Plavec and Berlec: Surface Anchoring on Lactococcus lactis ... Surface Anchoring on Lactococcus lactis by Covalent Isopeptide Bond

Display of recombinant proteins on the bacterial surface is an emerging research area with wide range of potential biotechnological applications. Because of its GRAS (generally recognized as safe) status, lactic acid bacterium Lactococcus lactis represents an attractive host for surface display and promising vector for in situ delivery of bioactive proteins. The present study focused on finding a new alternative approach for surface display on Lactoccocus lactis . We developed a system that enables the formation of irreversible isopeptide bonds on the surface of Lactococus lactis . This was achieved through the following two protein/peptide pairs, SpyCatcher/SpyTag and SnoopCatcher/SnoopTag. 1–3 Attachment of tagged model protein B domain to the cell surface of Lactococccus lactis displaying the corresponding catcher protein was demonstrated using flow cytometry. We demonstrated effective use of aforementioned protein anchors which thus represent a promising alternative to established approaches for surface display on Lactoccocus lactis .


Introduction
Display of recombinant proteins on bacterial surface offers a variety of possible biotechnological applications. Proteins-displaying bacteria can act as bioadsorbents, biosensors, biocatalysts or oral vaccines. They can be used in antibody production and in peptide screening. [4][5][6] Several lactic acid bacteria (LAB) are probiotics and are therefore considered valuable hosts in biotechnology due to their beneficial influence on health. 7,8 Because of the "generally recognized as safe" (GRAS) status which confirms their safety, LAB are attractive not only for industrial application but also therapeutically. 9 Display of heterologous proteins on the surface of LAB has already been exploited in therapy for the preparation of mucosal vaccines. [10][11][12] Moreover, beneficial effects in inflammatory bowel disease could be achieved when displaying binding molecules directed against pro-inflammatory molecules such as TNFα. [13][14][15][16][17] Different approaches can be exploited for displaying a protein on the bacterial surface. The protein to be dis-played is usually fused to an anchoring motif. 14,18 Five different types of surface anchoring domains have been described for LAB: transmembrane domains, LPXTG-type domains, lysin motif (LysM) domains, surface layer proteins and lipoprotein anchors. [19][20][21] The most frequently applied surface anchoring domains in prototype LAB, Lactococcus lactis, are the C-terminal part of endogenous AcmA, enabling non-covalent anchoring through peptidoglycan binding LysM repeats, [22][23][24][25] and the LPXTG sequence of M6 protein of Streptococcus pyogenes enabling covalent anchoring. 26,27 Despite these available options, alternative surface display approaches are being sought.
Recently, two peptide/protein pairs known as Spy-Tag/SpyCatcher, from Streptococcus pyogenes, and Snoop-Tag/SnoopCatcher, from Streptococcus pneumoniae have been developed. [1][2][3] Interaction between the peptide and the protein leads to the formation of an irreversible isopeptide bond. The reaction is high-yielding and fast while the bond is highly stable. It can survive extreme pH, high ionic strength and exposure to detergents. 1,28,29 Stable bond formation enables combinatorial assembly of multi-protein constructs, and opens an opportunity to use this approach in vaccine production, enzyme substrate channeling, antibody polymerization, cell signaling activation, and biomaterials. 1,30 The key properties of isopeptide binding are simple and fast procedure, irreversible and stable bond, specificity, and cysteine independence; the latter offering the possibility to use the approach in reducing environment. 3 The goal of the present study was to develop a system for surface display on recombinant LAB L. lactis by applying the isopeptide bond formation (Fig. 1a). This was achieved by preparing genetic constructs consisting of surface anchor, elements of SpyTag/SpyCatcher or SnoopTag/ SnoopCatcher pairs and model passenger protein B domain (Fig. 1b). 13 B domain is one out of five antibody-binding domains of staphylococcal protein A that can bind antibodies via their Fc region. 13 The fusion proteins were expressed either in L. lactis or in E. coli, and assembled on the surface of L. lactis, as confirmed by surface localization of B domain.

Molecular Cloning
Plasmid DNA was isolated with NucleoSpin Plasmid (Macherey and Nagel, Düren, Germany), with an additional lysozyme treatment step for L. lactis. Lactococci were transformed with electroporation using a Gene Pulser II apparatus (Biorad, Hercules, USA) according to the MoBiTec GmbH (Goettingen, Germany) instructions. Nucleotide sequencing was performed by GATC (Constance, Germany).
Gene for SpyTag in fusion with B domain for expression in E. coli was amplified from pSDBA3b by PCR using primers B-F-NcoI-Spy and B-R-XhoI, cloned to pGEM-T Easy and then to pET28a via restriction enzymes NcoI/ XhoI, yielding pET_SpyT_Bd. Gene for SpyTag in fusion with B domain for secretion from L. lactis was amplified from pSDBA3b by PCR using primers B-F-BamHI and B-R-Kpn-Sy-Xba. Gene for SnoopTag in fusion with B domain was prepared likewise using primers B-F-BamHI and B-R-Kpn-So-Xba. Both were first cloned to plasmid pGEM-T Easy and then to plasmid pSDBA3b via restriction enzymes BamHI/XbaI, yielding pSD_SpyT_Bd and pSD_SnT_Bd. Gene for B domain for secretion from L. lactis was amplified from pSDBA3b by PCR using primers B-F-BamHI and B-R-Xba, first cloned to plasmid pGEM-T Easy and then to plasmid pSDBA3b via restriction enzymes BamHI/XbaI, yielding pSD_Bd ( Table 2-4).
Genes for SpyCatcher and SnoopCatcher were designed and synthesized de novo as gBlocks (Table 4)   Genes for the surface display of SpyCatcher and SnoopCatcher were amplified from gBlocks using primer pairs Spy-F-Bam/Spy-R-Eco and Sno-F-Bam/Sno-R-Eco, respectively, and were cloned first to pGEM-T Easy and then to plasmid pSDBA3b via restriction enzymes EcoRI/ BamHI, yielding pSD_SpyC_AcmA and pSD_SnC_ AcmA, respectively. Gene for the display of SpyTag on the L. lactis surface was amplified from pSDBA3b by PCR using primers AcmA-F-Bam-SpyT and AcmA-R-Xba, cloned to pGEM-T Easy and then to pSD_SpyC_AcmA via re-  Table 3. Primers used in this study

Primer Relevant features or sequence Reference
This work striction enzymes BamHI/XbaI, yielding pSD_SpyT_ AcmA (

3. Expression of SpyTag-B Domain Fusion in E.coli
100 µL of overnight culture of E. coli BL21 (DE3) harboring plasmid pET_SpyT_Bd was diluted (1:100) in 10 mL of fresh LB medium and, to determine optimal expression conditions, various parameters were tested: incubation temperature 37 °C or 25 °C, induction at optical densities (A 600 ) 0.5 or 1.0, induction with IPTG in concentration of 0.5 and 1.0 mM.
Large-scale expression of SpyTag-B domain fusion was performed by diluting 10 mL of overnight culture of E. coli BL21 (DE3) harboring plasmid pET_SpyT_Bd in 1 L of fresh LB medium. The culture was grown to optical density A 600 = 0.5 at 37 °C. At that point, the expression of SpyTag-B domain fusion, additionally tagged with hexa-histidine (H6), was induced by the addition of 0.5 mM IPTG for 3 h at 37 °C; the conditions that were found to be the most effective in preliminary screen.

4. Purification of SpyTag-B Domain With
Hexa-histidine (H6) Tag The E. coli culture expressing SpyTag-B domain with hexa-histidine (H6) tag was centrifuged at 5000 × g for 20 min and the pellet resuspended in 20 mL of equilibration/ wash (Eq/W) buffer (50 mM NaH 2 PO 4 , 300 mM NaCl, pH 7.0). The cells were lysed with a cycle of freezing and thawing, and with 3 cycles of 5 min sonication with a UPS200S sonifier (Hielscher, Teltow, Germany). After cell lysis, the suspension was centrifuged at 15000 × g for 20 min and the supernatant stored. SpyTag-B domain with H6 tag was iso-lated with BD Talon metal affinity resin (BD Biosciences), using batch/gravity-flow column purification and imidazole elution (elution buffer: 50 mM NaH 2 PO 4 , 300 mM NaCl, 150 mM imidazole, pH 7.0) according to the manufacturer's instructions. Eluted fractions were analyzed by SDS-PAGE, pooled and concentrated by ultrafiltration using Amicon Ultra 1 KDa cut off (Merck Millipore; Darmstadt, Germany). Purified fusion protein was dialyzed against PBS.

6. Formation of Isopeptide Bond Between SpyTag and SpyCatcher / SnoopTag and SnoopCatcher
In order to enable binding of E. coli-expressed Spy-Tag-B domain to L. lactis with surface displayed SpyCatcher, we centrifuged 20 µL of the cell culture of L. lactis with surface displayed SpyCatcher for 5 min at 5000 × g at 4 °C, resuspended the pellet in 500 µL of purified E. coli-expressed SpyTag-B domain with concentration of 0.4 mg/ mL and incubated for 2 h at RT with constant shaking.
To enable binding of SpyTag-B domain from L. lactis conditioned medium to L. lactis with surface displayed Binding of SpyTag-B domain secreted from L. lactis to SpyCatcher displayed on L. lactis was also achieved during co-culturing of the two strains. 100 µL of overnight cultures of L. lactis NZ9000 harboring pSD_SpyT_Bd and pSD_SpyC_AcmA were concomitantly added to 10 mL of fresh GM-17 medium. Simultaneous expression of the two fusion proteins was induced with nisin. Similarly, binding of SpyCatcher-B domain secreted from L. lactis to SpyTag displayed on L. lactis was achieved by co-culturing L. lactis NZ9000 harboring pSD_SpyC_Bd and pSD_SpyT_AcmA, as well as binding of SnoopTag-B domain secreted from L. lactis to SnoopCatcher displayed on L. lactis by co-culturing L. lactis NZ9000 harboring pSD_SnC_AcmA and pSD_SnT_Bd.

7. SDS-PAGE and Western Blot
SDS PAGE was performed with a Mini-Protean II apparatus (Bio-Rad, Hercules, USA). Samples were thawed in an ice bath, briefly sonicated with UPS200S sonicator (Hielscher, Teltow, Germany), mixed with 2× Laemmli Sample buffer and DTT, and denatured by heating at 100 °C before loading. Page Ruler Plus (Fermentas, St. Leon-Rot, Germany) pre-stained standard was used for molecular weight comparison. Proteins were stained with Coomassie Brilliant Blue or transferred to polyvinylidene fluoride (PVDF) membranes (Immobilon-P, Millipore) using wet transfer at 100 V for 90 minutes. Membranes were blocked in 5% non-fat dried milk in TBS with 0.05% Tween-20 (TBST; 50 mM Tris-HCl, 150 mM NaCl, 0.05% Tween 20, pH 7.5) and incubated overnight at 4 °C with goat anti-protein A antibody (1:2000, Abcam) in 5% non-fat dried milk in TBST. Following three washes with TBST, membranes were incubated for 2 h with HRP conjugated secondary donkey anti-goat IgG (1:5000, Jackson ImmunoResearch) in 5% non-fat dried milk in TBST. After three further washes with TBST, membranes were incubated with Lumi-Light chemiluminescent reagent (Roche). Images were acquired using ChemiDoc MP Imaging System (BioRad).

8. Flow Cytometry
For flow cytometry 20 μL of cell culture in stationary phase was added to 500 μL of Tris-buffered saline (TBS; 50 mM Tris-HCl, 150 mM NaCl, pH 7.5) and centrifuged for 5 min at 5000 × g at 4 °C. The pellet was resuspended in 500 μL of TBS with 1 μL of fluorescein-5-isothiocyanate (FITC)-conjugated human IgG antibody (Jackson Immu-noResearch, West Grove, USA) that binds the B domain via Fc region. After 2 h of incubation at RT with constant shaking at 100 rpm, cells were washed three times with 200 μL 0.1% TBST and finally resuspended in 500 μL TBS. Samples were analyzed with a flow cytometer (FACS Calibur; Becton Dickinson, Franklin Lakes, USA) using excitation at 488 nm and emission at 530 nm in the FL1 channel. The geometric mean fluorescence intensity (MFI) of at least 20 000 bacterial cells in the appropriate gate was measured. The average of at least three independent experiments was considered. All the samples went through the same procedures of preparation for flow cytometry analysis.

9. Statistical Analyses
Statistical analyses were performed with GraphPad Prism 5.0 software. Student's t test was used to compare the significance of differences between B domain-displaying bacteria and control.

1. Expression and Purification of Recombinant SpyTag-B Domain from E. coli
Gene for SpyTag-B domain, possessing affinity for the Fc region of human IgG, was cloned into the pET28a plasmid (Fig. 1b) in order to express the protein in E. coli and obtain it in sufficient amount and purity. Recombinant SpyT-B domain fusion protein with H6 tag was produced in E. coli BL21 DE3. Various expression conditions (growth at 37 °C and 25 °C, induction at optical densities (A 600 ) 0.5 and 1, induction with IPTG concentration of 0.5 and 1.0 mM) were tested. The highest total amount of SpyT-B domain expression was achieved by growing the bacteria at 37 °C to A 600 = 0.5 or 1.0, followed by induction with 0.5 or 1 mM IPTG for 3 h at 37 °C (Fig. 2a). The majority of the fusion protein was produced in the soluble form as it could be detected in the soluble fraction (supernatant) of the cell lysate (Fig. 2b). SpyTag-B domain was isolated with immobilized metal affinity chromatography (IMAC) (Fig. 2b).

2. Binding of E. coli-expressed SpyTag-B Domain to L. Lactis with Surface Displayed SpyCatcher
SpyCatcher in fusion with Usp45 secretion signal 26 and the surface anchoring C-terminal domain of AcmA was displayed on the surface of L. lactis as previously reported for other proteins. 13,[35][36][37] Binding of SpyTag-B domain, isolated from E. coli, to recombinant L. lactis with surface displayed SpyCatcher was evaluated by flow cytometry using antibody recognizing B domain. Statistical-ly significant increase in MFI was observed when SpyT-B domain was incubated with L. lactis with induced Spy-Catcher expression, in comparison to control non-induced L. lactis cells, or induced L. lactis cells without the addition of SpyT-B domain (19.9%; Fig. 3).

Binding of SpyTag-B Domain from L. lactis Conditioned Medium to L. lactis with Surface Displayed SpyCatcher
SpyTag in fusion with B domain and Usp45 secretion signal (plasmid pSD_SpyT_Bd) was expressed in L. lactis under the control of NisA promoter 38 and secreted to the growth medium. The producer cells were removed and the conditioned medium containing SpyTag-B domain fusion protein was incubated with L. lactis cells with surface displayed SpyCatcher (plasmid pSD_SpyC_AcmA). Low extent of binding was observed with flow cytometry using antibody recognizing B domain. Small statistically significant increase in MFI was reported when SpyTag-B domain was incubated with SpyCatcher-displaying L. lactis, in comparison to empty plasmid pNZ8148-containing control. No difference was observed when non-tagged B domain was incubated with SpyCatcher-displaying L. lactis (Fig. 4).

4. Binding of SpyTag-B Domain Secreted from L. lactis to SpyCatcher Displayed on L. lactis During the Co-culturing of the Two Strains
We co-cultured L. lactis secreting SpyTag-B domain fusion (plasmid pSD_SpyT_Bd) with L. lactis displaying SpyCatcher (plasmid pSD_SpyC_AcmA) to achieve immediate bond formation between SpyCatcher and SpyTag protein/peptide pair. Binding was evaluated with flow cy-Plavec and Berlec: Surface Anchoring on Lactococcus lactis ... tometry using antibody recognizing B domain. Statistically significant increase in MFI (40.1%) was observed, when Spy-tagged B domain producing L. lactis was co-cultured with SpyCatcher-displaying L. lactis cells, in comparison to SpyCatcher-displaying L. lactis cells (Fig. 5).

6. Introducing SnoopCatcher and Tag:
Binding of SnoopTag-B Domain Secreted from L. lactis to SnoopCatcher Displayed on L. lactis During

Co-culturing of the Two Strains
We co-cultured L. lactis secreting SnoopTag-B domain fusion (plasmid pSD_SnT_Bd) with L. lactis displaying SnoopCatcher (plasmid pSD_SnC_AcmA), respectively, to achieve immediate bond formation, as demonstrated previously for SpyCatcher/SpyTag pair. Binding was evaluated with flow cytometry using antibody recognizing B domain. Statistically significant increase in MFI (21.1%) was observed, when Snoop-tagged B domain-producing L. lactis was co-cultured with Snoop-Tag-displaying L. lactis cells, in comparison to Snoop-Catcher-displaying L. lactis cells (Fig. 7). The location of interacting protein/peptide pair was reversed by engineering a strain of L. lactis to secrete the

Discussion
Isopeptide bond formation was applied to develop alternative surface display systems for LAB L. lactis by enabling a stable covalent bond between a peptide SpyTag and a protein SpyCatcher or, similarly, between a peptide SnoopTag and a protein SnoopCatcher. 1,2 The peptide/ protein pair has already been employed to stabilize enzymes, for modular vaccine production, vaccine optimization and formation of catalytic biofilms. 29,39 To test the feasibility of the isopeptide bond formation, we anchored one of the binding partners to the surface of L. lactis, by fusing it with Usp45 secretion signal and peptidoglycan-binding C-terminus of AcmA protein, as previously reported. 13,35 The second binding partner was fused to a reporter protein B domain that we previously applied for the assessment of surface display. 13,40 B domain fusion was isolated from E. coli or secreted from another recombinant L. lactis species (Fig. 1). Formation of the isopeptide bond resulted in the attachment of the B domain to the surface of L. lactis and was assessed by flow cytometry. We obtained statistically significant display of B domain on the surface of L. lactis with almost all the systems that were constructed.
Firstly, we expressed a fusion peptide SpyTag-B domain-His Tag in E. coli to obtain sufficient amount of the fusion protein following IMAC purification. Purified Spy-Tag-B domain-His Tag was added to SpyCatcher-displaying recombinant L. lactis. Statistically significant display of B domain, in comparison to the control, was determined. However, the binding was relatively weak and partially unspecific.
Secondly, we expressed SpyTag-B domain in L. lactis and directed it to the growth medium. After removal of the producer cells, the conditioned medium was incubated with SpyCatcher-displaying recombinant L. lactis. The display of B domain was lower than that achieved with Spy-Tag-B domain from E. coli and was not statistically significant. This could be due to the lower amount of the fusion protein in the conditioned medium. Moreover, there are several other factors in the medium that could hinder binding, for example pH value of the medium, and the presence of numerous other proteins and peptides. Even though SpyCatcher-SpyTag bond is claimed to be stable under a range of pH values (5-8) 29 it is possible that low pH of the conditioned medium hinders the bond formation. Additionally, numerous peptides in the conditioned medium might non-specifically interact with SpyCatcher.
Thirdly, we expected the formation of the isopeptide bond to be more probable if the Spy-tagged B domain was available immediately after induction of the surface display of SpyCatcher, as this would decrease the probability of unspecific interactions. Availability of Spy-tagged B domain was provided by co-culturing two species of L. lactis: one displaying SpyCatcher, and the other secreting Spy-Tag-B domain fusion. Thus achieved surface display of B domain was indeed higher than that achieved by the addition of SpyTag-B domain from E. coli or from the conditioned medium of L. lactis.
In the above examples the SpyCatcher was immobilized on the surface of L. lactis. To test the influence of the location of binding partners, we reversed the system by displaying SpyTag on L. lactis, and co-cultured the strain with L. lactis secreting SpyCatcher-B domain fusion protein. The display of B domain was again achieved; however due to relatively high unspecific binding of antibodies with SpyTag-displaying lactococci the display was not statistically significant.
Apart from SpyCatcher/SpyTag pair, the isopeptide bond can also be formed by combining SnoopCatcher and SnoopTag. We applied similar experimental setup as previously described for SpyCatcher/SpyTag by displaying SnoopCatcher on the surface of L. lactis and co-culturing the bacteria with a strain of recombinant L. lactis secreting SnoopTag-B domain fusion protein. Significant surface display of B domain was again observed; however there was no improvement over SpyCatcher/SpyTag pair.

Conclusion
In the present study we demonstrated, for the first time, the surface display of reporter protein on L. lactis by exploiting isopeptide bond-forming partners SpyCatcher and SpyTag, as well as SnoopCatcher and SnoopTag. The most effective display was obtained by anchoring Spy-Catcher to the bacterial surface, and co-culturing the bacteria with a lactococcal strain that secreted Spy-tagged reporter protein. This represents a proof-of-principle for a new, highly flexible surface display system for L. lactis that warrants further studies with an intention to improve the extent of surface display.