Bacillus subtilis spores displaying RBD domain of SARS-CoV-2 spike protein

2.2. Construction of strains carrying coat protein – RBD fusion

Four recombinant strains harboring the RBD-coat protein fusions were prepared, and the RBD domain was placed either on the N-terminus or C-terminus of the coat protein. B. subtilis chromosome integration vector pDG1662 was used and plasmids were constructed by Gibson assembly [31]. DNA fragments encoding for coat protein genes were PCR amplified using the B. subtilis PY79 genomic DNA as a template and the oligonucleotides listed in Table 2. The spike glycoprotein rbd gene was amplified by PCR from the pGBW-m4046887 plasmid (Addgene plasmid No. 145730). In all PCR reactions Phusion ® High-Fidelity DNA Polymerase was used according to the standard guidline (https://lnk.sk/fpzd).

Plasmid pDGCotZ-RBD, with the RBD protein on the C-terminus of CotZ, was constructed using a 250-bp P_cotYZ promoter region amplified with primers PcotYZ_47F and PcotYZ_cotZCL_R; the 441-bp cotZ gene was amplified with the cotZ_CL_F and cotZ_CL_R primers and the 805-bp SARS-CoV-2 spike glycoprotein rbd gene was amplified with the RBD_cotZ_CL_F and RBD_cotZCL_R primers. PCR fragments were inserted into a pDG1662 plasmid that had previously been cleaved with the BamHI and EcoRI restriction enzymes.

To construct plasmid pDGRBD-CotZ, which carries the RBD region of the spike protein on the N-terminus of CotZ, we prepared 3 PCR fragments: a 250-bp DNA fragment of the P_cotYZ promoter region, which was amplified by PCR with the PcotYZ_47F and PcotYZ_NL_R primers; an 805-bp fragment corresponding to the SARS-CoV-2 spike glycoprotein rbd gene amplified with primers RBD_NL_F and RBD_NL_R; and a 441-bp cotZ gene amplified with primers cotZ_NL_F and cotZ_NL_R. PCR fragments were inserted into a pDG1662 plasmid which had been previously cleaved with BamHI and EcoRI.

Plasmid pDGCotY-RBD for the C-terminal fusion of CotY and RBD was constructed using two PCR fragments: a 729-bp fragment containing the cotY gene and its upstream sequence, including the promoter region, was amplified with the PcotYZ_65F and PcotYZ_cotY_CL_R oligonucleotides and an 805-bp PCR fragment containing the SARS-CoV-2 spike glycoprotein rbd gene, amplified with RBD_cotY_CL_F and RBD_cotY_CL_R. These fragments were inserted into a pDG1662 plasmid which had been previously cleaved with BamHI and EcoRI.

Plasmid pDGRBD-CotY was prepared by the insertion of three PCR fragments into a pDG1662 plasmid which had been previously cleaved with BamHI and EcoRI: a 250-bp DNA fragment of the promoter region amplified by PCR using the PcotYZ_65F and PcotYZ_cotY_NL_R oligonucleotides; an 805-bp PCR fragment corresponding to the SARS-CoV-2 spike glycoprotein rbd gene, amplified by PCR using the RBD_NL_F and RBD_NL_R primers and a 486-bp fragment containing the cotY gene, amplified with cotY_NL_F and cotY_NL_R. All cloned sequences were confirmed by DNA sequencing.

B. subtilis competent cells were transformed with recombinant plasmids and the clones generated after the integration of the cloned DNA at the amyE (amylase) locus of B. subtilis were selected on agar plates supplemented with 5 μg/ml chloramphenicol.

2.3. Expression and purification of the receptor binding domain (RBD)

To express the SARS-CoV-2 spike glycoprotein RBD, we used expression plasmid pET28a. The rbd, part of the SARS-CoV-2 spike gene, was amplified with specific primers (Table 2) from plasmid pGBWm4046887 and, after cleavage with the BamHI and XhoI restriction enzymes, cloned into the corresponding cloning site of a pET28a plasmid. E. coli BL21 (DE3) competent cells were transformed with the expression plasmid pET28RBD and grown in LB medium supplemented with kanamycin at 16 °C until OD₆₀₀ ≈ 0.6. Protein expression was induced by the addition of 1 mM isopropyl-β-D-1-thiogalactopyranoside (IPTG). The cells were collected after overnight cultivation at 16 °C and stored at − 80 °C until used. To isolate the protein, bacteria were lysed by sonication and the lysate was centrifuged at 73,000g for 30 min. After centrifugation of the crude cell lysate, the protein was purified using a 1 ml Ni-column. Because we found that the RBD protein produced in E. coli was mostly insoluble, we supplemented the 50 mM Tris/HCl solubilization buffer, pH 9.0, containing 150 mM NaCl with 8 M urea.

2.4. Preparation of spores

The sporulation of wild-type PY79 and recombinant strains was induced by the exhaustion method in Difco sporulation medium (DSM) [32]. A single colony was used to inoculate LB medium and cultivated overnight at 37 °C. This culture was then used to inoculate liquid LB and the bacteria were grown until OD₆₀₀ ≈ 1.0. Subsequently, 200 μl of this culture was spread on DSM plates and left at 37 °C for 3 days, followed by growth at room temperature for another 3 days. Spores were harvested and purified by intensive washing with ice-cold water for at least a week to remove cell debris and vegetative cells, after which spores were washed at least twice a week, and stored at 4 °C until use.

2.5. Adsorption reaction

The RBD protein was purified using a buffer containing 8 M urea. To obtain native, non-denatured protein, the urea was removed using a centrifugal concentrator (Amicon® ultra-4 centrifugal filter unit). For the adsorption reaction, 4 μg of RBD was added to a suspension of 2 × 10⁹ B. subtilis WT PY79 spores in 0.15 M PBS buffer pH 4.0. After one hour of incubation at room temperature, the spores were centrifuged for 10 min at 13,000g. The pellet was then resuspended in 0.15 M PBS pH 4.0 and stored at 4 °C for further experiments [33].

2.6. Expression of fusion proteins in B. subtilis and western blot analysis

To analyze the expression level of the fusion proteins in sporulating B. subtilis cells, we used the western blot technique. Bacterial samples were collected at different time points after the onset of sporulation. Cell aliquots at OD₆₀₀ ≈ 40 (a bacterial culture with OD₆₀₀ ≈ 3.5 contained approx. 4 ×10⁸ CFU) were suspended in solubilization buffer (50 mM Tris/HCl, pH 9.0, 150 mM NaCl, 0.4 mg/ml lysosome, cOmplete Protease Inhibitor Cocktail (Sigma-Aldrich)), and incubated at 37 °C for 20 min. After sonication lysates were treated with SDS buffer (62.5 mM Tris-HCl pH 6.8, 4% SDS, 5% glycerol, 2% β-mercaptoethanol, 0.003% bromophenol blue) for 10 min at 95 °C. Proteins were resolved by SDS-PAGE electrophoresis and subsequently analyzed by western blot analysis using a mouse monoclonal antibody against RBD protein (Abcam ab277628).

2.7. Immunofluorescence microscopy

Spores were collected and used for immunofluorescence microscopy as described by Harry et al. [34] with the following modifications. Spores were blocked with 2% bovine serum albumin in PBS for 20 min at room temperature. Subsequently, samples were incubated overnight at 4 °C with the mouse anti-SARS-CoV-2 Spike Glycoprotein RBD antibody at a dilution 1:1000. Samples were then washed three times in PBS and incubated for two hours with the Alexa Fluor 594 labelled anti-mouse IgG (ThermoFisher, A11005) at room temperature in the dark at a dilution 1:500. Samples were washed three times in PBS and, subsequently, 1 μl was applied on a glass slide, and the sample was then observed and photographed with an Olympus BX63 microscope equipped with an Andor Zyla 5.5 sCMOS camera (Olympus Europa SE & Co. KG., Hamburg, Germany). Olympus CellP imaging software was used for image acquisition and analysis. The final adjustment of the fluorescent images was done using Fiji ImageJ (open-source software).

3.1. Construction of recombinant B. subtilis spores

To obtain recombinant B. subtilis spores displaying the SARS-CoV-2 spike glycoprotein RBD on their surface, we prepared four integrative plasmids carrying genes for two variants in which the RBD was fused to either the N-terminus or the C-terminus of the coat protein. Fusion protein production was controlled by the native P_cotYZ promoter to ensure the proper timing of expression during the formation of the spore coat. We also inserted the flexible peptide linker GGGGS between the coat protein and the RBD (Fig. 1). All gene fusions were integrated into the coding sequence of the non-essential amyE locus of the B. subtilis chromosome and the correct clones were selected by chloramphenicol resistance. All recombinant strains were used for further analysis. The recombinant strains and their isogenic parental strain PY79 showed comparable sporulation efficiency, indicating that the presence of the fusion did not significantly affect sporulation efficiency.

Open in a separate window

Fig. 1

Schematic representation of fusion proteins for spore display. Construction of four fusions carrying the genes of both hybrid variants, the N-terminal and C-terminal fusion of coat protein CotZ or CotY to RBD. Fusion protein expression was controlled by the native P_cotYZ promoter. The flexible peptide linker GGGGS was inserted between the coat protein and RBD.

3.2. Expression of fusion genes and western blot analysis

Next, we analyzed the production of hybrid proteins during sporulation. Proteins from the B. subtilis PY79 wild-type, CotZ-RBD, RBD-CotZ, CotY-RBD, and RBD-CotY strains were fractionated in 10% polyacrylamide gels containing SDS. The protein lysates obtained from the 2nd, 4th, 6th, 8th and 24th hours of sporulation were analyzed. An equal amount of cells corresponding to an OD₆₀₀ ≈ 40 was taken for analysis. As the time after the onset of sporulation increased, we could observe a gradual decrease in the amount of proteins in the SDS gels (Fig. 2) which corresponds to an increased proportion of spores in the cell culture and the lower solubility of the spore coat proteins compared to the proteins of vegetative cells. To test the expression level of the fusion genes, we performed a western blot analysis with the anti-SARS-CoV-2 spike glycoprotein RBD antibody. The pattern obtained from the recombinant strains varies significantly (Fig. 2), but in contrast to the wild-type, which served as a control and did produced no western blot signal (Fig. 2F), RBD signals are present in all recombinant strains 4 and 6 h after the onset of sporulation. Furthermore, we observed that the fusion proteins were rather degraded, but several bands corresponding to higher molecular weights were observed, most likely representing oligomerized forms of the hybrid proteins. The C-terminal fusion of CotZ and RBD resulted in the formation of large oligomers but at hours 8 and 24 these signals were lost. We assume that the hybrid CotZ-RBD was incorporated into the spore coat as part of the insoluble spore coat structure that cannot be solubilized using the common techniques [42], and thus, we were not able to detect them by western blotting. When RBD was attached to the N-terminus of CotZ, we did not see any bands corresponding to the larger proteins, suggesting that this arrangement could preclude the formation of CotZ-RBD oligomers. When RBD was attached to the N-terminus of CotY, in addition to the strong signals representing degradation products, a weak band corresponding to the high molecular weight oligomers characteristic of the CotY protein itself was detected on the SDS-PAGE gel. Using ImageJ (open-source software), we calculated that this macro structure represents at least 3% of the hybrid RBD-CotY protein. In contrast, the C-terminal fusion of CotY and RBD produced a wide range of signal sizes, even some with high molecular weight, but without the typical CotY pattern, which is characterized by the existence of discrete bands corresponding to high molecular weight oligomers at the top of the SDS gel [35]. These bands appear to be the result of adsorption of aggregated and/or degraded fusion protein on the spore surface. We hypothesize that the fusion protein was not incorporated into CotY structures: even after the 24 h of sporulation, when the total amount of solubilized protein is low, we observed a strong signal corresponding to the RBD protein which most likely resulted from the adsorption of RBD and which could be more easily released into solubilization buffer. Western blot analysis of intensively washed spores did not show any large oligomers of RBD-containing fusion proteins, only some digestion products. We hypothesize that if hybrid proteins form the structures with the intrinsic coat proteins, they will become part of the coat fraction that cannot be easily solubilized. We also do not anticipate that there would be a large number of hybrid protein molecules incorporated directly into the CotY or CotZ 2D crystals formed by these proteins within the spore coat [43]. Fig. 4 reflects an ideal model situation for the incorporation of the fusion protein CotY-RBD into the spore crust. However, the RBD protein is twice as big as either CotY or CotZ and we expect that there could be some steric hindrance in 2D crystal formation from these two proteins. This fact is also confirmed by the diversity of the western blot results (Fig. 2). Nevertheless, RBD is successfully displayed on the spore surface and, even when attached to the spore surface by adsorption, it is captured tightly enough that it cannot be removed by vigorous spore washing (Fig. 3A – WT-RBD).‬

Open in a separate window

Fig. 2

SDS-PAGE and western blot analysis of proteins extracted from spores of the B. subtilis wild-type (WT) strain and recombinant strains CotZ-RBD, RBD-CotZ, CotY-RBD, and RBD-CotY at different sporulation times. Cell lysates (cell cultures with OD₆₀₀ ≈ 3.5 contained approx. 4 ×10⁸ CFU and an aliquot of OD₆₀₀ ≈ 40 was used) were resolved on 10% SDS polyacrylamide gels (A-E). Fractionated proteins were subsequently transferred onto a nitrocellulose membrane and probed with a monoclonal antibody against the RBD protein (F-J). Cells from the onset of sporulation (lane 0) and the 2nd (lane 2), 4th (lane 4), 6th (lane 6), 8th (lane 8), and 24th (lane 24) hours of sporulation were analyzed. Molecular weight markers (in kilodaltons) are indicated. Stars mark the size of the fused proteins.

Open in a separate window

Fig. 3

Immunofluorescence microscopy of the B. subtilis PY79 wild-type spores, recombinant derivatives and spores with the SARS-CoV-2 spike glycoprotein RBD adsorbed on the surface and their fluorescence intensity. A: The left panel shows phase contrast images of wild-type strain PY79, CotZ-RBD, RBD-CotZ, CotY-RBD, RBD-CotY and WT-RBD and immunofluorescence microscopy images are shown in the right panel. The spores were incubated with the mouse anti-SARS-CoV-2 Spike Glycoprotein RBD antibody, followed by an incubation with the anti-mouse IgG- Alexa Fluor 594 conjugate. The scale bar represents 2 µm. B: Fluorescence intensity quantification of recombinant strains and spores with the SARS-CoV-2 spike glycoprotein RBD adsorbed on the surface compared to WT using ImageJ software. Significant p-values between WT and the recombinant samples from two-tailed unpaired t-tests are depicted above the plot; the threshold for statistical significance is as follows: stars indicate significantly different values, where ns (not significant) corresponds to p-value> 0.05, * to p ≤ 0.05, * * to p ≤ 0.01, * ** to p ≤ 0.001 and * ** * to p ≤ 0.0001. The size of the sample from each strain is 16 (n = 16).

Open in a separate window

Fig. 4

Model of the CotY-RBD fusion protein structure. A: Three-dimensional surface representations of the 2D CotY crystal into which the predicted AlphaFold 3D structure of CotY is superimposed. The three-dimensional reconstruction of the CotY crystals is adapted from Jiang et al. [35]. The RBD domain of the SARS-CoV-2 spike protein structure is adapted from Protein DataBank structure 6M0J. B: The monomer structure of the CotY protein as predicted by AlphaFold [46]. AlphaFold produces a per-residue confidence score (pLDDT) between 0 and 100. pLDDT ≥ 90 represents very high model confidence (dark blue), while the value 90>pLDDT ≥ 70 is classified as confident (light blue). 70>pLDDT ≥ 50 indicates low confidence (yellow), pLDDT<50 belongs to very low confidence (orange). Some regions with low pLDDT may be unstructured in isolation.

This work was supported by VEGA – Grant No. 2/0001/21 from the Slovak Academy of Sciences, a Grant from the Slovak Research and Development Agency under contract APVV-18–0104 to IB. We thank Jacob Bauer for critically reading the manuscript.