Intracellular invasion and survival of Brucella neotomae, another possible zoonotic Brucella species

In 1967, Brucella neotomae was first isolated from Neotoma lepida, the dessert wood rat, in Utah. With little infection data since its discovery, the zoonotic potential of this Brucella species is largely unknown. Recent reports of isolation from human cerebrospinal fluid, along with current literature suggest that B. neotomae has the ability to infect various hosts and cell types. In this report we extend the knowledge of B. neotomae ATCC 23459’s intracellular invasion and survival abilities to a variety of cell lines through gentamicin protection assays. Some of the phagocytic and epithelial cell lines from various mammalian species represent characteristics of some cell types that could be encountered by Brucella in potential hosts. It was found that B. neotomae ATCC 23459 exhibits generally lower intracellular bacterial CFUs compared to the mouse-passaged strain of B. neotomae ATCC 23459, B. suis 1330, and B. abortus 2308. Ultimately, these observations provide a small piece of the puzzle in the investigation of the breadth of B. neotomae’s pathogenic potential.


Introduction
The genus Brucella is composed of Gram-negative, non-sporulating, non-motile cocco-bacilli that lack a capsule from the alpha-proteobacteria group. This group is composed of a variety of bacterial species that are generally associated, but not obligated to specific hosts. The hosts of Brucella spp. range widely and include rodents, ungulates, marine mammals, and humans. There are currently 11 accepted species of Brucella: B. melitensis, B. abortus, B. suis, B. canis, B. inopinata, B. ovis, B. neotomae, B. pinnipedialis, B. ceti, and B. microti. Many are considered zoonotic with around 500,000 cases of human brucellosis reported worldwide each year [1].
The bacteria are generally spread through close contact with secretions from infected animals, aerosols, and the consumption of unpasteurized dairy products or undercooked meat. All contribute to a concern for public health [2]. Symptomatic patients exhibit flu-like symptoms for up to four weeks before spontaneous recovery followed by the onset of symptoms again (undulant) [1]. Most infected individuals exhibiting undulant fever recover after 4-12 months, but some develop a chronic infection even after treatment [3]. The disease is rarely fatal, but 2% of all untreated individuals die from complications of brucellosis making it an a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 intraperitoneal (IP) with 2.0 x 10 5 CFUs/mL. Seven days post inoculation the mice were euthanized via carbon dioxide asphyxiation, the spleens aseptically removed, and individually homogenized. Serial dilutions of homogenized spleens were plated on TSA and incubated for 2-3 days at 37˚C under 5.0% CO 2 atmospheric conditions. Single isolated colonies were grown in tryptic soy broth (TSB) and the protocol was repeated once more for each strain. Passaged stock cultures of the above strains were produced following established techniques [15,20]. The original B. neotomae ATCC 23459 and second passaged strains were used for all of the experiments performed in this paper.

Identity verification of Brucella spp. through presumptive tests
B. neotomae ATCC 23459, B. neotomae ATCC 23459 passage 2, B. abortus 2308, and B. suis 1330 were used for each presumptive test. Pure bacterial cultures were streaked for isolated colonies, and two to three isolated colonies were used for each presumptive test. One to two drops of oxidase reagent or 3% hydrogen peroxide were directly applied to the isolated colonies and observed for a reaction to occur within 30 seconds [10,18]. An acriflavin (Sigma A-8126) stock solution of 1:1000 wt/v was prepared and used immediately by transferring 2-3 isolated colonies into 30μL of the stock solution on a glass slide. A homogeneous solution was observed for agglutination [10,18]. Thirty microliters of 3-week post-infection mouse serum (VT-COHR sample collection) were mixed with 2-3 isolated colonies on a glass slide and examined for agglutination. A stock solution of 0.1% thionin (Sigma A-861340) and 0.1% basic fuchsin (Sigma A-215597) were prepared in distilled water and autoclaved. TSA was prepared and thionin or basic fuchsin dye (10mL stock dye/1000mL TSA) was added while the liquid medium was still warm. A single isolated colony, 24hrs old, was plated on the TSA plates containing a dye as well as on a TSA plate without dye. The plates were observed for growth after 24-48hrs at 37˚C under 5.0% CO 2 atmospheric conditions [10,18]. The Brucella ladder PCR protocol from López-Goñi, I., et.al. 2008 was followed. All polymerase chain reactions (PCR) were performed using established techniques with Platinum PCR SuperMix High Fidelity (Invitrogen) on a Gradient Mastercycler (Eppendorf) [25]. Brucella ladder oligonucleotides primers were purchased from Sigma-Genosys (Sigma-Aldrich, USA).

Collection of naïve murine peritoneal macrophages
Two 4-month-old female BALB/c mice (Envigo) were euthanized using carbon dioxide asphyxiation. Using a 16 gauge needle 200μL of Dulbecco's modified Eagle's medium (DMEM) (Sigma-Aldrich, USA) was injected into the peritoneal space. The medium was retrieved with care to prevent blood from contaminating the samples. The collected peritoneal macrophages were then placed in 50mL Falcon conical tubes. The process of washing the peritoneal space with DMEM was repeated 3 times per mouse. The cells were centrifuged at 3,000 revolutions per minute (rpm) for 5min. The white layer of peritoneal macrophages was resuspended in DMEM supplemented with 2mM Lglutamine, 2mM sodium pyruvate, 10% heatinactivated fetal bovine serum (FBS), 100 U/mL penicillin, and 100 μg/mL streptomycin (Sigma-Aldrich, USA) and counted under a microscope using a hemacytometer. Wells of a 24-well plate were seeded at 2x10 6 cells per well with peritoneal macrophages and incubated over night at 37˚C under 5% CO 2 atmospheric conditions before the gentamicin protection assay was performed.

Opsonization of Brucella species
Serum collected from female BALB/c mice 4 weeks post infection with B. neotomae ATCC 23459 was used to opsonize B. neotomae ATCC 23459 and B. abortus 2308 before infection with naïve intraperitoneal murine macrophages. To produce a bacterial suspension for opsonization, a culture of B. neotomae ATCC 23459 or B. abortus 2308 in stationary phase was centrifuged for 5min at 3,000 rpm. The bacterial pellet was resuspended in the appropriate cell culture media. The OD 600 of the cell culture media and bacterial suspensions was determined with a spectrophotometer, and then diluted to an OD 600 of 0.15 (10 9 CFUs/mL). Serial dilutions of the cell culture media and bacterial suspensions were plated on TSA plates to determine the CFUs used for each infection. One hundred μL of serum was then added to the suspension and incubated for 30 min shaking at 37˚C under 5% CO 2 atmospheric conditions before the gentamicin protection assays were performed.

Cellular culture
The cell lines were chosen based on the types of cells that Brucella would encounter in the host, with epithelial cells being the initial cell type and phagocytic cells being the ultimate site for replication. The species that the cell lines originate from represent potential zoonotic routes between wild/domestic animals and the public in the United States. Cell lines were cultured as monolayers in T 150 cell culture flasks (Corning) using aseptic technique and grown at 37˚C under 5% CO 2 atmospheric conditions [13]. The murine monocyte cell line (J774A.1), the bovine monocyte cell line (BM), and the human epithelial cell line (HeLa) were obtained from the culture collection at the Center for One Health Research laboratory building at Virginia Tech. These cell lines were cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 2mM Lglutamine, 2mM sodium pyruvate, 10% heat-inactivated fetal bovine serum (FBS), 100 U/mL penicillin, and 100 μg/mL streptomycin (Sigma-Aldrich, USA) [26][27][28]. The bovine epithelial cell line (MAC-T) was obtained from Dr. Isis Kanevsky's culture collection at Virginia Tech and was cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 2mM Lglutamine, 2mM sodium pyruvate, 5% heat-inactivated fetal bovine serum (FBS), 100 U/mL penicillin, and 100 μg/mL streptomycin (Sigma-Aldrich, USA) [29]. The human monocyte cell line (THP-1) � and the dog macrophage cell line (DH82) were cultured in RPMI 1640 medium supplemented with 2mM Lglutamine, 10% heat-inactivated fetal bovine serum (FBS), 100 U/mL penicillin, and 100 μg/mL streptomycin ( � THP-1 cells were activated/differentiated using 50 ng/mL phorbol myristate acetate (PMA) for 48hrs.) (Sigma-Aldrich, USA), and were also obtained from the culture collection at the Center for One Health Research [30,31]. The swine monocyte cell line (3D4/31) was obtained from Dr. XJ Meng's culture collection at Virginia Tech and was cultured in RPMI 1640 medium with 2 mM L-glutamine adjusted to contain 1.5 g/L sodium bicarbonate, 4.5 g/L glucose, 10 mM HEPES, 1.0 mM sodium pyruvate, 0.1 mM nonessential amino acids, 10% FBS, 100 U/mL penicillin, and 100 μg/mL streptomycin (Sigma-Aldrich, USA) [32]. The swine epithelial cell line (IPEC-J2) was obtained from Dr. Lijuan Yuan's culture collection at Virginia Tech and was cultured in DMEM supplemented with 5% FBS, 5μg/ml insulin, 5μg/ml transferrin, 5μg/ml selenium, 5ng/ml epidermal growth factor, 100 U/mL penicillin, and 100 μg/mL streptomycin (Sigma-Aldrich, USA) [33].

Gentamicin protection assay
Following laboratory protocols for each cell line, the following passage number were used for the gentamicin protection assays: J774A.1-9, DH82-6, THP-1-4, HeLa-5, 3D4/31-5, JPEC-1-3, BM-2, and MAC-T-7. Cell lines were seeded in triplicate per time point/bacterial strain in 24 well plates and incubated overnight in applicable antibiotic free media. The wells were seeded at 2x10 6 cells per well. After overnight incubation the cells were infected with a multiplicity of infection (MOI) of 1:100 for one hour at 37˚C under 5% CO 2 atmospheric conditions with 48hrs old bacterial cultures at the stationary growth phase in Tryptic soy broth (Sigma-Aldrich, USA). The Log 10 CFUs/mL of the inoculum can been seen in S1 Table. The bacterial suspensions used were produced as described above. After 1hr of incubation, the infected cells were washed three times with sterile PBS to remove non-phagocytized Brucella spp. Each well was filled with 2 mL of media containing 50 μg/mL gentamicin and incubated for 1hr. At specified time points over 48hrs post infection cells were washed three times with sterile PBS and then 200 μL of 0.1% Triton X-100TM (Sigma-Aldrich, USA) lysing agent was added to each well. After 5-10 minutes of incubation in the lysing agent, serial dilutions were plated on TSA plates and incubated for 2 days at 37˚C under 5.0% CO 2 atmospheric conditions to determine intracellular CFUs [13]. For each time point, a portion of the cell culture media was plated on TSA plates to determine if there were free-living bacterial cells in the wells.

Statistical analysis
All group and time-point data were compared using a One-way ANOVA plus a Dunnett's correction post hoc test, with both tests significant <0.05. B. neotomae ATCC 23459 (BN) was used as the species that the other species were compared to. Differences were considered to be significant when p values were less than 0.05.

Identification and characterization of B. neotomae
B. neotomae ATCC 23459 (BN) and B. neotomae ATCC 23459 passage 2 (BNP2) were characterized using several presumptive and biochemical tests. B. abortus 2308 and B. suis 1330 were used as control organisms for each test. All tested Brucella species grew small circular colonies after 24-48hrs of incubation at 37˚C under 5.0% CO 2 atmospheric conditions. The organisms were Gram negative short rods that exhibited the smooth phenotype when a homogenous suspension was made in a 0.001% solution of acriflavin. The organisms were oxidase negative, catalase positive, grew in the presence of 100 μg/mL of thionin, and did not grow in the presence of 100 μg/mL of basic fuchsin (Table 1). Representative photos of the growth/lack of growth on thionin and basic fuchsin plates can be seen in S1-S3 Figs. Although the B. neotomae strains exhibited growth on thionin dye that differed from the literature, it has been shown that Brucella strains can have some variability in regards to biochemical assay results [20,34]. The organisms still maintained a Brucella ladder PCR banding pattern similar to B. neotomae in the literature (Fig 1) [25]. After passage through BALB/c mice the splenic bacterial CFUs were: 6.96 (B. neotomae), 6.58 (B. abortus 2308), and 7.19 (B. suis 1330) Log 10 CFUs/spleen.  [10]. The recovered Log 10 CFUs/mL of intracellular BN at each time point (2 hours, 6 hours, 24 hours, and 48 hours post infection) over 48hrs was significantly lower than the Log 10 CFUs/mL of intracellular B. abortus 2308 recovered over 48 hours (Fig 2). It is important to note that no bacterial growth was observed from all plated infected cell culture supernatant for all the described gentamicin protection assays.

Uptake and survival of Brucella strains compared to B. neotomae ATCC 23459 in phagocytic cells
The results from each phagocytic cell line tested can be seen in Figs 6-10, and a summary of all the statistical results can be seen in S1-S4 Tables. BNP2 was

Discussion
Cell culture is widely used as a representative model for microorganism's infectivity and is especially important in helping predict the range of pathogenicity of atypical Brucella species [16,[35][36][37]. Invasion and replication in epithelial cells and macrophages are vital in Brucella infections, although they are just one aspect to virulence [16,37]. Other aspects important to virulence are the type IV secretion system and the LPS O-side chain [11,13,23,35,36]. Current literature supports that B. neotomae possesses these virulence factors [4,16,35]. B. neotomae has also been shown to have type IV dependent intracellular survival and a similar late  CFUs were determined over 48hrs in BM as described. Error bars represent the standard deviation of the triplicates from a single experiment. � represents when intracellular CFUs from more than one species tested statistically differs from BN with any p value lower than 0.05.
https://doi.org/10.1371/journal.pone.0213601.g007 endoplasmic reticulum associated phagosome stage to that of other virulent Brucella spp. [11]. Ultimately, it was expected to see B. neotomae strains have the ability to survive within phagocytic and epithelial cell lines in a similar manor to other virulent species, especially with current literature showing that it can infect humans, swine, rodents, and bison [11,12,14,20,24,37]. Our intracellular invasion and survival profiles of B. neotomae add to the evidence in current literature warranting further characterization of its pathogenic ability. Our data aligns with current literature and extends the known intracellular uptake and survival profiles of the Brucella species to include a breadth of cell lines [38][39][40][41][42][43][44][45][46].
The mucous membranes of the respiratory and digestive tracts are the most common sites of entry for Brucella species, with epithelial cell invasion a crucial initial stage of infection [37,39,42,43]. We show that B. neotomae can persist in cultured epithelial cells, although the other strains tested generally maintained higher intracellular CFUs. Mammary tissue has been considered a known route of infection for Brucella, but a major reduction in intracellular bacterial numbers was observed in the bovine mammary epithelial cells (MAC-T). More in depth CFUs were determined over 48hrs in 3D4/31 as described. Error bars represent the standard deviation of the triplicates from a single experiment. � represents when intracellular CFUs from more than one species tested statistically differs from BN with any p value lower than 0.05.
https://doi.org/10.1371/journal.pone.0213601.g008 experiments with various strains and types of mammary tissues need to be conducted to confirm this observation [35]. It would also be worth while to determine other tissue types that B. neotomae can be isolated from in various species and whether those infected tissues can be a source of infection to other potential hosts. It is important to note that current literature has shown MAC-T cells have the ability to clear intracellular organisms, which could possibly explain this observation [35,45].
After crossing the epithelial cell border, phagocytic cells have been shown to be an important site for Brucella intracellular replication within host tissues [33,37,42,45]. Although internalized in low numbers, enough cells are able to activate virulence genes and establish a replication niche to produce chronic infection [33,42,45]. Our data shows that Brucella 2308, B. suis 1330, and BNP2 generally had higher intracellular bacterial CFUs in phagocytic cell lines compared to BN.
B. abortus 2308 and BN exhibited a downward trend in naïve murine peritoneal macrophages, which brings into question the validity of the common practice of testing Brucella species in cultured cell lines. Further biological replicates need to be done along with determining the intracellular uptake and survival profiles in non-immortalized cell culture to better understand this phenomenon. It is of importance to take into account that survival in phagocytic cells can be negatively affected by opsonization [35]. This may contribute to the naïve murine peritoneal macrophage intracellular profiles observed as non-opsonized Brucella infection of naïve macrophages in current literature shows higher intracellular replication [12,13,35,43,44]. B. neotomae antibodies are grouped under B. abortus-B. suis, but no studies have been done to determine their effects on BN intracellular survival [10-12, 18, 25, 46]. We have shown that opsonization with different mouse sera can significantly impact initial uptake into the cell and further investigation needs to be done to understand this effect on intracellular uptake and survival.
The limitations of cell culture are important to take into account when interpreting cell culture data like ours. Cell culture can differ genetically and phenotypically from primary cell culture, among other limitations including a lack of biological stratification and influence that the host provides. Although this is important to take into consideration, there is a large body of evidence in current literature that depicts the intracellular profiles of Brucella and it is common practice to characterize them in immortalized cell culture [23,35,39,44,47,[48][49][50][51]. CFUs were determined over 48hrs in DH82 as described. Error bars represent the standard deviation of the triplicates from a single experiment. � represents when intracellular CFUs from more than one species tested statistically differs from BN with any p value lower than 0.05. https://doi.org/10.1371/journal.pone.0213601.g010 Furthermore, it is of interest to highlight the general significantly higher intracellular bacterial CFUs exhibited by BNP2 compared to the parent strain BN. As stated, it has been shown that passage through rodents produces variations in virulence of B. neotomae. It was important to compare the classical virulent and passaged strains to the B. neotomae ATCC strain to observe any differences. There were some significant differences between the parent and passaged B. neotomae strains at various time points, but more experiments need to be completed to better understand this. In particular, BNP2 was statistically higher than BN in HeLa, JPEC-1, and most macrophage cell cultures [20].
In conclusion, this work shows that B. neotomae possesses the ability to invade and survive in cultured phagocytic and epithelial cells derived from a wide array of species. In general, B. abortus 2308, B. suis 1330, and B. neotomae passage 2 had statistically higher intracellular bacterial CFUs compared to B. neotomae ATCC 23459. Our data, along with the current literature warrants that the currently accepted thought that B. neotomae lacks pathogenic ability needs to be investigated [7,9,25,47]. Future research is needed to better understand B. neotomae's ability (genotypically and phenotypically) to infect a variety of hosts, the tissues in which it replicates, the infectious dose and host immune status required to cause disease, the routes of exposure to cause infection, the host's immunological response to B. neotomae, and the affects of passaging on the species.