Expression cassette and plasmid construction for Yeast Surface Display in Saccharomyces cerevisiae

Develop a Cell Surface Display system in Saccharomyces cerevisiae, based on the construction of an expression cassette for pYES2 plasmid. The construction of an expression cassette containing the α-factor signal peptide and the C-terminal portion of the α-agglutinin protein was made and its sequence inserted into a plasmid named pYES2/gDαAgglutinin. The construction allows surface display of bovine herpesvirus type 5 (BoHV-5) glycoprotein D (gD) on S. cerevisiae BY4741 strain. Recombinant protein expression was confirmed by dot blot, and indirect immunofluorescence using monoclonal anti-histidine antibodies and polyclonal antibodies from mice experimentally vaccinated with a recombinant gD. These results demonstrate that the approach and plasmid used represent not only an effective system for immobilizing proteins on the yeast cell surface, as well as a platform for immunobiologicals development.


Introduction
Cell Surface Display is a technique developed for recombinant protein expression in heterologous systems, such as bacterial, insect and yeast cells (Bertrand et al. 2016). Its application allows immobilization of peptides, whole proteins, or small fractions of antibodies on the cell surface by binding with an anchor protein (Tanaka and Kondo 2015). The attached target retain activity and stability, allowing them to interact with specific molecules present in the extracellular environment (Ueda 2016). Saccharomyces cerevisiae is the most used yeast for Cell Surface Display, being used for different purposes such as: the production of bioethanol, chemicals synthesis, and more recently, development of oral vaccines (Ç elik and Ç alık 2012; Parapouli et al. 2020).
The commercial availability of plasmid vectors for Yeast Surface Display (YSD) is generally based on options that depend of two main anchor proteins: a-agglutinin and a-agglutinin. The a-agglutinin anchoring system consists of its Aga1p and Aga2p subunits forming a protein complex, both being linked via two disulfide bonds, in which Aga1p subunit is responsible for cell wall attachment via GPI (Glycosylphosphatidylinositol) anchor and Aga2p is used to fuse target proteins (Kuroda and Ueda 2013). On the other hand, a-agglutinin system was developed based only on the C-terminal half of a-agglutinin, which is a fraction of the GPI-anchored cell wall protein able to immobilize fused proteins and expose them to the extracellular environment (Kuroda and Ueda 2013;Ueda 2016). The application of these systems are directly related to recombinant protein immobilization and their stability in YSD (Yang et al. 2019). Thus, new plasmids able to facilitate protein anchoring should be sought as possible alternative to implement this method. This techniques success depends on the choice of a plasmid, which should be able to allow the protein expression in the correct folding (similar to natural form) and requires an easy way to reproduce it (Routledge et al. 2016). YSD plasmid vectors are usually composed of a eukaryotic promoter, a multiple cloning site, selection markers, secretion factor, eukaryotic origin of replication (if not an integrative plasmid), and may contain tags to facilitate protein identification or purification (Ç elik and Ç alık 2012; Baghban et al. 2019). One of the alternatives for choosing the right plasmid for YSD is the construction of an expression cassette that has all the necessary components and allows the integration with conventional plasmids (Nasser et al. 2003).
Immunogenic potential of glycoprotein D (gD) of bovine herpesvirus type 5 (BoHV-5) has been studied by our group for some years, expressed recombinantly (rgD) using Pichia pastoris and Escherichia coli expression systems. In previous studies, our group have showed that rgD used as a vaccine antigen, was able to induce high levels of neutralizing antibodies titers in mice and cattle (Araujo et al. 2018). An ELISA test (Enzyme-Linked Immunosorbent Assay) composed of rgD demonstrated protein recognition by the serum of naturally infected animals (Dummer et al. 2016). Based on the knowledge using gD in our laboratory, we chose it as a model protein for the present work.
In this study, we first constructed an expression cassette using the C-terminal half of a-agglutinin anchor protein for YSD, in which several components were added for an efficient heterologous protein expression. It was constructed for pYES2 (Invitrogen) plasmid insertion, which allowed its use for transformation of S. cerevisiae BY4741. Glycoprotein D gene sequence was inserted on the cassette and protein expression was induced by GAL1 promoter, then rgD was immobilized on yeast surface by a-agglutinin and exposed to extracellular medium. We were able to identify rgD expressed and displayed on S. cerevisiae using Dot blot and Immunofluorescence techniques, what confirmed the effectiveness in establishment of YSD protocol.

Construction of pYES2/gDaAgglutinin displaying vector
Using Vector NTI Advance software (Invitrogen), the construction of gDaAgglutinin cassette was made aiming its expression on plasmid pYES2. The first cassette component added was the a-factor signal peptide sequence, which directs the processed protein for secretion into the extracellular medium. A sequence encoding 6 histidine amino acids (6XHistag) and a linker composed by Ser/Gly amino acids was also added. Finally, the sequence of the Cterminal half of a-agglutinin and the stop codon (TAA) were added to the construct.
The gene sequence of Bovine herpesvirus type 5 glycoprotein D was obtained from GenBank (accession number AAA67359.1). This protein is composed of transmembrane regions, cytoplasmic domain, signal sequence and extracellular domain. However, only the extracellular domain (a region composed of 311 amino acids) was selected and used in this construction. The fragment used was based on Dummer et al. (2009) studies. Never the less, it was submitted to codon optimization based on S. cerevisiae codon usage. Nucleotide sequence corresponding to that region was used to construct gDaAgglutinin expression cassette, which was deposited on NCBI (National Center for Biotechnology Information), accession number MW556769. Nucleotide sequences added to the expression cassette are shown in Table 1. The final sequence was sent for synthesis by GenOne Company (Rio de Janeiro, Brazil), which also constructed the plasmid pYES2/gDaAgglutinin vector for YSD.
The length comparison between plasmids pYES2 and pYES2/gDaAgglutinin was performed by electrophoresis in a 0.8% w/v agarose gel, on TBE buffer (Tris-Borate-EDTA). Saccharomyces cerevisiae BY4741 strain (MATa his3-1 leu2 met15 ura3) was selected and transformed by electroporation following MicroPulser TM (BioRad) electroporator protocol. Different concentrations of both plasmids (100 ng, 1500 ng, 3000 ng and 6000 ng) were added to 40 lL of competent cells (1 9 10 10 cells/mL) and subjected to electroporation in 0.2 cm cuvettes, with pulses of 5 ms and 1.5 kV voltage in the MicroPulser TM electroporator. After electroporation, the electroporated material was suspended in 1 mL of 1 M sorbitol, and 100 lL was plated on Sc-U agar medium [2 g/L agar, 2 g/L dextrose, 6.7 g/L yeast nitrogen base w/o amino acids and 1.9 g/L yeast synthetic drop-out medium supplement (Sigma Aldrich)], being incubated over 72 h at 30°C.
Recombinant colonies of S. cerevisiae BY4741 were selected and cultured in liquid Sc-U medium with glucose during 20 h at 30°C, under agitation (150 rpm). Upon reaching a biomass of 3 (O.D. 600 nm ), the culture was centrifuged (1.5009g during 5 min) in a DTR16000 centrifuge (DAIKI), washed three times and suspended in liquid Sc-U containing galactose for GAL1 promoter induction, in a final biomass of 0.4 D.O. at 600 nm. The induction step was maintained over 24 h in the same galactose-containing medium culture conditions, and later centrifuged (1.5009g for 5 min) to separate biomass. All stages of the expression process were also performed with a BY4741 sample transformed with pYES2 plasmid as a control culture. The possibility of extracellular expression or nonattachment on yeast surface was evaluated through medium supernatant purification. The supernatant from S. cerevisiae BY4741 culture expressing immobilized gD was subjected to purification by affinity chromatography using a 5 9 1 mL HisTrap TM column (GE Healthcare). Purified samples were recovered using Ä KTA TM wash buffer (2.34 g/L monobasic sodium phosphate, 29.2 g/L sodium chloride, 34 g/L imidazole) in different imidazole concentrations.
Detection of rgD expression on S. cerevisiae BY4741 cell surface

Dot blot
Biomass was diluted for a cell concentration of 10 7 cells/mL, then samples were sonicated with 6 s pulses at 3 s interval, on an ice bath for 10 min. It was collected 10 lL of each lysate, which were added over a nitrocellulose membrane (Hybond TM ECLTM, Amersham Biosciences) and incubated in a blocking buffer containing 5% w/v skim milk powder and PBS-T (phosphate buffered saline added of 0,5% v/v Tween TM 20) for 1 h at room temperature. Washing steps were performed between the next steps, using PBS-T. It was then incubated with anti-histidine monoclonal antibodies (Invitrogen) or anti-rgD mouse serum (1:3.000), for 1 h at 37°C. The membrane was washed and incubated (1 h at 37°C) with horseradish peroxidase conjugated mouse IgG antibody (1:4.000 dilution) (Sigma Aldrich). Washing step was repeated and membranes were placed in the chromogenic substrate solution, containing 6 lg of DAB (3,3 0diaminobenzidine), 50 mM Tris-HCl, 0.3% nickel sulfate solution and 10 lL of hydrogen peroxide (H 2 O 2 ). Biomass samples from the control culture BY4741 pYES2, BY4741 pYES2/gDaAgglutinin and their supernatants were tested against recognition by anti-histidine mAb, following the adaptation of the dot blot protocol published at Nizoli et al. (2009).

Indirect immunofluorescence assay
Samples of BY4741 pYES2 and BY4741 pYES2/ gDaAgglutinin cultures, both at a concentration of 10 7 cells/mL, were fixed (30 min at 30°C) over indirect immunofluorescence slides. After that, slides were incubated in methanol at 4°C for 10 min and then blocked with PBS ? fetal bovine serum 10% v/v, for 30 min at 30°C in a dark humid chamber. The primary antibody, anti-histidine (Invitrogen) or anti-rgD mouse serum (both 1:100 concentration) were applied and incubated for 2 h at 30°C. After this step, slides were maintained for 18 h at 4°C in a dark humid chamber. Secondary mAb anti-mouse IgG conjugated to FITC (fluorescein isothiocyanate) (Sigma Aldrich) was applied in a 1:80 dilution and incubated for 90 min at 30°C. Antigen-antibody reaction was visualized using fluorescence microscope (Olympus BX51).

pYES2/gDaAgglutinin plasmid construction for Yeast Surface Display
Plasmid pYES2 had originally 5856 bp (Fig. 1a), while total length prediction for pYES2/gDaAgglutinin plasmid on Vector NTI Advance software (Invitrogen), after inserting the cassette in its sequence, was 7985 bp (Fig. 1b). Schematic representation of expression cassette components can be seen in Fig. 1c.
It was possible to verify by agarose gel electrophoresis that the synthetized plasmid presented approximately 7.9 kb, confirming the addition of * 2.0 kb to plasmid pYES2, corresponding to the insertion of gDaAgglutinin cassette (Fig. 2a). When the plasmid pYES2/gDaAgglutinin was digested using EcoRI restriction enzyme, a fragment of approximately 900 bp was observed, referring to the gD gene sequence included in the expression cassette (Fig. 2b).

Confirmation of recombinant protein expression immobilized by YSD in S. cerevisiae
Protocols applied for transformation and culture of S. cerevisiae BY4741 had their effectiveness proven with the yeasts culture on Sc-U agar, where only transformed cells were able to grow (Fig. S1). Biomass and supernatant samples from BY4741 pYES2 control culture and BY4741 pYES2/gDaAgglutinin culture were tested against their recognition by mAb anti-histidine, which recognizes the 6xHis tag added to the expression cassette construction. Cell lysate of BY4741 pYES2/gDaAgglutinin culture revealed reaction positivity, demonstrating effectiveness on cassette construction and recombinant protein expression. Thus, in addition to presenting recombinant gD on yeast cells, the glycoprotein is also found in the culture supernatant (Fig. 3a). Dot blot was applied for a quick and easy screening of recombinant protein expression, in which was possible Fig. 1 Schematic representation of plasmid vector pYES2 and the new plasmid pYes2/gDaAgglutinin. a Structure of plasmid pYES2 used to construct plasmid pYes2/gDaAgglutinin. b plasmid pYes2/gDaAgglutinin construction. c Expression cassette, the sequence of BoHV-5 glycoprotein D (green color) and the C-terminal half of a-Agglutinin (orange) are highlighted, the other components are the secretion signal a-Factor, 6xHis (histidine) tag, the linker composed of Ser/Gly amino acids and, finally, stop codon. Flanking gD sequence there are restriction sites for EcoRI enzyme ( Source Vector NTI Advance Software) Fig. 2 Electrophoresis on 0.8% agarose gel of plasmid vector pYes2/gDaAgglutinin and its restriction with EcoRI enzyme. a Size comparison between pYES2 and pYes2/gDaAgglutinin, 1 marker Lambda DNA/HindIII (Thermo Scientific), 2 pYes2/ gDaAgglutinin with 7.9 kb, 3 pYes2 with 5.8 kb. b pYes2/ gDaAgglutinin digestion with EcoRI enzyme, 1 1 kb marker DNA plus ladder (Thermo Scientific), 2 undigested pYes2/ gDaAgglutinin-control, 3 EcoRI digested pYes2/gDaAgglutinin plasmid to observe rgD recognition in BY4741 pYES2/ gDaAgglutinin cells by polyclonal antibodies from mice immunized with purified rgD (produced by Dummer et al. 2009, in P. pastoris) (Fig. 3b).
Detection of rgD by indirect immunofluorescence was performed using the same antibodies describe in dot blot, however a 1:100 dilution was used as a primary antibodies. FITC fluorescence was observed only in cell surface of clones that had pYES2/gDa-Agglutinin plasmid, while control samples resulted in absence of fluorescence detection (Fig. 4), therefore recognition of recombinant protein displayed in S. cerevisiae BY4741 by immunofluorescence confirmed that YSD was successfully executed.

Discussion
Recombinant protein expression on S. cerevisiae surface is a methodology that has been consolidated in basic and applied research, arousing more and more interest in improving the technique (Jeong et al. 2019;Lei et al. 2020). Choosing eukaryotic cells as expression platforms proves to be attractive because there is the ability to express proteins with the original fold, an important factor when there is a need for the maintenance of conformational epitopes in recombinant protein structure. Thus, the choice of S. cerevisiae is justified for this work, in which the glycoprotein D of BoHV-5 requires the correct assembly to attain its structure, similar to when it is expressed by the virus, and important for vaccines and diagnostic tests development.
Plasmid pYES2 allows the intracellular expression of recombinant proteins in S. cerevisiae, however if Fig. 3 Confirmation of gD expression by S. cerevisiae. a Dot blot test performed using anti-histidine antibodies. 1 BY4741 pYES2 culture, 2 BY4741 pYes2/gDaAgglutinin, 3 BY4741 pYes2/gDaAgglutinin Supernatant purified, 4 Reaction positive controls. b Dot blot using anti-rgD mice sera. 1 BY4741 pYES2, 2 BY4741 pYes2/gDaAgglutinin, 3 purified rgD expressed in Pichia pastoris (positive control) Fig. 4 Indirect immunofluorescence for gD recognition on S. cerevisiae surface. a Test using Anti-histidine mAb, 1 BY4741 Pyes2 cells observed in bright-field, 2 BY4741 pYES2 observed under FITC fluorescence excitation filter, 3 BY4741 pYes2/ gDaAgglutinin in bright-field, 4 BY4741 pYes2/gDaAgglutinin observed under FITC fluorescence excitation filter; scale bar 20 lm. b Mice Polyclonal sera anti-rgD was used, 1 BY4741 pYES2 cells observed in bright-field, 2 BY4741 pYES2 observed under FITC fluorescence excitation filter, 3 BY4741 pYES2/gDaAgglutinin in bright-field, 4 BY4741 pYES2/ gDaAgglutinin observed under FITC fluorescence excitation filter; scale bar 10 lm. Olympus BX51 fluorescence microscope, 495 nm filter for FITC excitation, magnification of 9 40 (image a) and 9 100 (image b) the objective is its use in YSD method, there is a need to add new components that enable the recombinant protein to enter the secretory pathway and subsequent immobilization on cell surface. When digesting pYES2/gDaAgglutinin using EcoRI restriction enzyme, a fragment of approximately 900 bp was observed, encoding to gD gene sequence. This fragment was not observed in pYES2 control plasmid, confirming that the plasmid can be used for expression of other proteins, as long as the gene sequence should be flanked by EcoRI sites and its insertion in the region destined for gD. Lezzi et al. (2012) expressed the Agaricus bisporus tyrosinase intracellularly, and Chen et al. (2011) expressed monellin using pYES2, adding the DNA sequence of the signal peptide a-factor to their constructions. Due to the applicability of pYES2 plasmid, it was chosen in the present study as an alternative system for YSD in S. cerevisiae. In this work, we performed the construction of an expression cassette for pYES2, containing the a-factor signal peptide and the C-terminal portion of a-agglutinin. These two components were responsible for processing and secretion to the extracellular medium and immobilizing the recombinant protein on yeast cell wall by GPI-anchoring (Glycosylphosphatidylinositol) protein, respectively (Bertrand et al. 2016).
It has been reported a low display efficiency and protein instability for some YSD systems, such as a-agglutinin system, which may be explained by the need of protein complex formation via disulfide bonds between Aga1p and Aga2p, a linkage that impacts on protein expression and is sensitive under specific stress conditions (Goyal et al. 2011;Tanaka and Kondo 2015;Yang et al. 2019). Aiming to maximize recombinant protein immobilization, several other anchor proteins have been tested, one being a-agglutinin, one of the most commonly used because of its proven stability and superior display efficiency (Goyal et al. 2011). In our study, we observed that a-agglutinin fragment was able to immobilize rgD on cell surface successfully, confirming its effectiveness in cell wall attachment and recombinant protein stability, since it was identified by different techniques for recombinant protein expression analysis.
In order to facilitate rgD expression and its correct protein conformation, codon optimization for expression in yeasts was essential for the present work. As noted by Kaishima et al. (2016) in the expression of GFP (green fluorescent protein) variants, it was observed that the codon optimization leads to a higher concentration and more effective expression of GFP protein. The choice for yeast usual codons provides greater and easier expression of target protein (Sharp and Cowe 1991), what suggests a facilitated and favored expression of immobilized gD in S. cerevisiae and its recognition by mice antibodies immunized with rgD.
The designed cassette had a linker connecting the C-terminal portion of a-agglutinin to gD, consisting of a flexible structure with no significant changes in target protein conformation, facilitating the molecular stability and protein function (Reddy Chichili et al. 2013). One may suggest that its presence can be a promising tool to increase the access to the immobilized protein (Tanaka and Kondo 2015). In the present study, since there are no purification tags in pYES2 original composition, the sequence that codes for 6-histidines amino acids was added as a cassette component. This insertion proved to be effective, since mAb anti-histidine was able to recognize the recombinant protein on yeast surface, being demonstrated by dot blot (Fig. 3) and Immunofluorescence (Fig. 4). Dot blot revealed that rgD could be identified on biomass sample, as well as in the culture supernatant (Fig. 3a, b). This fact was also observed by Baptista (2013), who when immobilizing a-amylase, detected its presence in the culture supernatant. This is usually related to protein debris that detaches from the cell wall, protein units that do not adequately attach to the cell surface and protein proteolysis (Harnpicharnchai et al. 2010;Baptista 2013). Even though a recombinant protein expression to the extracellular medium was not the objective in this work, supernatant purification was performed as a control step for YSD, in case rgD was not found on S. cerevisiae surface.
Indirect immunofluorescence assay was used to confirm the rgD expression by YSD technology. It is important highlight that there were no considerable impacts on molecule structure with 6xHis tag insertion and rgD immobilization, since C-terminal portion of a-agglutinin was able to immobilize rgD on S. cerevisiae surface and thereby displaying it for an easy recognition by specific antibodies generated in mice immunized with soluble ''free'' rgD (Santos et al. 2018) (Fig. 4b). Recombinant protein detection through dot blot and immunofluorescence using serum from these mice suggests that epitopes remained on displayed gD, which demonstrates the potential of S. cerevisiae BY4741 pYES2/gDaAgglutinin yeast for future immunobiological applications.
In conclusion, gDaglutinin expression cassette was constructed in silico and inserted on pYES2 plasmid, being able to conduct the target protein rgD to the surface of S. cerevisiae BY4741 strain. Finally, we can conclude that the construction of pYES2-target protein-aAgglutinin opens a new perspective as a Yeast Surface Display tool to be used in the most diverse biotechnological areas.