Human super antibody to viral RNA-dependent RNA polymerase produced by a modified Sortase self-cleave-bacteria surface display system

Abstract

Background: RNA-dependent RNA polymerase (RdRp) is a good target of anti-RNA virus agents; not only it is pivotal for the RNA virus replication cycle and highly conserved among RNA viruses across different families, but also lacks human homolog. Recently, human single-chain antibody (HuscFv) that bound to thumb domain of hepatitis C virus (HCV) RNA-dependent RNA polymerase (functionalized NS5B protein) was produced and engineered into cell-penetrating antibody (super antibody) in the form of cell-penetrating peptide (penetratin, PEN)-linked HuscFv (PEN-HuscFv34). The super antibody was produced and purified from inclusion body (IB) of a pen-huscfv34-vector-transformed Escherichia coli. The super antibody inhibited replication of alpha- and beta- coronaviruses, flaviviruses, and picornaviruses that were tested (broadly effective); thus, it has high potential for developing further towards a pan-anti-RNA virus agent. However, production, purification, and refolding of the super antibody molecules from the bacterial IB are laborious and hurdles to large-scale production. Therefore, in this study, Sortase-self-cleave method and bacteria surface display system were combined and modified for the super antibody production.

Methods and results: BL21 (DE3) ΔA E. coli, a strain lacking predominant outer membrane protein (OmpA) and ion and OmpT proteases, that displayed a membrane-anchored fusion protein, i.e., chimeric lipoprotein (Lpp')-OmpA', SUMO, Sortase protease, Sortase cleavage site (LPET↓G) and PEN-HuscFv34-6× His was generated. The soluble PEN-HuscFv34-6× His with glycine at the N-terminus could be released from the E. coli surface, simply by incubating the bacterial cells in a Sortase-cleavage buffer. After centrifugation, the G-PEN-HuscFv34-6× His could be purified from the supernatant. The purified G-PEN-HuscFv34-6× retained original cell-penetrating ability (being super antibody) and the broadly effective anti-RNA virus activity of the original IB-derived-PEN-HuscFv34.

Conclusion: The functionalized super antibody to RNA virus RdRp was successfully produced by using combined Sortase self-cleave and bacterial surface display systems with modification. The display system is suitable for downstream processing in a large-scale production of the super antibody. It is applicable also for production of other recombinant proteins in soluble free-folding form.

Keywords: Bacterial surface display; Cell-penetrating peptide (CPP); Human single-chain antibody variable fragment (HuscFv); RNA viruses, SARS-CoV-2; RNA-dependent RNA polymerase (RdRp); SUMO; Sortase; Super antibody.

Fig. 1

Preparation of E. coli transformants that displayed LPP′-OmpA′-SUMO-Δ59 Sortase A-LPETG-PEN-HuscFv34-6× His on the surface. (a) Diagram of a DNA construct coding for Lpp′-OmpA′-SUMO-Δ59 Sortase A-LPETG-PEN-HuscFv34. (b) Amplicons of Lpp′-OmpA′ (456 bp), SUMO (288 bp), Δ59 Sortase A (447 bp) and LPETG-pen-huscfv34 (840 bp). (c) Amplicons of Lpp′-OmpA′-Δ59 Sortase A-LPETG-pen-huscfv34 contig (∼2000 bp, arrowhead) in the DH5α competent E. coli clones 1–15 that had been transformed with Lpp′-OmpA′-Δ59 Sortase A-LPETG-pen-huscfv34-pET28b⁺ vector (lanes 1–15, respectively); lanes M of (b) and (c) are DNA size standard; lane N of (c) is negative DNA template control; lane P of (c) is positive DNA template control; numbers at the left of (b) and (c) are DNA sizes in bp. (d) Flow cytometric analysis of transformed BL21 (DE3) ΔA competent E. coli clones 1–5 that were stained by mouse monoclonal anti-HuscFv34 antibody and goat anti-mouse Ig-Alexa Fluor 488 conjugate for detection of Lpp′-OmpA′-SUMO-Δ59 Sortase A-LPETG-PEN-HuscFv34-6× His fusion protein on the cell surface. Cells were counter-stained by DAPI to indicate nuclei. Q1 and Q2, transformed E. coli cells displaying Lpp′-OmpA′-SUMO-Δ59 Sortase A-LPETG-PEN-HuscFv34-6× His fusion protein; Q3 and Q4, E. coli cells that did not display the fusion protein. Q2 and Q3, intact bacterial cells with nuclei; Q1 and Q4, bacterial cells without nuclei or dead cells

Fig. 2

Indirect ELISA titers of mAbs in spent culture supernatants of hybridomas raised against HuscFv34. HuscFv34 (1 μg per well) was used as antigen in the indirect ELISA. Homogenate of original BL21 (DE3) E. coli was used as control antigen. Mouse immune serum (mouse anti-HuscFv34 pAb) at dilution 1:5000 served as positive antibody control. X axis, culture supernatant of hybridoma clones; Y axis indirect ELISA OD 405 nm against blanks (wells added with PBS instead of antibody). The dotted line is the arbitrarily cut-off level (OD 405 nm = 0.2) between positive and negative indirect ELISA results

Fig. 3

Optimal IPTG concentration in induction of LPP′-OmpA′-SUMO-Δ59 Sortase A-LPETG-PEN-HuscFv34-6× His expression on the bacterial surface. The transformed BL21 (DE3) ΔA competent E. coli was grown under induction of different IPTG concentrations (0.025, 0.05, 0.1, 0.5 and 1.0 mM) at 37 °C. The bacterial cells were fixed with 4% paraformaldehyde, washed, incubated with mouse mAb to HuscFv34, washed, added with goat anti-mouse Ig-Alexa Fluor 488 conjugate, washed, stained with DAPI, washed, and resuspended in 1% paraformaldehyde. The PEN-HuscFv34-6× His displayed on bacterial cells was determined by flow cytometry. Q1 and Q2, transformed E. coli cells displaying Lpp′-OmpA′-SUMO-Δ59 Sortase A-LPETG-PEN-HuscFv34-6× His fusion protein; Q3 and Q4, E. coli cells that did not display the fusion protein. Q2 and Q3, intact bacterial cells with nuclei; Q1 and Q4, bacterial cells without nuclei or dead cells

Fig. 4

Suitable temperature for growing 0.025 mM IPTG-induced transformed bacteria to express the fusion protein. The transformed BL21 (DE3) ΔA competent E. coli was grown under 0.025 mM IPTG induced condition at 16 and 37 °C for 5 h. The cells were then stained as in Fig. 3 and analyzed for percent surface displayed LPP′-OmpA′-SUMO-Δ59 Sortase A-LPETG-PEN-HuscFv34-6× His by flow cytometry. The transformed bacteria cultured at 37 °C expressed higher percentage of the cells displaying the fusion protein on surface (61.1%) than when they were grown at 16 °C (28.1%)

Fig. 5

Large-scale culture of transformed BL 21 (DE3) ΔA E. coli for preparing G-PEN-HuscFv34-6× His. (a) The percentage of intact bacterial cells (grown in 1000 mL broth for 5 h) that displayed the fusion protein on their surface as determined by flow cytometry (Q2 of the right panel). (b) Left sheet: Western blot patterns and intensities of G-PEN-HuscFv34-6× His bands (between 25 and 35 kDa, arrowhead) in supernatants of the bacteria in (a) after incubating in Sortase cleavage buffer for 0, 1, 2, 3, 4 and 5 h (lanes 1–6, respectively). Right sheet, Western blot patterns of the proteins in the supernatants that were bound by anti-SUMO (arrowhead); this band should be SUMO-Δ59 Sortase-LPET- that still linked to LPP′-OmpA′ on the membrane of lysed bacterial cells. (c) Left sheet: Lane 1 is SDS-PAGE and CBB stained pattern of the purified G-PEN-HuscFv34-6× His (arrowhead) from the supernatant of the transformed bacteria suspended in Sortase cleavage buffer for 2 h; middle and right sheet show Western blot patterns of the purified G-PEN-HuscFV34-6× His probed with anti-His antibody (lane 2) and mouse monoclonal anti-HuscFv34 (lane 3). Lanes M of (b) and (c) are protein molecular masses in kDa

Fig. 6

Biocompatibility and cell-penetrating ability of the G-PEN-HuscFv34-6× His. (a) Percent viability of cells: Vero (blue bars)/Vero E6 (olive bars) after treatment with different concentrations of G-PEN-HuscFv34-6× His at 37 °C in 5% CO₂ incubator for 24 h. (b) Left block: Intracellular localization of the G-PEN-HuscFv34-6× His after adding 0.5 μM to the Vero E6 cells, incubated, fixed, permeated and probed with mAb to HuscFv34 followed by goat anti-mouse Ig-Alexa Fluor 488 and DAPI. The G-PEN-HuscFv34-6× His appears green; the nuclei stained blue; the cell contours are indicated by dotted lines. Right block, negative control (cells in medium instead of the G-PEN-HuscFv34-6x His and probed with mAb to HuscFv34)

Fig. 7

Fold-differences in viral RNAs recovered from G-PEN-HuscFv34-6× His treated-infected cells compared to medium treated-infected cells. Vero E6 cells were infected with Betacoronavirus, i.e., SARS-CoV-2 (Wuhan wildtype or B1.1.529 Omicron variant) while Vero cells were infected with flaviviruses (DENV serotypes 1–4, ZIKV) or Alphacoronavirus (PEDV) for 1 h. The fluids in wells were discarded and the infected cells were treated with G-PEN-HuscFv34-6× His (0.5 μM) for 18 h (SARS-CoV-2 and PEDV) or 48 h (DENV and ZIKV). Infected cells in medium alone served as controls. The cells were harvested for RNA extraction and the viral RNAs were quantified by qRT-PCR. ∗∗∗∗, p < 0.0001