BanLec-eGFP Chimera as a Tool for Evaluation of Lectin Binding to High-Mannose Glycans on Microorganisms

Fluorescently labeled lectins are useful tools for in vivo and in vitro studies of the structure and function of tissues and various pathogens such as viruses, bacteria, and fungi. For the evaluation of high-mannose glycans present on various glycoproteins, a three-dimensional (3D) model of the chimera was designed from the crystal structures of recombinant banana lectin (BanLec, Protein Data Bank entry (PDB): 5EXG) and an enhanced green fluorescent protein (eGFP, PDB 4EUL) by applying molecular modeling and molecular mechanics and expressed in Escherichia coli. BanLec-eGFP, produced as a soluble cytosolic protein of about 42 kDa, revealed β-sheets (41%) as the predominant secondary structures, with the emission peak maximum detected at 509 nm (excitation wavelength 488 nm). More than 65% of the primary structure was confirmed by mass spectrometry. Competitive BanLec-eGFP binding to high mannose glycans of the influenza vaccine (Vaxigrip®) was shown in a fluorescence-linked lectin sorbent assay (FLLSA) with monosaccharides (mannose and glucose) and wild type BanLec and H84T BanLec mutant. BanLec-eGFP exhibited binding to mannose residues on different strains of Salmonella in flow cytometry, with especially pronounced binding to a Salmonella Typhi clinical isolate. BanLec-eGFP can be a useful tool for screening high-mannose glycosylation sites on different microorganisms.


Introduction
Lectins are carbohydrate-binding proteins involved in many biological processes like recognition and binding of glycans, host-pathogen interactions, cell targeting, cell-cell communication, induction of apoptosis, cancer metastasis, and differentiation [1,2]. Many different lectins are known and their glycan-binding specificities characterized (consortium for functional glycomics at http://www.functionalglycomics.org/). Antimicrobial activities of lectins toward viruses, fungi, and bacteria have been reported [3], with plant lectins as the well-characterized group.
The first report on the isolation of banana lectin (BanLec) was from Musa paradisiac [4]. It belongs to the jacalin-related superfamily (mJRL) of lectins, which recognizes and binds to high mannose glycans [5]. It is homodimeric in solution, with a molecular mass of about 26 kDa. Each subunit consists of twelve β-strands arranged in a β-Prism-I fold [6,7]. BanLec is an IgG4-inducing antigen [8] and a potent inhibitor of human immunodeficiency virus (HIV) replication [9,10]. BanLec binds to the glycosylated proteins in the viral envelope and inhibits its cellular entry by blocking glycan-mediated interactions with the The three-dimensional (3D) model of the BanLec-eGFP chimeric structure is designed from the crystal structure of a recombinant banana lectin, Protein Data Bank (PDB) entry 5EXG [10] (chain A containing 141 amino acids) and the crystal structure of an enhanced green fluorescent protein, PDB entry 4EUL [27] (chain A containing 239 amino acids) by applying the molecular modeling and molecular mechanics to optimize the structure. All modifications in the sequence of BanLec-eGFP chimera relative to the used crystal structures, such as replacing the residues, removing the excessive, and adding the missing residues, were done in Discovery Studio 19.1.0. [28] The N-terminal part (MVS) and Cterminal fragment (GMDELYK) were modeled in a random coil conformation but ended up with a few turns and bends.
A glycine-serine-glutamate-phenylalanine (GSEF) linker was constructed between the BanLec and eGFP to prevent steric obstacles. The GSEF linker was modeled as a coil extension to the β-strand secondary structure of the C-terminus of eGFP in Discovery Studio 19.1.0 [28]. The computer-generated BanLec-eGFP chimeric design was obtained by connecting the three fragments and adding the hydrogen atoms, followed by the full structural minimization in NAMD 2.9 [29] with CHARMM22 force field [30] in implicit water model for 300 ps. The optimized resulting structure was strain-free, fully minimized, and relaxed without any bond, angle, or torsion angle restraints, and van der Waals atomic clashes. The final BanLec-eGFP structure comprised 384 amino acids, which encode for the BanLec, linker GSEF, and enhanced green fluorescent protein (eGFP). The strategy of cloning eGFP in the vector that already contains the BanLec gene sequence was conducted using NdeI (Thermo Fisher Scientific, Waltham, MA, USA) and BamHI (Thermo Fisher Scientific) restriction enzymes.

Expression of BanLec-eGFP
Escherichia coli BL21 (DE3)-pLysS (Agilent Technologies Inc., La Jolla, CA, USA) strain was employed as host for the expression of recombinant chimera. Bacteria were grown in LB medium containing 100 µg/mL of ampicillin and 25 µg/mL of chloramphenicol until optical density (OD) at 600 nm of 0.6-0.8 was reached. The BanLec-eGFP synthesis was induced by adding isopropyl-D-thiogalactopyranoside (IPTG) (VWR International, LLC, Radnor, PA, USA) at a concentration of 1 mM, following incubation for 5 and 12 h at 25 • C. After expression, the cell pellet obtained by centrifugation (3000× g for 20 min) was resuspended in 150 mM NaCl, 20 mM tris(hydroxymethyl)aminomethane (TRIS) HCl pH 7.6, and sonicated by the ultrasonic horn. The supernatant obtained by centrifugation (3000× g for 20 min) was used for downstream processing of the expressed chimera.

Purification of Recombinant BanLec-eGFP
BanLec-eGFP was purified by combining affinity chromatography on branched α-1,6 and α-1,3 glucan polymer (Sephadex G-75 superfine, column volume (CV) = 10 mL) and ion-exchange chromatography (HiTrap ANX column, 1 mL, GE Healthcare, Little Chalfont, UK). After pre-equilibration with 20 mM TRIS HCl pH 7.6, 150 mM NaCl (TRIS buffered saline, TBS), the matrix (10 mL) was incubated with the protein solution at room temperature for 1.5 h. After the transfer to a column and thorough washing with five column volumes (CV) of TBS, bound BanLec-eGFP was eluted from the matrix with 0.3 M glucose in TBS and collected in tubes (1.5 mL).
After affinity chromatography, collected samples were purified by ion-exchange chromatography connected to Äkta Purifier (GE Healthcare) using a combination of equilibration buffer (10 mM TRIS HCl, pH 8.0) and elution buffer (10 mM TRIS HCl, 0.5 M NaCl, pH 8.0), at room temperature and a flow rate of 1 mL/min. The column pressure limit was 0.3 MPa. Bound proteins were eluted using a linear gradient of elution buffer (0-100% (v/v), for 10 column volumes). The BanLec-eGFP was eluted at 50% of the elution buffer.
The wild type BanLec (GenBank accession number EU0556441) and BanLec H84T mutant were produced and purified according to the same protocol as BanLec-eGFP.

Analysis of the BanLec-eGFP Structure by CD Spectroscopy, Fluorescent and Mass Spectrometry
For assessing the secondary structures of the recombinant BanLec-eGFP, Circular dichroism (CD) spectra were recorded at 25 • C on Jasco J-815 circular dichroism spectropolarimeter (JASCO Inc., Easton, PA, USA) in the wavelength range from 185 to 260 nm in a cuvette with a path length of 0.1 mm. The CD data were collected in intervals of 0.1 nm, at a velocity of 50 nm/min. BanLec-eGFP was dialyzed against 20 mM TRIS HCl, pH 7.6 before spectral analysis. Purified BanLec-eGFP (0.92 mg/mL, 80 µL) was used for the study.
The fluorescence of BanLec-eGFP was measured using a spectrofluorimeter FLUORO MAX-4 Jobin Yvon (HORIBA, Kyoto, Japan). The emission spectra were recorded between 400-650 nm. The concentration of BanLec-eGFP used for the analysis was 0.92 mg/mL in 20 mM TRIS HCl buffer, pH 7.6.
To confirm the amino acid sequence, BanLec-eGFP was analyzed by mass spectrometry. The protein sample was separated by SDS-PAGE, followed by Coomassie Blue staining of the gel. The protein band corresponding to BanLec-eGFP was excised and destained by 100 mM NH 4 HCO 3 /50% acetonitrile (ACN). The gel pieces were then reduced for 40 min by 20 mM dithiothreitol and subsequently alkylated with 50 mM iodoacetamide for 30 min in the dark. The samples were digested with sequencing grade trypsin according to the manufacturer's instructions (Promega, Mannheim, Germany). Before MS analysis, samples were cleaned over the C18 zip tip (Millipore, MA, USA). For MS analyses of the tryptic fragments, α-cyano-4-hydroxycinnamic acid as a matrix was used. Tryptic maps were obtained on matrix assisted laser desorption ionization-time of flight/time of flight (MALDI-TOF/TOF) Ultraflextreme mass spectrometer (Bruker) operating in a reflectron positive mode and a peptide calibration standard mixture (Bruker 206195) was used for mass calibration.
The obtained peptide masses were searched against the NCBI protein sequence database using the MASCOT program (http://www.matrixscience.com/search_form_ select.html).

Competitive FLLSA (Fluorescence-Linked Lectin Sorbent Assay)
Recombinant BanLec-eGFP chimera was employed in competitive FLLSA for the influenza virus detection to test its applicability in high-mannose binding assays. The assay was performed by inhibiting the BanLec-eGFP binding to the immobilized hemagglutinin from a Vaxigrip™ vaccine (Sanofi Pasteur, Lyon, France) with monosaccharides (mannose or glucose), wild type BanLec, and BanLec H84T mutant. Black 96-Well Immuno plates (Thermo Fisher Scientific) were coated with hemagglutinin (0.5 µg per well) in 15 mM Na 2 CO 3 /35 mM NaHCO 3 , pH 9.5 overnight at 4 • C. Plates were washed with tPBS (phosphate buffered saline, PBS with 0.05% Tween 20). Then, the plates were blocked with 3% bovine serum albumin (BSA) in tPBS for 1 h at 37 • C. After washing, BanLec-eGFP (50 µg per well) was mixed without (control) or with different amounts of monosaccharides (0.01, 0.1, 1, 10, 100, 1000, 5000 mM) or BanLec inhibitors (BanLec wt, or BanLec H84T; 0.01, 0.1, 1, and 10 µg per well), which were added to the wells, and incubated in the dark for 2 h at room temperature. The FLLSA was performed in triplicate and the emission was measured at 535 nm on a fluorescent ELISA reader. Percent of inhibition was calculated as and IC50 value was determined as a concentration of inhibitor at 50% of inhibition.

Assessment of Fluorescence-Linked Lectin Binding to Bacterial Cells with Flow Cytometry
For the assessment of recombinant BanLec-eGFP binding to bacterial cells, 7 strains of proteobacteria, species Salmonella enterica, subspecies enterica were used. The bacteria used were: serovar Enteritidis ATCC 13076; serovar Enteritidis clinical isolate E; serovar Typhimurium isolate B; serovar Typhimurium ATCC 14028; serovar Typhimurium isolate 2865; serovar Typhi clinical isolate 1243; and serovar Typhi clinical isolate 12. The bacteria were cultivated in nutritious broth for 24 h and washed twice in PBS, by centrifugation at 2000× g for 15 min. Cell density was adjusted to be the same for all strains, giving optical density of 0.15 at 620 nm in a 100 µL volume in ELISA plate. The staining with BanLec-eGFP was done by aliquoting the bacteria and adding BanLec-eGFP. Incubation period was 15 min, at 4 • C, in the dark; followed by washing with PBS and centrifugation 2000× g for 15 min two times. Finally, the bacteria were resuspended in PBS with 0.4% formaldehyde (PBS-FA).
Inhibition with monosaccharides was performed on serovar Typhimurium isolate 2865. The bacterial suspensions were centrifuged for 15 min 2000× g and PBS was removed. The bacteria were vortexed and different amounts of monosaccharides were added. This was followed by the addition of 1 µg BanLec-eGFP. After incubation for 15 min, at 4 • C, in the dark, the bacteria were washed twice by adding PBS, centrifuged as before, and resuspended in PBS-FA. Each concentration point was done in quadruplicate.
The staining with BanLec-B (biotin labeled BanLec) was done the same way, except that after washing the unbound BanLec-B 0.2 µg streptavidin-phycoerythrin (PE) (eBioscience) was added and after 15 min of incubation, the bacteria were washed as before.
The signal was analyzed with FACSVerse with BD FACSuite software (BD Biosciences). The graphs were drawn and results analyzed with OriginPro 8 and GraphPad Prism software.

Design, Cloning, Expression, and Purification of Recombinant BanLec-eGFP
The model of BanLec-eGFP chimera designed in silico and fully optimized with the CHARMM22 force field [30] is depicted in Figure 1. Both carbohydrate-binding sites of BanLec are highlighted, according to [32], for further analysis and discussion. The binding site 1 consists of the GG loop (14-15), GKFLD loop (129-133) with additional S16 andV88 residues, while the binding site 2 includes the GG loop (59-60), GDVVD loop (34-38) with additional T61 and F131 residues, which interact with the bound glycans stabilizing the protein-receptor complex.
Enhanced green fluorescent protein (eGFP) is built up as the β-barrel tertiary structure with the barrel's fluorescent chromophore. The chromophore is generated by cyclization of T66, Y67, and G68 backbone atoms, and its central position is shown in Figure 1A. On the other hand, the β-strands with loops and turns define the tertiary structure of BanLec.  BanLec-eGFP chimera structure appears to be stable over the time of simulation, mostly retaining the secondary and tertiary structure of both domains. The position of eGFP should not affect the binding capacity of BanLec for the glycans.
Gene for eGFP was cloned in the pET-23b-BL vector, and the resulting protein was expressed in E. coli BL21 (DE3)-pLysS. The addition of IPTG induced the expression of BanLec-eGFP. The produced recombinant protein was purified by combining affinity chromatography on a Sephadex G-75 superfine matrix and ion-exchange chromatography on an ANX column. The BanLec-eGFP chimeric protein was confirmed with Western blot using the polyclonal rabbit anti-rBanLec antibodies and alkaline phosphatase labeled anti-rabbit IgG antibodies ( Figure 1B).
The yield of BanLec-eGFP after purification was about 16 mg of protein per liter of cell culture. The purified protein showed a band of approximately 42 kDa in SDS-PAGE under reducing conditions. The theoretical molecular mass of BanLec-eGFP calculated by the Compute pI/Mw tool (https://web.expasy.org/compute_pi/) was 41,925.32 Da.

Physical-Chemical Characterization of BanLec-eGFP
Physical-chemical characterization of BanLec-eGFP chimeric protein was performed by fluorescent spectroscopy, CD spectroscopy, and mass spectrometry. Fluorescent spectroscopy analysis revealed an emission maximum of BanLec-eGFP at 509 nm ( Figure 1C).
Secondary structures of BanLec-eGFP were analyzed by CD spectroscopy in the far-UV region from 185-260 nm ( Figure 1D). The protein's secondary structures were predicted by the K2D program (http://cbdm-01.zdv.uni-mainz.de/~andrade/k2d2/), which revealed the presence of 1.64% of α-helix and 40.58% of β-strand in the BanLec-eGFP. The spectra of BanLec-eGFP showed well-defined secondary structures in the construct.
A peptide mass fingerprint was employed to confirm the primary structure of the BanLec-eGFP. MS spectrometry data were compared with theoretically obtained data using amino acid sequences (Source: https://web.expasy.org/peptide_cutter/). Peptides are shown in Table 1. By mass analysis, more than 65% of the primary sequence was confirmed ( Figure 1E). While only one amino acid V88 belonging to the binding site 1 is experimentally verified, all amino acid residues from the binding site 2 are confirmed by the MS analysis.

Competitive Inhibition of BanLec-eGFP Binding to Influenza Vaccine High-Mannose Glycans
BanLec-eGFP reactivity towards high-mannose glycans present on the influenza virus vaccine was assessed by competitive FLLSA inhibition with monosaccharides mannose and glucose, and wt BanLec and BanLec H84T. This assay was done with two controls, without hemagglutinin (min) and inhibitor (max). The IC50 value for wt BanLec was 9.12 × 10 −8 M, while IC50 for BanLec H84T was 2.35 × 10 −7 M, indicating 2.6 higher inhibition with wt BanLec in the employed experimental settings. In accordance with our previous findings, mannose was more efficient inhibitor for BanLec carbohydrate binding ( Figure 2B) than glucose [11]. rus vaccine was assessed by competitive FLLSA inhibition with monosaccharides mannose and glucose, and wt BanLec and BanLec H84T. This assay was done with two controls, without hemagglutinin (min) and inhibitor (max). The IC50 value for wt BanLec was 9.12 × 10 −8 M, while IC50 for BanLec H84T was 2.35 × 10 −7 M, indicating 2.6 higher inhibition with wt BanLec in the employed experimental settings. In accordance with our previous findings, mannose was more efficient inhibitor for BanLec carbohydrate binding ( Figure 2B) than glucose [11].

Flow Cytometric Detection of BanLec-eGFP binding to bacteria
The binding of BanLec-eGFP was detected to certain strains of Salmonella. As can be seen in Figure 3, the binding was not uniform and showed differences between the strains, with especially pronounced binding to serovar Typhi isolate 12 and serovar Typhimurium isolate 2865, Figure 3A,B. The binding was also detected to Salmonella enterica subsp. enterica serovar Typhimurium ATCC 14028 and serovar Typhi isolate 1243, Figure 3C,D.

Flow Cytometric Detection of BanLec-eGFP binding to bacteria
The binding of BanLec-eGFP was detected to certain strains of Salmonella. As can be seen in Figure 3, the binding was not uniform and showed differences between the strains, with especially pronounced binding to serovar Typhi isolate 12 and serovar Typhimurium isolate 2865, Figure 3A,B. The binding was also detected to Salmonella enterica subsp. enterica serovar Typhimurium ATCC 14028 and serovar Typhi isolate 1243, Figure 3C,D. Essentially no binding was detected for the other strains tested, Figure 3E-G. The dependence of relative fluorescence (FITC-A Median) on the concentration of BanLec-eGFP used to stain different strains of Salmonella is shown in Figure 3H. Essentially no binding was detected for the other strains tested, Figure 3E-G. The dependence of relative fluorescence (FITC-A Median) on the concentration of BanLec-eGFP used to stain different strains of Salmonella is shown in Figure 3H. The binding of BanLec-eGFP was completely dependent on lectin-carbohydrate interaction as complete inhibition could be achieved with both mannose and glucose, Figure  4A,B. Galactose could not inhibit this interaction even at a high concentration of 0.1 M. Inhibition with mannose gave IC50 ≤ 1.7 mM, which was substantially lower than that obtained with glucose IC50 = 7.4 mM, Figure 4B. The binding of BanLec-eGFP was completely dependent on lectin-carbohydrate interaction as complete inhibition could be achieved with both mannose and glucose, Figure 4A,B. Galactose could not inhibit this interaction even at a high concentration of 0.1 M. Inhibition with mannose gave IC50 ≤ 1.7 mM, which was substantially lower than that obtained with glucose IC50 = 7.4 mM, Figure 4B. The binding of BanLec-eGFP was completely dependent on lectin-carbohydrate interaction as complete inhibition could be achieved with both mannose and glucose, Figure  4A,B. Galactose could not inhibit this interaction even at a high concentration of 0.1 M. Inhibition with mannose gave IC50 ≤ 1.7 mM, which was substantially lower than that obtained with glucose IC50 = 7.4 mM, Figure 4B.  In order to compare the binding of BanLec-eGFP to Salmonella enterica serovar Typhimurium isolate 2865 to that which would be obtained by using BanLec-B and streptavidin-PE, a correlation was made, Figure 4C. A correlation coefficient of 0.875 was obtained between the two staining procedures, with very little optimization, showing that the correlation was highly significant (p = 0.0019).

Discussion
In this study, recombinant BanLec-eGFP chimera was in silico designed, produced, and tested for application in the binding assay for high mannose structures present on microorganisms.
The chimera retained the secondary structures of both individual proteins and the structure was stable over the time of simulation, mostly retaining the secondary and tertiary structures of both domains, and further on as demonstrated by functional assays. The position of eGFP did not affect the binding capacity of BanLec for the glycans. This was evaluated by testing the binding of BanLec-eGFP chimera to influenza virus antigens and to different strains of Salmonella. These microorganisms were selected for analysis because of the presence of mannose on their surface and because of the medical relevancy.
Viruses utilize envelope proteins for receptor recognition at the entry point into the host cell. The antiviral activity of lectins is based on binding to oligosaccharides attached to the viral envelope glycoproteins and inhibiting viral fusion and entry [33].
Various lectins have been fuzzed with fluorescent proteins as fluorescent probes for the detection of specific glycan structures present on the cell-surface or intracellular conjugates, making them versatile primary detection reagents. The neurocan-GFP was successfully employed for the detection of hyaluronan tissue distribution [24]. Fluorescently labeled Aspergillus oryzae lectin was a useful probe for the detection of the core fucose in N-glycans, but can be also useful vehicles for delivery of substances into the cells [26].
Vaxigrip ® , Sanofi Pasteur IIV3, used in this study is a trivalent vaccine, containing antigens from A/H1N1, A/H3N2, and B strains, inactivated A and B split-virion influenza vaccine, (15 µg of haemagglutinin for each strain) [41]. By employing BanLec-eGFP in the competitive FLLSA, it was shown that wt BanLec exhibits higher binding for high mannose glycans present on the employed influenza vaccine than the H84T BanLec mutant, contrary to some literature data [14,38].
In order to assess the binding potential of BanLec-eGFP to highly mannosylated microorganisms, based on our experimental data and literature data, we decided to use Sallmonella. Previously we demonstrated that selectivity of rBanLec to fungal β-glucans can be achieved in enzyme linked lectin sorbent assay (ELLSA) [42]. The ability of BanLec to bind fungal β-glucans has been demonstrated previously by Goldstein et al. [43]. On that occasion [42] Salmonella was not included in the analysis. After conducting ELLSA to test the binding of BanLec to Salmonella, we detected approximately equal binding to that obtained with yeast cells.
Salmonella is a well-known bacterial pathogen that causes both food-borne gastroenteritis in humans and animals and typhoid fever. There are currently two species of Salmonella, with Salmonella enterica being medically relevant. Within the species of S. enterica individual subspecies and serovars have been identified. Different serovars are differentiated by the structures of their flagellum, carbohydrates, and lipopolysaccharides (LPS) [44,45]. Certain strains of Salmonella are known to contain mannose-rich LPS, and the binding of Mannosebinding lectin (MBL) to this LPS has been investigated [46][47][48]. Studies on the effect of MBL binding to S. enterica serovar Typhimurium have revealed the bactericidal properties of MBL [39,40]. The inhibitory effect of MBL on the motility of Salmonella occurs by affecting the energy source required for motility and the signaling pathway of chemotaxis [48].
The binding of MBL to a strain of Salmonella producing a mannose-rich LPS was previously demonstrated with the similar principle employed in this study, by the binding of fluorochrome conjugated anti-MBL serum [46]. The approach used in this study is a direct one, with the novel BanLec-eGFP chimera, and the results indicate that BanLec-eGFP acts in a similar fashion as MBL, which remains to be corroborated by parallel testing.
Important to note is that we have obtained differences between individual serovars and strains. In particular, serovar Typhi clinical isolate 12, showed highest binding. The tested strains from serovar Enteritidis showed no binding, while variations between strains in the serovar Typhimurium have been noticed. MBL does not bind to nonpathogenic mutants of Salmonella enterica serovar Montevideo that lack the mannose-rich O-polysaccharide within LPS [46] leading to the question whether mannosylation influences the pathogenicity of these organisms and in what way. The novel BanLec-eGFP chimera represents an excellent novel tool for such analysis.
In this study, we have demonstrated by inhibition experiments with monosaccharides that the binding of BanLec-eGFP to microorganisms occurs through the interaction of the lectin with the saccharide component. Moreover, we have demonstrated that there is no significant influence of the addition of GFP to the specificity of BanLec.
Each lectin molecule is characterized by affinity, specificity, and selectivity, and the comparison of different lectins can be done within basic research studies. We previously compared BanLec with other mannose binding lectins such as Concanavalin A and Lens culinaris agglutinin (LCA) and found profound differences [42]. In biotechnology, often relying on the knowledge obtained through basic research, a more targeted approach, with a specific purpose should always be applied. Based on the results presented above and other literature data, further studies including the comparison of MBL with BanLec should be performed.
In conclusion, BanLec-eGFP was produced in high yield, with retained secondary structures in both domains. The protein is stable and offers the advantage of constant emission capacity, unlike the chemically labeled lectin molecules, which increases reproducibility.

Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.

Data Availability Statement:
The data that support the findings of this study are available from the corresponding author M.G.-J., upon reasonable request.