al. 2023). In the meantime, there are reports of COVID-19 patients suffering from an increase in the level of Alzheimer’s disease (AD) progression (Chen et al. 2022; Patel et al. 2022) as well as an elevation in the level of β-amyloid (Aβ) in the blood (Hsu et al. 2021), interacting directly with fibrin and fibrinogen to cause blood hypercoagulation (Grobbelaar et al. 2021) and diabetes, which is linked to islet amyloid polypeptide (IAPP) (Young et al. 2015). All of these reports are associated with amyloidotic fibrillar aggregates. One crucial question is which SARS-CoV-2 structural proteins can form amyloid, as this knowledge is vital for understanding the pathogenesis of the disease, including long COVID-19 (Leung et al. 2023).

Amyloid fibril formation has a connection with the increased content of β-structure in a protein structure, resulting in its fibrillar aggregation (Jimenez et al. 1999). Remarkably, the SARS-CoV-2 proteins comprise various amyloid-forming sequences (Milton 2023; Tayeb-Fligelman et al. 2023); among them, the spike protein contains RBD, with a secondary β-structure content of 34.1% (Brindha and Kuroda 2022; Nystrom and Hammarstrom 2022). RBD is a vaccine candidate and a target for drug development (Fitzgerald et al. 2021); however, the amyloidogenicity of RBD is still unclear, and the main related question is the possibility of the SARS-CoV-2 RBD amyloidogenic characteristics and its difference among various SARS-CoV-2 strains. Due to the rapid mutation of SARS-CoV-2, certain species presently do not exist, and only Omicron subvariants have become prevalent worldwide (Farahat et al. 2022; Parums 2023). The present research, for the first time, was conducted with a particular emphasis on Omicron and Delta Plus RBDs, aiming to investigate the potential of RBD in the formation of amyloid fibrils and to identify the impact of small ubiquitin-like modifier (SUMO) fusion Tag on this phenomenon. Intending to elucidate the underlying reasons for the variations in amyloid formation potential within the RBD across different strains of SARS-CoV-2, we compared the amyloidogenic potential of RBD among two significant variants of SARS-CoV-2 (Delta Plus and Omicron strains) using the FoldAmyloid algorithm (Garbuzynskiy et al. 2010), an in silico method of predicting amyloidogenic regions in a protein sequence. As fusion tags can provide control over the aggregation of an amyloidogenic protein (Kronqvist et al. 2022), the impact of the SUMO tag, which can enhance protein solubility due to its external hydrophilic surface and inner hydrophobic core (Butt et al. 2005), was examined on the Thioflavin T (ThT) fluorescence intensity of RBD amyloid-like fibrils formation. The findings will be presented based on the analysis of purified RBD of Delta Plus and RBD of Omicron variant with and without SUMO tag, isolated from Escherichia coli (E. coli) (Fig. 1) using ThT kinetics assay, Fourier-transform infrared spectroscopy (FTIR) and transmission electron microscopy (TEM).

Fig. 1
figure 1

Schematic workflow for evaluating the amyloidogenic potential of RBD

Full size image

Materials and methods

Materials

All the materials used in the present study were of analytical grade and purchased from Merck (Germany). E. coli T7 SHuffle cells were obtained from Novagen (USA).

In silico assessment of amyloidogenic potential of RBD

Two SARS-CoV-2 variants, including Delta Plus (B.1.617.2.1) and Omicron (B.1.1.529) with and without SUMO tag, were selected to compare the amyloidogenic potential of RBD in these strains and identify amyloidogenic regions in this domain. For this purpose, Protein Data Bank (https://www.rcsb.org/; accessed on 1 February 2021) was used for sequence determination of the SARS-CoV-2 spike RBDs (residues 319-541 aa). The FoldAmyloid program (http://bioinfo.protres.ru/fold-amyloid/) was also applied to evaluate the amyloid-forming ability of the RBDs (Nystrom and Hammarstrom 2022). This method predicts the amyloidogenic regions of the RBD by assessing the packing density and probability of backbone-backbone hydrogen bond formation. The amyloid-forming ability of the RBD was determined by dividing the percentage of the number of amino acid residues in the consensus amyloidogenic regions by the total number of amino acid residues on the protein. This calculation was conducted according to a protocol developed previously (Aksenova et al. 2022).

Plasmid constructs

The optimized Delta Plus (B.1.617.2.1) and Omicron (B.1.1.529) RBD genes were synthesized and cloned by Proteogenix (France) into NdeI and XhoI restriction sites of the expression vector pET-26b(+), which encodes 319–541 amino acid fragments of SARS-CoV-2 surface glycoprotein containing a 6 × histidine tag at the N-terminus. To create the SUMO-RBD structure, we inserted the DNA encoding RBD (319-541 aa) of SARS-CoV-2 Omicron spike into the pET-21a(+) vector containing the SUMO fusion tag (Fig. 2). The resulting recombinant expression vectors were then transformed into T7 SHuffle competent cells using the CaCl2-mediated method (Dagert and Ehrlich 1979).

Fig. 2
figure 2

The recombinant plasmid construct. a pET-26b(+) Delta Plus RBD, b pET-26b(+) Omicron RBD, and c pET-21a(+) Omicron SUMO-RBD

Full size image

Protein expression and purification

T7 SHuffe cells transformed with the pET-26b (+) and pET-21a (+) vectors, containing the RBD and SUMO-RBD genes, respectively, were cultivated in a Luria–Bertani (LB) medium containing 100 μg/mL of antibiotic resistance markers at 30 °C overnight. Subsequently, 1 mL of the preculture medium was utilized to inoculate 100 mL of the LB medium. When the optical density at 600 nm (OD600) reached about 0.5, the protein expression was induced with 0.5 mM of isopropyl β-D-1-thiogalactopyranoside (IPTG). After 18 h of induction at 22 °C, the cells were harvested by centrifugation (4000×g, 20 min). Tris buffer (50 mM of Tris–HCL and 300 mM of NaCl; pH 7.5) was added to the cells and lysed by sonication on ice at 15 × 10 s, with a 50% duty cycle at 70% power. The resulting suspension was centrifuged at 12,000×g at 4°C for 20 min. The inclusion bodies were solubilized in a Tris–HCl buffer (pH 7.5) comprising 1 mM of dithiothreitol (DTT) and 8.0 M of urea. The unfolded protein was loaded on a Ni-Sepharose column (Amersham Biosciences, UK) that had been equilibrated with a 50-mM Tris buffer (pH 7.5) consisting of 10 mM of imidazole and 8.0 M of urea. Bound proteins were eluted with the same buffer used for column equilibration, containing 8 M of urea and 300 mM of imidazole. The protein purity was confirmed by SDS-PAGE according to the Laemmli method (Cleveland et al. 1977), and protein concentration was determined by the Bradford method (Bradford 1976).

ThT fluorescence

The benzothiazole dye ThT, a commonly employed method for diagnosing amyloid fibrils (Bolder et al. 2007), is known for its fluorescence enhancement when interacting with amyloid fibrils (Khurana et al. 2005). The ThT assay was carried out for SARS-CoV-2 Delta Plus and Omicron RBDs, which had a 6 × histidine tag at the N-terminus. The assay was also conducted for the SARS-CoV-2 Omicron RBD with a SUMO fusion tag. The RBD stock solution, supplemented with 0.4 mg/mL of RBD and 0.65 mg/mL of SUMO-RBD, was prepared to equalize the molar concentration of RBDs in a 50 mM Tris–HCl buffer (pH 7.5) containing 8.0 M urea. Subsequently, this solution was loaded into a dialysis bag with a molecular weight cut-off of 10 kDa. The solution was then dialyzed against a 50-mM Tris–HCl buffer (pH 8.5) under stirring conditions at 4 °C for 30 h. At regular intervals, 60-μL samples of each RBD strain were collected before dialysis, and after 10 and 30 h of dialysis. These samples were then mixed with 440 μL of a 25-mM phosphate buffer (pH 6.0) containing 25 μM of ThT at 25 °C. The samples were measured for fluorescence in a range of between 450 and 550 nM at a 440-nM excitation wavelength, using a PerkinElmer Life Sciences LS-55 fluorimeter (Waltham, MA, USA) equipped with a thermostated cell compartment attached to a Haake F8 water bath (Karlsruhe, Germany), with a 2 × 10-mM path length cuvette.

Fourier-transform infrared (FTIR) spectroscopy

A sample of 4 mg/mL RBD in 8.0 M urea was dialyzed against a 50-mM Tris–HCl buffer (pH 8.5) with continuous stirring at 4 °C for 30 h. The sample underwent alternating centrifugation at 8050 g for 20 min at 25 °C. The resulting pellet was then resuspended in small amounts of D2O, reaching a concentration of 12 mg/mL. The samples were placed on a CaF2 window within a semi-permanent liquid cell, utilizing a 25 μM spacer. FTIR spectra were recorded within the range of 400 to 4000 cM−1 at 25 °C. Background subtraction was performed using a Jasco FTIR 4200 spectrophotometer (Tokyo, Japan) under a continuous flow of N2. The amide I band (1600–1700 cM−1) was isolated and baseline-corrected. The resulting spectrum was analyzed using Casa XPS software.

TEM

An EM 900A transmission electron microscope (Carl Zeiss NTS, Oberkochen, Germany) was used to achieve transmission electron micrographs. A small drop (~ 5 μL) of each RBD sample comprising RBD (0.4 mg/mL) in a 50-mM Tris–HCl buffer (pH 8.5), was deposited onto a copper 400 mesh grid covered with a perforated polymer film and a thin carbon layer. The images were visualized at a 40,000 × magnification.

Results

Amyloidogenic potential of SARS-CoV-2 RBD with and without tag

Amyloidogenicity is a significant factor affecting the interaction between hosts and pathogens (Ezzat et al. 2019). Therefore, the present study predicted the presence of amyloidogenic regions in Delta Plus (B.1.617.2.1) and Omicron (B.1.1.529) RBD amino acid sequences using the FoldAmyloid program. These amyloidogenic regions in polypeptide chains play a crucial role in forming and aggregating amyloids. Compared to SARS-CoV-2 wild type, Delta Plus (B.1.617.2.1) contained three RBD mutations (Starr et al. 2021). Predicting amyloidogenic regions in RBD from its sequence demonstrated the amyloidogenic potential of the Delta Plus strain as 22% (Fig. 3a), slightly higher than the Wuhan variant, which showed the potential of 18%. However, the Omicron variant with 15 mutations (Cao et al. 2022) indicated a 32% potential for amyloid formation compared to the Wuhan strain, mainly due to significant enhancement of the amyloidogenic potential in residues 369–381, 389–396, and 487–498 (Fig. 3b). In addition, the amyloid-forming ability of the Omicron RBD along with a SUMO tag was lower (22%) than the Omicron RBD without fusion (32%) (Fig. 3c).

Fig. 3
figure 3

The amyloidogenic regions of RBDs in two major SARS-CoV-2 variants, including a the Delta Plus [green sections] and b the Omicron strains [blue sections], predicted by FoldAmyloid. c The yellow and blue sections represent the SUMO and Omicron RBD regions in SUMO-RBD, respectively, Asterisks indicate the presence of amino acid mutations in the amyloid regions

Full size image

Protein expression and purification

The most suitable conditions for protein expression were identified by investigating parameters such as temperature, induction time, and IPTG concentration. The effect of these factors on protein expression was qualitatively examined by SDS-PAGE (data not shown). The SARS-CoV-2 Delta plus and Omicron strain RBDs with a 6 × histidine tag and SARS-CoV-2 Omicron RBD with a SUMO fusion tag expressed at high levels in T7 SHuffle cells with 0.5 mM of IPTG at 22 °C for 18 h. Subsequently, the unfolded RBD proteins were purified (Fig. 4).

Fig. 4
figure 4

SDS-PAGE showing the expression and purification of recombinant RBD and SUMO-RBD from E. coli. a Lane M: protein markers; Lanes 1, 3, and 5: The Omicron SUMO-RBD (45 kDa), Omicron RBD (27 kDa), and Delta Plus RBD (27 kDa) inclusion bodies, respectively. Lanes 2, 4, and 6: soluble expression of total bacterial proteins of Omicron SUMO-RBD, Omicron RBD, and Delta Plus RBD, respectively. b Lane M: protein markers; Lane 1: purified SUMO-RBD. c Lane M: protein markers; Lane 1: purified Delta Plus RBD, and Lane 2: purified Omicron RBD

Full size image

ThT-positive aggregates of RBDs

Based on the ThT assay, which was accomplished using unfolded RBD with 8.0 M of urea, there were not any aggregates in the protein to bind ThT. Actually, no fluorescence emission at 485 nm was observed in the ThT fluorescence spectrum achieved. However, after the dialysis against Tris–HCl, the resulting samples were explored to bind the ThT dye and caused a significant fluorescence emission at 485 nM (Fig. 5).

Fig. 5
figure 5

ThT fluorescence spectra of Delta Plus (0.4 mg/mL) and Omicron (0.4 mg/mL) RBDs. RBD after 30 h of dialysis (red line), RBD after 10 h of dialysis (blue line), and RBD in the presence of 8 M urea (black line)

Full size image

The results of ThT assay also indicated that both the Delta Plus and Omicron strains of SARS-CoV-2 formed amyloid-like fibrils after 10-h dialysis and reached the highest levels of amyloid accumulation after 30 h. The fluorescence intensity of amyloid fibrils in the Omicron strain was slightly higher than that of the Delta Plus strain, which supports in silico prediction of this study, showing the presence of specific mutations in RBD regions of these strains.

The ThT assay was conducted for the RBD of the Omicron strain along with a SUMO fusion tag. The presence of the SUMO fusion tag effectively reduced the fluorescence intensity of amyloid accumulations in the Omicron SUMO-RBD (Fig. 6).

Fig. 6
figure 6

ThT fluorescence spectra of Omicron RBD (0.4 mg/mL) and Omicron SUMO-RBD (0.65 mg/mL). Omicron RBD after 30 h of dialysis (red line), Omicron SUMO-RBD after 30 h of dialysis (green line), and RBD in the presence of 8 M urea (black line)

Full size image

FTIR spectroscopy

FTIR is a well-established analytical technique for analyzing the secondary structure of proteins (Susi and Byler 1986). In the amide I region (1700−1600 cM−1), various secondary structures lead to variations in the C=O stretching frequency, influenced by their distinct molecular shapes and hydrogen bonding patterns (Kong and Yu, 2007). Mathematical techniques, such as Fourier self-deconvolution (FSD), can resolve overlapping bands, enabling a quantitative estimation of protein secondary structure (Barth 2007).The Amide I region of the resultant FTIR spectrum of RBD amyloid-like fibrils revealed a peak at 1657 cM−1, indicative of the α-helical secondary structure. The bands observed at 1621 and 1635 cM−1 can be attributed to inter- and intra-molecular β-sheet structures, respectively. Furthermore, the existence of β-turn structures is evidenced by the peaks at 1668 and 1690 cM−1 (Dong et al. 1992). After the analysis, it was determined that the secondary structure of the RBD amyloid-like fibrils consists of 48.4% β-sheet, 29.5% β-turn, and 5.6% α-helix (Fig. 7).

Fig. 7
figure 7

The FTIR spectrum (dotted line) and the peak deconvolution of the FTIR spectra for RBD amyloid-like fibrils, and summary of the secondary structure content obtained from FTIR spectrum analysis

Full size image

TEM results

Analyzing the morphological appearance of the RBD aggregates, formed after dialysis against Tris–HCl, by TEM displayed the presence of elongated, uncomplicated fibrils (Fig. 8a and b). Fibrils often include two or more thinner filaments coiling around one another, resulting in a twisted ultrastructural appearance with regular periodicity. The fibril diameter was between 9 and 11 nM, a characteristic of amyloid fibrils. The morphology of amorphous aggregates was visualized by the rapid removal of urea through a dilution technique (Fig. 8c), and that of soluble proteins was examined by TEM (Fig. 8d).

Fig. 8
figure 8

TEM images showing β-sheet-rich aggregates with fibrillar, amyloid-like morphology formed from RBD after 30 h of dialysis (a and b), amorphous aggregates (c), and soluble RBD proteins (d) with the Protein Data Bank (PDB) code: 6M0J (available at https://www.rcsb.org/)

Full size image

Discussion

RBD is a vaccine candidate against COVID-19 and a target for drug development (Yang et al. 2021). A large-scale Bacterially Expressed SARS-CoV-2 RBD could yield low-cost antigen for a subunit vaccine (Brindha et al. 2022; Liu et al. 2022), but it is crucial to note that its amyloidogenic properties may affect its effectiveness in vaccine development (Leung et al. 2023). Therefore, it is important to discover methods to decrease the aggregation of RBD amyloid-like fibrils. In a previous study, Sarr et al. demonstrated that fusion tags can reduce the aggregation of amyloid fibrils (Sarr et al. 2018). Solubility tags such as SUMO fusion tags can prevent the formation of inclusion bodies by stabilizing the proteins during expression (Butt et al. 2005). Therefore, the ThT assay was carried out for the RBD of the Omicron strain along with a SUMO fusion tag. The presence of the SUMO fusion tag effectively declined the ThT fluorescence intensity of amyloid accumulations in the RBD of the Omicron strain (Fig. 6), which is consistent with in silico predictions of the present study. This behavior can be attributed to the three-dimensional structure and solubility properties of the SUMO fusion tag, which have the potential to enhance the overall solubility of the fused protein (Butt et al. 2005). Furthermore, it is well-established that SUMO tags play a crucial role in facilitating proper protein folding. By promoting correct folding, proteins are less susceptible to misfolding and aggregation (Marblestone et al. 2006). Moreover, the bulky nature of the SUMO fusion tag may create steric hindrance, thereby limiting the interaction between amyloidogenic regions in the RBD (Lee et al. 2008).

According to Nystrom and Hammarstrom’s report, the full-length folded S-protein did not form amyloid fibrils. However, when the S protein was co-incubated with the protease neutrophil elastase (NE) in vitro for 24 h, it formed amyloid-like fibrils (Nystrom and Hammarstrom 2022). The key finding of this study is that the RBD was converted only into amyloid-like fibrils after the removal of urea (Fig. 9). Mutations can influence the severity of amyloid fibril formation, contingent upon various factors (Chiba et al. 2003; Greenbaum et al. 2005). For instance, amino acid mutations may destabilize the conformation of the protein, making it more likely to misfold and aggregate into amyloid fibrils (Marin-Argany et al. 2015). Furthermore, mutations can either accelerate or decelerate the rate of fibril formation, thereby influencing the overall severity of protein aggregation (Stefani and Dobson 2003). The ThT fluorescence intensity of amyloid accumulations differs among the strains of the SARS-CoV-2 virus can be attributed to specific mutations found in their RBD regions. The presence of amino acid mutations in the amyloid regions is indicated by an asterisk (*) in Fig. 3.

Fig. 9
figure 9

Single-step formation of RBD amyloid-like fibrils

Full size image

Conclusion

Overall, the study describes a cost-effective method for producing the SARS-CoV-2 RBD using E. coli T7 SHuffle cells, which offer low production costs, short generation times, and scalability. The RBD was successfully expressed with and without a SUMO tag and purified using a Ni Sepharose column. Biophysical techniques used in the present research found that the SARS-CoV-2 RBD can assemble itself by forming aggregates containing amyloid-like fibrils. Furthermore, in silico analysis of our study suggests that differences in the ThT fluorescence intensity of amyloid accumulations in Delta Plus and Omicron strains of the SARS-CoV-2 could be due to specific mutations in their RBD regions. The amyloidogenic properties of the RBD are a significant challenge for vaccine development. The SARS-CoV-2 RBD forms amyloid-like fibrils only after the removal of urea. Strategies such as incorporating fusion tags are necessary to reduce amyloid formation, and the findings of this research reveal that the presence of a SUMO fusion tag can significantly decrease the formation of RBD amyloid-like fibrils. Hence, this question is raised: what other factors can reduce the formation of amyloids, and which strategies can be utilized in the vaccine industry? Our study has limitations, primarily due to the need for experimental validation under diverse physiological conditions to fully understand the impact and behavior of RBD amyloid-like fibrils in vivo. It would be better for future research to focus on in vivo studies to assess the effects of RBD-formed amyloids on amyloid-related diseases and explore other inhibitors of amyloidogenesis.