Entrance and Survival of Brucella pinnipedialis Hooded Seal Strain in Human Macrophages and Epithelial Cells

Marine mammal Brucella spp. have been isolated from pinnipeds (B. pinnipedialis) and cetaceans (B. ceti) from around the world. Although the zoonotic potential of marine mammal brucellae is largely unknown, reports of human disease exist. There are few studies of the mechanisms of bacterial intracellular invasion and multiplication involving the marine mammal Brucella spp. We examined the infective capacity of two genetically different B. pinnipedialis strains (reference strain; NTCT 12890 and a hooded seal isolate; B17) by measuring the ability of the bacteria to enter and replicate in cultured phagocytes and epithelial cells. Human macrophage-like cells (THP-1), two murine macrophage cell lines (RAW264.7 and J774A.1), and a human malignant epithelial cell line (HeLa S3) were challenged with bacteria in a gentamicin protection assay. Our results show that B. pinnipedialis is internalized, but is then gradually eliminated during the next 72 – 96 hours. Confocal microscopy revealed that intracellular B. pinnipedialis hooded seal strain colocalized with lysosomal compartments at 1.5 and 24 hours after infection. Intracellular presence of B. pinnipedialis hooded seal strain was verified by transmission electron microscopy. By using a cholesterol-scavenging lipid inhibitor, entrance of B. pinnipedialis hooded seal strain in human macrophages was significantly reduced by 65.8 % (± 17.3), suggesting involvement of lipid-rafts in intracellular entry. Murine macrophages invaded by B. pinnipedialis do not release nitric oxide (NO) and intracellular bacterial presence does not induce cell death. In summary, B. pinnipedialis hooded seal strain can enter human and murine macrophages, as well as human epithelial cells. Intracellular entry of B. pinnipedialis hooded seal strain involves, but seems not to be limited to, lipid-rafts in human macrophages. Brucella pinnipedialis does not multiply or survive for prolonged periods intracellulary.


Introduction
Brucellosis is an infectious disease that affects a wide range of mammalian species and is regarded as the world's most common bacterial zoonotic disease [1]. For decades, the genus Brucella included six species with different preferred terrestrial mammalian hosts, four of which are pathogenic for humans [2]. Recently four additional species have been included [3][4][5]. The occurrence of human disease is dependent on animal reservoirs, including wildlife [6].
Brucella spp. were isolated from marine mammals for the first time in 1994 [7,8] and validly published as members of genus Brucella with the names B. pinnipedialis (pinnipeds; seals, sea lions and walruses) and B. ceti (cetaceans; whales, dolphins, and porpoises) in 2007 [3]. Marine mammal brucellae have since been serologically indicated in and isolated from pinnipeds and cetaceans from multiple locations; however gross pathology in association with Brucella infection in marine mammals is exclusively found in cetaceans [9,10]. The results from experimental infections in various animal species are diverging and the zoonotic potential of the marine mammal brucellae is largely unknown [7,[11][12][13]. However, reports of human disease exist [14][15][16]. Interestingly, none of the naturally infected human cases reported contact with marine mammals, but consumption of raw seafood was noted [15,16]. All three Brucella-isolates from the naturally acquired infections shared an identical genotype (sequence type 27), to date only being found in marine mammals in Pacific waters [17]. The infectious cycle of marine mammal brucellae is to a large extent uncertain and unknown hosts or carriers may exist. Brucella melitensis has been isolated from Nile catfish (Clarias gariepinus) [18], while B. ceti and B. pinnipedialis have been isolated from lungworms in cetaceans [19] and pinnipeds [20] respectively, pointing at other possible reservoirs in the marine ecosystem besides pinnipeds and cetaceans. In light of the extensive use of marine resources, including products from marine mammals, increased knowledge about Brucella spp. in the marine environment and their possible implications in human disease, is an important issue in the "One Health" perspective.
Brucella spp. are facultative intracellular bacteria that can survive and replicate within membrane-bound compartments in phagocytes and epithelial cells [21][22][23]. Studies of the mechanisms of bacterial invasion and intracellular multiplication involving the marine mammal brucellae are sparse and previously only investigated by Maquart and coworkers (2009). They observed different infection dynamics between the marine mammal brucellae during macrophage infection in vitro [24]. Brucella pinnipedialis reference strain (NCTC 12890), isolated from a common seal (Phoca vitulina), was able to infect and multiply in human and murine macrophage cell lines to the same extent as the virulent terrestrial Brucella spp. In contrast, B. pinnipedialis isolated from hooded seal (Cystophora cristata) showed absence of invasive potential. This particular strain displayed a high prevalence in the declined Northeast Atlantic hooded seal stock [25] and has been isolated from various organs of young seals; however Brucella-associated pathology has not been identified [26]. The Northeast Atlantic stock of hooded seal is now 10 -15 % of the 1946 population and has been stable at this low level since the 1980s [27]. How B. pinnipedialis hooded seal strain (HS) may affect the population dynamics is unknown, as we lack knowledge about the strains ability to establish infection and cause pathology. Entering macrophages after a transient initial bacteremia protects brucellae from antibodies and complement during dissemination in the host [28]. It seemed intriguing that the hooded seal strain could be isolated from multiple organs [26] without being able to escape an immune response by entering host cells. Recent studies have now confirmed that B. pinnipedialis HS is able to enter hooded seal alveolar macrophages, but is eliminated within 48 h post infection [29]. Furthermore, an age-dependent pattern of anti-Brucella antibodies, indicating exposure early in life with a subsequent clearance of infection, is identified in both hooded [30] and harbor seals [31].
To increase our general knowledge about the marine members of a bacterial genus that occurs at the animalhuman interface and to predict possible implications in human disease, extended information about intracellular entry, trafficking and multiplication of B. pinnipedialis in human cells should be obtained. The hooded seal strain is in whole genome analysis found to differ from the other marine brucellae with respect to genome size and GC content [32]. Knowing that the strains isolated from the human incidents also differ from the other marine Brucella spp., we aimed to study the infection biology of the hooded seal strain. The relationship between B. pinnipedialis HS and human macrophages was explored by adding an inhibitor study, as well as immunocytochemistry and transmission electron microscopy, to the conventional macrophage infection assays. Macrophage activation and cellular integrity following bacterial entry and elimination were investigated to further study the interaction of B. pinnipedialis with phagocytes and epithelial cells.  [26]; from now on entitled B17). The strains were kept at -80 °C on Microbank™ beads (Pro-Lab Diagnostics, Round Rock, USA). Before each assay a bead was taken from one specific Microbank™ batch and plated on sheep blood agar at 37 °C in a 5 % CO 2 atmosphere for 2 -4 days, the strains were thereafter plated again on sheep blood agar and grown at 37°C in a 5 % CO 2 atmosphere for 48 h for B. suis 1330 and 96 h for B. pinnipedialis. Dilutions of the bacteria in sterile PBS were used to prepare the final infective solution. We verified the expression of smooth surface antigens for the Microbank™ batches by crystal violet staining and agglutination with antiserum to smooth B. abortus [33,34]. The identity of the strains was verified before and after the gentamicin protection assay by PCR and gel electrophoresis with the marine mammal brucellae specific primers designed to amplify the bp26 gene [35] and the primer sets Bruce 11, 18 and 45 from the MLVA-15 assay [36,37].

Gentamicin protection assay
THP-1 cells were differentiated into macrophage-like cells by the use of PMA as previously described [38]. In brief, THP-1 cells were seeded (3.5 x 10 5 cells/well) in 24 well plates and cultured for 24 h in complete RPMI containing 50 ng/ml PMA. RAW264.7, J774A.1 and HeLa S3 cells were seeded (2 x 10 5 cells/well) in 24 well plates and cultured in the appropriate complete medium for 24 h prior to infection. The cells were maintained in their respective culture medium without penicillin/ streptomycin during the Brucella infection assays. The macrophage cell lines were challenged with Brucella spp. with a multiplicity of infection (MOI) of 50 and the HeLa cells with a MOI of 500 for 1 h at 37 °C, supplemented with 5 % CO 2 . The exposure was synchronized by centrifugation at 230 x g for 10 min in room temperature (RT). This stage was ended by first rinsing the wells twice with PBS, and refilling with 1 ml of complete culture medium containing 100 µg/ml gentamicin to kill extracellular bacteria. After 1 h the medium was replaced with complete culture medium containing 10 µg/ml gentamicin, and the cells were incubated for the desired period of time (1.5, 7, 24, 48, 72, and 96 h). Before harvesting of intracellular bacteria the cells were washed three times with PBS to remove remaining antibiotics. The cell membranes were disrupted using 300 µl/well of 0.1 % Triton X-100 in PBS followed by incubation at 37 °C for 10 min. Sterile cell scrapers were used to ensure complete detachment of all cells and the lysate was repeatedly pipetted to aid cell membrane disruption. TSA agar plates were inoculated in duplicate with 100 µl each of lysate in serial dilutions and evaluated for the presence of colony forming units (CFU)/well. Supernatants were also plated to control the efficiency of extracellular bacterial killing by gentamicin. The MOI applied was verified for each assay.

Immunocytochemistry
THP-1 and HeLa cells were seeded on glass coverslips in 12 well plates (2.0 -2.5 x 10 5 cells/well). The THP-1 cells were PMA-treated for differentiation. After 24 h the cells were challenged with B17 at a MOI of 50 (THP-1) or 500 (HeLa) as described in the gentamicin protection assay. Infected cells were incubated with 75 nM LysoTracker Red for 1 h prior to fixation. The cells were fixed for 15 min at RT using 4 % paraformaldehyde, 0.02 M sucrose, in PBS, pH 7.2 at desired time points following bacterial exposure and washed once in PBS before permabilization in 0.1 % Triton X-100 for 4 min. Immune labeling was done using rabbit polyclonal anti-Brucella spp. antibody, diluted at 1:100. Secondary antibody was Alexa Fluor 488 goat-anti-rabbit IgG, diluted at 1:500. The fluorescent DNA dye DRAQ5, diluted at 1:1000, was used for visualization of cell nuclei. Confocal microscopy was performed using a Zeiss LSM510 META system (Carl Zeiss, Obercochen, Germany) equipped with a 40X 1.2NA water immersion lens. Three sequential channels were recorded using the following excitation and emission parameters: Alexa 488 was excited at 488 nm and fluorescence collected through a 500 -550 nm BP filter; LysoTracker Red was excited at 543 nm and fluorescence collected through a 565 -615 nm BP filter; DRAQ5 was excited at 633 nm and fluorescence collected in the META detector from 644 -700 nm.

Immune transmission electron microscopy
THP-1 and HeLa S3 cells were seeded in 6 well plates at 3.5 x 10 5 and 1.5 x 10 5 cells/well, respectively, and allowed to grow for 48 h. The cells were challenged with B17 as described for the gentamicin protection assay and fixed at 1.5, 7 and 24 h post infection using 4 % formaldehyde, 0.1 % glutaraldehyde in 100 mM sodium phosphate buffer, pH 7.2. Samples were processed for cryoimmune electron microscopy as previously described [39]. Cryosections were labeled using rabbit polyclonal anti-Brucella antibody, diluted at 1:80. Positive labeling was detected by 10 nm protein A-gold complexes. The dried sections were examined in a JEOL JEM 1010 transmission electron microscope (JEOL, Tokyo, Japan) operating at 80 kV. Control experiments were routinely included in parallel by omission of the primary antibodies. Sections of fixed B17 functioned as positive control.

Inhibitor assays
Monolayers of PMA-treated THP-1 cells (3.5 x 10 5 cells/well) were cultured in 24 well plates and pre-incubated for 45 min at 37 °C in complete culture medium containing increasing concentrations of the cholesterol-scavenging drug β-methyl cyclodextrin, as previously described [40]. After inhibitor treatment, macrophages were washed once with complete culture medium and infected with Brucella spp. (MOI = 50). Bacterial uptake was determined using the gentamicin protection assay.

Cellular integrity and macrophage activation
Potential toxic cell damage due to inhibitor treatment or Brucella spp. infection was measured by quantitatively determining the release of lactate dehydrogenase using the CytoTox 96 Non-Radioactive Cytotoxicity Assay. Production of nitric oxide (NO) following infection, indicating macrophage activation, was assessed using the Griess Reagent System. Both assays were performed according to the manufacturer's instructions. Absorbances were read using an Epoch Microplate Spectrophotometer (BioTek, Winooski, USA). LPS from Escherichia coli (O111:B4) at 0.5, 1, and 10 µg/ml was used as a positive control for NO measurements.

Statistical analysis
Group data were compared using Student's t test for independent samples. Differences were considered significant for p values of < 0.05.

Results
Verification of the identity of the strains PCR and gel electrophoresis showed bands of the expected size for all strains used in the study [35][36][37] (data not shown). All strains expressed smooth surface antigens before and after execution of the gentamicin protection assay.

Brucella pinnipedialis HS enters macrophages and epithelial cells in vitro
The results from the gentamicin protection assay revealed that both B. pinnipedialis strains (12890 and B17) were able to enter human macrophage-like cells in culture (THP-1) ( Figure  1A). Challenging the cells with a MOI of 50 lead to the retrieval of 3.92 log CFU for B17 after 1.5 h, compared to 4.08 log CFU retrieval for the reference strain 12890. B17 were also taken up in two murine macrophage cell lines (RAW264.7 and J774A.1) with the retrieval of 4.28 log CFU and 3.50 log CFU at the 1.5 h time interval respectively (Figures S1 and S2). A known pathogenic strain; B. suis 1330, was included as a control for the ability of the macrophage cell types to support internalization as well as intracellular multiplication. While being challenged with the same MOI of 50, the amount of protected bacteria retrieved for B. suis 1330 after 1.5 h pi was 3.35 log CFU and 3.24 log CFU for THP-1 cells ( Figure 1A) and RAW264.7 ( Figure S1) respectively. Additionally, both 12890 and B17 entered HeLa cells. Challenging the cells with a MOI of 500 lead to a retrieval of 4.60 and 4.65 log CFU at 1.5 h pi respectively ( Figure 1B). No bacterial growth was observed from the plated supernatant controls. The intracellular localization of B17 in THP-1 and HeLa cells was confirmed using confocal microscopy (Figure 2A and B, video S1). Immune labeling with a rabbit anti-Brucella antibody and Lysotracker Red revealed colocalization of intracellular bacteria with lysosomes at both 1.5 and 24 h pi. Brucella bacteria were observed in 37 % of the THP-1 cells at 1.5 h pi; 67 % of the infected cells had one bacteria, 15 % two bacteria, while 18 % had more than two bacteria (total THP-1 cells, n = 498). The corresponding numbers in HeLa cells were 61 % infected cells; 40 % had one bacteria, 21 % two bacteria, while 39 % had more than two bacteria (total HeLa cells, n = 151). Immune electron microscopy further verified intracellular presence of B17 in THP-1 and HeLa cells (Figure 3).

Brucella pinnipedialis HS is gradually eliminated from cultured macrophages and epithelial cells
Following cellular entry during infection in THP-1 cells, B17 showed a large drop in intracellular bacterial numbers with one logarithmic decrease at 7 h (3.09 log CFU) and additionally one logarithmic reduction at 24 h (1.92 log CFU) pi ( Figure 1A). At 72 and 96 h pi bacteria were still present intracellulary in some wells, but were completely eliminated in others, dividing the results in two groups. At 72 h pi, three of total nine wells displayed elevated bacterial numbers with mean 2.98 log CFU in the elevated group versus mean 1.48 log CFU in the low group. At 96 h pi, three of total nine well were elevated with mean 3.12 log CFU versus mean 0.57 log CFU. Although the total mean log CFU decreased from 24 to 96 h pi, this effect caused a high standard deviation for the 72 and 96 h time  points. In murine macrophages, intracellular B17 was completely eliminated by 48 h (J774A.1) and 72 h (RAW264.7) pi ( Figures S2 and S1). The variability between wells at later stages of the infection was not observed in these latter cell lines. Contrary to published data [24], strain 12890 did not multiply intracellulary in THP-1 cells and displayed the same pattern as B17 but with less variability between wells at 72 h pi ( Figure 1A).
In HeLa cells, the intracellular bacterial numbers retrieved at 24 h pi were lowered with almost two logarithmic decreases for B17 (2.87 log CFU) and more than two logarithmic decreases for 12890 (1.93 log CFU) compared to the initial numbers at 1.5 h pi ( Figure 1B). From 48 h pi the total mean log CFU retrieved slowly decreased for B17, however some wells had higher retrievable bacterial counts at 72 h (three of total nine; mean 2.65 log CFU versus mean 1.4 log CFU) and 96 h (one of total nine; 3.28 log CFU versus mean 0.8 log CFU) pi compared to others. For 12890, the total mean intracellular bacterial counts were similar from 24 h to 96 h pi, but due to inter well differences the standard deviations are large. At 72 h pi, six of total nine wells still displayed elevated bacterial numbers with mean 3.12 log CFU in the elevated group versus mean 0.06 log CFU in the low group. At 96 h pi, still five of total nine wells were elevated with mean 3.41 log CFU versus mean 0.09 log CFU.
Although B17 was statistically different from 12890 at various time points during infection (7, 24, and 48 h pi in the THP-1 cell line and 24 h pi in HeLa cells), the outcome of infection was similar. Brucella suis 1330, serving as a positive control, showed the classical infection pattern with a slight drop in CFU at 7 h, followed by high intracellular multiplication from 24 h reaching 6.1 log CFU after 72 h in the THP-1 cell line ( Figure  1A). Intracellular multiplication of B. suis 1330 was also confirmed in murine macrophages (RAW264.7) ( Figure S1).

Intracellular B. pinnipedialis do not cause macrophage NO-production or increased cell death
Nitrite, as an indicator of NO-production, was not detected to be released from murine macrophages (RAW264.7) challenged with B. pinnipedialis reference strain (12890) or HS (B17) at 1.5, 24 or 48 h pi. Measurements of nitrite were not found applicable in the THP-1 cell line, this in line with previous reports [41]. No release of lactate dehydrogenase was detected following infection with either 12890 or B17, suggesting minimal cytotoxicity induced by intracellular bacterial localization and degradation (results not shown).

Internalization of B. pinnipedialis HS depends on lipid raft-associated mechanisms
Marine mammal brucellae express a smooth type lipopolysaccharide (LPS) [42,43]. To evaluate if entry of B17 into human macrophages was associated with lipid raft components as reported for virulent terrestrial smooth Brucella spp. [44][45][46][47], THP-1 cells were pretreated with a cholesterolscavenging drug (β-methyl cyclodextrin) before carrying out the gentamicin protection assay. Brucella suis 1330 was incorporated as a positive control showing more than 90 % reduction in retrieved CFU in the two highest concentrations of inhibitor (10 and 5 mM), and a 87 % reduction in the 2.5 mM concentration compared to untreated exposed cells measured after 1.5 h (Figure 4). Entrance of B17 was reduced with 85, 73, and 60 % respectively (Figure 4). No release of lactate dehydrogenase was detected following inhibitor treatment of the cells, suggesting minimal cytotoxicity induced by this procedure.

Discussion
This work reports on the ability of B. pinnipedialis HS (B17) to enter human and murine macrophages, as well as an epithelial cell line in vitro. As these findings contrast the previous report regarding the behavior of hooded seal strains in human macrophages [24], our results from the gentamicin protection assays were verified by confocal and transmission immune electron microscopy. In line with the preceding report [24], B. pinnipedialis reference strain (12890) also entered human macrophages and HeLa cells, however, divergent to the prior report [24], the strain did not multiply intracellulary. When further exploring the interaction of B. pinnipedialis with phagocytes and epithelial cells, we revealed that although this marine mammal brucellae behave dissimilarly to pathogen terrestrial brucellae (lack of intracellular multiplication), it cannot be compared to engineered rough mutants with respect to macrophage entry and activation, or induction of cell death.
Virulent smooth Brucella spp. as well as the naturally rough species internalize into macrophages via cholesterol-rich lipid rafts, including the raft-associated class A scavenger receptor (SR-A) [40,[44][45][46]. This stealthy entry allows pathogenic Brucella spp. to limit early fusion with the endosome-lysosome pathway. Although internalized in low numbers, this enables the bacteria to activate their specific virulence genes involved in building a definite replication niche [48]. In order to gain more information regarding the possible pathogenicity of B17, a property that seems to be influenced by the mode of entry along with other bacterial traits, we explored if this marine brucellae has the capacity of stealthy entry. To this end we investigated the mode of B17 cellular entry mechanism by blocking lipid rafts. The number of bacteria entering inhibitortreated cells was reduced with 85 % at the most. This was significantly different from untreated cells, but not as pronounced as for B. suis 1330 (more than 90 % reduction of bacterial cell entry). These results suggest that B17 can enter human macrophage-like cells via lipid raft-as well as non-lipid raft-associated mechanisms. Entering via lipid raft-associated mechanisms would give B17 an opportunity to activate possible virulence genes; however, the lack of such genes and/or cellular activation caused by other entry mechanisms may disrupt reaching or initiation of the replicative niche [49].
Although probably not the only determinant, bacterial mechanisms of cellular entry have been linked to the particular composition of LPS as different internalization mechanisms are observed for virulent smooth Brucella spp. and engineered rough mutants [48,50]. Recently, LPS from the smooth B. suis biovar 2 has been shown to lack some O-antigen epitopes previously thought to be present in all smooth brucellae [51,52]. The marine mammal Brucella spp. express a smooth-type LPS that show heterogeneity with regard to their O-polysaccharide common epitope content and average size [42,43]. B17 was previously shown to differ from the other marine mammal brucellae in whole genome analysis [32]. If structural differences in smooth strain LPS affect entry mechanisms, this could be a possible explanation for the variances observed between B. suis 1330 and B17 in the inhibitor assay.
Compared to the pathogenic strains known to be the cause of human brucellosis (B. melitensis/B. abortus/B. suis), both B17 and 12890 entered the cell lines in higher numbers (range log CFU 3.92 -4.65) than observed for B. suis (log CFU 3.35) in THP-1 cells. The number of intracellular bacteria is influenced by their mode of entry. In murine macrophages, B. abortus engineered rough mutants are showing a higher rate of infection compared to wild type B. abortus, both regarding the number of cells infected as well as the number of bacteria detected/cell [50]. This may be due to a more unspecific entry mechanism, as rough mutants are shown to enter via non-lipid raft associated mechanisms [48,50]. These rough mutants induce a strong degree of macrophage activation following infection, such as activation of NF-κB, increased expression and production of cytokines and inducible nitric oxide synthase (iNOS); an event that may lead to elimination of intracellular bacteria [47,49,53]. Additionally, multiplication of engineered rough mutants is attenuated due to their cytopathic effect on the macrophages, leading to cellular lysis and release of the bacteria, and not only due to intracellular killing as previously interpreted [50,54]. Confocal microscopy of THP-1 cells infected with B17 revealed a moderate rate of infected cells (37 %) at 1.5 h pi. Although certain individual cells were heavily loaded (> 10 bacteria/cell), only 18 % of the infected cells harbored more than 2 bacteria. This pattern resembles what is previously shown for wild type B. abortus [50], with a slightly higher rate of total infected cells. Along with the results from the inhibitor study, this supports the hypothesis of B17 entering the cells in similar manners as pathogenic terrestrial brucellae. The observed reduction of intracellular bacteria in the later stages after exposure for B. pinnipedialis HS is believed to be due to cellular activation and subsequent bacterial elimination, and not cell death. Morphological changes in the cell culture following exposure were not observed microscopically and release of lactate dehydrogenase was not detected following infection of the cells, suggesting minimal cytotoxicity induced by intracellular entry and degradation of bacteria. On the other hand, production of NO as a consequence of cell activation following infection with B. pinnipedialis HS in murine macrophages could not be detected. This in line with previous reports of B. suis shown not to induce expression of iNOS nor release of NO from murine macrophages [49]. The cellular response triggered by B. pinnipedialis is still unidentified, and future investigations including a broader panel of cytokines are suggested.
In vitro, virulent Brucella spp. ensure their survival by overriding the intracellular response following bacterial infection. By forming the Brucella-containing vacuole they are able to circumvent normal lysosomal destruction and facilitate intracellular proliferation [55][56][57]. This is demonstrated in B. suis infected THP-1 cells where the initial log value at 1.5 h (3.35 log CFU) was doubled by 48 h of infection (6.01 log CFU). Both B17 and 12890 were reduced quite rapidly the first 24 h after exposure in both macrophages and epithelial cells. Colocalization of B17 with lysosomal compartments was observed at both 1.5 and 24 h pi in THP-1 cells, weakly at 1.5 h and more pronounced at 24 h pi in HeLa cells. Additionally, at 24 h pi bacterial debris appears scattered throughout the infected cells indicating degradation of bacteria, as previously reported for an attenuated B. abortus strain [22,58]. Both findings support the observed reduction of intracellular bacteria retrieved at the later stages after infection.
Although the infection dynamics of both B17 and 12890 displays a declining mean CFU, in some wells a higher number of intracellular bacteria were retrieved at 72 and 96 h than the corresponding numbers at 24 and 48 h pi. This shows that not all bacteria were eliminated equally efficiently, suggesting that in heavily bacterial-loaded cells a fraction of the intracellular bacteria may escape the lysosomal degradation process in order to proliferate intracellulary, as pointed out by others [59]. The observed divergence between wells was only noted in the THP-1 and HeLa cell lines, and not in the murine macrophages. This could reflect the bactericidal efficiency of the cell type, as HeLa cells are not professional phagocytes and PMA-induced THP-1 cells show increased macrophage maturation 5 days post treatment [60]. In this work, the gentamicin protection assay was initiated 24 h after PMAtreatment. Furthermore, activation of macrophages by either phorbol ester or 1.25-dihydroxyvitamin D 3 is different [60,61] and caution should prevail when comparing results.
A relationship between MOI and a possible bacterial overload was not observed, as MOI 300, which was tested in the THP-1 cell line (results not shown), did not increase inter well differences compared to MOI 50, which was used for THP-1 cells throughout this study.
In conclusion, this work shows that B. pinnipedialis HS is able to enter macrophages and HeLa cells in vitro. The entry mechanism in macrophages involves, but seems not to be limited to, membrane lipid rafts. Although able to survive for up to 96 h intracellulary, none of the strains were able to multiply at levels comparable to B. suis 1330 in the cell lines tested. A lack of/or low capacity to replicate and survive for prolonged periods within host cells, particularly macrophages, abolishes the ability to produce chronic infections as reported for terrestrial pathogenic brucellae [28]. The observed elevated bacterial numbers in some wells, most pronounced in HeLa cells, could imply that B. pinnipedialis may cause acute, transient infections in some cell types. An increased rate of survival in cell lines with reduced bactericidal competence gives room to speculate if changes in the cellular immune competence might support different infection dynamics.
In the Arctic region seal meat and other raw nutriments from marine mammals are consumed and for indigenous people it constitutes a large part of the diet. Consumption of raw bone marrow from reindeer has been documented to be a source of B. suis biovar 4 contamination in Alaska [62]. Reports of brucellosis in this area are sparse [63], however there may be an underreporting due to the unspecific symptoms of human brucellosis and non-diagnosed incidents could possibly occur [64]. In light of the close relationship to the marine environment and exposure to marine species/products, either occupational, or by lifestyle, further understanding of the pathogenesis of marine mammal brucellae in target species, as well as humans, is warranted. Although the results presented here indicate a low zoonotic potential of the strains tested, variations between strains as well as host immunity may influence the outcome of infection. Future research should aim to identify potential virulent traits in marine mammal brucellae that may have implications in the establishment of disease in marine mammals and humans. Figure S1. Infection dynamics of Brucella spp. in murine macrophages. RAW264.7 cells were challenged with B. suis 1330 and B. pinnipedialis HS (B17) in a gentamicin protection assay as described in materials and methods. B17 entered murine macrophages but was not able to multiply. Intracellular bacteria were eliminated within 72 h pi. Brucella suis 1330 showed the classical infection pattern with intracellular entry, a slight drop in intracellular bacterial numbers at 7 h pi and multiplication from 24 h pi. Error bars correspond to the standard error. Each indicator represents the mean of four replicate wells from one assay (B. suis 1330) or six replicate wells from two separate assays (B17). (TIF) Figure S2. Infection dynamics of B. pinnipedialis HS in murine macrophages. J774A.1 cells were challenged with B. pinnipedialis HS (B17) in a gentamicin protection assay as described in materials and methods. B17 entered murine macrophages but was not able to multiply. Intracellular bacteria were eliminated within 48 h pi. Error bars correspond to the standard error. Each indicator represents the mean of three replicate wells from one representative assay. (TIF) Video S1.

Supporting Information
Brucella pinnipedialis HS in human macrophages. THP-1 cells were cultured on glass coverslips and challenged with B. pinnipedialis HS (B17) in a gentamicin protection assay as described in the materials and method section. Cells were incubated with LysoTracker Red (red) for 1 h before fixed at 24 h after exposure and immune labeled with anti-Brucella antibody 1:100, and Alexa 488 goat-anti-rabbit 1:500 (green). DRAQ5 was used for visualization of the cell nuclei (blue). Three-dimensional animation of a confocal zstack spanning the central volume of an infected macrophage, showing colocalization between bacteria and lysosomes at 24 h pi. Bacterial cells are situated at different levels intracellulary, many near the THP-1 nucleus. The animation was created in Volocity ver. 6.2.1 (Perkin-Elmer). (MOV)