**3. Discussion**

Physiological and pathological immune responses are a continuum in which platelets are recognized as innate immune effector cells. Their activation stimulates interactions with endothelial cells and myeloid leukocytes in many pathologic inflammatory syndromes, as well as consequences in acute inflammation [28–31]. Platelets also have signaling functions in endothelial cells. These functions also contribute to critical inflammatory and immune responses [32,33].

Brain microvascular endothelium activation and BBB dysfunction is a significant contributor to the pathogenesis of a variety of brain pathologies [2], many of them of microbial origin [9,18,34]. We have previously described the ability of *B. abortus*to induce inflammation in the cerebral parenchyma, which leads to the activation of the endothelial cells that form the BBB [9]. In this study, we elucidated the role of platelets in brain microvascular endothelial cell activation mediated by *B. abortus*. Platelets enhance HBMEC activation in the context of *B. abortus*infection. These results correlate with the reported ability of other bacterial species to activate platelets and harm endothelial cells [35,36].

Interestingly, HBMEC activation does not require direct contact between platelets and brain endothelial cells, since supernatants of *B. abortus*-stimulated platelets recapitulated the HBMEC activation observed in the presence of platelets. Furthermore, HBMEC activation by secreted factors from *B. abortus*-stimulated platelets is su fficient to induce transmigration of both monocytes and neutrophils. Moreover, *B. abortus*-stimulated platelets also activate the HMEC-1 cell line and primary culture of HUVEC, underscoring the ability of *B. abortus*-stimulated platelets to activate any endothelium.

A long time ago, it was demonstrated that activated platelets increase CCL-2 secretion and ICAM-1 expression on HUVECs [37]. This indicates that activated platelets are able to change the chemotactic and adhesive properties of endothelial cells, increasing the ability to attract monocytes and neutrophils. Under physiological conditions, endothelial cells of the vasculature of non-inflamed tissues have as main functions the maintenance of blood fluidity and the control of vascular permeability [33]. Under these conditions, resting endothelial cells do not interact with circulating leukocytes since the proteins necessary for this interaction are mainly retained inside the cell [38]. Under acute inflammatory conditions, such as those induced by *B. abortus* infection, the vascular endothelium is rapidly activated, mobilizing these adhesion molecules to the extracellular membrane [33]. In accordance with this, we demonstrated that, although the infection with *B. abortus* induces a mild activation of HBMEC, HMEC-1, and HUVEC, the presence of platelets during the infection enhances its activation state upregulating the expression of ICAM-1 and CD40, thus stressing the amplifying role of platelets on endothelial inflammation [39]. In line with these results, other authors have shown that the presence of activated platelets significantly induces the expression of E-Selectin (CD62E), CD106 (VCAM-1), and ICAM-1 on the surface of HUVEC cells, even in the absence of others inflammatory agents [40]. In addition to the increase in adhesion molecules, we have demonstrated that the activation of HBMECs by supernatants from *B. abortus*-stimulated platelets increase IL-6, IL-8, and CCL-2 secretion. These results are in agreemen<sup>t</sup> with those previously published describing that HUVEC secrete IL-8 and CCL-2 after co-incubation with activated platelets [40]. In turn, in vivo experiments have shown that platelets are one of the first cellular components arrested in the inflamed endothelium, promoting their activation and allowing the subsequent arrest of leukocytes [41].

Platelet activation was also induced by exposure to heat-killed *B. abortus*, which indicated that it was not dependent on bacterial viability and suggests that it was elicited by a structural bacterial component. Our laboratory has been investigating for years the role of lipoproteins in inflammation generated by *Brucella*. We have described that *Brucella* LPS does not produce cellular activation, however, *Brucella* lipoproteins produce activation of several cell types [6,8,25,26]. Thus, we hypothesized that *B. abortus* lipoproteins might be the structural components involved in the observed phenomenon. L-Omp19, a prototypical *B. abortus* lipoprotein, recapitulated platelet stimulation and concomitant HBMEC activation. Acylation of Omp19 was required for its biological activity since U-Omp19 had no e ffect on platelet stimulation. The genome of *B. abortus* possesses no less than 80 genes encoding putative lipoproteins [42], and many of them are expressed in the outer membrane of the bacterium [43]. In this context, we posit that any surface-exposed *Brucella* lipoprotein may be significant beyond in vitro assays and not one lipoprotein but rather a combination of them may contribute to the platelet activation elicited by *B. abortus*.

Our recent work revealed a physiological mechanism employed by *B. abortus* to traverse the BBB. *Brucella* is incapable of traversing the BBB by itself, despite the ability to invade and replicate in

endothelial cells of the brain microvasculature. Instead, it could cross a BBB model in vitro as a consequence of naturally migrating monocytes carrying viable bacteria, which serve as source of de novo infection to astrocytes and microglia [5]. Interestingly, we have also demonstrated that activated *B. abortus*-infected glial cells were able to increase the transmigration of monocytes through the secretion of inflammatory mediators [9]. These mediators would escalate the entering of infected cells from the peripheral circulation, increasing the infection and the subsequent BBB dysfunction through a pathological vicious circle. The capacity of secreted factors from *B. abortus*-stimulated platelets to increase neutrophil and monocyte transmigration through microvascular endothelial cells demonstrated in this paper would worsen this situation (Figure 8).

**Figure 8.** *B. abortus*-stimulated platelets (1) secret factors (2) that induce HBMEC activation, leading to ICAM-1 and CD40 upregulation, increasing the secretion of IL-6, IL-8, and CCL-2 (3), and promoting neutrophils and monocytes to traverse a polarized HMBEC monolayers (4). Platelet-induced activation would escalate the entering of infected cells from the peripheral circulation and the subsequent infection of glial cells (5), worsening the inflammatory signs of neurobrucellosis.

The mitogen-activated protein kinase (MAPK) pathway has been associated to several biological processes such as cell activation and proliferation, cell differentiation, and apoptosis [44]. In particular, the Erk1/2 pathway is involved in HBMEC activation [27] and endothelial permeability [45]. Experiments of pharmacological inhibition determined that Erk1/2 was involved in HBMEC activation induced by supernatants from *B. abortus*-activated platelets. In particular, it was involved in ICAM-1 upregulation and enhanced the transmigration of neutrophils. Since MAPK inhibitors, such as pyridinyl imidazole drugs, have been identified as putative drugs for anti-inflammatory therapies in the CNS [46], the data presented in this paper sugges<sup>t</sup> that inhibiting such molecules (Erk1/2) may represent a pharmaceutical strategy to restrict BBB deterioration, thereby potentially reducing the morbidity associated with neurobrucellosis.

In summary, the results presented here describe a mechanism whereby *B. abortus*-stimulated platelets can induce HBMEC and other endothelial cell activation, promoting neutrophils and monocytes to traverse the BBB. Moreover, this could contribute to increase the infection of glial cells, generating and/or deteriorating neurobrucellosis and the inflammatory response motivated by glial activation (Figure 8).

### **4. Materials and Methods**

### *4.1. Ethics Statement*

Human platelets, monocytes, and neutrophils were isolated from the blood of healthy adult donors in agreemen<sup>t</sup> with the guidelines of the Ethical Committee of the Instituto de Medicina Experimental (protocol number: 20160518-M). All adult blood donors provided their informed consent prior to the study.

### *4.2. Bacteria and Lipoproteins*

*B. abortus* S2308 was cultured in tryptic soy broth supplemented with yeas<sup>t</sup> extract (Merck, Buenos Aires, Argentina). The number of bacteria on stationary-phase cultures was determined by comparing the optical density at 600 nm with a standard curve. All live *Brucella* manipulations were performed in biosafety level 3 facilities located at the Instituto de Investigaciones Biomédicas en Retrovirus y SIDA (INBIRS, Buenos Aires, Argentina). To obtain heat-killed *B. abortus* (HKBA), bacteria were washed five times for 10 min each in sterile phosphate-buffered saline (PBS), heat-killed at 70 ◦C for 20 min, aliquoted, and stored at −70 ◦C until used. The total absence of *B. abortus* viability after heat killing was verified by the absence of bacterial growth on tryptic soy agar.

*B. abortus* lipidated-outer membrane protein 19 (L-Omp19) and unlipidated-Omp19 (U-Omp19) were obtained as described [31]. Both recombinant proteins contained less than 0.25 endotoxin U/μg of protein as assessed by Limulus Amebocyte Lysates (Associates of Cape Cod Inc., Falmouth, MA, USA).

### *4.3. Cell Lines*

HBMECs were isolated from a brain biopsy of an adult female with epilepsy as previously described [47]. These cells were positive for factor VIII–Rag, carbonic anhydrase IV, and Ulex europaeus agglutinin I. They took up fluorescently labeled low-density lipoprotein and expressed g-glutamyl transpeptidase, thus demonstrating their brain endothelial cell properties [47]. HBMECs were subsequently immortalized by transfection with SV40 large T Ag and maintained their morphological and functional characteristics for at least 30 passages [48]. The cells are polarized and exhibit a transendothelial electric resistance (TEER) of at least 100 ohms/cm<sup>2</sup> [49]. Cells (passage < 30) were cultured in tissue culture flasks in Roswell Park Memorial Institute (RPMI) medium 1640 (Life Technologies, Grand Island, NE, USA) supplemented 10% with heat-inactivated fetal bovine serum (FBS) (Life Technologies), 10% NuSerum IV (Becton Dickinson, Bedford, OH, USA), 1% modified Eagle's medium nonessential amino acids (Life Technologies), sodium pyruvate (1 mM), L-glutamine (2 mM), 1% MEM vitamin solution (Life Technologies), penicillin (100 U/mL) and streptomycin (100 μg/mL). Human microvascular endothelial cells (HMEC-1) were obtained from ATCC® (CRL-3243™, Manassas, VA, USA). Cells were grown in Dulbeccos Modified Eagle's (DMEM) medium (Life Technologies) containing 10% FBS (Natocor, Córdoba, Argentina), 10 μg/mL hydrocortisone, 1 ng/mL epidermal growth factor (BD Pharmingen, San Diego, CA, USA), L-glutamine (2 mM), penicillin (100 U/mL), and streptomycin (100 μg/mL). All cell cultures were incubated at 37 ◦C in a humidified atmosphere of 5% CO2. Human umbilical vascular endothelial cells (HUVECs) were obtained as described previously [10,50]. Briefly, umbilical vascular tissue was treated with collagenase for digestion. Cells were seeded until confluence on 1% gelatin-coated 25 cm<sup>2</sup> tissue culture flasks and identified by their cobblestone morphology and von Willebrand factor (VWF) antibody (Immunotech, Ocala, FL, USA) binding. Cells were grown in RPMI 1640 medium (Gibco) supplemented with 10% FBS (Life Technologies), heparin (100 μg/mL), endothelial cell growth factor (50 μg/mL), sodium pyruvate (2 mM), L-glutamine (2 mM), penicillin (100 U/mL), and streptomycin (100 μg/mL) at 37 ◦C

in a humidified 5% CO2 incubator. HUVECs used for experiments were kept between the first and third culture passage.

### *4.4. Platelet Purification and Stimulation*

Platelets were obtained from whole blood from healthy adult human donors as described previously [14]. Briefly, blood samples were collected into tubes containing sodium citrate (Merck) and centrifuged. The platelet-rich plasma was collected and centrifuged in presence of 75 nM prostaglandin I2 (Cayman Chemical, Ann Arbor, MI, USA). Platelets were then washed with RPMI 1640 medium. Finally, platelets were resuspended in RPMI 1640 medium. Platelets were incubated with *B. abortus* (1 × 10<sup>7</sup>/mL) (PLT:B.a. ratio of 1:1) for 24 h in RPMI 1640 medium with 10% FBS (Life Technologies) and L-glutamine (2 mM). In addition, platelets were incubated with HKBA (1 × 10<sup>8</sup> bacteria/mL), L-Omp19, U-Omp19 (both 500 ng/mL), or thrombin (0.1 U/mL) (Sigma Aldrich, St. Louis, MO, USA). Then, supernatants were collected, sterilized by filtration, ultracentrifuged when mentioned (at 100,000 × *g* for 5 h at 4 ◦C), and stored at −70 ◦C until they were used.

### *4.5. Endothelial Cell Treatment*

HBMEC, HMEC-1, and HUVEC were cultured in 48 wells plate (5 × 10<sup>4</sup>/0.2 mL). To co-culture infection, platelets were added (cell:platelets ratio, 1:100) and endothelial cells–platelets cultures were infected by *B. abortus* (multiplicity of infection of 100). In all cases, the infection was performed for 2 h in medium containing no antibiotics. Then, cells were maintained for 24 h in the presence of antibiotics (100 μg/mL gentamicin and 50 μg/mL streptomycin) to kill the remaining extracellular bacteria. For experiments with platelet-conditioned media, HBMEC, HUVEC, and HMEC-1 cells were treated with 0.2 mL of diluted supernatants from *B. abortus*-stimulated platelets for 24 h. Culture supernatants from *Brucella*-infected astrocytes and recombinant human IL-1β were used as control. In all cases, cells were harvested to determine cell surface molecule expression by flow cytometry. Supernatants from stimulated endothelial cells were collected and stored at −70 ◦C until they were used.

### *4.6. Erk1*/*2 Signaling Pathway*

HBMECs were treated with Erk1/2MAPK pharmacological inhibitor PD98059 (50μM) (Calbiochem, San Diego, CA, USA) or vehicle (dimethyl sulfoxide) 2 h before the stimulation with supernatants and the inhibitor were kept throughout the experiment, based on previous report [7].

### *4.7. Measurement of Cytokine and Chemokine Concentrations*

Human IL-6, IL-8, and CCL-2 concentrations were quantified in supernatants harvested from HBMECs and HMEC-1 treated with supernatants from *B. abortus*-stimulated platelets by Sandwich ELISA using paired cytokine-specific mAbs according to the manufacturer's instructions (BD Pharmingen).

### *4.8. Determination of Cell Surface Molecules by Flow Cytometry*

ICAM-1 and CD40 surface expression was determined by flow cytometry. For this, treated HBMECs or HMEC-1 were washed and stained with a PE-labeled antibody (Ab) against human ICAM-1 (CD54) (clone HA58, BD Pharmingen), PE-labeled Ab against human CD40 (clone 5C3; BioLegend, San Diego, CA, USA) or the PE-labeled isotype-matched control Ab (BD Pharmingen). Labeled cells were analyzed on a FACSCalibur flow cytometer (BD Biosciences, San Diego, CA, USA), and data were processed using FlowJo software.

### *4.9. Neutrophil and Monocytes Transendothelial Migration Assay*

Peripheral blood mononuclear cells (PBMCs) and neutrophils were separated by Ficoll-Hypaque (GE Healthcare, Uppsala, Sweden) gradient centrifugation. Human neutrophils were isolated by sedimentation of erythrocytes in 6% dextran and hypotonic lysis as previously described [26]. Monocytes were then purified from PBMCs by Percoll (GE Healthcare) gradient. Both types of cells were resuspended in RPMI 1640 supplemented with 10% FBS. Cell purity was 90% as determined by flow cytometry for both populations. Viability of cells was more than 95% in all the experiments as measured by trypan blue exclusion test.

HBMEC monolayers were established from 20,000 cells per insert on 3-μm pore size membrane Transwell plates of 6.5-mm diameter insert (Corning-Costar, Acton, MA, USA) previously treated with rat tail collagen (50 mg/mL in 1% acetic acid) (BD Biosciences) and neutralized in a saturated atmosphere of ammonium hydroxide. After 5 days, when cellular confluence was reached TEER and passive diffusion of horseradish peroxidase was measured as an indication of monolayer integrity [5]. Then, monolayers were incubated for 24 h with supernatants from *B. abortus*-stimulated platelets. Supernatants from platelets alone as well as non-treated HBMECs were used as negative control. Culture supernatants from *Brucella*-infected astrocytes and recombinant human IL-1β were used as positive control. After that, monolayers were washed and neutrophils or monocytes (1 × 10<sup>5</sup> cells) were added to the upper chamber in fresh medium. Plates were incubated for 3 h at 37 ◦C in 5% CO2 and transmigrated cells to the lower chamber were counted on a hemocytometer.

### *4.10. Statistical Analysis*

Results were analyzed with one-way ANOVA followed by Tukey post-test using the GraphPad Prism 5.0 software.

**Author Contributions:** A.M.R., A.T., P.B., and G.H.G. conceived and designed the experiments. A.M.R., A.T., A.P.M., M.C.M., and M.V.D. performed the experiments. K.S.K. supported the work with key suggestions and helped with data interpretation. A.M.R. and G.H.G. analyzed the data and wrote the manuscript. G.H.G. supervised experiments, interpreted the data, and supervised the manuscript. All authors reviewed the manuscript. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by grants from the Agencia Nacional de Promoción Científica y Tecnológica (ANPCYT-Argentina) [PICT 2014-1111, 2014-1925, 2015-0316, 2016-0356, 2016-1493, 2017-1393, 2017-1905], by grants from CONICET (Argentina) [PIP 0200], by gran<sup>t</sup> UBACYT from University of Buenos Aires [20020170100320BA] (Argentina).

**Acknowledgments:** We thank Horacio Salomón and the staff of the Instituto de Investigaciones Biomédicas en Retrovirus y SIDA (Universidad de Buenos Aires) for allowing us to use the biosafety level 3 laboratory facilities. A.M.R., M.V.D., P.B., and G.H.G. are members of the Research Career of CONICET (Argentina). A.T. and M.C.M. are recipients of fellowships from Consejo Nacional de Investigaciones Científicas y Técnicas (Argentina).

**Conflicts of Interest:** The authors declare no conflict of interest.
