Lactococcus lactis IBB477 presenting adhesive and muco-adhesive properties as a candidate carrier strain for oral vaccination against influenza virus.

In the gastrointestinal tract (GIT), adhesion is a prerequisite for bacterial colonization. Lactococci can be used in functional food (probiotics) and health-related applications (mucosal vaccines, therapeutic drug delivery), both potentially involving adhesive properties. A candidate lactic acid bacterium for influenza antigen delivery through the GIT should display the ability to adhere. The present work probes the interactions between Lactococcus lactis and mucins using pig gastric mucin (PGM) as a model. Two strains were used for the optimization of the screening method for adhesion: L. lactis subsp. cremoris IBB477 persistent in the GIT of germ-free rats, and the low-adhering control strain MG1820. High adhesion to bare and mucin-coated polystyrene of IBB477 in comparison with MG1820 was observed. We searched for genetic determinants potentially involved in the adhesion/muco-adhesion of IBB477, identifying two such genes: prtP and a gene coding for a protein with MUB and MucBP domains. Based on its persistence in the GIT and adhesive properties, L. lactis IBB477 is a candidate carrier strain for expression of influenza haemagglutinin (HA) protein for induction of mucosal immune response.


INTRODUCTION
Lactococci can be used in functional food (probiotics) and health-related applications (mucosal vaccines, therapeutic drug delivery), both potentially involving adhesive properties. The main reason for using the mucosal route of vaccination is that most infections (gastrointestinal, respiratory or genital infections), including influenza, affect a mucosal surface or start from one. Thus, adhering to the mucosal surface of the gastrointestinal tract (GIT) is tightly correlated with bacterial pathogenic activity and at the same time reflects the protective capabilities of "good" bacteria, e.g., lactic acid bacteria (LAB), including those serving as carriers for delivery of antigens and immunomodulatory factors. Furthermore, this route of vaccine application is often required for induction of a protective immune response. The mucosal vaccine should be targeted to the mucosal inductive sites and should be protected from physical elimination and enzymatic digestion (Holmgren & Czerkinsky, 2005). The mucosal delivery systems include live bacterial vectors and among them those that use commensal bacteria, such as lactobacilli or certain streptococci and staphylococci. We have shown that using Lactococcus as a producer of an influenza virus antigenic protein, haemagglutinin, is a promising approach (Szatraj et al., 2014).
Surface proteins of Gram-positive bacteria play a crucial role in the bacterial adhesion to host tissues. In L. lactis, only three surface proteins with a direct implication in adhesion have been described to date: the chromosomally-encoded sex factor aggregation protein CluA (Godon et al., 1994), and two plasmid-encoded proteins, serine proteinase PrtP (Reid & Coolbear, 1999) and protein YghE2 displaying pilin characteristics (Meyrand et al., 2013). The PrtP proteinase is responsible for enhanced cell surface hydrophobicity and adhesion to solid surfaces (Habimana et al., 2007), whereas CluA is involved in mediating cell-to-cell contact and DNA transport during conjugation (Stentz et al., 2006). Recently, the role of pilin in adhesion has been demonstrated by inactivation of the pilin gene yhgE2, which led to the loss of adhesion of L. lactis to Caco-2 cells (Meyrand et al., 2013). In lactobacilli, the most common proteins that have been shown to promote adhesion to mucus are those containing MUB and MucBP domains (Van Tassell & Miller, 2011;Roos & Jonsson, 2002). Based on bioinformatics studies, MUB domain-containing proteins have been identified and characterized in nine LAB species, including two Lactococcus representatives (Boekhorst et al., 2006).
The present work was focused on the adhesive and muco-adhesive properties of L. lactis subsp. cremoris IBB477. This strain, originally isolated from raw cow milk in Poland, persists in the GIT of germ-free rats and was shown to be tetracycline-resistant (Boguslawska et al., 2009). The adhesion of the IBB477 strain to an abiotic polystyrene surface was quantified and genetic determinants encoding putative adhesion/muco-adhesion capacity of IBB477 were identified. These studies were carried out with an intention to use IBB477 as a candidate for
Mucins and PGM-coated polystyrene plates. Type III mucin from porcine stomach (PGM) (lyophilized powder, cat. no. M1778, Sigma-Aldrich) was dissolved at 10 mg/mL in phosphate buffered saline (PBS) pH=7.5 just before use. Adhesion of L. lactis to PGM was determined on polystyrene microtiter 96-well plates (cat. no. 163320, Thermo Fischer Scientific Nunc A/S, Roskilde, Denmark) coated with 200 µL of PGM solution (10 mg/mL) and incubated overnight at 4°C, with gentle agitation on a platform rocker shaker. After incubation, the wells were washed three times with PBS and five times with sterile MilliQ-grade water to remove loosely bound material. The plates were air-dried and used directly after preparation.
Adhesion assay under static conditions. Adhesion of bacterial cells to bare polystyrene (PS) or PGM-coated polystyrene (PS+PGM) was tested on the microtiter plates, using the technique described by Christensen et al. (1985;1995), as follows. Bacteria from overnight cultures diluted to OD 660 nm of 1.0 were harvested by centrifugation at 9 000 × g for 1 min and resuspended in an equal volume of PBS. A volume of 100 µL of bacterial suspension was added to each well (eight for each strain). After a 3-h incubation under static conditions at 30°C the wells were carefully washed three times with 200 µL of sterile MilliQ-grade water to remove unbound bacteria. Bound cells were stained with crystal violet (cat. no. 109218, Merck, Darmstadt, Germany) (100 µL per well) at room temperature for 10 min and rinsed three times with water as above, and once with 200 µL of 96% ethanol to remove excess stain. Finally, stained bacteria were suspended in 200 µL of 96% ethanol and optical density was determined at 583 nm on a Synergy HT Multi-Detection Reader (BioTek Instruments Inc., Winooski, VT). The average value of eight measurements was calculated after rejecting extreme results. Bacterial adhesion was determined in three independent experiments and the results are presented as means ± standard deviations. A statistical analysis was performed using Welch t-test. Each microtiter plate included the negative control strain L. lactis MG1820, low-adhesive to PS and PS+PGM (Dague et al., 2010;Le et al., 2011) and blank wells with PBS.
Primer design and PCR amplification. Identification of genes cluA, prtP, yhgE2 and the gene coding for a MucBP domain-containing protein in L. lactis IBB477 was performed by PCR in a 25-µL reaction volume, with 0.625 U of Taq polymerase (Fermentas), 200 µM deoxynucleoside triphosphate mix (Fermentas), 0.4 µM specific primers (Table 1) and 2 µL of the template DNA. DNA templates for PCR amplification were obtained as follows: supernatants were obtained from a single bacterial colony suspended in 100 µL of sterile MilliQ-grade water, mixed with 50 mg of sterile glass beads (Sartorius Stedim Biotech, Goettingen, Germany), and disrupted in a Mini-Beadbeater apparatus (BioSpec Products, Inc., Bartlesville, OK, USA). The PCR mixture was pre-denaturated at 94°C for 5 min and incubated for 30 cycles, each consisting of 3 steps (94°C for 30 s, 55°C for 1 min, and 70°C for 2 min), with a final cycle of 70°C for 7 min.
DNA sequencing. The complete gene coding for the MucBP domain-containing protein from IBB477 was amplified with primers dopUC19F and dopUC19R (Table 1) and cloned into pUC19 vector. In order to facilitate sequence analysis of long cloned DNA fragment, the Deletion Kit for Kilo-Sequencing (TaKaRa Bio Inc., Dalian, China) was used according to manufacturer's instructions. The sequencing products were analysed on an ABI 3730/xl Genetic Analyzer in the Laboratory of DNA Sequencing and Oligonucleotide  Synthesis, Institute of Biochemistry and Biophysics, Polish Academy of Sciences. Nucleotide sequences were compared with those referenced in the GenBank database using the online BLAST program (http:// www.ncbi.nlm.nih.gov/BLAST/). Nucleotide sequence analysis. Sequence similarity was detected with BLAST program. Signal peptides were predicted with SignalP 4.0 (Petersen et al., 2011), sortase recognition sites with LPxTG-type motifs were predicted with the CW-PRED method (Litou et al., 2008) (http://biophysics.biol.uoa.gr/CW-PRED/index.jsp) and MUB domains were predicted with a developed HMMs (Boekhorst et al., 2005;. Other domains in MucBP domain-containing proteins were identified using Pfam database (Punta et al., 2012).
Nucleotide sequence accession number. The complete nucleotide sequence of the mub gene coding for the LPXTG-anchored mucus-binding protein (in this work also mentioned as the MucBP domain-containing protein) from IBB477 determined in this study has been deposited in the GeneBank under accession number JX845572.

Adhesion of L. lactis IBB477 to bare and PGM-coated polystyrene plates
As the L. lactis IBB477 strain persists in the GIT of germ-free mice (Boguslawska et al., 2009) we speculated that it should display a strong adhering potential to the mucous layer of the GIT. The binding ability of IBB477 strain to PS was tested from early to late exponential growth phases, corresponding to OD 660 nm of 0.4, 0.7 and 1, and also in the post-exponential (stationary) growth phase (overnight culture) (Fig. 1). Irrespective of the growth phase the adhesion level was the same. Therefore, L. lactis cells from overnight cultures diluted to OD 660 nm of 1 were used in subsequent experiments.
Next, the microtiter plate method was used to quantitate the adhesive properties of IBB477 to PS used as a standard abiotic surface, and to PS+PGM. Adhesion was expressed as the optical density (OD 583 nm ) of stained cells (Fig. 2). As expected, the IBB477 strain adhered significantly better than the low-adhering MG1820 control strain, with the p-value < e -11 (95% CI = 0.54-0.76) and p-value < e -16 (95% CI = 0.077-0.080) for PS and PS+PGM, respectively. IBB477 showed about 2-fold better adherence compared with MG1820 strain to PS and ca. 9-fold better adherence to PS+PGM.

Searching for genetic determinants encoding adhesion and muco-adhesion capacity of L. lactis IBB477
The genes were selected for analysis based on the current knowledge pertinent to adhesion of bacteria from the genus Lactococcus (Giaouris et al., 2009). Amplification of cluA and yhgE2 genes gave the expected DNA products of 2022 bp and 813 bp only for positive controls, MG1363 (cluA + ) and TIL448 (yhgE2 + ), which suggested that these genes are absent in the IBB477 genome. The PCR reaction with the prtPF and prtPR primers resulted in the generation of DNA products of 895 bp for the control strain J60011 (prtP + ) as well as for IBB477. The PCR product obtained for IBB477 was positively verified as prtP by DNA sequencing and comparison of the sequence against the NCBI nucleotide database.
In addition, amplification of the gene coding for the MucBP domain-containing protein resulted in the formation of the expected PCR product of 1858 bp for the control strain MG1363. However, for IBB477 this DNA product was approximately 500 bp larger. Its sequence analysis confirmed amplification of a gene coding for a protein containing MucBP domains. The complete sequence of this gene was determined with the aid of the Deletion Kit for Kilo-sequencing. The amino acid (aa) sequence was subsequently analyzed and compared with aa sequences of corresponding genes or pseudo-genes in sequenced lactococcal genomes (Table 2). It appears that the protein encoded by the gene of 3240 bp from the IBB477 genome contains four MucBP domains as well as four partly overlapping but larger MUB domains, postulated to be present only in LAB (Boekhorst et al., 2006). Furthermore, other structures such as an LPxTGtype motif, C-terminal anchor and Gram-positive anchor, predicted to be involved in binding to extracellular components, were identified in this protein as well ( Table 2). The complete nucleotide sequence of the gene has been deposited in the GeneBank as mub gene coding for the LPXTG-anchored mucus-binding protein.  Adhesion to mucosa is essential for the bacterial persistence in the host and may promote beneficial health effects such as stimulation of the innate immune system to generate effective adaptive immunity in the case of vaccines against infections. Thus a candidate lactic acid bacterium for influenza antigen delivery through the GIT should display adhering properties. The L. lactis IBB477 strain was selected due to its unexpected ability to persist in the GIT of germ-free rats (Boguslawska et al., 2009) indicating that it should possess adhesive properties. We focused on the interaction between L. lactis and mucins. To this end, the microtiter plate method, aiming at fast screening of adhesive properties on a large panel of strains, was first optimized and then used to analyze adhesion of IBB477 to bare and PGM-coated polystyrene plates. Under both conditions the adhesion of IBB477 was much higher than that observed for the low-adherent MG1820; however, the difference was more pronounced on PS+PGM (ca. 9-fold), than on PS (ca. 2-fold).
Next we focused on potential genetic determinants involved in adhesion/muco-adhesion of L. lactis IBB477. To this end, we selected for PCR amplification all the genes coding for surface proteins, comprising adhesins and aggregation proteins that have been described to date in L. lactis. We found that the cluA and yhgE2 were absent in the IBB477 genome, while the prtP gene, coding for a serine proteinase shown to enhance cell surface hydrophobicity and mediate adhesion to solid surfaces, was present (Habimana et al., 2007). This gene has been found in only a few strains out of the 22 sequenced LAB Liu et al., 2010). The presence of PrtP was also explored using pangenome comparative genome hybridization analysis in 39 L. lactis strains (Liu et al., 2010). According to the authors, prtP is mainly present in L. lactis subsp. cremoris, although several L. lactis subsp. lactis strains may also harbour this gene. This statement corroborates the taxonomic position of IBB477 that has been identified as L. lactis subsp. cremoris. Furthermore, we revealed the presence of a gene coding for a MucBP domain-containing protein in the genome of IBB477, which was named mub gene coding for the LPXTG-anchored mucus-binding protein. The nucleotide sequence analysis demonstrated that this gene is longer than that found in other L. lactis subsp. cremoris genomes. Indeed, this gene from IBB477 contains four MucBP and four partly overlapping but larger MUB domains, which are postulated to play an important role in the adherence to mucus. It is proposed that each additional copy tends to increase the cell affinity for mucins (Boekhorst et al., 2006). Furthermore, bioinformatics analysis showed that the protein encoded by this gene contains an LPxTG anchoring motif, and C-terminal and Gram-positive anchors, all predicted to be involved in binding to extracellular components. The role in adhesion of the prtP and mub genes in IBB477 will be further analyzed with knockout mutants. The highest number of MUB and MucBP domains among the L. lactis subsp. cremoris strains sequenced so far seems likely to enhance the muco-adhesion capacity of IBB477 and its potential to induce a protective immune response when applied as an oral vaccine against avian flu.