Design and Characterization of a Novel Tool for the Antigenic Enrichment of Actinobacillus pleuropneumoniae Outer Membrane

Production and isolation of recombinant proteins are costly and work-intensive processes, especially in immunology when tens or hundreds of potential immunogens need to be purified for testing. Here we propose an alternative method for fast screening of immunogen candidates, based on genetic engineering of recombinant bacterial strains able to express and expose selected antigens on their outer membrane. In Actinobacillus pleuropneumoniae, a Gram-negative porcine pathogen responsible for extensive economic losses worldwide, we identified a conserved general secretion pathway (GSP) domain in the N-terminal part of the outer membrane protein ApfA (ApfA stem: ApfAs). ApfAs was used as an outer membrane anchor, to which potential immunogens can be attached. To enable confirmation of correct positioning, ApfAs, was cloned in combination with the modified acyl carrier protein (ACP) fluorescent tag ACP mini (ACPm) and the putative immunogen VacJ. The chimeric construct was inserted in the pMK-express vector, subsequently transformed into A. pleuropneumoniae for expression. Flow cytometry, fluorescence imaging and mass spectrometry analysis were employed to demonstrate that the outer membrane of the transformed strain was enriched with the chimeric ApfAs-ACPm-VacJ antigen. Our results confirmed correct positioning of the chimeric ApfAs-ACPm-VacJ antigen and supported this system’s potential as platform technology enabling antigenic enrichment of the outer membrane of A. pleuropneumoniae.


Introduction
Since its inception, genetic engineering has enabled protein expression and purification in a variety of microorganisms [1,2]. One of the possible applications of this technique is the expression of recombinant antigens for vaccine development, defined as immunogens. In the case of bacterial vaccines, these antigens are often proteins located on the outermost surface of the bacteria [3,4].
The identification of useful protein immunogens is not an easy task, despite the contribution offered by the advent of bioinformatic screening tools for these purposes [5][6][7]. One of the main challenges is the identification of conserved proteins able to offer protection against ideally all known variants of the bacterial species/types targeted [8,9]. Even when conserved candidates are successfully identified, prediction tools offer disappointingly little help regarding the actual potential of these

ApfA s is a Conserved OM Domain
Analysis by PANTHER™ identified amino acid residues 14-67 of the ApfA protein (ApfA s ) as a conserved domain of general secretion pathway (GSP) proteins (PANTHER ID: PTHR30093). Further in silico analysis corroborated the inferred OM localization of ApfA (PSORTb 3.0), also supported by the prediction of a cytoplasmic domain at the N-terminal of ApfA (residues 1-11; PHOBIUS) and the folding of residues 16-66 to form a hydrophilic alpha-helix (I-TASSER).
Analysis by protein BLAST (non-redundant protein sequences database) revealed a significant homology of the ApfA s domain with peptidase-dependent pilins from several Enterobacteriaceae species. The following entry is reported here as representative of the analysis: MULTISPECIES: prepilin peptidase-dependent pilin [Enterobacteriaceae] (NCBI accession number: WP_000360918.1; 82% query coverage, 51,79% identity).

ApfA s -ACP m -VacJ is Expressed and Localized on A. pleuropneumoniae OM
Fluorescence imaging indicated that pMK_apfa s -ACP m -vacJ cells emitted higher fluorescent signal as compared to the wild type (wt) (Figure 1) when AcpS and fluorescent coenzyme A (CoA) are provided in solution. Under similar conditions, flow cytometry analysis detected a significant increase (p < 0.01) between 30-60% of the geometric mean of the emitted green fluorescence signal from the pMK_apfa s -ACP m -vacJ cells when compared to the wt ( Figure 2).   Sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) analysis showed no qualitatively detectable differences in banding patterns of the OM fractions of pMK_apfa s -ACP m -vacJ and wt cells ( Figure 3). MS analysis of the OM fraction of pMK_apfa s -ACP m -vacJ cells assigned 49 total spectra (11 exclusive unique peptides, 19 exclusive unique spectra) to VacJ, as compared to 35 total spectra (10 exclusive unique peptides, 17 exclusive unique spectra) assigned to VacJ for the OM fraction of wt cells (Table S1), indicating that the concentration of VacJ was about 29% higher for the pMK_apfa s -ACP m -vacJ OM fraction as compared to the wt. On the contrary, no significant differences in total spectra assigned to ApfA were detected between OM fractions of pMK_apfa s -ACP m -vacJ and wt strains (Table S1). No spectra were assigned to the ACP m tag in the OM fractions of both pMK_apfa s -ACP m -vacJ and wt strains (Table S1).

Discussion
The in silico design of the ApfAs-ACPm-VacJ chimera was a multi-step process that can be divided in the identification of the three domains comprising the final protein: (i) an OM anchoring domain (ApfAs); (ii) a domain for post-expression localization and quantification of the chimeric antigen; (iii) a putative immunogen to be tested (VacJ).
(i) ApfA is a fimbrial subunit protein involved in the initial stages of colonization of A. pleuropneumoniae whose role in pathogenesis and potential as an immunogen have previously been investigated [28][29][30][31][32][33]. As a well-characterized fimbrial protein, we hypothesized that ApfA would be exposed and anchored on A. pleuropneumoniae OM. This was subsequently supported by the in silico analysis performed in the current study, leading us to conclude that the ApfAs domain would likely fold into an OM ''stem'' (hence the ''S'' in ApfAs) and could thus act as an anchoring domain within a chimeric open reading frame (ORF) designed to expose selected antigens on the surface of A. pleuropneumoniae.
Due to the high conservation across Gram-negative bacteria of the general secretion pathway (GSP) domain identified in the ApfA sequence, we also hypothesized that a delivery system based on ApfA could offer a broader applicability and be employed for the selective OM enrichment of bacteria other than A. pleuropneumoniae. Additionally, the lack of significant homology between the ApfAs domain and a selection of other proteins previously employed in membrane display strategies (OM proteins OmpA, LamB and PhoE; fimbrial and flagellar proteins FimH, PapA and flagellin; Autotransporter protein AidA) [34][35][36] supported the novelty of the strategy proposed. ApfAs was thus selected as a suitable OM anchor.

Discussion
The in silico design of the ApfA s -ACP m -VacJ chimera was a multi-step process that can be divided in the identification of the three domains comprising the final protein: (i) an OM anchoring domain (ApfA s ); (ii) a domain for post-expression localization and quantification of the chimeric antigen; (iii) a putative immunogen to be tested (VacJ).
(i) ApfA is a fimbrial subunit protein involved in the initial stages of colonization of A. pleuropneumoniae whose role in pathogenesis and potential as an immunogen have previously been investigated [28][29][30][31][32][33]. As a well-characterized fimbrial protein, we hypothesized that ApfA would be exposed and anchored on A. pleuropneumoniae OM. This was subsequently supported by the in silico analysis performed in the current study, leading us to conclude that the ApfA s domain would likely fold into an OM ''stem" (hence the ''S" in ApfA s ) and could thus act as an anchoring domain within a chimeric open reading frame (ORF) designed to expose selected antigens on the surface of A. pleuropneumoniae.
Due to the high conservation across Gram-negative bacteria of the general secretion pathway (GSP) domain identified in the ApfA sequence, we also hypothesized that a delivery system based on ApfA could offer a broader applicability and be employed for the selective OM enrichment of bacteria other than A. pleuropneumoniae. Additionally, the lack of significant homology between the ApfA s domain and a selection of other proteins previously employed in membrane display strategies (OM proteins OmpA, LamB and PhoE; fimbrial and flagellar proteins FimH, PapA and flagellin; Autotransporter protein AidA) [34][35][36] supported the novelty of the strategy proposed. ApfA s was thus selected as a suitable OM anchor.
(ii) The next feature of the chimeric antigen to be addressed was the need for post-expression localization and quantification. For this purpose, we relied on a method known as covalent labelling. This technique is based on the ability of certain enzymes to recognize specific peptide tags, catalyzing then the covalent transfer of substrates on the tagged protein [37]. The main reason behind the selection of this technique over autofluorescent-based detection here was the comparative advantage offered by covalent labelling when designing membrane-bound recombinant proteins. For covalent labelling in fact, both the enzyme and the substrate need to be externally supplied by the user. Contrary to autofluorescent-based detection, this ensures that only the exported fraction of the recombinant protein will be detected, as the labelling enzyme in particular will not be able to access intracellular compartments.
The selected tag for covalent labelling was the ACP-tag, a small tag recognized by the acyl carrier protein synthase (AcpS) for the covalent transfer of fluorescent derivatives of coenzyme A (CoA). [38]. Albeit relatively small, the full length of the ACP-tag is 77 amino acids, resulting in a 9-kDa domain. The sheer size of the ACP tag clearly hampers its usefulness for post-expression quantification and localization of recombinant immunogens, due to the likely alteration of the antigenic profile of the immunogen that the inclusion of such large domain would induce. Interestingly, it has been demonstrated that just eight residues of the ACP-tag are required for labelling by AcpS [39]. This shortened ACP-tag, defined here as ACP-mini (ACP m ), presents two advantages over the full ACP-tag, both related to its smaller size. First and foremost, a domain of just eight amino acids is far less likely to affect trafficking and antigenicity of the engineered protein it is included in [20]. Secondly, the 24 nucleotides that encode for the ACP m tag can be easily included as 5 overhangs in the primers used during cloning, reducing the number of inserts needed to be stitched together for the creation of the recombinant ORF, thus increasing transformation efficiency. These considerations led us to select the ACP m tag as a detection domain for the chimeric antigen discussed here.
(iii) The last step of the design consisted of identifying a native A. pleuropneumoniae immunogen to be tested. Ideally, this immunogen should possess the following characteristics: (1) proven immunogenicity, in order to provide immunological relevance to the chimera; (2) predicted OM subcellular localization, to increase the chances of seamless transport of the chimera through cellular membranes; (3) lack of toxicity, allowing high level of expression of the chimera in the host strain.
Protein homologs belonging to the VacJ lipoprotein family are part of the conserved VacJ/Yrb ATP-binding cassette (ABC) transport system involved in phospholipid translocation through the OM [40]. We previously demonstrated that the VacJ-like homolog expressed in A. pleuropneumoniae strains presents a predicted OM subcellular localization, is easily expressible as recombinant protein and is highly immunogenic when administered in vivo as part of an immunization regime [32,33]. For these reasons, VacJ was selected as putative immunogen to be included in the chimera design described in this study.
Quantification by fluorescent labelling indicated that the ApfA s -ACP m -VacJ chimera is expressed and localized on the OM of pMK_apfa s -ACP m -vacJ cells. When using a fluorescent CoA-based labelling system though it is important to keep in mind that Acetyl-CoA is a major component of several metabolic pathways in bacteria, such as the biosynthesis of fatty acids and phospholipids. Accordingly, many Gram-negative bacteria actively import CoA to the cytoplasm and possess a native acetyl-transferase activity due to the expression of a homolog of the recombinant AcpS enzyme used for ACP-tag labelling [38]. This represents one of the main disadvantages when using ACP-tag labelling in bacterial strains, including A. pleuropneumoniae, as the accurate estimation of the expression level of the tagged recombinant protein can be skewed by a high background due to: (i) translocation to the cytoplasm of the fluorescent CoA substrate administered extracellularly during labelling; (ii) Incorporation of fluorescent CoA substrate into untagged proteins by native AcpS proteins. Both these processes represent a central part of the normal metabolism of unmodified bacterial strains and thus an inherent shortcoming of this labelling system. To overcome this problem, the ACP tag may need to be replaced with similar tags and fluorescent derivatives [41] in order to achieve a more stringent quantification of the levels of expression of the recombinant chimera.
To overcome this potential issue, protein assays were performed on the OM fractions of pMK_apfa s -ACP m -vacJ and wt cells to verify the actual concentration of the ACP m -VacJ chimera. Although SDS-PAGE showed no qualitatively detectable differences in banding patterns of the OM fractions of pMK_apfa s -ACP m -vacJ and wt cells, MS analysis suggested an increase of about 29% in the concentration of VacJ for the pMK_apfa s -ACP m -vacJ OM fraction as compared to the wt. This confirmed the level of enrichment of the ACP m -VacJ chimera previously observed by fluorescent labelling. Interestingly, no significant differences in ApfA concentration were detected between OM fractions of pMK_apfa s -ACP m -vacJ and wt strains, possibly due to the technical challenge posed by extraction and purification of trans-membrane protein domains such as the N-terminal of ApfA s [42]. As expected, the small size of the ACP m tag prevented its identification by the MS methodology employed in this study.

In Silico Functional Prediction of ApfA Domains
A list of the genetic sequences, vectors and primers used in this study is provided in Table 1. The protein ApfA was selected for the identification of OM domains due to its well-characterized OM localization. The ApfA peptide sequence was analyzed by different online tools for functional prediction: (i) PANTHER™ scoring [43,44]; (ii) PHOBIUS [45]; (iii) I-TASSER [46]; (iv) PSORTb [7]. Nucleotides 1-205 from the apfA open reading frame were selected as ''stem" domain (apfA s ) for OM anchoring of the chimeric antigen. The amino acid sequence of the apfA s domain was analyzed by protein BLAST [47] to assess inter-species homology. Standard parameters were used unless otherwise stated. Table 1. Sequences and primers used in this study.

Construction and Transfer of the apfAs-ACPm-vacJ Expression Vector
apfAs and vacJ nucleotide sequences (444 and 747 nucleotides respectively) were amplified from genomic DNA using AccuPrime Taq DNA high-fidelity Polymerase (ThermoFisher Scientific, Waltham, MA, USA). The pMK-express vector was linearized by reverse amplification using the same polymerase, in order to replace the native GFPmut3 gene with the recombinant apfas-ACPm-vacJ ORF by ligation-independent cloning (LIC). apfAs and vacJ amplicons were joined with the linearized vector by three-point ligation-independent cloning, according to the In-Fusion protocol (Clontech Laboratories, Mountain View, CA, USA). The sequence of the ACPm tag was introduced in frame between apfAs and vacJ sequences by PCR-based tagging, including the 24 nucleotides of the tag into

Construction and Transfer of the apfA s -ACP m -vacJ Expression Vector
apfA s and vacJ nucleotide sequences (444 and 747 nucleotides respectively) were amplified from genomic DNA using AccuPrime Taq DNA high-fidelity Polymerase (ThermoFisher Scientific, Waltham, MA, USA). The pMK-express vector was linearized by reverse amplification using the same polymerase, in order to replace the native GFPmut3 gene with the recombinant apfa s -ACP m -vacJ ORF by ligation-independent cloning (LIC). apfA s and vacJ amplicons were joined with the linearized vector by three-point ligation-independent cloning, according to the In-Fusion protocol (Clontech Laboratories, Mountain View, CA, USA). The sequence of the ACP m tag was introduced in frame between apfA s and vacJ sequences by PCR-based tagging, including the 24 nucleotides of the tag into the 15 bp overhangs of apfA-Rev and vacJ-Fwd primers ( Table 1). The resulting pMK_apfa s -ACP m -vacJ construct was transformed into E. coli Stellar competent cells (Clontech Laboratories, Mountain View, CA, USA) according to the In-Fusion protocol, with selection of recombinant clones on brain heart infusion (BHI) plates supplemented with 75 µg mL −1 kanamycin. Positive clones were analyzed by PCR and sequencing to verify the acquisition of the vector and the presence of the correct recombinant ORF. The pMK_apfa s -ACP m -vacJ construct was extracted and purified using a miniprep kit (Qiagen, Hilden, Germany) and subsequently transformed into E. coli electrocompetent S17-1 λpir cells, prepared and transformed as described in Reference [48]. Positive S17-1 λpir clones were then used for transferring the pMK_apfa s -ACP m -vacJ vector into A. pleuropneumoniae wt cells by conjugation, following the protocol described in Reference [49]. A. pleuropneumoniae pMK_apfa s -ACP m -vacJ clones were analyzed by PCR to verify the acquisition of the recombinant pMK_apfa s -ACP m -vacJ vector.

Outer Membrane Isolation
A. pleuropneumoniae serotype 8 MIDG2331 [50] wt and the pMK_apfa s -ACP m -vacJ transformed strain were inoculated into 5 mL of BHI broth containing 20 mg L −1 β-NAD (β nicotinamide adenine dinucleotide) (Merck Millipore, Burlington, Massachusetts, United States) in sterile 50 mL centrifuge tubes. Next, the cultures were incubated overnight (≥16 h) at 37 • C with agitation (200 rpm) in an aerobic atmosphere. 10 µL of the cultures were inoculated into 50 mL fresh BHI broth containing 20 mg L −1 β-NAD in 250 mL Erlenmeyer conical flasks. After incubation for 12-14 h, the cultures were adjusted to an OD 600 of 1 and centrifuged for 1 h at 2600× g. The pelleted cells were then resuspended, washed once in 10 mmol L −1 HEPES of pH 7.4 and stored at −80 • C. After thawing, 15 µl of 1 mmol L −1 MgCl 2 and standard units of DNAse I and lysozyme were added to the cells. The cells were lysed using a bead beater (Precellys Minilys, Bertin Technologies, Montigny-le-Bretonneux, France) (3 cycles of 3200× g for 3 min). The lysate was then centrifuged for 5 min at 29,000× g at 4 • C in a tabletop centrifuge to remove cell debris. The supernatant was moved to a fresh 2 mL tube and centrifuged for 1 h at 29,000× g at 4 • C to collect membranes. The pelleted membranes were resuspended thoroughly in 0.2 mL 10 mmol L −1 HEPES of pH 7.4. The inner membrane was then solubilized with 0.2 mL of 0.2% lauroyl sarcosine in 10 mmol L −1 HEPES (pH 7.4) for 30 min at room temperature. Following this, the solution was centrifuged for 1 h at 29,000× g at 4 • C to pellet the OM. The pellet was then washed once with 10 mmol L −1 HEPES (pH 7.4) and resuspended in 30 µL of 10 mmol L −1 HEPES.

SDS-PAGE
For SDS-PAGE, a standard protocol was applied. 10 µL of sample buffer, Laemmli, 2x concentrate loading buffer and 10 µL of supernatant and OM fractions of A. pleuropneumoniae serotype 8 MIDG2331 wt and pMK_apfa s -ACP m -vacJ strain were loaded onto 4-20% gradient SDS-Page gel (Bio-Rad). The Precision Plus Protein™ Dual Color Standard (Bio-Rad, Hercules, CA, USA) was included for molecular weight estimation. The proteins were run at 150 V for 45 min and then stained with Coomassie Blue. Imaging was performed using a Gel doc ™ XR+ device with Image Lab ™ software (Bio-Rad, Hercules, CA, USA).

Liquid Chromatography-Mass Spectrometry
Protein concentrations of inner and OM fractions were determined by Pierce™ bicinchoninic acid (BCA) Protein Assay Kit (ThermoFisher Scientific, Waltham, MA, USA). For liquid chromatography-mass spectrometry (LC/MS) analysis, all samples were adjusted to the protein concentration of 0.4 mg mL −1 in 10 mmol L −1 HEPES of pH 7.4. The samples were digested with 10 µg trypsin (sequencing grade) (Promega, Madison, WI, USA) overnight at 37 • C. The digestion was stopped by adding 5 µL 50% formic acid and the generated peptides were purified using a ZipTip C18 (Merck Millipore, Burlington, MA, USA) according to the manufacturer's instructions and dried using a Speed Vac concentrator Plus (Eppendorf, Hamburg, Germany). The tryptic peptides were analyzed using an Ultimate 3000 nano-UHPLC system connected to a Q Exactive mass spectrometer (ThermoFisher Scientific, Waltham, MA, USA) equipped with a nano electrospray ion source.

Protein Identification by Scaffold Viewer
LC/MS data were analyzed using Scaffold TM (Scaffold_4.8.9, Proteome Software Inc., Portland, Oregon, United States). The threshold for peptide identification was set at >95.0% probability (Peptide Prophet algorithm) [51] with Scaffold delta-mass correction. The threshold for protein identification was set at >99.9% probability in addition to at least one identified peptide. Protein probabilities were assigned by the Protein Prophet algorithm [52]. Proteins that contained similar peptides and which could not be differentiated based on tandem mass spectrometry (MS/MS) analysis alone were grouped to satisfy the principles of parsimony.

Flow Cytometry
Inoculation loops containing stored (-80 • C) A. pleuropneumoniae serotype 8 MIDG2331 [50] wt and the pMK_apfa s -ACP m -vacJ plasmid containing strain were inoculated into 5 mL of BHI broth containing 20 mg L −1 β-NAD in sterile 50 mL centrifuge tubes. Next, the cultures were incubated overnight (≥16 h) at 37 • C with agitation (200 rpm) in an aerobic atmosphere. 10 µL of the cultures were inoculated in 50 mL fresh BHI broth containing 20 mg L −1 β-NAD in 250 mL Erlenmeyer conical flasks. After incubation for 12-14 h, the OD 600 of the cultures were adjusted to 0.1. Cells from 500 µL of the OD 600 adjusted cultures were harvested (5000× g, 4 min) and re-suspended in 100 µl of a staining solution. For the preparation of the staining solution, 10 −2 mol L −1 MgCl 2 , 0.625 nmol L −1 ACP synthase (NEB, Ipswich, MA, USA) and 0.25 nmol L −1 fluorescent CoA 488 (NEB, Ipswich, MA, USA) were added in BHI broth. Incubation was performed in a 1 mL Eppendorf tube in heating block at 37 • C for 1 h. The stained cells were harvested (5000× g, 4 min) and re-suspended in 200 µL of sterile filtered (0.22 µm sterile filters, Sartorius, Göttingen, Germany) Dulbecco's phosphate-buffered saline (ThermoFisher Scientific, Waltham, MA, USA) and analyzed by flow cytometry (BD Accuri ® C6, BD Biosciences, San Jose, CA, USA). Following acquisition, the data obtained were analyzed using the FlowJo V10 software (BD Biosciences, San Jose, CA, USA). Three independent experiments were performed. For each data set, the statistical difference between the mean green fluorescence values of wt and pMK_apfas-ACPm-vacJ strains was determined by t-test (p < 0.01).

Fluorescence Microscopy
Cultures of A. pleuropneumoniae serotype 8 MIDG2331 wt and pMK_apfa s -ACP m -vacJ strain were prepared as described for flow cytometry. After staining, the cells were harvested (5000× g, 4 min) and re-suspended in 50 µl of filtered (0.22 µm sterile filters) Dulbecco's phosphate-buffered saline (#141900, ThermoFisher Scientific, Waltham, MA, USA) and analyzed with fluorescence microscopy (Leica DM5000B, Leica Microsystems, Wetzlar, Germany). For the acquisition of images, phase contrast microscopy with 100× magnification and an oil-immersion objective lens were used. For fluorescence imaging, filter cube I3 with band pass 450-490 nm excitation filter, 510 nm dichromatic mirror and long pass 515 suppression filter were used. The cell cultures were mixed in a 1:1 ratio with 50% sterile glycerol on the glass slides just before the microscopy analysis in order to reduce the mobility of the cells for the acquisition of images. Three independent experiments were performed.

Conclusions
These results represent a valid proof of concept indicating that ApfA s could be used as a fusion partner to direct translocation and exposure of small proteins on A. pleuropneumoniae OM. This would ideally result in the generation of recombinant strains which, once inoculated in the host as part of an immunization regime, could elicit a lasting humoral response against the antigens selected for ApfA s -mediated OM enrichment. Additionally, the cross-species conservation of some ApfA s domains leads us to speculate that the expression platform described in this study could be used for the construction of recombinant chimeras aimed at antigen enrichment of the OM of other Gram-negative species of interest. Nonetheless, all these assumptions are preliminary in nature and further testing will be necessary to verify these hypotheses.