Incorporation of Actinobacillus pleuropneumoniae in Preformed Biofilms by Escherichia coli Isolated From Drinking Water of Swine Farms

Actinobacillus pleuropneumoniae, the etiological agent of porcine pleuropneumonia, represents one of the most important health problems in the swine industry worldwide and it is included in the porcine respiratory disease complex. One of the bacterial survival strategies is biofilm formation, which are bacterial communities embedded in an extracellular matrix that could be attached to a living or an inert surface. Until recently, A. pleuropneumoniae was considered to be an obligate pathogen. However, recent studies have shown that A. pleuropneumoniae is present in farm drinking water. In this study, the drinking water microbial communities of Aguascalientes (Mexico) swine farms were analyzed, where the most frequent isolated bacterium was Escherichia coli. Biofilm formation was tested in vitro; producing E. coli biofilms under optimal growth conditions; subsequently, A. pleuropneumoniae serotype 1 (strains 4074 and 719) was incorporated to these biofilms. Interaction between both bacteria was evidenced, producing an increase in biofilm formation. Extracellular matrix composition of two-species biofilms was also characterized using fluorescent markers and enzyme treatments. In conclusion, results confirm that A. pleuropneumoniae is capable of integrates into biofilms formed by environmental bacteria, indicative of a possible survival strategy in the environment and a mechanism for disease dispersion.


Incorporation of Actinobacillus pleuropneumoniae in Preformed Biofilms by Escherichia coli Isolated From Drinking Water of Swine Farms INTRODUCTION
Actinobacillus pleuropneumoniae is a Gram-negative coccobacillus, pleomorphic, facultative anaerobe, non-spore forming, encapsulated (1) and a member of the Pasteurellaceae family (2)(3)(4). A. pleuropneumoniae is the etiological agent of porcine pleuropneumonia; one of the most important health problems in the swine industry worldwide, and along with other porcine respiratory pathogens, this pathogen is also included in the Porcine Respiratory Disease Complex (5)(6)(7). Isolates can be classified into two biotypes depending on their requirement for nicotinamide adenine dinucleotide (NAD-dependant and NAD-independant). There are 16 recognized serovars (8). In Mexico, swine pleuropneumonia is widespread (6,7,(9)(10)(11). Infection usually occurs through air or by direct contact. The microorganism is able to colonize the tonsils and to adhere to the alveolar epithelium. In general, the initial step is the bacterial colonization and adhesion to host cells (12).
Biofilms are microorganisms three-dimensional complex communities embedded in an extracellular matrix, where displayed characteristic phenotypes are similar to the free-living organisms, also known as planktonic (13)(14)(15)(16)(17)(18). Biofilms have a dynamic structure in which a multitude of metabolic interactions between neighboring cells are developed (19). Naturally, the dominant growth of microorganisms is through multi-species consortia, regulated by a variety of important intra-and interspecific interactions in development, composition, structure, and function (20)(21)(22). These bacterial microbial communities constitute a multi-species society, with its own "rules and patterns of behavior" (23).
Recently, (24) described the involvement of biofilm formation during the infection process of A. pleuropneumoniae. However, few studies have been done on its ability to survive outside of the pig so as to be considered an obligate pathogen. Assavacheep and Rycroft (25) investigated the survival of A. pleuropneumoniae under controlled laboratory conditions. In aqueous suspension, survival was improved by the presence of NaCl and mucin; as well as lowered temperature. Our group has detected the presence of A. pleuropneumoniae in drinking water from pig farms in Mexico using antibodies and a specific PCR for the gene of the ApxIV toxin (6,26). Subsequently, we evaluated the ability of A. pleuropneumoniae to form multispecies biofilms with other swine bacterial pathogens in the absence of pyridine compounds (nicotinamide mononucleotide [NMN], riboside nicotinamide [NR], or nicotinamide adenine dinucleotide [NAD]) that are essential for growth of A. pleuropneumoniae (27). A. pleuropneumoniae was able to grow with all species tested in the absence of pyridine compounds. Furthermore, A. pleuropneumoniae was able to form strong biofilms when mixed with Streptococcus suis, Bordetella bronchiseptica, or Staphylococcus aureus. Notably, in the presence of Pasteurella multocida, and Escherichia coli, A. pleuropneumoniae was able to form a two-species biofilm, although this was weaker than the biofilms formed with other bacteria (27).
In this study, Escherichia coli strains isolated from the microbial community of drinking water of swine farms of the State of Aguascalientes were characterized and evaluated to explore their possible interaction with A. pleuropneumoniae to form two-species biofilms, suggesting a possible mechanism used by A. pleuropneumoniae to survive in the drinking water in pig farms, and in the environment. Furthermore, changes in the composition of the extracellular matrix during the formation of these two-species biofilms were also characterized.

Sampling of Drinking Water in Swine Farms
The study was performed in the swine farm of the Universidad Autonoma de Aguascalientes, Aguascalientes, Mexico. The farm is used for breeding and fattening pigs for teaching-learning purposes. Random water samples were aseptically obtained from drinkers located on the floor of barnyard from select areas on the farm (26). Water samples were taken directly from the drinkers at its deepest zone with 50 ml sterile Corning tubes. Samples were stored at room temperature until used for bacteria isolation.

Isolation of Bacteria From Drinking Water Samples
Samples were centrifuged at 10,000 × g for 10 min to recover the bacteria and the supernatant was discarded. The obtained pellets were re-suspended in the remaining volume. Dilutions were made in distilled water in the order 10 3 , 10 4 , 10 5 , 10 6 , plated on BHI (agar brain-heart infusion, Bioxon, Mexico) and incubated at 37 • C for 24 h. Colonies of each bacterium were plated alone on BHI agar and incubated at 37 • C for 24 h. All isolated bacteria were stored in glycerol 30% and stored at −80 • C.

Characterization of Isolates
Once isolation of bacteria from drinking water was made, morphological characterization of the colonies and the biochemical tests such as Gram stain, catalase and oxidase were performed. All the strains were evaluated by Api NE biochemistry test (BioMérieux, France), according to the manufacturer's instructions.

Confirmation of E. coli Isolation
Escherichia coli isolation was confirmed by PCR as previously reported (28) by de presence of uidA gene, which encodes the beta-glucuronidase enzyme. Phylogenetic group of each strain was identified (29). Escherichia coli isolated from drinking water were screened for the presence of selected virulence genes usually associated with the E. coli strains responsible for extra-intestinal infections, including: fyuA (yersiniabactin receptor), kpsMTII (capsular polysaccharide genes), and papC [P fimbriae, (30)]. In order to detect the genes agn43 [antigen 43, (31)], fimH [minor component of type 1 fimbriae, (32)], hlyA [haemolysin, (33)], and afa [afimbrial adhesins, (30)] a multiplex PCR was designed with the following conditions: 94 • C for 5 min followed by 40 cycles of 30 sec at 94 • C, 1 min at 60 • C and 1 min at 68 • C with a final elongation step at 72 • C for 10 min. For the sequences of the primers see Table 1. The amplification products were observed by electrophoresis in 1.5% agarose gel stained with 1 µg ethidium bromide ml −1 .

Integration of Actinobacillus pleuropneumoniae in Biofilms Formed by Escherichia coli
To analyze the incorporation of A. pleuropneumoniae in preformed E. coli biofilms, a previously described methodology with several modifications was used (27). For this test, two strains of A. pleuropneumoniae (reference strain 4074 and swine isolated strain 719), both belonging to serotype 1 and biotype 1, were used. Briefly, overnight cultures of A. pleuropneumoniae grown in BHI broth plus NAD (15 µg/ml) and E. coli grown in LB culture media, were diluted 1/100 in LB broth plus glycerol (0.20%). A volume (200 µl) was aliquoted by triplicate in wells of a sterile 96-well microtiter plate (Costar R 3599, Corning, NY, USA) using the following template: 100 µl A. pleuropneumoniae in BHI (glycerol 0.20%) plus 100 µl E. coli in LB (glycerol 0.20%) and incubated 24 h at 37 • C. Wells containing sterile broth or A. pleuropneumoniae (100 µl of bacteria plus 100 µl of LB-glycerol 0.20%) were used as blank and negative control, respectively (A. pleuropneumoniae it is unable to grow and form biofilms under these conditions). Wells containing E. coli ATCC 25922 and L17608 (100 µl of bacteria plus 100 µl of LB-glycerol 0.20%) were used as positive controls for biofilm formation.

Colony Forming Unit (CFU) Counts of Mono and Two-Species Biofilms
To confirm the presence of A. pleuropneumoniae and E. coli in the biofilms, the colony forming units (CFU) were counted, using selective growth media and colony morphology. The CFU test was performed as previously described (27,37) with modifications. Briefly, the medium was carefully removed from each well by pipetting and washed with 200 µl of sterile water. Twenty microliters of NaCl 0.85% were added. A tip was used to scrape the bottom and completely disintegrate the biofilm, taking 20 µl to perform serial dilutions in saline solution 0.85% (from 10 −2 to 10 −7 ). Finally, 100 µl of the dilution were plated on BHI, BHI plus NAD, and Blood agar plus NAD (A. pleuropneumoniae causes beta-hemolysis), incubated 24 h at 37 • C, and the CFU count was performed.

Confocal Laser Scanning Microscopy (CLSM)
In order to study the morphology of mono and two-species biofilms, E. coli biofilms with or without A. pleuropneumoniae 719 were prepared as described above and stained with FilmTracer

Enzymatic Treatments of Two-Species Biofilms
The enzymatic treatment assays were performed as described previously (3) for proteinase K and DNase I, and (38) for cellulase. Biofilms were prepared as described above and 50 µL of proteinase K (500 µg/mL in 50 mM Tris-HCl pH 7.5, 1 mM CaCl 2 ), 50 µL of DNase I (500 µg/mL in 150 mM NaCl, 1 mM CaCl 2 ), or 50 µL of cellulase (40 µU/ml in 100 mM C 2 H 3 NaO 2 , 50% DMSO) were added directly to the biofilms. Samples with proteinase K or DNase I were incubated for 1 h at 37 • C, and with cellulose were incubated 30 min at 37 • C. Control wells were treated with 50 µL of the buffer without the enzyme. Biofilms were washed and stained with crystal violet and the absorbance was measured at 590 nm.

Phenotype Assay: Congo-Red and Calcofluor
Congo-red and calcofluor assays were performed in order to determine the production of fimbriae-curli and cellulose, and were performed as described previously (38). For the assay, a 2 µl drop of bacterial culture was taken from liquid medium, and was placed on Luria-Bertani salt-free plates (LB; Difco Laboratories, Detroit, MI), containing 0.02% of Congo-red (Sigma R , C-6767) and 0.002% of Coomassie brilliant blue G (Sigma R , F3546-5G) for fimbriae detection, and containing 0.02% calcofluor (fluorescent brightener 28, Sigma-Aldrich R F-3543) dissolved in 1 mM HEPES for cellulose detection. For the two-species assay, a 1:1 dilution of the bacterial cultures (E. coli plus A. pleuropneumoniae) was performed in respective culture media. Twelve strains were placed per plate with a centimeter of distance between each drop. After seeding, the plates were allowed to dry for 5 to 10 min faceup and then incubated for 48 hat 30 • C. The fluorescence of the colonies was verified by UV light illumination (360 nm) after overnight incubation at 30 • C. E. coli CFT073 and E. coli ATCC 25922 were used as positive and negative controls, respectively (39).

Scanning Electron Microscopy
The two-species biofilm formed by A. pleuropneumoniae 719-E. coli ATCC 25922 was observed under electron microscopy (SEM). Mono-species biofilms of A. pleuropneumoniae and E. coli were used as positive controls and were grown as described previously. The two-species biofilm was prepared as described above. Samples were processed as described Loera-Muro et al. (26), and were observed with a Jeol LV-5900 scanning electron microscope. The bacteria dimensions were measured with the microscope software. The experiment was repeated three times, measured between 3 and 5 bacteria in three different fields.

Statistical Analysis
Statistical significance analyses (p-value < 0.05) of differences in biofilms were determined by Two-way ANOVA followed

Identification of Bacteria From Drinking Water of Swine Farms
A total of 10 samples of drinking water from pig farm were obtained. After performing the isolation of bacteria from water, 52 colonies were selected for identification, 63.46%

Colony Forming Unit Counts of Mono and Two-Species Biofilms
In the case of the CFU count, the numbers of A. pleuropneumoniae and E. coli found in two-species biofilms were similar in almost all cases (Figure 3). No significant differences between populations were shown. These results suggest strongly that A. pleuropneumoniae has the ability to be incorporated into biofilms produced by environmental bacteria, which supports that A. pleuropneumoniae is using the multi-species biofilms as a survival strategy in the environment, at least for 72 h of interaction. For all subsequent assays, A. pleuropneumoniae reference strain 719 was selected because this strain isolated from pigs has a high capacity for biofilm formation.

Two-Species Biofilms Matrix Composition
Confocal laser scanning microscopy (CLSM) and the dye FM-143 was carried out to visualize biofilm morphology from both, mono-and two-species biofilm. Observation of several fields on each sample evidenced by increments in most of the two-species biofilms formed with different E. coli strains and the A. pleuropneumoniae isolates, as compared to the monospecies biofilms of E. coli (Figure 4). As a whole, these images are in accordance with the results obtained by the crystal violet technique. Likewise, it was observed that the biofilms morphology had some changes (Figures 4, 5). Otherwise, biofilm matrix components were characterized by CLSM in combination with different dyes directed mainly toward PGA, eDNA, and proteins. These three macromolecules were detected in the extracellular matrix (Figure 6). In some cases, the production of proteins, PGA and eDNA, stained with SYPRO Ruby, WGA, and BOBO-3; respectively, showed an increase in biofilms formed by two-species compared to mono-species biofilms (Figure 6). The data obtained from the SYPRO Ruby stain, which labels most classes of proteins, showed a protein increase in the two-species biofilms E. coli 14-2 and EcATCC with A. pleuropneumoniae 719. The isolates 8-4, 13-1, 13-2, 13-5, 14-1, and EcATCC were stained with WGA, suggesting the increase in the presence of PGA or at least in the presence of N-acetyl-D-glucosamine and N-acetylneuraminic acid residues in the biofilm matrix. BOBO-3 iodide that stains extracellular DNA showed an increase only among the isolates 14-2 and EcATCC.
Congo-red has been used extensively to supplement nutrient agar to distinguish the production of the extracellular matrix components cellulose and curli fimbriae from non-cellulose curliated bacteria. Likewise, the phenotype on calcofluor plates served as an indicator of cellulose production. In this work, the presence of few fimbriae-curli forming, and cellulose producer strains was observed (13-1, 13-2, 13-5, 14-1, 14-2, and 14-5, Figure 7). Also, changes were observed when the E. coli strains were together to A. pleuropneumoniae in the cellulose and curli production. These results confirm changes in the production of extracellular matrix components in two-species biofilms as compared to the E. coli mono-species.
To determine the structural roles played by the compounds forming the extracellular matrix, enzymatic treatments were performed on the two and mono-species biofilms (Figure 8). Treatment with proteinase K, DNase I and cellulose provoked reduction in all two-species biofilms. It was more important that the effect observed in the mono-species biofilms. These results indicate that when biofilms of two-species are being formed, the cellulose, as well as proteins and eDNA, take a structural function as occur in mono-species biofilms.

Scanning Electron Microscopy
Actinobacillus pleuropneumoniae and E. coli two-species biofilms were analyzed by SEM (Figure 9). It was possible to observe the presence of two populations in the two-species biofilms, an abundant population of larger bacteria, and a minor population of smaller bacteria (p < 0.01, Figure 9D). It was also interesting to observe fimbriae-like or curli-like structures, and their promotion of interaction between all the bacteria present in the biofilm. Moreover, these structures appear more abundant in the E. coli mono-species biofilm ( Figure 9B) than in the E. coli-A. pleuropneumoniae two-species biofilm ( Figure 9C).

DISCUSSION
Porcine pleuropneumonia caused by A. pleuropneumoniae, is one of the most important porcine respiratory diseases which is spread by direct contact between the carrier-infected pig FIGURE 6 | Escherichia coli strains in mono-or two-species biofilms with A. pleuropneumoniae (strain 719) observed by confocal laser scanning microscopy. Images show the mono-species biofilms of E. coli isolates and two-species biofilms of E. coli isolates and A. pleuropneumoniae 719 in LB media stained with wheat-germ agglutinin (WGA)-Oregon green, SYPRO Ruby, and BOBO-3 (all from Invitrogen, Eugene, OR). Scale bar 30 µm.
Frontiers in Veterinary Science | www.frontiersin.org 8 August 2018 | Volume 5 | Article 184 FIGURE 7 | Congo-red and calcofluor binding assays on mono or two-species colonies formed by E. coli alone or with A. pleuropneumoniae. Congo-red assay evidenced curli-producing bacteria when E. coli was growing on congo-red-supplemented nutrient agar. Moreover, changes in cellulose production were detected by the calcofluor assay. The plus sings indicate differences on production observed in the colonies of one or two-species. The crosses indicate the qualitative production of cellulose seen in each of the biofilms (+ low, ++ medium, + + + medium high and + + ++ high). and an uninfected pig or by aerosols (1)(2)(3)(4)(5). The indirect route of transmission via surface was not considered very important and therefore the ability of A. pleuropneumoniae to survive in the environment outside of its host is not yet known (25). Previous studies from our group demonstrated that A. pleuropneumoniae is able to grow in unsuitable environments when forming multi-species biofilms with other respiratory pathogens of pigs that are also part of the porcine respiratory disease complex (6,17,27). In this study, a total of 10 samples of drinking water were taken from a swine farm in the State of Aguascalientes where previously Loera-Muro et al. (26) detected the presence of A. pleuropneumoniae in drinking water using a specific PCR for apxIV gene. We found that some microorganisms that form the microbial community of drinking water of swine farms were bacteria such as Escherichia, Enterobacter, Pseudomonas, Photobacterium, Salmonella, Ochrobactrum, Pasteurella, Cryseumonas, Kluyvera, Citrobacter, and Buttiauxella. The more abundant culturable bacterium isolated from samples of drinking water from a swine farm in the State of Aguascalientes was E. coli. All E. coli isolates belonged to the group of extra-intestinal pathogens (ExPEC). ExPEC are facultative pathogens, which can reside in the gastrointestinal tract of a certain fraction of the human and animal population. They possess several virulence traits that allow them to colonize different niches including urogenital tract resulting in urinary tract infections (UTIs), meningitis and sepsis in animals and humans (40). In pigs, these pathogens could cause fatal pneumonia, severe septicaemia and haemorrhagia; thus, they also represent a latent risk for human health (41)(42)(43). The presence of ExPEC strains may indicate a zoonotic potential risk posed by swine farms to cause infections by ExPEC stains in both, pigs and humans, mainly farm workers.
To seek out interactions during biofilm formation in twospecies biofilms between the swine respiratory pathogen A. pleuropneumoniae (strain 4074 and 719), and E. coli isolated from drinking water, different approaches were undertaken. An increase in biofilm formation was evidenced by crystal violet staining, when two-species biofilms were compared to those obtained with the E. coli mono-species assay. This result suggests an interaction between both bacteria affecting bacterial distribution and probably biomass production as reported by others (44). Furthermore, by applying the methodology of (36), for the qualitative determination of E. coli's ability to form biofilm, this increment was observed. When A. pleuropneumoniae was combined with any of the E. coli strains, the biofilm classification changed from non-adherent to weakly adherent or moderately adherent. In addition, we were able to recover A. pleuropneumoniae from the biofilm in most cases, which was unexpected considering biofilm was cultured in optimal conditions for E. coli growth, but not for A. pleuropneumoniae. Thus, the ability of A. pleuropneumoniae to form two-species biofilm with E. coli isolated from drinking water was confirmed. This interaction occurs because E. coli possibly supplies some nutrients that promote A. pleuropneumoniae growth (27).
There are several reports on the advantages obtained during multi-species biofilm formation. Liu et al. (45) determined the capacity to incorporate the bacteria E. coli O157:H7 in pre-formed biofilms with bacteria obtained during the fresh produce processing environments. When co-cultured with E. coli O157:H7, Burkholderia caryophylli, and Ralstonia insidiosa exhibited increases in biofilm biomass, which were around 180 and 63%, respectively; as well as in the thickness of the biofilm. Biyikoglu et al. (46) reported that Actinomyces oris and Veillonella parvula promoted biofilm growth of all Fusarium nucleatum strains tested in their study. Both studies reported similar effects: increases in biofilms when they are formed by multiple species, such as in our case. The results presented here are in accordance with the study carried out by Bridier et al. (47), where pathogenic Staphylococcus aureus grown in mixed biofilm with the Bacillus subtilis ND medical strain, was protected from peracetic acid (PAA), an oxidizing agent, thus enabling its persistence in the environment. Standar et al. (37) also showed that two-species combinations of Streptococcus mitis with either Streptococcus mutans or Aggregatibacter actinomycetemcomitans favored bacterial interactions influencing biofilm mass, biofilm structure and cell viability. The result reported by Standar et al. (37) is similar to that observed in our study where E. coli in presence of A. pleuropneumoniae favored an increase in biofilm formation, allowing it to survive even under conditions unfit for its development. Likewise, the integration of pathogenic bacteria in biofilms formed by other bacteria was shown by Stewart et al. (48), where Legionella pneumophila 130b persisted within a two-species biofilm formed by Klebsiella pneumoniae and Flavobacterium sp., or by K. pneumonia, and P. aeuroginosa. Furthermore, the authors reported that Legionella pneumophila 130b was able to colonize biofilms formed by single-species such as K. pneumoniae and Pseudomonas fluorescens, and persist in the environment. Finally, (49) using the chinchilla otitis media model concluded that the biofilm formation and persistence on the middle-ear mucosal surface by pneumococcal is facilitated by Haemophillus influenzae coinfection. In this study, A. pleuropneumoniae was able to colonize and incorporate into biofilms formed by E. coli, which might allow it survive in hostile conditions, outside of its host, persisting in the environment as a source for transmission to other pigs. Considering our results, an interaction is evidenced, between bacteria, A. pleuropneumoniae and the E. coli environmental isolates. It is unknown whether the increase in biofilm produced when going from mono to two-species is due to the incorporation of A. pleuropneumoniae into these biofilms, or a more complex interaction is, causing E. coli to over-produce biofilm, via the generation of extracellular matrix components, like cellulose, curli, antigen 43, DNA, β-1,6-N-acetylglucosamine (β-1,6-GlcNAc), capsule sugars, and colanic acid (50,51). Our working model considers that A. pleuropneumoniae is incorporated into E. coli biofilms, thus in order to survive and grow in this hostile environment, at least for 72 h, it promotes an increment in biofilm, followed by interactions between both bacteria, resulting in the final increment seen in the two-species biofilms. Moreover, it was also possible to observe that the components of the extracellular matrix in the two-species biofilms changed their function, promoting greater structural stability to the biofilm. In the enzymatic assays a decrease in the biofilm formed by E. coli and A. pleuropneumoniae was seen, when compared to the biofilms formed only by E. coli strains. This change in the structural function of components in the extracellular matrix when going from mono-species to multi-species biofilm had already been reported by our group previously with A. pleuropneumoniae (27). However, little is known with regards to other bacterial species (52,53).
In conclusion, our data suggests that A. pleuropneumoniae has the ability to integrate and form multi-species biofilms with environmental bacteria, which could allow it to survive outside of the host, specifically in water, establishing relationships with bacteria from the microbial community of water such as E. coli; therefore suggesting a possible mechanism for porcine pleuropneumonia persistence or transmission.

AUTHOR CONTRIBUTIONS
FR-C directed the biofilms experiments with Escherichia coli, mono species, and di-species. AL-M advised the biofilm experiments with Actinobacillus pleuropneumoniae, mono species, and di-species. NV-P, CB-G, and AM-F conducted the experiments with biofilms for both species, and they analyzed them by confocal microscopy. FA-G, JH, MJ, and RO advised the management of bacterial strains. F-AG also advised the microbiological analysis. AG-B proposed the research line, is the responsible of the project that support this work, directed five thesis involved in the work.