Production of Outer Membrane Vesicles by the Human Pathogen Burkholderia Cenocepacia: Purification and Preliminary Characterization Rui Filipe Brás Teles Martins

Home , Neisseria, Neisseria meningitidis

Production of outer membrane vesicles by the human pathogen Burkholderia cenocepacia: purification and preliminary characterization Rui Filipe Brás Teles Martins

Abstract Cystic Fibrosis (CF) is an autossomal recessive disease resulting from mutations in the cystic fibrosis transmembrane conductance regulator (CFTR) gene, leaving affected patients more susceptible to infection by bacteria. One group of bacteria, the Burkholderia cepacia complex (Bcc), is responsible for dangerous infections in these patients, due to their resistance to multiple antibiotics being therefore necessary to develop alternative treatments. Outer membrane vesicles (OMV) have been used for the development of vaccines against pathogens such as Neisseria meningitidis and Burkholderia pseudomallei, with an OMV vaccine against N. meningitidis already available in the market. This study aimed at the development of a scalable production method of OMVs, in order to allow their use in vaccine development studies. OMVs from B. cenocepacia K56-2 were isolated and purified by size exclusion chromatography (SEC), however the final purity was not high, with OMVs still contaminated with pili and flagella. This prompted the development of an alternative strategy to obtain purified OMVs in which the strain used for OMV production was changed to B. cenocepacia AU1054 non piliated strain. With this strain it was possible to obtain contaminant-free OMVs after the isolation process, which was confirmed by transmission electron microscopy (TEM). OMVs obtained from B. cenocepacia AU1054 were used to perform cytotoxic assays in 16HBE14o- human bronchial epithelial cells and in Galleria melonella. 16HBE14o- cells suffered reduced survival upon exposure to AU1054-derived OMVs, however this result was not reproduced in G. melonella, with all injected caterpillars surviving the assay.

Keywords: Outer membrane vesicles, Cystic fibrosis, purification, Burkholderia cepacia complex, Size exclusion chromatography, cytotoxicity

Introduction leads to the development of problematic Cystic Fibrosis (CF) is an autossomal recessive infections [21]. disease resulting from mutations in the cystic Bcc is a group of 20 closely related species of fibrosis transmembrane conductance regulator bacteria, to which Burkholderia cenocepacia, (CFTR) [53], among other problems CF patients among other species, belongs [15]. suffer from thicker mucus in the lungs [3; 62], Infection with Bcc bacteria usually causes which leaves these patients more susceptible to reduced lung function, and in a subset of patients bacterial infections [62; 66]. Among the bacterial it may cause necrotizing pneumonia and a rapidly infections these patients may suffer it is included fatal septicaemia termed “Cepacia Syndrome” infection by species belonging to the [21]. The severity of Bcc infection is explained by Burkholderia cepacia complex (Bcc), which several characteristics of this group of bacteria, despite the small percentage of infections in CF such as the ability of resisting exposure to a wide patients being caused by these bacteria [36], range of antibiotics [16; 37], ability to form biofilms [12], among others [13; 46; 64], which

1 facilitate the infection with Bcc strains and difficult ranging from 50 to 250nm and produced by eradication of the bacteria. Furthermore, these organisms from the three branches of the tree of species are easily spread by contact between life. They are produced by Gram-negative and infected CF patients [25; 36], saline solutions Gram-positive bacteria [28; 50], Fungi [48], [14], contaminated hospital equipment [32] and archaea [18], and even in parasites [55], in which even from the environment [38]. seems to be an evolutionary conserved Due to the severity of infections caused in CF mechanism [35]. These vesicles have several patients and the poor outcome of infected biological functions which provide an advantage patients, as well as the resistance to multiple to OMV-producing bacteria, they provide antibiotics, it becomes interesting to investigate resistance to antibiotics [42; 52], provide possible alternative treatments and mechanisms protection from environmental stresses [44], to avoid the infection. deliver virulence factors [28; 65], among others There have been several attempts for the [6; 34]. development of a vaccine against species of the OMVs may be isolated from in vitro cultures. Bcc, several strategies have been employed in For the isolation of OMVs the most usual process order to achieve this objective [9; 13], with the is to perform a bacterial culture until late most promising results being achieved with outer logarithmic phase, followed by a series of membrane proteins (OMP) [5; 43]. However centrifugations and filtrations in order to remove strategies employed were not able to achieve the bacteria and concentrate the vesicles [28]. desired efficiency. However in order to obtain purified OMVs it is In the last decade there has been an increasing necessary to apply a purification step, usually interest in the use of outer membrane vesicles OMVs are purified by density gradient (OMV) for the development of vaccines, since centrifugation [28]. In order to evaluate the purity these vesicles exhibit some characteristics which of the OMV preparation the method most make them interesting candidates for vaccine commonly used is Transmission Electron development, such as the presence of multiple Microscopy (TEM), which allows to visualize the antigens, being acellular and stability at 4ºC [19; vesicles and other contaminants, if they are 51]. Studies in vaccine development using OMVs present [28; 41; 44]. have been performed in several bacteria species This study will evaluate the purity of a B. [2; 27; 50], with one OMV-based vaccine against cenocepacia OMV preparation. Followed by N. meningitids already in the market [24] and optimization of OMVs production and purification studies for a vaccine against B. pseudomallei methods. Purified OMVs will be used for well advanced [47; 49]. However there are no cytotoxicity assays, in order to perform a studies regarding development of OMV vaccines preliminary evaluation of OMV cytotoxicity. against Bcc species. OMVs are formed by the blebbing of the outer Materials and methods membrane of bacterial cell wall, they do not result 2.1 Culture conditions and bacterial strains- 10 from cell death or lysis, and are produced during different bacterial strains were used in this work (Table 1), the strains were maintained frozen in glycerol at - all phases of growth [7; 44; 69]. OMVs are 80ºC. When necessary strains were defrosted to Luria- spherical with a bilayered membrane, with sizes Bertani (LB) plates.

OMV production and purification- Vesicles were was reached. Cells were removed by centrifugation purified by a method adapted from Huang et al. [29] (10,000 x g, 10 min). Supernatants were filtered and Allan et al. [1]. Bacteria were grown in 2 L of LB through a 0.45µm cellulose membrane (Milipore) and medium, they were incubated with an initial OD640nm of concentrated using centrifugal filter Amicon® (Merck

0.1 and the culture was stopped when OD640nm Millipore) with 100,000 Da NMWL. reached 1.5. Cells were removed by pelleting (10,000 Proteomic analysis- OMVs were analyzed by 2-DE x g, 10 min.). Supernatants were filtered through a (CBMA- Centre of Molecular Environmental Biology, 0.45µm cellulose membrane (Milipore) and Department of Biology, University of Minho, Braga, concentrated via a 100-kDa tangential filtration Portugal) using the immobilized pH gradient technique concentration unit (GE Healthcare) to approximately 70 (IPG). OMV samples were resuspended in 7 M urea, 2 mL. The OMVS present in the retentate were pelleted M thiourea, 1% (w/v) dodecylmaltoside prior to loading by ultracentrifugation (150,000 x g, 3h), resuspended in the first dimension of IPG. IPG strips used were 24 in 1.5 mL of PBS (pH 7.4). Ressuspended vesicles cm non-linear, with pH range from 3 to 10. Isoelectric were filtered through a 0.2 µm polyethersulfone focusing (IEF) was performed during 3 hours at 7,000 membrane (Whatman, GE Healthcare). V. The completed first-dimensional strip was subjected Size exclusion chromatography (SEC)- Sephacryl to a second dimension using 12% dodecyl sulfate S-1000 Superfine chromatography (GE Healthcare polyacrylamide gel (SDS-PAGE), separation was Life Sciences ) was performed using an Äkta™ Purifier performed at 110V. Gels were stained with silver 10 system, using Tricorn™ 5/100 or Tricorn™ 10/600 nitrate and scanned on a Typhoon Trio laser scanner columns. 1.5mL of OMV solution were loaded using a (GE Healthcare). 1 mL loop with a flow rate of 0.5 mL/min. The flow- Proteins spots were excised manually and identified through and elute were continuously collected in 0.5 by mass spectrometry (MS) (Proteomic and Analytical mL fractions in a FRAC 950 Fraction collector (GE Biochemistry Laboratory, Department of Biology, Healthcare). University of York, York, UK ). Proteins in gel spots Optimization of growth parameters, conditions were digested with trypsin, tryptic peptides analyzed by and strains used- In order to optimize the OMVs matrix-assisted laser desorption ionization (MALDI)- production process the OMVs production was MS and MS/MS using a Bruker ultraflex III (Bruker- evaluated in different Bcc isolates. The selected Daltonic, USA) tandem time of flight (TOF/TOF) mode. isolates were cultured in 150 mL of LB medium. The For identification MS spectra were searched using initial OD640nm was 0.1 and the culture was maintained MASCOT software (Matrix Science) using data from until OD640nm reached 1.5. Analysis of OMV production the NCBI non-redundant protein database. in different points of the growth was performed using Classification into functional categories was performed the same protocol, differing only in the volume used using the Burkholderia genome database (500 mL) and the length of the culture (4h, 10h, 16h (http://www.burkholderia.com), TIGR database and OD640nm 1.5). (http://www.tigr.org/), the Role Category Lists, KEGG OMVs production under different growth conditions database (http://www.genome.jp/kegg/) and NCBI and mediums was performed using a volume of 500 (http://www.ncbi.nlm.nih.gov/). Subcellular prediction mL of LB. Two different mediums were used, regular was performed using the bioinformatic tool PSORTb LB medium and LB medium with 300 mM NaCl. version 2.0.4 [23].

Culture was performed until OD640nm reached 1.5, OMVs quantification (BCA assay)- In this study initial OD640nm was 0.1. Total protein quantification was used as an indicator of Microaerophilic conditions were achieved using an the OMV present in the analyzed sample. Total protein anaerobic box (Mitsubishi Gas Chemical Company) quantification was performed by a bicinchoninic acid and GENbox microaerophilic generator (BioMérieux), (BCA) assay kit (Pierce). Standard solutions of bovine bacteria were incubated over-night, until a OD640nm 1.5 albumin serum (BSA) ranging from 25 to 450 µg/mL

Table 1- Characteristics of the Bcc isolates used in this work.

Strain name Other name Source, location Relevant information Source Burkholderia cenocepacia K56-2 LMG 18863 CF patient, ET-12 lineage BCESM+; J. J. LiPuma Canada cblA+; recA-A J2315 LMG 16656 CF patient, UK ET-12 lineage BCESM+; G. Doring cblA+; recA-A AU1054 LMG 24506 CF patient, USA recA-B PHDC lineage BCCM™/LMG MCO-3 LMG 24308 Maize recA-B BCCM™/LMG rhizosphere, USA HI2424 LMG 24507 Soil, USA recA-B PHDC lineage BCCM™/LMG Burkholderia multivorans ATCC17616 Soil, USA ATCC D2095 CF patient, D. P. Speert Canada HI2229 LMG 17588 CF patient, USA BCESM-; cblA- J. J. LiPuma Burkholderia cepacia PC783 LMG 1222 Onion, USA J. J. LiPuma Burkholderia dolosa AU0654 LMG 18943 CF patient, USA J. J. LiPuma BCESM: Burkholderia cenocepacia epidemic strain marker – cblA: cable pilus gene – recA: recombinase A gene – PHDC: Philadelphia-District of Columbia clone – ET: Electrophoretic type were used to establish a calibration curve. The assays and plate’s absorbance was read at 590nm in a were performed in 96 wells microplates (Orange SpectroStarnano (BMG Labtech) microplate reader. Scientific) with 200 µL of BCA reagent and 10 µL of Galleria melonella killing assays- Galleria sample. Absorbance was read at 562nm in a melonella killing assays were based on the methods Spectrostarnano (BMG Labtech) microplate reader. previously described [45]. A micrometer was adapted MTT assay- 16HBE14o- human bronchial epithelial to control the volume of a mycrosyringe and inject 3.5 cells were maintained in fibronectin/vitrogen coated µL (0.2 µg) or 10 µL (1µg) of purified OMVs in each flasks in Minimum Essential Medium with Earle’s salt caterpillar, via the hindmost left proleg, previously (MEM) (Gibco) supplemented with 10% fetal bovine sanitized with 70% (v/v) ethanol. Following injection serum (FBS) (Lonza), 0.292 g/L L-glutamine (Sigma larvae were placed in petri dishes and incubated in the Aldrich) and Penicilin/Streptomycin 100 U/mL (Gibco), dark, at 37ºC. Survival was followed during a 96 hour in a humified atmosphere at 37ºC and 5% CO2. Cells period, larvae were considered dead when no were seeded in a 96 well microplate (Orange response to touch was displayed. Control larvae were Scientific) at a density of 7.5x103 cells per well. 0.5 µg, injected with PBS (pH 7.4). 1 µg and 2 µg of B. cenocepacia AU1054-derived Transmission Electron Microscopy (TEM)- 10µL OMVs were added to 100 µL of MEM medium and of purified OMV preparation were negatively stained by incubated in the same conditions for 24 hours, after freshly prepared 2% (v/v) uranyl acetate and applied to which 20 µL of 5 mg/mL [3-(4,5 dimethylthiazol-2-yl-2,5 200-mesh carbon-coated cooper grids (Electron tetrazolium bromide)] (MTT) were added in each well Microscopy Sciences), being visualized with a JEOL and incubated for 3 hours in the same conditions. 1200-EX transmission electron microscope. Following the incubation MTT was removed and 150 µL of 40 mM HCl in isopropanol were added and incubated for 15 min., the crystals were resuspended

Results B. cenocepacia K56-2-derived OMVs purity analysis OMVs were isolated from cultures of B. cenocepacia K-56-2 in order to evaluate the purity of the OMVs by TEM. From a 2 L culture of B. cenocepacia, it was possible to obtain, following isolation of OMVs, a yield of 0.1 ±0.03 mg of OMVs/L of initial culture. The OMVs obtained were analyzed by TEM, allowing to confirm the presence of OMVs with diameters ranging from 10 nm to 300 nm, with OMVs with 50 nm being more commonly observed (Fig.1). It Fig. 1- TEM analysis of OMV samples obtained from B. cenocepacia K56-2 in vitro growth is also possible to observe the presence of Optimization of OMVs production contaminants such as flagella and cable pili The process of OMVs production was (Fig.1). optimized, by analyzing OMVs production by Proteomic analysis of B. cenocepacia K56-2- different strains of the Bcc, under different derived OMVs incubation periods, and under different growth Three different samples of purified OMVs were conditions. The strains analyzed were chosen subjected to a proteomic analysis. 33 different from the four clinically most relevant Bcc species proteins were identified from 73 spots (Fig.2). [8; 16]. From the strains tested B. cenocepacia The proteins identified were divided into 8 K56-2 achieved the best yields (Fig.3). different functional groups with the most abundant groups being cell envelope biogenesis and outer membrane, energy conversion and amino acid transport and metabolism. It was possible to observe the presence of a large number of cytoplasmic proteins which is expected since OMVs incorporate proteins from multiple bacterial compartments. It is also possible to observe the presence of cable pilus and flagella, which corroborates TEM results. A set of proteins identified in the proteomic analysis is of particular interest. This group is composed by proteins which are immunogenic. In Fig. 2- Proteome reference 2-DE gel of B cenocepacia K56-2- this study 6 of 15 immunogenic proteins were derived OMVs. On the left is represented the scale of the molecular weight (higher on top to lower on bottom), and on identified [54]. the top is represented the pH scale (from 3 to 10). The circles and numbers indicate excised protein spots which were analyzed by MALDI-TOF

For the optimization of culture time both B.

0,5 cenocepacia K56-2 and B. multivorans HI2229 0,4 were tested. The OMVs produced in cultures 0,3 stopped at different points suggested that longer 0,2 incubation times lead to increased OMVs 0,1 production, however the increase in longer yield OMV(mg/L)

…

incubation times may be due to cell lysis, which 3

2 2

- - was not analyzed in this study. The final optimization step was the growth

conditions. Two stresses, which are present in B. cepacia B. PC783

CF lung, were tested, namely osmotic stress [33] dolosa AU0654 B.

B. multivorans multivorans B. ATCC

B. cenocepacia B. J2315

B. multivorans multivorans B. D2095

B. cenocepacia B. K56

B. multivorans multivorans B. HI2229 B. cenocepacia B. MCO represented by LB medium with 300 mM NaCl, cenocepacia B. HI2424 B. cenocepacia B. AU1054 Strains tested and reduced oxygen concentration [68] represented by microaerophilic conditions. Fig. 3- Graphic representation of the OMVs yield (mg of OMVs/L of initial culture) produced by the different strains The growth conditions tested yielded different used in the strain screening results, while the use of LB with 300 mM of NaCl To further purify the OMVs a SEC step was resulted in reduced OMVs production compared applied to the samples, to increase purity of the with LB medium, the use of microaerophilic preparation. conditions resulted in an increase of OMV The OMVs were loaded into a Tricorn™ 5/100 production. with 30 cm of length. The use of this column did Despite the results obtained in the optimization not allow complete separation of the OMVs and of OMVs production process, the conditions were the contaminants present in the sample, with only maintained. The strains being used was the best a partial separation being achieved. producer. The increase in the total protein In order to increase the separation between the produced in longer incubation periods may be OMVs and the contaminants in the sample, a caused by cell lysis. longer column was used. The reasoning behind Table 2 OMV yields (mg/L of initial culture) obtained following the employment of a longer column was that the culture of B. cenocepacia K56-2, under different growth conditions. increased retention time in the column would Growth conditions OMV yield allow for a more complete separation. (mg/L) Broth Stiring Oxygen The longer column used was a Tricorn™ LB 250 aerobic 0.9 10/600 with 60 cm of length. OMVs were also 0.3M NaCl 250 aerobic 0.7 loaded into the column. The separation of the LB LB 60 microaerophilic 1.5 samples was greater using the longer column, Size exclusion chromatography of B. however complete separation was not achieved, cenocepacia K56-2-derived OMVs with an small overlap between the OMVs The results obtained in TEM and proteomic fractions and contaminants fraction, thus not analysis of B. cenocepacia K56-2-derived OMVs achieving complete separation between the suggested that further purification of the samples samples and contaminants. was required.

Production of B. cenocepacia AU 1054- It is also important to notice that cable pilus or derived OMVs flagella were not observed in these preparations Since the use of SEC for the purification of B. (Fig.4). cenocepacia K56-2-derived OMVs did not achieved the desired results, a different strategy was employed for the achievement of purified OMVs. Taking into consideration the observations from the TEM analysis of B. cenocepacia K56-2- derived OMVs, in which Cable pili and flagella were the main contaminants, we reasoned that the use of a strain not capable of producing cable pili would reduce the contamination of the OMV preparation. Fig. 4- Transmission electron microscopy image from a To achieve this objective the strain used for sample of B. cenocepacia AU1054-derived OMVs. OMVs production was switched to B. In vitro cytotoxic effects of B. cenocepacia cenocepacia AU 1054, which is CblA-, thus not AU1054-derived OMVs being able to produce cable pili. This strain was Following the isolation of purified OMVs from used for production of OMVs, the protocol used B. cenocepacia AU1054 it became possible to for production and isolation of B. cenocepacia perform a preliminary evaluation of cytotoxic AU1054-derived OMVs was the same as the one potential of OMVs. Initial cytotoxic assays were used for B. cenocepacia K56-2. From the 2 L of performed in vitro by performing MTT assays in initial culture and following the isolation of OMVs 16HBE14o- human bronchial epithelial cells. from B. cenocepacia AU 1054-derived OMVs, it Exposure to OMVs revealed a reduction in cell was possible to obtain 0.04 ±0.003 mg of OMV/L survival, which increased with an increase in the of initial culture. amount of OMVs to which cells were exposed, Transmission Electron Microscopy of B. reaching almost 50% in the highest quantity of cenocepacia AU1054-derived OMVs OMVs (Fig. 5). B. cenocepacia AU1054-derived OMVs were 100 analyzed by TEM to confirm the presence of 80 vesicles and evaluate the purity of the OMVs. TEM analysis of OMVs isolated from B. 60

cenocepacia AU1054 revealed the presence of 40 Survival (%) Survival OMVs with diameters between 100 and 150 nm. 20 It was also possible to observe vesicles with 0 smaller diameters (75 to 100nm) and large 0,5 1 2 diameters (175 to 225 nm) were less commonly OMVs (ug) observed (Fig. 3). Fig. 5- Cell survival after exposure to purified B. cenocepacia AU1054-derived OMVs during a 24 hour period.

In vivo cytotoxic effects of B. cenocepacia step for the purification of isolated OMVs based AU1054-derived OMVs on a previously described protocol for OMVs Following the in vitro cytotoxic assays purification in N. meningitidis [59]. Testing of SEC performed with OMVs, the cytotoxic effects of efficiency in the purification of OMVs revealed OMVs were also tested in Galleria melonella. The that the process was not capable of achieving the results obtained from the in vivo testing of desired results, therefore a different strategy was cytotoxic effects were different from the MTT employed. assays. Exposure to purified OMVs in G. As said before the proteomic analysis of melonella did not result in reduced survival, in B.cenocepacia K56-2-derived OMVs revealed the fact all larvae injected with OMVs survived the presence of 6 of 15 immunoproteins across B. testing period (Table 3). cenocepacia and B. multivorans species, which Table 3- Number of Galleria mellonella caterpillars alive after are good candidates for vaccine development, a 96 hours period after being injected with a OMV sample isolated from the strain B. cenocepacia AU1054 namely chaperonin GroEL, elongation factor Tu, Sample Caterpillars Caterpillars OMPA/MotB domain-containing protein, porin, inoculated alive after 96 alkyl hidroperoxide reductase/Thiol specific hours antioxidant and phospopyruvate hydratase [54]. Control 10 10 Besides, the 33 proteins identified in the proteomic analysis may also be encountered in AU1054 10 10 OMVs derived from other organisms, such as P. (0.2µg) aeruginosa [11], N. lactamica [60] and N. meningitidis [67]. AU1054 10 10 In the optimization of the OMVs production (1µg) process, when optimizing culture length a B. Discussion multivorans strain was used, this strain is a The development of a vaccine against Bcc clinical isolate, and was used in order to have a species has been subject to various studies, representative of the two clinically relevant however they were not able to achieve enough species. The results suggested that using longer protection [13]. incubation periods lead to increased OMVs Recently OMVs have been employed in the yields, however the increase in OMVs production development of vaccines due to their may be caused by cell lysis. Since the cell lysis characteristics [51]. was not evaluated, the incubation length was Using B. cenocepacia K56-2 we were able to maintained, this incubation period is similar to the produce and isolate OMVs. Assessment of purity period used by other authors, which stop the of isolated OMVs by TEM and proteomic analysis cultures at lat log phase/early stationary [4; 28]. revealed the presence of contaminants, requiring Growth conditions have an important effect in further purification. Several authors use density vesicle production [40; 41]. Taking that gradients for OMV purification [4; 63], however observation into consideration it was necessary this purification process is not easily scalable. to evaluate the effect of growth conditions in Taking those observations into consideration OMV production. Two different conditions were we employed a size exclusion chromatography tested, osmotic stress and microaerophilic

8 conditions, which, as previously stated, represent without the presence of contaminants, thus the conditions encountered by bacteria in host’s achieving the objective of producing purified lungs. Literature reports regarding OMV OMV preparations using an alternative strategy. production under osmotic stress and The evaluation of cytotoxic effects both in vitro microaerophilic conditions are scarce, however and in vivo revealed some differences, while the some reports may be found. Regarding OMV in vitro model exhibited reduced cell survival, the production under osmotic stress Fulsundar et al. in vivo model did not exhibited reduced survival. have studied the effect of different stresses in This discrepancy between the models used may vesicle production in Acineobacter bayly, be explained by the lipopolyssacharide (LPS) concluding that LB medium supplemented with toxicity, which may cause some toxicity [51]. 500 mM of NaCl for the culture of A. bayly, Work performed by Elmi et al., revealed that resulted in a small increase in vesiculation [22]. Campylobacter jejuni 11168H-derived OMVs The results obtained in the present study are not revealed dose-dependent cytotoxic effects in in agreement with the results obtained by Caco-2 intestinal epithelial cells and in G. Fulsundar et al. In respect to reduced oxygen mellonella [20], while in this work only in availability, Tran et al., have evaluated the 16HBE14o- human bronchial epithelial cells was release of Shiga toxin by a strain of E. coli. They possible to observe cytotoxic effects derived from evaluated the release of Shiga toxin in OMVs exposure, which may be explained by the microaerobic conditions, and concluded that the higher OMVs dose used in the work performed release of Shiga toxin was reduced under these by Elmi et al [20]. conditions [56]. Since Shiga toxin is released by Conclusions and future work OMVs [17], it may be possible that in E. coli Some aspects of this work should be further strain used also releases Shiga toxin via OMVs. investigated. The effect of osmotic stress and The results obtained in the present study, once microaerophilic conditions should be further again, are not in agreement with the literature, studied, repeating the assays performed, in order having achieved higher OMV yields in to achieve a more solid conclusion, since microaerophilic conditions. literature reports are not in agreement with the For the development of an alternative results obtained. purification strategy, and taking into account the The cytotoxic effects of OMVs should also be results obtained in TEM and proteomic analysis subjected to a more in-depth analysis, evaluating of B. cenocepacia K56-2, the strain used for the effect of exposure to higher amounts of OMVs production was replaced with B. OMVs. cenocepacia AU 1054, which does not produce One interesting strategy which could be used in cable pilus or flagella, facilitating the purification the future to overcome the low OMVs yields of OMVs. obtained is the use of genetic engineering The yields obtained by B. cenocepacia AU techniques to cause a mutation in the rmpM 1054 are lower than the yields obtained by B. gene, which would lead to a mutant with cenocepacia K56-2, however TEM analysis increased vesiculation [57], thus increasing the revealed the presence of OMVs in B. OMVs yields. cenocepacia AU 1054 OMVs preparations,

This work lead to the development of a strategy Burkholderia dolosa among persons with cystic fibrosis. FEMS Microbiol. Lett. 228, 57–62 (2003). to obtain purified OMVs from B. cenocepacia. 9. Choh, L.-C. et al. Burkholderia vaccines: are we moving forward? Front. Cell. Infect. Microbiol. 3, 5 This strategy allows the production of OMVs (2013). 10. Choi, D.-S. et al. Proteomic analysis of outer suitable for the development of future studies in membrane vesicles derived from Pseudomonas aeruginosa. Proteomics 11, 3424–3429 (2011). the development of OMVs-based vaccines. 11. Conway, B. D., Chu, K. K., Bylund, J., Altman, E. & Speert, D. P. Production of Exopolysaccharide Acknowledgements by Burkholderia cenocepacia Results in Altered Cell-Surface Interactions and Altered Bacterial We would like to acknowledge Dr. Pedro M. Clearance in Mice. 4, 957–966 (2004). 12. Conway, B. D., Venu, V. & Speert, D. P. Biofilm Santos at CBMA- Centre of Molecular and Formation and Acyl Homoserine Lactone Production in the Burkholderia cepacia Complex Environmental Biology, Department of Biology, Biofilm Formation and Acyl Homoserine Lactone Production in the Burkholderia cepacia Complex. University of Minho, Braga, Portugal, for J. Bacteriol. 184, 5678–5685 (2002). performing the proteomic analysis. 13. Corbett, C. R., Burtnick, M. N., Kooi, C., Woods, D. E. & Sokol, P. A. An extracellular zinc We would also like to acknowledge Dr. António metalloprotease gene of Burkholderia cepacia. Microbiology 149, 2263–2271 (2003). Pedro Alves de Matos, at Centro de Investigação 14. Cunha, V. et al. Exceptionally High Representation of Burkholderia cepacia among B Interdisciplinar Egas Moniz (CiiEM), Egas Moniz– . cepacia Complex Isolates Recovered from the Major Portuguese Cystic Fibrosis Center ᰔ. 45, Cooperativa de Ensino Superior, Caparica, 1628–1633 (2007). 15. Depoorter, E. et al. Burkholderia: an update on Portugal, for performing the Transmission taxonomy and biotechnological potential as antibiotic producers. Appl. Microbiol. Biotechnol. Electron Microscopy. 100, 5215–5229 (2016). 16. Drevinek, P. & Mahenthiralingam, E. Burkholderia cenocepacia in cystic fibrosis: epidemiology and References molecular mechanisms of virulence. Clin. Microbiol. Infect. 16, 821–830 (2010). 1. Allan, N. D., Kooi, C., Sokol, P. a & Beveridge, T. 17. Dutta, S. et al. Release of Shiga toxin by J. Putative virulence factors are released in membrane vesicles in Shigella dysenteriae association with membrane vesicles from serotype 1 strains and in vitro effects of Burkholderia cepacia. Can. J. Microbiol. 49, 613– antimicrobials on toxin production and release. 624 (2003). Microbiol. Immunol. 48, 965–9 (2004). 2. Alaniz, R. C., Deatherage, B. L., Lara, J. C. & 18. Ellen, A. F. et al. Proteomic analysis of secreted Cookson, B. T. Membrane vesicles are membrane vesicles of archaeal Sulfolobus immunogenic facsimiles of Salmonella species reveals the presence of endosome typhimurium that potently activate dendritic cells, sorting complex components. Extremophiles 13, prime B and T cell responses, and stimulate 67–79 (2009). protective immunity in vivo. J. Immunol. 179, 19. Ellis, T. N. & Kuehn, M. J. Virulence and 7692–7701 (2007). Immunomodulatory Roles of Bacterial Outer 3. Bashkar, K. R. et al. Profound increase in Membrane Vesicles. Microbiol. Mol. Biol. Rev. 74, viscosity and aggregation of pig gastric mucin at 81–94 (2010). low pH. Am. J. Physiol. 261, 827–832 (1991). 20. Elmi, A. et al. Campylobacter jejuni outer 4. Bauman, S. J. & Kuehn, M. J. Purification of outer membrane vesicles play an important role in membrane vesicles from Pseudomonas bacterial interactions with human intestinal aeruginosa and their activation of an IL-8 epithelial cells. Infect. Immun. 80, 4089–98 response. Microbes Infect. 8, 2400–2408 (2006). (2012). 5. Bertot, G. M. et al. Nasal immunization with 21. Folkesson, A. et al. Adaptation of Pseudomonas Burkholderia multivorans outer membrane aeruginosa to the cystic fibrosis airway: an proteins and the mucosal adjuvant evolutionary perspective. Nat. Rev Microbiol 10, adamantylamide dipeptide confers efficient 841–51 (2012). protection against experimental lung infections 22. Fulsundar, S. et al. Gene transfer potential of with B. multivorans and B. cenocepacia. Infect. outer membrane vesicles of Acinetobacter baylyi Immun. 75, 2740–2752 (2007). and effects of stress on vesiculation. Appl. 6. Beveridge, T. J. Bacteriolytic effect of membrane Environ. Microbiol. 80, 3469–3483 (2014). vesicles from Pseudomonas aeruginosa on other 23. Gardy, J. L. et al. PSORTb v.2.0: Expanded bacteria including pathogens : conceptually new prediction of bacterial protein subcellular Bacteriolytic Effect of Membrane Vesicles from localization and insights gained from comparative Pseudomonas aeruginosa on Other Bacteria proteome analysis. Bioinformatics 21, 617–623 Including Pathogens : Conceptually New A. 178, (2005). 2767–2774 (1996). 24. Gorringe, A. R. & Pajon, R. Bexsero: A 7. Beveridge, T. J. Sructure of fram-negative cell multicomponent vaccine for prevention of walls and their derived mebrane vesicles. J. meningococcal disease. Hum. Vaccines Bacteriol. 181, 4725–4733 (1999). Immunother. 8, 174–183 (2012). 8. Biddick, R., Spilker, T., Martin, A. & LiPuma, J. J. 25. Govan, J. R. W. et al. Evidence for transmission Evidence of transmission of Burkholderia of Pseudomonas cepacia by social contact in cepacia, Burkholderia multivorans and cystic fibrosis . The Lancet. 342, 15–19 (1993).

26. Govan, J. R. W., Hughes, J. E. & Vandamme, P. 44. McBroom, A. J. & Kuehn, M. J. Release of outer Burkholderia cepacia: Medical, taxonomic and membrane vesicles by Gram-negative bacteria is ecological issues. J. Med. Microbiol. 45, 395–407 a novel envelope stress response. Mol. Microbiol. (1996). 63, 545–558 (2007). 27. Holst, J. et al. Properties and clinical performance 45. Mil-Homens, D., Bernardes, N. & Fialho, A. M. of vaccines containing outer membrane vesicles The antibacterial properties of docosahexaenoic from Neisseria meningitidis. Vaccine 27, 3–12 omega-3 fatty acid against the cystic fibrosis (2009). multiresistant pathogen Burkholderia 28. Horstman, A. L. & Kuehn, M. J. Enterotoxigenic cenocepacia. FEMS Microbiol. Lett. 328, 61–69 Escherichia coli Secretes Active Heat-labile (2012). Enterotoxin via Outer Membrane Vesicles *. 275, 46. Mil-Homens, D. & Fialho, A. M. A BCAM0223 12489–12496 (2000). mutant of Burkholderia cenocepacia is deficient in 29. Huang, W. et al. Immunization against multidrug- hemagglutination, serum resistance, adhesion to resistant Acinetobacter baumannii effectively epithelial cells and virulence. PLoS One 7, protects mice in both pneumonia and sepsis (2012b). models. PLoS One 9, 16–19 (2014). 47. Nieves, W. et al. A naturally derived outer- 30. Isles, A. et al. Pseudomonas cepacia infection in membrane vesicle vaccine protects against lethal cystic fibrosis: an emerging problem. J Pediatr pulmonary Burkholderia pseudomallei infection. 104, 206–210 (1984). Vaccine 29, 8381–8389 (2011). 31. Kadurugamuwa, J. L. & Beveridge, T. J. 48. Oliveira, D. L. et al. Characterization of yeast Virulence Factors Are Released From extracellular vesicles: Evidence for the Pseudomonas-Aeruginosa in Association With participation of different pathways of cellular Membrane-Vesicles During Normal Growth and traffic in vesicle biogenesis. PLoS One 5, (2010). Exposure To Gentamicin - a Novel Mechanism of 49. Petersen, H., Nieves, W., Russell-Lodrigue, K., Enzyme-Secretion. J. Bacteriol. 177, 3998–4008 Roy, C. J. & Morici, L. a. Evaluation of a (1995). Burkholderia pseudomallei Outer Membrane 32. Kamiya, A. BRIEF Microbial contamination and Vesicle Vaccine in Nonhuman Primates. Procedia disinfectants. Am. J. Infect. Control 24, 389–395 Vaccinol. 8, 38–42 (2014). (1996). 50. Rivera, J. et al. Bacillus anthracis produces 33. Kilbourn, J. P. Bacterial content and ionic membrane-derived vesicles containing composition of sputum in cystic fibrosis. The biologically active toxins. Proc. Natl. Acad. Sci. U. Lancet. 311, 334 (1978). S. A. 107, 19002–7 (2010). 34. Lappann, M., Otto, A., Becher, D. & Vogel, U. 51. Rosenthal, J. A., Chen, L., Baker, J. L., Putnam, Comparative proteome analysis of spontaneous D. & DeLisa, M. P. Pathogen-like particles: outer membrane vesicles and purified outer Biomimetic vaccine carriers engineered at the membranes of neisseria meningitidis. J. Bacteriol. nanoscale. Curr. Opin. Biotechnol. 28, 51–58 195, 4425–4435 (2013). (2014). 35. Lee, E. Y. et al. Gram-positive bacteria produce 52. Schaar, V., Nordström, T., Mörgelin, M. & membrane vesicles: Proteomics-based Riesbeck, K. Moraxella catarrhalis outer characterization of Staphylococcus aureus- membrane vesicles carry β-lactamase and derived membrane vesicles. Proteomics 9, 5425– promote survival of Streptococcus pneumoniae 5436 (2009). and Haemophilus influenzae by inactivating 36. LiPuma, J. J. et al. Person-to-person transmision amoxicillin. Antimicrob. Agents Chemother. 55, of Pseudomonas cepacia between patients with 3845–3853 (2011). cystic fibrosis . The Lancet.336, 1094–1096 53. Sheppard, D. N. & Welsh, M. J. Structure and (1990). function of the CFTR chloride channel. Physiol. 37. LiPuma, J. J. Burkholderia cepacia: Management Rev. 79, S23–S45 (1999). issues and new insights. Clin. Chest Med. 19, 54. Shinoy, M. et al. Immunoproteomic analysis of 473–486 (1998). proteins expressed by two related pathogens, 38. LiPuma, J. J., Spilker, T., Coenye, T. & Gonzalez, Burkholderia multivorans and burkholderia C. F. An epidemic Burkholderia cepacia complex cenocepacia, during human infection. PLoS One strain identified in soil. Lancet 359, 2002–2003 8, (2013). (2002). 55. Silverman, J. M. et al. Proteomic analysis of the 39. LiPuma, J. J. The changing microbial secretome of Leishmania donovani. Genome epidemiology in cystic fibrosis. Clin. Microbiol. Biol. 9, R35 (2008). Rev. 23, 299–323 (2010). 56. Tran, S. L., Billoud, L., Lewis, S. B., Phillips, A. D. 40. MacDonald, I. A. & Kuehna, M. J. Stress-induced & Schüller, S. Shiga toxin production and outer membrane vesicle production by translocation during microaerobic human colonic Pseudomonas aeruginosa. J. Bacteriol. 195, infection with Shiga toxin-producing E.coliO157: 2971–2981 (2013). H7 and O104: H4. Cell. Microbiol. 16, 1255–1266 41. McMahon, K. J., Castelli, M. E., Vescovi, E. G. & (2014). Feldman, M. F. Biogenesis of outer membrane 57. Van Der Ley, P. et al. Modification of lipid a vesicles in Serratia marcescens is biosynthesis in Neisseria meningitidis lpxL thermoregulated and can be induced by mutants: Influence on lipopolysaccharide activation of the Rcs phosphorelay system. J. structure, toxicity, and adjuvant activity. Infect. Bacteriol. 194, 3241–3249 (2012). Immun. 69, 5981–5990 (2001). 42. Manning, A. J. & Kuehn, M. J. Contribution of 58. van de Waterbeemd, B. et al. Improved OMV bacterial outer membrane vesicles to innate vaccine against Neisseria meningitidis using bacterial defense. BMC Microbiol 11, 258 (2011). genetically engineered strains and a detergent- 43. Makidon, P. E. et al. Induction of immune free purification process. Vaccine 28, 4810–4816 response to the 17 kDa OMPA Burkholderia (2010). cenocepacia polypeptide and protection against 59. van de Waterbeemd, B. et al. Improved pulmonary infection in mice after nasal Production Process for Native Outer Membrane vaccination with an OMP nanoemulsion-based Vesicle Vaccine against Neisseria meningitidis. vaccine. Med. Microbiol. Immunol. 199, 81–92 PLoS One 8, 1–11 (2013). (2010).

60. Vaughan, T. E. et al. Proteomic analysis of Neisseria lactamica and N eisseria meningitidis outer membrane vesicle vaccine antigens. Vaccine 24, 5277–5293 (2006). 61. Veerman, E. C. et al. Viscosity of salivary mucins: effect of pH and ionic strength and role of sialic acid. J. Biol. Bucc.17, 297–306 (1989) 62. Veerman, E. C. I. et al. Sulfated glycans on oral mucin as receptors for Helicobacter pylori. Glycobiology 7, 737–743 (1997). 63. Vida, A. et al. Neutralization of Neisseria meningitidis outer membrane vesicles. Inflamm. Res. 60, 801–5 (2011). 64. Visser, M. B., Majumdar, S., Hani, E. & Sokol, P. A. Importance of the Ornibactin and Pyochelin Siderophore Transport Systems in Burkholderia cenocepacia Lung Infections Importance of the Ornibactin and Pyochelin Siderophore Transport Systems in Burkholderia cenocepacia Lung Infections. Infect. Immun. 72, 2850–2857 (2004). 65. Wai, S. N. et al. Vesicle-mediated export and assembly of pore-forming oligomers of the enterobacterial ClyA cytotoxin. Cell 115, 25–35 (2003). 66. Wanke, C. A., Cronan, S., Goss, C., Chadee, K. & Guerrant, R. L. Characterization of binding of Escherichia coli strains which are enteropathogens to small-bowel mucin. Infect. Immun. 58, 794–800 (1990). 67. Williams, J. N. et al. Proteomic analysis of outer membranes and vesicles from wild-type serogroup B Neisseria meningitidis and a lipopolysaccharide-deficient mutant. Infect. Immun. 75, 1364–1372 (2007). 68. Worlitzsch, D. et al. Effects of reduced mucus oxygen concentration in airway Pseudomonas infections of cystic fibrosis patients. J. Clin. Invest. 109, 317–325 (2002). 69. Zhou, L., Srisatjaluk, R., Justus, D. E. & Doyle, R. J. On the origin of membrane vesicles in Gram- negative bacteria. FEMS Microbiol. Lett. 163, 223–228 (1998).