Costa‑Hurtado et al. Vet Res (2019) 50:69 https://doi.org/10.1186/s13567-019-0687-2 RESEARCH ARTICLE Open Access Haemophilus parasuis VtaA2 is involved in adhesion to extracellular proteins Mar Costa‑Hurtado*†, Laura Garcia‑Rodriguez†, Sergi Lopez‑Serrano and Virginia Aragon Abstract Haemophilus parasuis is part of the microbiota of the upper respiratory tract in swine. However, virulent strains can cause a systemic disease known as Glässer’s disease. Several virulence factors have been described in H. parasuis including the virulence-associated trimeric autotransporters (VtaAs). VtaA2 is up-regulated during infection and is only found in virulent strains. In order to determine its biological function, the vtaA2 gene was cloned with its native promotor region in pACYC184, and the transformed Escherichia coli was used to perform functional in vitro assays. VtaA2 was found to have a role in attachment to plastic, mucin, BSA, fibronectin and collagen. As other VtaAs from H. parasuis, the passenger domain of VtaA2 contains collagen domains. In order to examine the contribution of the collagen repeats to VtaA2 function, a recombinant vtaA2 without the central collagen domains was obtained and named vtaA2OL. VtaA2OL showed similar capacity than VtaA2 to adhere to plastic, mucin, BSA, fibronectin and plasma but a reduced capacity to adhere to collagen, suggesting that the collagen domains of VtaA2 are involved in collagen attachment. No function in cell adhesion and invasion to epithelial alveolar cell line A549 or unspecific binding to pri‑ mary alveolar macrophages was found. Likewise VtaA2 had no role in serum or phagocytosis resistance. We propose that VtaA2 mediates adherence to the host by binding to the mucin, found in the upper respiratory tract mucus, and to the extracellular matrix proteins, present in the connective tissue of systemic sites, such as the serosa. Introduction immune system by degradation of IgA [7], resistance to Haemophilus parasuis is a Gram-negative bacterium of phagocytosis by alveolar macrophages [8] and resistance the Pasteurellaceae family. It is an early colonizer of the to serum complement killing [9]. Thus, H. parasuis is upper respiratory tract and part of the normal microbiota able to reach internal organs, where it replicates in ser- of healthy pigs [1]. Under certain circumstances, viru- osal surfaces. Presence of systemic bacteria produces a lent strains of H. parasuis can produce Glässer’s disease, strong inflammation in the host [10]. Massive fibrin exu- which is characterized by fibrinous polyserositis, includ- dates, increased permeability by disruption of adherens ing arthritis, and meningitis [2]. Glässer’s disease repre- junctions, invasion of endothelial cells, and unbalanced sents a big economic burden for the pig industry [3] and inflammatory response are probably key steps in the prevention and control continue to be challenging since development of systemic disease, which causes the char- the pathogenicity of H. parasuis is not fully characterized. acteristic lesions of Glässer’s disease, including meningi- Comparison between virulent and non-virulent strains tis [10–13]. allowed the description of some mechanisms of the Antimicrobial treatment remains useful to control pathogenesis of H. parasuis. Strains of H. parasuis dis- Glässer’s disease at the farm level. However, the rise of play mechanisms for mucosal colonization by adhering antimicrobial resistances is a serious threat to animal and and invading epithelial and endothelial cells [4–6]. H. public health. Global efforts have been adopted to reduce parasuis virulence is associated to evasion of the innate the use of antibiotics in veterinary medicine. Besides, the use of antimicrobials may affect the colonization by H. parasuis and modify disease dynamics and the develop- *Correspondence: mar.costa@irta.cat † Mar Costa-Hurtado and Laura Garcia-Rodriguez are equal contributors ment of immune responses [14–16], which may finally IRTA, Centre de Recerca en Sanitat Animal (CReSA, IRTA-UAB), Campus de affect the immunocompetence of the host. la Universitat Autònoma de Barcelona, 08193 Bellaterra, Spain © The Author(s) 2019. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creat​iveco​mmons​.org/licen​ses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/ publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Costa‑Hurtado et al. Vet Res (2019) 50:69 Page 2 of 9 Virulence-associated trimeric autotransporters Table 1 Plasmids used in this work and their main (VtaA) are surface exposed proteins of the Type V characteristics secretion system involved in host–pathogen interac- Plasmids Description Reference tion [17, 18] and promising vaccine candidates [19]. Trimeric autotransporters are composed of an N‐termi- pACYC184 low copy number, CmR, TetR, ATCC number 37033 4245 bp nal signal peptide for transport across the inner mem- pLGR-vtaA2 pACYC184 with a 4016 bp insert, This work brane; a C‐terminal translocator domain that creates a including vtaA2 and its promo‑ pore with a β‐barrel structure in the outer membrane; tor region and a passenger domain that is secreted to the bacterial pLGR-vtaA2OL pACYC184 with a 2514 bp insert, This work surface [20]. The passenger domains of all the VtaAs including vtaA2OL and the promotor region of vtaA2 possess motifs and repeats characteristic of adhesins, pEGFP gfp, AmR Clontech hemagglutinins and invasins, with various centrally located copies of triple helix collagen-like repeats [17]. Cm: chloramphenicol, Tet: tetracycline, Am: ampicillin, R: resistant. Virulent H. parasuis strains have several copies of vtaA in their genome and some of them have been found to Table 2 Primers used in PCR in this work be immunogens [17, 21], demonstrating that these pro- teins are expressed during infection, and supporting Primer Primer sequence (5′–3′) their role in infection. Some vtaAs have been associated HPNK_01698_F GGA​GTC​ATG​C TC​AAC​AGG​C T specifically to virulent strains [17, 22, 23], but so far a HPNK_01698_R CAG​CCT​C TT​CCT​TCC​T TC​CG function has been assigned only to VtaA8 and VtaA9, BamHI-vtaAFa GCGC​GGA​TCC​TCT​TAG​T TT​TGT​GTA​ACT​C TTG​ with a role in phagocytosis resistance [18]. Through BamHI-vtaARa GCGC​GGA​TCC​T TC​TAA​T TT​ATA​GGT​GCT​AGA​T TA​C transcriptomic studies, vtaA8 and vtaA9 (together with OL-vtaA2F TGG​CGA​TCT​AGG​ACC​TAC​GGG​TCC​AAC​GGG​ AC other vtaA) were found to be up-regulated during lung OL-vtaA2R CCG​TAG​GTC​C TA​GAT​CGC​CAT​T TA​GTT​CAT​T infection [24]. Likewise, up-regulation of vtaA2 tran- vtaA2-865F AGC​AGT​GGC​AAG​GTA​GGC​ scription in lung was detected after 2 h and after 1, 2 vtaA2-2655R CTC​CAT​CGG​TAA​ACC​GCT​G and 3 days post-infection [24], indicating that VtaA2 is a BamHI site in primers is underlined. involved in for lung infection. VtaA2 comprises 1225 amino acids and presents a slightly different structure compared to VtaA8 and Cloning of vtaA2 and protein expression VtaA9. Although they display the same basic structure, A fosmid genomic library of H. parasuis virulent strain the number of LbR-YadA-like repeats and specially the Nagasaki previously produced [18] was used for vtaA2 number of collagen repeats differ between VtaA2 and amplification. In order to detect clones containing those found in VtaA8 or 9 [17, 25], suggesting a differ- the individual vtaA genes, a specific PCR previously ent function. described was used [23]. To identify fosmids carrying In the present study, we evaluated the function of vtaA2, a PCR to detect a vtaA2 fragment was performed VtaA2 by cloning vtaA2 from the virulent strain Naga- with primers HPNK_01698_F and HPNK_01698_R saki of H. parasuis with its native promotor in an E. (Nagasaki genome number ANKT01000000; Table 2) coli background. In vitro assays showed that VtaA2 [24]. Gene vtaA2 was detected in 2 fosmids and one of has adhesion features that could be important for host them was used as template to PCR-amplify vtaA2 with infection. Platinum™SuperFI™ DNA polymerase (Invitrogen, Bar- primers BamHI-vtaAF and BamHI-vtaAR [18] and celona, Spain) (Table 2). Amplicon size was confirmed Materials and methods and the product cloned into pACYC184 to produce Bacterial strains and plasmids pLGR-vtaA2. Plasmid pLGR-vtaA2 was introduced in Plasmids used in this work are indicated in Table 1. Plas- electrocompetent E. coli BL21 (DE3). mid pACYC184 is a low copy number vector, at about 15 In order to determine the role of the collagen-like copies per cell, and contains the p15A origin of replica- repeats present in the passenger domain of VtaA2, a tion. The unique BamHI site in the plasmid was used as recombinant VtaA2 without the central protein frag- cloning site. ment containing the collagen domains was produced. Escherichia coli BL21 (DE3) was used as the host for The 5′ terminus and 3′ terminus of the gene were ampli- recombinant plasmids and was grown in Luria–Bertani fied from the fosmid containing the vtaA2 with overlap- (LB) medium, supplemented with 30 µg/mL chloram- ping primers (Table 2). The PCR fragment containing phenicol (Cm30). the promoter region and nucleotides 1 to 1050 of vtaA2 Costa‑Hurtado et al. Vet Res (2019) 50:69 Page 3 of 9 (amplified with BamHI-vtaAF and OL-vtaA2R) and pLGR-vtaA2 and pLGR-vtaA2OL. Bacterial suspensions the second fragment containing nucleotides 2551 to were diluted down 1:100 in fresh LB broth with Cm30 3678 of the vtaA2 gene (amplified with OL-vtaA2F and and were used to inoculate wells in triplicate (100 µL/ BamHI-vtaAR) were purified and ligated, yielding pLGR- well). vtaA2OL. To confirm that the two gene fragments were Plates were incubated in static conditions or under in-frame, primers vtaA2-865F and vtaA2-2655R (down- shaking (100 rpm) for different time-points (3 h, 6 h, 12 h stream and upstream of the deleted gene section, Table 2) and 24 h). After incubation, plates were washed once by were used for sequencing. water immersion and attached bacteria were stained with Expression of the corresponding protein was assessed 0.1% (w/v) crystal violet for 2 min. After three washes in in the clones. E. coli BL21 containing pLGR-vtaA2 or water by immersion, plates were allowed to dry at 42 °C. pLGR-vtaA2OL were grown in LB-Cm30 up to late-loga- Stain in the wells was solubilized with 70% ethanol and tion was performed with TRIzol™ Reagent (Invitrogen) rithmic phase. Bacteria were pelleted and protein extrac- the released dye was quantified by measuring the absorb- ance at 590 nm with a microplate reader (Power Wave according to the manufacturer’s instructions. Protein HT Microplate Spectrophotometer, BioTek, Winooski, extracts were quantified by the BCA protein assay kit USA). ­ uPAGE® Novex 4–12% Bis–Tris (Pierce-Thermo Fisher Scientific, Sant Cugat del Vallès, Gel, followed by a transfer to a Hybond™ ECL™ nitro- Spain), loaded in a N Cellular adhesion and invasion assays (ATCC​® CCL-185™) was used for cell adhesion and inva- Epithelial cell line A549 from human lung carcinoma cellulose membrane (GE Healthcare, Barcelona, Spain). After blocking with 1% casein 0.05% Tween20 in PBS, sion experiments. Cells were seeded in 96-well plates membranes were incubated with the monoclonal anti- (SPL, Life Sciences) at a concentration of 6 × 104 cells/ body 69C6 [18] at 2 ng/µL in blocking solution. Proteins well in Dulbecco’s modified Eagle’s medium (DMEM; VtaA2 and its truncated version VtaA2OL were detected Hyclone GE Healthcare) supplemented with 10% fetal by chemiluminescence after incubation with peroxidase- bovine serum (FBS) and 1% l-glutamine; complete conjugated goat anti-mouse immunoglobulin (Sigma- DMEM (CDMEM). Plates were incubated overnight at Aldrich, Madrid, Spain) diluted 1:10 000. 37 °C with 5% ­CO2. Bacterial suspensions of E. coli BL21 pACYC184 and Attachment assays E. coli BL21 pLGR-vtaA2 were prepared to reach an The role of VtaA2 in attachment was evaluated on plastic ­OD600 of 0.3 (Spectrophotometer VIS7200), which corre- and on different protein substrates. Various time points sponded approximately to 1 ­ 08 CFU/mL. Protein expres- (6 h, 12 h, 24 h, 48 h) were tested under static or shacking sion under conditions was confirmed by immunoblot as conditions. Attachment assays were performed in 96-well described above. Bacterial concentration was confirmed plates and attachment was measured by crystal violet by serial dilution and plating. Duplicate wells with epi- staining as previously described with some modifications thelial cells were inoculated with 100 µL of each bacterial [26]. First, flat-bottom 96-well polystyrene plates (SPL, suspension and plates were incubated at 37 °C with 5% Life Sciences, Naechon- Myeon, Korea) were coated with ­CO2 for 1 h. After incubation, wells were washed three 200 µL of 25 µg/mL mucin (Sigma Aldrich), fibronec- times with PBS to eliminate unbound bacteria. To release tin (Invitrogen) or collagen (type I, Sigma Aldrich) in bacteria, epithelial cells in each well were lysed with 100 carbonate-bicarbonate buffer (pH 9.6) in triplicate. Por- µL of sterile Milli-Q water and scrapped. Bacteria were cine plasma was also used. Plasma was obtained by cen- counted by plating serial dilution of the lysates. This assay trifugation of heparinized blood from a healthy animal was performed twice. and used for coating at 1:2 or 1:10 dilution. Sampling To test cell invasion, the adhesion assay was performed was performed under institutional authorization and as above for adhesion but with an additional step after the followed good veterinary practices. According to Euro- 1 h incubation to allow bacteria to invade the cells. The pean (Directive 2010/63/EU) and Spanish (Real Decreto additional step consisted of 2 h incubation with 5 µg/mL 53/2013) normative, this procedure did not require spe- penicillin G and 100 µg/mL gentamicin to kill extracel- cific approval by an Ethical Committee. Bovine serum lular bacteria. After washes, cells were lysed and bacteria albumin (BSA; Sigma Aldrich) was used as non-specific were counted as in the standard adhesion procedure. protein control and some wells remained uncoated. After overnight incubation at 4 °C, wells were washed once Interaction with porcine alveolar macrophages (PAMs) with sterile PBS. Bacterial suspensions were prepared The ability of the E. coli clone carrying VtaA2 to attach to in PBS (­OD600 of 0.3) using overnight agar cultures on PAMs was also tested. All procedures involving animals LB-Cm30 of E. coli BL21 pACYC184 (negative control), followed EU and Spanish normative (Directive 2010/63/ Costa‑Hurtado et al. Vet Res (2019) 50:69 Page 4 of 9 EU and Real Decreto 53/2013). PAMs were collected Tenfold serial dilutions of each well were made before from bronchoalveolar fluid lavages from healthy pigs and after incubation, and bacteria were quantified by that were euthanized for this purpose following good plating onto LB-Cm30. The counts of viable bacteria veterinary practices. According to European (Directive before and after serum treatment were compared. 2010/63/) and Spanish (Real Decreto 53/2013) norma- tive, this procedure did not require specific approval by Results an ethical committee. PAMs were isolated and stored at VtaA2 is involved in attachment to animal proteins −150 °C following standard procedure [8]. A fosmid containing vtaA2 was used as template to Cell adhesion assays to PAMS were performed on ice amplify and subsequently clone vtaA2 with its putative to avoid active phagocytosis. Cells were thawed, gently promotor region in pACYC184, yielding pLGR-vtaA2. E. washed of cryopreservant and seeded in 96-well plates at coli BL21 (pLGR-vtaA2) showed a growth in LB-Cm30 a concentration of 5 × 105 cells per well. Plates were incu- similar to the E. coli with the empty plasmid pACYC184 bated at 37 °C with 5% ­CO2 overnight to allow the attach- (not shown). Compared to the empty plasmid control, ment of the cells to the wells. Bacterial suspensions were BL21 (pLGR-vtaA2) exhibited more capacity to attach to made in CDMEM (­ OD600 of 0.3) and 100 µL of these sus- plastic and to all the purified proteins tested (Figure 1), pensions were used to inoculate triplicate wells. After a indicating that VtaA2 promoted attachment. Attach- centrifugation during 10 min at 100 × g plates were incu- ment to plastic (Figure 1A), mucin (Figure 1B), collagen bated on ice for 1 h. Bacterial concentration in the initial (Figure 1C), fibronectin (Figure 1D) and BSA (Figure 1E) inoculum was confirmed by dilution and plating. After by the vtaA2 clone was detected as soon as 6 h of incu- incubation, PAMs were washed three times with PBS to bation. However, no attachment to swine plasma was eliminate unbound bacteria, and 100 µL of 0.1% saponin observed, indicating no ability of VtaA2 to bind to plasma (Sigma Aldrich) in PBS was added to the wells to lyse the components (Figure 1F). PAMs. Serial dilutions were plated in LB agar-Cm30. The assay was repeated three independent times. Phagocytosis assays were performed as described Collagen domains of VtaA2 are involved in collagen before [8]. PAMs were seeded in 6-well plates at a con- attachment centration of 5 × 105 PAMs/well in CDMEM. After To examine the role of the collagen domains of VtaA2 in incubation to allow attachment of the PAMs, (overnight the attachment observed, a truncated version of vtaA2 incubation) wells were inoculated with bacteria at a mul- lacking the central portion of the gene that encoded tiplicity of infection (MOI) of 200. E. coli clones were the collagen domains was cloned in pACYC184 to yield labelled using different strategies to ensure consistent pLGR-vtaA2OL. Similarly, BL21 (pLGR-vtaA2OL) dis- fluorescence of all the bacterial population, using fluo- played comparable level of attachment to plastic, mucin, rescein isothiocyanate (FITC), SYTO9 green fluorescent fibronectin and BSA than the E. coli with the complete nucleic acid stain (Thermo Fisher Scientific) or transfor- vtaA2 (Figure 1). However, a reduction in the attachment mation with pEGFP (plasmid carrying the green fluores- to collagen was observed in the truncated VtaA2OL (Fig- cent protein [GFP] gene; Table 1). After incubation at ure 1C), indicating that the collagen domains of VtaA2 37 °C for different times, wells were washed to eliminate may play a role in attachment to collagen through colla- unbound bacteria and PAMs with associated bacteria gen-collagen interaction. were detected by flow cytometry in a MACS Miltenyi cytometer (MILTENYI Biotec, Madrid, Spain). Assays VtaA2 does not promotes cell adhesion or invasion were performed in duplicate and were repeated using To explore if VtaA2 could play a role in adhesion to epi- PAM from different animals. thelial cells, an assay with the pulmonar epithelial cell line A549 was performed. After 1 h of incubation to Serum resistance assay allow the attachment of the bacteria to the cells, no dif- To test if VtaA2 was involved in serum resistance, an ferences were observed between the control E. coli BL21 in vitro assay with rabbit serum was performed as (pACYC184) and the clone carrying vtaA2 (Figure 2), described before [9] with some modifications. Briefly, indicating that VtaA2 does not have a direct effect on overnight colonies of E. coli BL21 (pACYC184) and E. bacterial adhesion to epithelial cells. In agreement with coli BL21 (pLGR-vtaA2) were used to prepare bacterial the lack of effect on cell adhesion, we did not detect a role suspensions. Bacteria and serum were mixed in 96 well of VtaA2 in cellular invasion (not shown). In fact, nei- plates to give a final quantity of bacteria in the wells of ther BL21 (pACYC184) nor BL21 pLGR-vtaA2 were able ­105–106 CFU and a final concentration of 50% serum. to invade epithelial A549 cells, and internalized bacteria Mixtures were incubated at 37 °C for 30 min and 1 h. were not detected in any of the cases. Costa‑Hurtado et al. Vet Res (2019) 50:69 Page 5 of 9 3.0 3.0 2.5 Abs, 590 nm Plastic A 2.5 Mucin B 2.0 2.0 1.5 1.5 1.0 1.0 0.5 0.5 0.0 0.0 0 5 10 15 20 25 30 0 5 10 15 20 25 30 3.0 3.0 2.5 Collagen C 2.5 Fibronectin D Abs, 590 nm 2.0 2.0 1.5 1.5 1.0 1.0 0.5 0.5 0.0 0.0 0 5 10 15 20 25 30 0 5 10 15 20 25 30 3.0 3.0 E F 2.5 BSA 2.5 Plasma Abs, 590 nm 2.0 2.0 1.5 1.5 1.0 1.0 0.5 0.5 0.0 0.0 0 5 10 15 20 25 30 0 5 10 15 20 25 30 Time of incubation, h Time of incubation, h Figure 1 Kinetics of attachment to different substrates of E. coli clones expressing VtaA2 and VtaA2OL. Wells were coated with mucin (B), collagen (C), fibronectin (D), BSA (E), or plasma (F), or were left uncoated (plastic A). Plates were inoculated and incubated under static conditions. Attachment was measured at different time points. Black circles: BL21 (pLGR-vtaA2); red triangles: BL21 (pLGR-vtaA2OL); and white circles BL21 (pACY184) (control). Bars represent the standard deviation of the mean ­Abs590 from triplicate wells. VtaA2 does not play a role in the interaction with porcine between them. On one hand, BL21 (pLGR-vtaA2) trans- alveolar macrophages (PAMs) formed with pEGFP did not consistently express GFP, Assays to evaluate the role of VtaA2 in phagocytosis unlike the empty plasmid control, making this labelling resistance were performed. Unfortunately, fluorescent inadequate for this study. On the other hand, FITC did labelling of E. coli BL21 (pLGR-vtaA2) and the control not label properly the control BL21 (pACYC184), and E. coli was challenging due to differences in staining 5 times more compound was needed to reach similar Costa‑Hurtado et al. Vet Res (2019) 50:69 Page 6 of 9 (Figure 4), indicating that VtaA2 does not play a role in 109 phagocytosis. 108 Additionally, attachment to PAMs was examined by incubation of the clones with PAMs on ice during 1 h. CFU / mL 107 High levels of attachment to the surface of PAMs were found and no differences were detected in E. coli BL21 106 (pLGR-vtaA2) and the empty plasmid control (not 105 shown), indicating that VtaA2 does not to play a role in adhesion to macrophages. 104 pACYC184 pLGR-vtaA2 Figure 2 VtaA2 has no role in the adhesion to lung epithelial VtaA2 is not involved in serum resistance cells A549 cells. Adhesion of the clones E. coli BL21 (pLGR-vtaA2) and (pACYC184) to lung epithelial cells A549 after 1 h of incubation Both BL21 (pACYC184) and BL21 (pLGR-vtaA2) (black bars). Concentration of bacteria in the wells at the beginning of showed similar survival rate after incubation with rab- the assay are shown in gray bars. Bars represent the mean ± standard bit serum. A reduction of 5 log was observed for both deviation of the concentration of bacteria from duplicate wells. clones after 30 min of incubation with 50% serum, and Results are representative of 2 independent experiments. no recovery of live bacteria was detected after 1 h of incubation (not shown). Therefore, this suggests that VtaA2 has no effect on evasion of killing by serum fluorescence intensity than the BL21 expressing VtaA2. complement. When FITC-labelled bacteria were used, similar asso- ciation with PAMs was observed in the vtaA2 clone and the control (Figure 3). To confirm these results, the Expression of VtaA2 and VtaA2OL nucleic acid stain SYTO9 was used to label the bacteria. Proteins of approximately the expected sizes of VtaA2 Equivalent mean of fluorescence intensity in the bacte- (120.99 kDa) and VtaA2OL (75.61 kDa) were detected ria was obtained with this dye (BL21 control: 12.9; BL21 in the bacterial pellet of the corresponding clones with vtaA2: 10.63). Again, the percentage of PAMs grown under the conditions used in the functional associated with the fluorescent vtaA2 clone was simi- assays (Figure 5), confirming that the lack of phenotype lar to the percentage of PAMs with the control bacteria found in cell adhesion, invasion or phagocytosis were not due to the lack of protein production. Figure 3 VtaA2 has no role in the bacterial interaction with Figure 4 VtaA2 has no role in the bacterial interaction PAMs; bacteria stained with FITC. E. coli BL21 clones (pLGR-vtaA2; with PAMs; bacteria stained with SYTO9. E. coli BL21 clones dark grey bars) and empty vector pACYC184 (black bars) were stained (pLGR-vtaA2; dark grey bars) and empty vector pACYC184 (black with FITC, incubated for 1 h or 3 h with PAMs and the interaction bars) were stained with SYTO9, incubated for 1 h with PAMs and the with the macrophages was analyzed by flow cytometry. Results are interaction with the macrophages was analyzed by flow cytometry. shown as the percentage of fluorescent macrophages. Macrophages Results are shown as the percentage of fluorescent macrophages. without bacteria were included as negative (light grey bars). Bars Macrophages without bacteria were included as negative (light grey represent the mean ± standard deviation. bars). Bars represent the mean ± standard deviation. Costa‑Hurtado et al. Vet Res (2019) 50:69 Page 7 of 9 are also up-regulated during lung infection [24], would participate in phagocytosis resistance [18]. In addition, the capacity of VtaA2 to bind to ECM molecules suggest that VtaA2 is also important once H. parasuis spreads to internal organs. Therefore, we could hypothesize a dual role of VtaA2 in H. parasuis pathogenesis, in the initial colonization of the respiratory tract and in the posterior invasive infection. Mucin is the major structural component of mucosa secretion that overlays the upper respiratory tract of swine. Mucins are a family of heavily O-glycosylated proteins that are the major organic components of the mucus layer, the protective layer covering the epithelial cells in many human and animal organs [28]. Pathogens that infect mucosal surfaces share two main goals: (1) to overcome the mucus barrier and (2) to interact with the underlying epithelial cells and cause disease [29]. Mucus layers consist mainly of gel-forming mucin-type glycoproteins [28]. Most mucins have a high sialic acid content. Interestingly, systemic isolates of H. parasuis present the sialyltransferase-encoding gene lsgB gene, which is involved in the terminal sialylation of the lipoo- Figure 5 VtaA2 and VtA2OL are produced under in vitro ligosacharide (LOS), with roles in host cell adhesion and conditions. Detection by immunoblot of VtaA2 and its truncated serum resistance [30, 31]. LOS sialylation is a molecu- version VtaA2OL from E. coli BL21 (pLGR-vtaA2) and E. coli BL21 lar mimicry mechanism extensively used by bacteria to (pLGR-vtaA2OL) respectively. Both proteins were extracted from bacterial pellets grown in LB-Cm30 or incubated with CDMEM. evade the host immune system. We could hypothesize Molecular size standards in kDa are shown on the left. Protein that, by binding to mucin through VtaA2, H. parasuis extraction from E. coli BL21 transformed with pACYC184 empty could enhance its interaction to mucin, facilitating LOS plasmid was used as negative control of expression. sialyation by using mucin as source of sialic acid, and pro- moting H. parasuis virulence. In addition, prior infection of piglets by Bordetella bronchiseptica, which induces an Discussion accumulation of mucus [32], results in an increased colo- Bacterial pathogens adhere to epithelial cells or host pro- nization by H. parasuis [33]. VtaA2 could play a key role teins in order to prevent clearance by the host. Adhesion in such scenario by promoting attachment to mucin in is achieved by adhesins, which are bacterial surface struc- the abundant mucus. tures that confer ability to bind to surrounding cells, pro- Inflammatory responses due to multiple pathogens can teins or tissues. Virulent strains of H. parasuis have the disrupt the mucus layer and cause epithelial damage in the ability to overcome the host clearing system, and after airway mucosa, allowing access to intruding bacteria to the an initial colonization of the upper respiratory tract, can underlying epithelium and to ECM proteins, and explain- reach the lower respiratory tract, multiply and spread ing the exacerbation of clinical signs [34, 35]. Virulent systemically [27]. However, how H. parasuis infection strains of H. parasuis can be detected in pig lung, nasal tur- is established at systemic sites is not fully understood binates, trachea including the submucosa [36], indicating yet. Some VtaAs, including VtaA2, are associated with that, as other respiratory pathogens, H. parasuis may have virulent strains of H. parasuis [17]. The involvement of the capacity to bind and degrade ECM proteins in order VtaA2 in H. parasuis virulence was supported by its up- to adhere to and invade host tissues [37]. Collagens are regulation during lung infection [24], but until now no the most abundant proteins in the lung ECM [38]. Colla- specific function was assigned to this protein. Here we gen can be associate with cartilage, including trachea, pre- have shown that VtaA2 is able to promote attachment to sent in the connective tissues or as part of the basal lamina several host molecules in a time-dependent manner. The that anchors epithelial cells to the mesenchyme [39]. The ability of VtaA2 to bind to mucin could be an indication attachment of the host cells to components of the ECM, of the role of this protein in the initial attachment to the such as collagen, is also mediated by fibronectin [40]. Mul- upper respiratory tract. Later, in the lung, VtaA2 could tiple bacterial fibronectin and collagen binding proteins still promote attachment, while VtaA8 and VtaA9, which have been described, including trimeric autotransporters Costa‑Hurtado et al. Vet Res (2019) 50:69 Page 8 of 9 [41–43]. Therefore, expression of the trimeric autotrans- 2. Vahle JL, Haynes JS, Andrews JJ (1995) Experimental reproduction of Haemophilus parasuis infection in swine: clinical, bacteriological, and porter VtaA2 in the bacterial surface could be a key factor morphologic findings. J Vet Diagn Invest 7:476–480 for Glässer’s disease progress by attachment to both colla- 3. Holtkamp D, Rotto H, Garcia R (2007) Economic cost of major health chal‑ gen and fibronectin. lenges in large US swine production systems. Swine News Newslett 30:4 4. Bouchet B, Vanier G, Jacques M, Auger E, Gottschalk M (2009) Studies on Interestingly, VtaA2OL showed a reduced capacity to the interactions of Haemophilus parasuis with porcine epithelial tracheal bind to collagen, suggesting that the loss of these collagen cells: limited role of LOS in apoptosis and pro-inflammatory cytokine domains in VtaA2 are essential for collagen attachment, by release. Microb Pathog 46:108–113 5. Aragon V, Bouchet B, Gottschalk M (2009) Invasion of endothelial cells by collagen-collagen interaction. On the other hand, VtaA2 systemic and nasal strains of Haemophilus parasuis. Vet J 186:264–267 includes other binding domains to host molecules since the 6. Frandoloso R, Martinez-Martinez S, Gutierrez-Martin CB, Rodriguez-Ferri attachment to other proteins, such as mucin or fibronectin, EF (2012) Haemophilus parasuis serovar 5 Nagasaki strain adheres and invades PK-15 cells. Vet Microbiol 154:347–352 is not affected when the collagens domains were removed. 7. Mullins MA, Register KB, Bayles DO, Butler JE (2011) Haemophilus parasuis In fact, other domains with adhesion function previously exhibits IgA protease activity but lacks homologs of the IgA protease described in VtaAs [17] are characteristic of cell surface genes of Haemophilus influenzae. Vet Microbiol 153:407–412 8. Olvera A, Ballester M, Nofrarias M, Sibila M, Aragon V (2009) Differences in adhesion molecules from other pathogenic bacteria [44, phagocytosis susceptibility in Haemophilus parasuis strains. Vet Res 40:24 45]. Further characterization of the VtaA2 structure would 9. Cerdà-Cuéllar M, Aragon V (2008) Serum-resistance in Haemophilus be needed to elucidate other functional domains, confor- parasuis is associated with systemic disease in swine. Vet J 175:384–389 10. Costa-Hurtado M, Olvera A, Martinez-Moliner V, Galofre-Mila N, Martinez mational structures or other factors that can affect the P, Dominguez J, Aragon V (2013) Changes in macrophage phenotype binding affinity for host receptors. after infection of pigs with Haemophilus parasuis strains with different With the dramatic rise of antimicrobial resistances and levels of virulence. Infect Immun 81:2327–2333 11. Nedbalcova K, Satran P, Jaglic Z, Ondriasova R, Kucerova Z (2006) Haemo- the corresponding threat to human and animal health, it is philus parasuis and Glässer’s disease in pigs: a review. Vet Med 51:168–179 crucial to find alternative approaches to block the ability of 12. Bouchet B, Vanier G, Jacques M, Gottschalk M (2008) Interactions of bacteria, such as H. parasuis, to cause disease at the early Haemophilus parasuis and its LOS with porcine brain microvascular endothelial cells. Vet Res 39:42 stages of infection with the design, for instance, of anti- 13. Hua K, Li Y, Zhou H, Hu X, Chen Y, He R, Luo R, Zhou R, Bi D, Jin H (2018) adhesion vaccines. Haemophilus parasuis Infection disrupts adherens junctions and initializes EMT dependent on canonical Wnt/beta-catenin signaling pathway. Front Cell Infect Microbiol 8:324 Abbreviations 14. Vilalta C, Galofre N, Aragon V, Perez de Rozas AM, Fraile L (2012) Effect CDMEM: complete DMEM; Cm: chloramphenicol; DMEM: Dulbecco’s modified of marbofloxacin on Haemophilus parasuis nasal carriage. Vet Microbiol Eagle’s medium; ECM: extracellular matrix; FITC: fluorescein isothiocyanate; 159:123–129 GFP: green fluorescent protein; LB: Luria–Bertani; LOS: lipooligosaccharide; 15. Olvera A, Cerda-Cuellar M, Nofrarias M, Revilla E, Segales J, Aragon V PAMs: porcine alveolar macrophages; PBS: phosphate buffered saline; VtaA: (2007) Dynamics of Haemophilus parasuis genotypes in a farm recovered virulence-associated trimeric autotransporter. from an outbreak of Glasser’s disease. Vet Microbiol 123:230–237 16. Macedo N, Rovira A, Torremorell M (2015) Haemophilus parasuis: infection, Acknowledgements immunity and enrofloxacin. Vet Res 46:128 Nuria Galofré-Milà is acknowledged for her technical work. The authors are 17. Pina S, Olvera A, Barcelo A, Bensaid A (2009) Trimeric autotransporters also grateful to the Centres de Recerca de Catalunya (CERCA) Programme. of Haemophilus parasuis: generation of an extensive passenger domain repertoire specific for pathogenic strains. J Bacteriol 191:576–587 Authors’ contributions 18. Costa-Hurtado M, Ballester M, Galofre-Mila N, Darji A, Aragon V (2012) MCH and LGR carried out the experiments and drafted the manuscript. SLS VtaA8 and VtaA9 from Haemophilus parasuis delay phagocytosis by alveo‑ assisted in flow cytometry experiments. VA conceived the study and its design, lar macrophages. Vet Res 43:57 coordinated it, and drafted the manuscript. All authors read and approved the 19. Olvera A, Pina S, Perez-Simo M, Aragon V, Segales J, Bensaid A (2011) final manuscript. Immunogenicity and protection against Haemophilus parasuis infec‑ tion after vaccination with recombinant virulence associated trimeric Funding autotransporters (VtaA). Vaccine 29:2797–2802 This study was supported by the Ministry of Economy and Competitiveness 20. Henderson IR, Navarro-Garcia F, Desvaux M, Fernandez RC, Ala’Aldeen (MINECO) of the Spanish Government (Grant Number AGL2016-77361-R). D (2004) Type V protein secretion pathway: the autotransporter story. MCH is a recipient of a Juan de la Cierva fellowship, from MINECO (Spain). The Microbiol Mol Biol Rev 68:692–744 funding institution did not participate in the study. 21. Olvera A, Pina S, Perez-Simo M, Oliveira S, Bensaid A (2010) Virulence- associated trimeric autotransporters of Haemophilus parasuis are anti‑ Competing interests genic proteins expressed in vivo. Vet Res 41:26 The authors declare that they have no competing interests. 22. Olvera A, Pina S, Macedo N, Oliveira S, Aragon V, Bensaid A (2012) Identification of potentially virulent strains of Haemophilus parasuis using Received: 14 May 2019 Accepted: 23 August 2019 a multiplex PCR for virulence-associated autotransporters (vtaA). Vet J 191:213–218 23. Galofre-Mila N, Correa-Fiz F, Lacouture S, Gottschalk M, Strutzberg-Minder K, Bensaid A, Pina-Pedrero S, Aragon V (2017) A robust PCR for the dif‑ ferentiation of potential virulent strains of Haemophilus parasuis. BMC Vet References Res 13:124 1. Møller K, Kilian M (1990) V factor-dependent members of the family 24. Bello-Orti B, Howell KJ, Tucker AW, Maskell DJ, Aragon V (2015) Metatran‑ Pasteurellaceae in the porcine upper respiratory tract. J Clin Microbiol scriptomics reveals metabolic adaptation and induction of virulence 28:2711–2716 factors by Haemophilus parasuis during lung infection. Vet Res 46:102 25. Bateman A, Coin L, Durbin R, Finn RD, Hollich V, Griffiths-Jones S, Khanna A, Marshall M, Moxon S, Sonnhammer EL, Studholme DJ, Yeats C, Costa‑Hurtado et al. Vet Res (2019) 50:69 Page 9 of 9 Eddy SR (2004) The Pfam protein families database. Nucleic Acids Res 37. Singh B, Fleury C, Jalalvand F, Riesbeck K (2012) Human pathogens utilize 32:D138–D141 host extracellular matrix proteins laminin and collagen for adhesion and 26. Merritt JH, Kadouri DE, O’Toole GA (2005) Growing and analyzing static invasion of the host. FEMS Microbiol Rev 36:1122–1180 biofilms. Curr Protoc Microbiol. https​://doi.org/10.1002/97804​71729​259. 38. Booth AJ, Hadley R, Cornett AM, Dreffs AA, Matthes SA, Tsui JL, Weiss K, mc01b​01s00​ Horowitz JC, Fiore VF, Barker TH, Moore BB, Martinez FJ, Niklason LE, White 27. Aragon V, Segales J, Oliveira S (2012) Glässer’s disease. In: Zimmerman ES (2012) Acellular normal and fibrotic human lung matrices as a culture JJ, Karriker LA, Ramirez A, Schwartz KJ, Stevenson GW (eds) Diseases of system for in vitro investigation. Am J Respir Crit Care Med 186:866–876 swine, 10th edn. Wiley-Blackwell, Ames, pp 760–769 39. Shoulders MD, Raines RT (2009) Collagen structure and stability. Annu 28. Derrien M, van Passel MW, van de Bovenkamp JH, Schipper RG, de Vos Rev Biochem 78:929–958 WM, Dekker J (2010) Mucin-bacterial interactions in the human oral cav‑ 40. Lodish H, Berk A, Zipursky SL, Matsudaira P, Baltimore D, Darnell J (2000) ity and digestive tract. Gut Microbes 1:254–268 Noncollagen components of the extracellular matrix. In: Freeman WH, ed. 29. Naughton J, Duggan G, Bourke B, Clyne M (2014) Interaction of microbes Molecular cell biology. 4th edn, New York with mucus and mucins: recent developments. Gut Microbes 5:48–52 41. Nomura R, Ogaya Y, Nakano K (2016) Contribution of the Collagen-bind‑ 30. Martinez-Moliner V, Soler-Llorens P, Moleres J, Garmendia J, Aragon V ing proteins of Streptococcus mutans to bacterial colonization of inflamed (2012) Distribution of genes involved in sialic acid utilization in strains of dental pulp. PLoS ONE 11:e0159613 Haemophilus parasuis. Microbiology 158:2117–2124 42. Henderson B, Nair S, Pallas J, Williams MA (2011) Fibronectin: a multid‑ 31. Wang H, Liu L, Cao Q, Mao W, Zhang Y, Qu X, Cai X, Lv Y, Chen H, Xu X, omain host adhesin targeted by bacterial fibronectin-binding proteins. Wang X (2018) Haemophilus parasuis alpha-2,3-sialyltransferase-mediated FEMS Microbiol Rev 35:147–200 lipooligosaccharide sialylation contributes to bacterial pathogenicity. 43. Caswell CC, Oliver-Kozup H, Han R, Lukomska E, Lukomski S (2010) Scl1, Virulence 9:1247–1262 the multifunctional adhesin of group A Streptococcus, selectively binds 32. Dugal F, Belanger M, Jacques M (1992) Enhanced adherence of Pas- cellular fibronectin and laminin, and mediates pathogen internalization teurella multocida to porcine tracheal rings preinfected with Bordetella by human cells. FEMS Microbiol Lett 303:61–68 bronchiseptica. Can J Vet Res 56:260–264 44. Qin W, Wang L, Lei L (2015) New findings on the function and potential 33. Brockmeier SL (2004) Prior infection with Bordetella bronchiseptica applications of the trimeric autotransporter adhesin. Antonie Van Leeu‑ increases nasal colonization by Haemophilus parasuis in swine. Vet Micro‑ wenhoek 108:1–14 biol 99:75–78 45. Linke D, Riess T, Autenrieth IB, Lupas A, Kempf VA (2006) Trimeric 34. Zanin M, Baviskar P, Webster R, Webby R (2016) The Interaction between autotransporter adhesins: variable structure, common function. Trends respiratory pathogens and mucus. Cell Host Microbe 19:159–168 Microbiol 14:264–270 35. Paulsson M, Riesbeck K (2018) How bacteria hack the matrix and dodge the bullets of immunity. Eur Respir Rev 27:180018 36. Bello-Orti B, Costa-Hurtado M, Martinez-Moliner V, Segales J, Aragon V Publisher’s Note (2014) Time course Haemophilus parasuis infection reveals pathological Springer Nature remains neutral with regard to jurisdictional claims in pub‑ differences between virulent and non-virulent strains in the respiratory lished maps and institutional affiliations. tract. Vet Microbiol 170:430–437 Ready to submit your research ? Choose BMC and benefit from: • fast, convenient online submission • thorough peer review by experienced researchers in your field • rapid publication on acceptance • support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations • maximum visibility for your research: over 100M website views per year At BMC, research is always in progress. Learn more biomedcentral.com/submissions