Effects of Pseudomonas aeruginosa on Microglial-Derived Extracellular Vesicle Biogenesis and Composition

The packaging of molecular constituents inside extracellular vesicles (EVs) allows them to participate in intercellular communication and the transfer of biological molecules, however the role of EVs during bacterial infection is poorly understood. The goal of this study was to examine the effects of Pseudomonas aeruginosa (P. aeruginosa) infection on the biogenesis and composition of EVs derived from the mouse microglia cell line, BV-2. BV-2 cells were cultured in exosome-free media and infected with 0, 1.3 × 104, or 2.6 × 104 colony forming units per milliliter P. aeruginosa for 72 h. The results indicated that compared with the control group, BV-2 cell viability significantly decreased after P. aeruginosa infection and BV-2-derived EVs concentration decreased significantly in the P. aeruginosa-infected group. P. aeruginosa infection significantly decreased chemokine ligand 4 messenger RNA in BV-2-derived infected EVs, compared with the control group (p ≤ 0.05). This study also revealed that heat shock protein 70 (p ≤ 0.05) and heat shock protein 90β (p ≤ 0.001) levels of expression within EVs increased after P. aeruginosa infection. EV treatment with EVs derived from P. aeruginosa infection reduced cell viability of BV-2 cells. P. aeruginosa infection alters the expression of specific proteins and mRNA in EVs. Our study suggests that P. aeruginosa infection modulates EV biogenesis and composition, which may influence bacterial pathogenesis and infection.


Introduction
Pseudomonas aeruginosa (P. aeruginosa) is a Gram-negative, opportunistic pathogen that contributes to chronic airway infections in cystic fibrosis patients [1]. Moreover, P. aeruginosa infections have been

Pseudomonas aeruginosa Strain
The P. aeruginosa laboratory strain PAO1 was generously gifted by Dr. Jessica Scoffield (University of Alabama at Birmingham) [15]. Pseudomonas isolation agar and Luria-Bertani broth were routinely used to culture PAO1 at 37 • C.

Pseudomonas aeruginosa Infection on Microglial Cells
BV-2 cells (500,000) were seeded in T-25 flasks (Corning) and infected with 0 (control; no infection) and 2.6 × 10 4 CFU/mL P. aeruginosa at 0.1 optical density (OD) in RPMI-1640 media with exosome-free FBS. Bacterial cells were prepared from an overnight culture and then subcultured to 0.1 OD. The bacterial pellet was obtained after centrifugation at 14,000 rpm for 10 min and was resuspended in RMPI-1640 medium. The culture medium was collected at 72 h (i.e., the experimental time point).

Cell Viability by Trypan Blue Exclusion Assay
BV-2 cells were examined for viability after control (no infection) or infection with 2.6 × 10 4 CFU/mL P. aeruginosa (0.1 OD) at 72 h. All media was removed. The cells were gently scraped with a cell scraper and non-sterile 1× phosphate buffered saline (PBS) solution (1.5 mL). The solution was collected and gently mixed in 2.0-mL Eppendorf tubes. The cells were then counted, and viability was assessed using 0.4% trypan blue dye in a Countess™ Automated Cell Counter (Thermo Fisher Scientific).

EV Isolation
At various time points post-inoculation, groups of EVs were isolated from RPMI exosome-free cell culture media. After the media was collected, it was filtered through a 3-mL syringe with an attached 25-mm syringe filter (0.22-µm pore size). PBS (1×) was added to the media and centrifuged at 32,000 rpm for 70 min in a SW41Ti swinging bucket rotor at 4 • C using a Beckman Coulter Optima L-70K Ultracentrifuge. The supernatant was removed and approximately 500 µL resuspended EVs were collected from each sample. Following isolation, the EVs were quantitated using the Bradford-Lowry protein quantitation procedure (Bio-Rad Laboratories, Hercules, CA, USA) [17].

Quantification of EVs
EV size distribution and particle number per milliliter were quantified using NanoSight tracking analysis (NTA) (Nanosight NS300-LM10, Malvern Instruments, Inc., Malvern, UK) [18]. Briefly, the EV samples were diluted in PBS (1×) (1:1000) and placed inside a 0.3-mL disposable syringe. NTA visually analyzes and quantitates particle size by measuring Brownian motion as it relates to particle size in fluids. Five independent experiments were analyzed, and the results were calculated as mean ± standard deviation of the mean values.

RNA Isolation from EVs
Total RNA was isolated from BV-2 cell-derived EVs using IsoMag Carbon-Based RNA Purification Magnetic Beads (Life Magnetics Inc., SKU LMCC250, Detroit, MI, USA). Prior to isolation, RNase was used to remove any residual RNA to avoid contamination. One day before infection, 5 × 10 5 cells were plated in T-25 flasks. On the day of treatment, the medium was aspirated from each flask. Each flask was then rinsed with PBS and 6 mL exosome-free FBS RPMI-1640 media was added. The cells were treated with P. aeruginosa, and control wells with exosome-free FBS RPMI-1640 medium only were maintained. The media fractions were collected after infection treatment and the EVs were extracted as described in the EV isolation section. RNA was isolated from the exosomal fraction. Briefly, the EV fraction was lysed in 200 µL lysis buffer. The lysate was homogenized completely via vortexing to ensure that the solution was non-viscous. Seventy-five microliters of Life Magnetics beads were added directly to the exosomal lysate and were well-mixed using pipetting to achieve a uniform bead distribution. Neutralization buffer (125 µL) was then added to the suspension and it was thoroughly mixed using pipetting; 400 µL binding buffer was added during mixing. After thorough mixing, the lysate-bead solution was placed on the magnetic stand for 120 s and the clear supernatant was aspirated. The beads were washed using a two-step process; 750 µL and 500 µL wash buffer were used for the first and second washes, respectively. After the second wash, the mixture was transferred to a new 1.5-mL centrifuge tube, it was thoroughly mixed using pipetting, and placed on the magnetic stand for 30 s. The wash buffer was then removed completely and the beads were left to dry on the bench for 10-15 min at room temperature. After drying, 30 µL elution buffer was added to the beads, mixed, and incubated at 65 • C for 5 min, followed by centrifugation at 10,000× g for 2 min. The eluent was collected and quantified using a Nanodrop (NanoDrop 2000c, ThermoFisher Scientific, Waltham, MA, USA) spectrophotometer.

cDNA Synthesis of Messenger RNAs
Total RNA (100 ng) isolated from the EVs was used to synthesize messenger RNA (mRNA)-specific cDNA using the mRNA cDNA synthesis kit, with poly (A) polymerase tailing (Applied Biological Materials (ABM), catalog no. G903), according to the manufacturer's protocol. The cDNA was amplified using BrightGreen 5× qPCR master mix (ABM catalog no. MasterMix-5C). One microliter of each primer specific to mRNA CCL4 (TTCCTGCTGTTTCTCTTACACCT f, AAGGACGACAAAGAGAATGAGGA r 10 µM, IDT) and CCL5 (TTTGCCTACCTCTCCCTCG f, AAACGGATGGAGAGGGAGC 10 µM, IDT), along with the following control primers: 36B4 f (GGAGCCAGCGAGGCCACACTGCTG), 36B4 r (CTGGCCACGTTGCGGACACCCTCC), TBP f (ACCCTTCACCAATGACTCCTATG), TBP r (TGACTGCAGCAAATCGCTTGG), HPRT f (GTTAAGCAGTACAGCCCCAAA) and HPRT r (AGGGCATATCCAACAACAAACTT) were used in the amplifications. The quantitative polymerase chain reaction (qPCR) analysis was performed as follows: pre-incubation, 95 • C 10 min, followed by 40 cycles of 95 • C 10 s, 63 • C 15 s, and 72 • C for 5 s. The fold change in mRNA expression was calculated using the delta-delta CT (2-∆∆CT) method. The CT values for the control and treated samples were normalized to the CT values of 36B4, TBP, and HPRT [19]. The fold change in mRNA expression for each treated sample compared with the control sample was determined. For qPCR, gene expression was calculated from four replicates for the respective treatments and from four independent experiments.

Evaluation of EV Proteins Via Enzyme-Linked Immunosorbent Assay
Enzyme-linked immunosorbent assay (ELISA) was performed to determine the proteins found in or on the EVs before and after P. aeruginosa infection. EVs (40 µg) were bound in 100 µL bicarbonate buffer (pH 9.5) in 96-well plates and incubated at 4 • C overnight. The following day, each plate was washed three times in 200 µL washing buffer (0.05% Tween-20 in PBS). The plate was then blocked using 100 µL blocking solution (5% non-fat dry milk and 0.05% Tween-20 in 1× PBS) for 1 h. After blocking, the primary antibodies anti-HSP70 (Santa Cruz, catalog no. sc-32239), anti-HSP 90β Thermo Scientific, catalog no. 37-9400), anti-cleaved caspase 3 (R & D, catalog no. MAB835), anti-LAMP-1 (DSHB, catalog no. 1D4B), anti-IL-6 (DSHB), anti-IL-1β (Bioss antibodies, catalog no. bs-0812R-FR), and anti-TSG101 (Thermo Fisher, catalog no. MA1-23296) were added to each sample with blocking buffer and incubated at room temperature for 2 h. After incubation, each plate was washed, horseradish peroxidase secondary antibody (Dako) was added, and the plate was incubated at room temperature for 30 min. The ELISAs were developed using SigmaFast o-phenylenediamine dihydrochloride peroxidase substrate (Sigma Aldrich) and read at a 405-nm wavelength using a Genemate UniRead 800 microplate reader.

EV Transfer Assay
BV-2 cells were plated at 500,000 cells per T175 flask. Following 24 h, each flask was washed with 1× PBS and filled with RMPI medium/exosome-free FBS. The BV-2 cells were then treated with BV-2-derived EVs from control cells and P. aeruginosa-infected cells at 10 µg and 40 µg [21]. At 72 h, the medium was discarded and cells were collected in 1× PBS. The solution was collected and gently mixed in a 50-mL centrifuge tube. The cells were then counted, and viability was assessed using 0.4% trypan blue dye in a Cell Countess Automated Cell Counter (Thermo Fisher Scientific).

Statistical Analysis
The statistical analysis was performed using standard functions provided with Microsoft Excel Software (e.g., calculation of mean and standard deviation values). Excel was also used to perform the unpaired two-tailed t tests. For statistical comparisons between study groups for cell transfer anaylsis, two-way ANOVA was used followed by post Tukey testing. Data are displayed as mean ± standard deviation or standard error (as indicated). The probabilities of error were: (* indicated in figures) p ≤ 0.05, (**) p ≤ 0.01, (***) p ≤ 0.001, and (****) p ≤ 0.0001.

Pseudomonas aeruginosa Infection Reduces Microglial Cell Viability
To examine the effect of P. aeruginosa on the microglia cell line, BV-2, cells were infected with 0, 1.3 × 10 4 , or 2.6 × 10 4 CFU/mL of P. aeruginosa for 24, 48, and 72 h in exosome-free media. Our objective was to determine the effect of P. aeruginosa on microglial-derived EVs. To determine the optimal infection concentration, we performed a series of dose and time-dependent response experiments [14]. The results indicated that a P. aeruginosa concentration of 2.6 × 10 4 CFU/mL was an optimal sub-lethal infection concentration to determine cellular responses for the subsequent experiments.
The trypan blue exclusion assay was performed to quantitate the numbers of viable cells present in BV-2 cell suspensions after P. aeruginosa infection. At 72 h, with 2.6 × 10 4 CFU/mL infection, cell viability was significantly decreased by approximately 12% (p ≤ 0.01), compared with the control cells ( Figure 1). Our results indicated that by 72 h, P. aeruginosa infection significantly reduced the viability of microglial, BV-2 cells.

Pseudomonas aeruginosa Infection Reduces BV-2-Derived EVs
To determine the effect of P. aeruginosa on BV-2-derived EVs, BV-2 cells were cultured in exosome-free media and infected with P. aeruginosa (2.6 × 10 4 CFU/mL) for 72 h. The EVs released

Pseudomonas aeruginosa Infection Reduces BV-2-Derived EVs
To determine the effect of P. aeruginosa on BV-2-derived EVs, BV-2 cells were cultured in exosome-free media and infected with P. aeruginosa (2.6 × 10 4 CFU/mL) for 72 h. The EVs released into the cultured media were isolated using a series of high-speed ultracentrifugation steps [9] and were then characterized. EV presence was also confirmed with ELISA using EV markers, and TSG101 and LAMP-1 proteins (Figure 2A,B). LAMP-1 protein is a model EV marker. It is a marker for late endosomes and was used to confirm that the EVs completed endocytosis and were transported to endosomal organelles [22,23]. We found a significant decrease in the presence of LAMP-1 in the EVs derived after infection, compared with the control (Figure 2A). EV marker, TSG101, was also present within both groups of BV-2-derived EVs ( Figure 2B).
To determine the effect of P. aeruginosa on BV-2-derived EVs, BV-2 cells were cultured in exosome-free media and infected with P. aeruginosa (2.6 × 10 4 CFU/mL) for 72 h. The EVs released into the cultured media were isolated using a series of high-speed ultracentrifugation steps [9] and were then characterized. EV presence was also confirmed with ELISA using EV markers, and TSG101 and LAMP-1 proteins (Figure 2A,B). LAMP-1 protein is a model EV marker. It is a marker for late endosomes and was used to confirm that the EVs completed endocytosis and were transported to endosomal organelles [22,23]. We found a significant decrease in the presence of LAMP-1 in the EVs derived after infection, compared with the control (Figure 2A). EV marker, TSG101, was also present within both groups of BV-2-derived EVs ( Figure 2B).
EV size was determined using NTA ( Figure 2C-E). Illustrative histograms of control-derived EVs and EVs derived at 72 h infection revealed that control BV-2-derived EVs had a mean concentration of 4.02 × 10 9 ± 6.86 × 10 7 particles/mL. Pseudomonas aeruginosa-derived EVs had a mean concentration of 4.07 × 10 8 ± 7.13 × 10 6 particles/mL ( Figure 2C,D). The NTA revealed strong peaks at the mean EV size at 159 ± 0.4 nm in the control microglial-derived EVs and 131.1 ± 0.4 nm for P. aeruginosa microglial-derived EVs ( Figure 2C,D). This result indicated there was a decrease in EV size when infection was present. Infection with P. aeruginosa significantly (p ≤ 0.05) decreased the EV particle number by 1.77 × 10 7 particles/mL ( Figure 2F). EV size was determined using NTA ( Figure 2C-E). Illustrative histograms of control-derived EVs and EVs derived at 72 h infection revealed that control BV-2-derived EVs had a mean concentration of 4.02 × 10 9 ± 6.86 × 10 7 particles/mL. Pseudomonas aeruginosa-derived EVs had a mean concentration of 4.07 × 10 8 ± 7.13 × 10 6 particles/mL ( Figure 2C,D). The NTA revealed strong peaks at the mean EV size at 159 ± 0.4 nm in the control microglial-derived EVs and 131.1 ± 0.4 nm for P. aeruginosa microglial-derived EVs ( Figure 2C,D). This result indicated there was a decrease in EV size when infection was present. Infection with P. aeruginosa significantly (p ≤ 0.05) decreased the EV particle number by 1.77 × 10 7 particles/mL ( Figure 2F).

Cytokine Response to Pseudomonas aeruginosa Infection in EVs
Small RNAs participate in disease initiation and progression [24,25]. EVs carry a variety of mRNAs, such as the chemokine ligands. We further evaluated the effects of P. aeruginosa infection on the release of CCL4 and CCL5 mRNAs into BV-2 cell-derived EVs (Figure 3). BV-2 cell-derived EVs were collected at 72 h post-inoculation with P. aeruginosa and evaluated for the presence of CCL4 and CCL5 mRNA. Pseudomonas aeruginosa infection elicited no change in the presence of CCL5 mRNA in EVs, compared with the control (Figure 3). CCL4, showed a significant decrease (p ≤ 0.05) in expression post-inoculation in BV-2 cell-derived EVs when compared to the control EVs.

Cytokine Response to Pseudomonas aeruginosa Infection in EVs
Small RNAs participate in disease initiation and progression [24,25]. EVs carry a variety of mRNAs, such as the chemokine ligands. We further evaluated the effects of P. aeruginosa infection on the release of CCL4 and CCL5 mRNAs into BV-2 cell-derived EVs (Figure 3). BV-2 cell-derived EVs were collected at 72 h post-inoculation with P. aeruginosa and evaluated for the presence of CCL4 and CCL5 mRNA. Pseudomonas aeruginosa infection elicited no change in the presence of CCL5 mRNA in EVs, compared with the control (Figure 3). CCL4, showed a significant decrease (p ≤ 0.05) in expression post-inoculation in BV-2 cell-derived EVs when compared to the control EVs.

Pseudomonas aeruginosa Induces Cellular Stress Response
As molecular chaperone proteins, HSPs facilitate the synthesis and folding of proteins. HSPs can be expressed in response to stressful environments. These chaperone proteins are involved in the direct maintenance of protein structure and control apoptotic pathways [26]. Mammalian cells express high levels of stress proteins after exposure to bacteria or bacterial proteins [26]. To determine how BV-2 cells responded to the stress of P. aeruginosa infection, we evaluated HSP70 and HSP90β in BV-2 cell lysates. Both HSP70 and HSP90β were downregulated in response to P. aeruginosa infection ( Figure 4A,B). We found that both HSP70 (p ≤ 0.05), anti-death and immunoregulatory proinflammatory protein, and HSP90β (p ≤ 0.001), immunomodulator and pro-immune responsive, were both significantly upregulated in expression in the infection groups compared with the control groups ( Figure 4C,D), in EVs. The results indicated that P. aeruginosa greatly affected the expression of stress-related proteins inside the cell and within the released EVs.

Pseudomonas aeruginosa Induces Cellular Stress Response
As molecular chaperone proteins, HSPs facilitate the synthesis and folding of proteins. HSPs can be expressed in response to stressful environments. These chaperone proteins are involved in the direct maintenance of protein structure and control apoptotic pathways [26]. Mammalian cells express high levels of stress proteins after exposure to bacteria or bacterial proteins [26]. To determine how BV-2 cells responded to the stress of P. aeruginosa infection, we evaluated HSP70 and HSP90β in BV-2 cell lysates. Both HSP70 and HSP90β were downregulated in response to P. aeruginosa infection ( Figure 4A,B). We found that both HSP70 (p ≤ 0.05), anti-death and immunoregulatory pro-inflammatory protein, and HSP90β (p ≤ 0.001), immunomodulator and pro-immune responsive, were both significantly upregulated in expression in the infection groups compared with the control groups ( Figure 4C,D), in EVs. The results indicated that P. aeruginosa greatly affected the expression of stress-related proteins inside the cell and within the released EVs.

Expression of Immune Regulators are Present in EVs
Interleukins are pro-inflammatory cytokines that are vital for cellular signaling and inflammatory feedbacks [27]. The immune system depends essentially on interleukins. IL-6 acts as a

Expression of Immune Regulators are Present in EVs
Interleukins are pro-inflammatory cytokines that are vital for cellular signaling and inflammatory feedbacks [27]. The immune system depends essentially on interleukins. IL-6 acts as a first line of defense system for many cell types (e.g., stromal [28], epithelial [28], hematopoietic [28], and microglial cells) [29] by releasing signals to nearby and distant cells. Gram-negative bacteria stimulate IL-6 production in microglial cells [30]. IL-6 release enhances survival rates of infected cells [30]. Here, we found IL-6 in BV-2 control cell lysate and P. aeruginosa-infected cell lysate ( Figure 5A). We also found IL-6 in both groups of EVs ( Figure 5B). A representative dot blot image of IL-6 expression in cell lysate can be found in Supplementary Figure S1.

Pseudomonas Aeruginosa Decreases Samm50 Expression in Cell Lysates
Caspases are modulators of cell death in response to stimuli (e.g., infections, stress) [32]. We evaluated cleaved caspase-3 expression in cell lysates and EVs. The presence of cleaved caspase-3 indicates the manifestation of apoptotic pathways within the cell as it responds to the bacterial infection [33]. Cleaved caspase-3 was expressed in cell lysates and slightly decreased as a result of P. aeruginosa infection ( Figure 6A). However, in EVs there was relatively no change between the control and the P. aeruginosa-infected groups ( Figure 6B). Samm50 is a component of the sorting and assembly machinery complex (SAM). It is in the outer membranes of mitochondria and has an essential role in the maintenance of mitochondrial cristae structure [34,35]. Samm50 targets contaminants and bacterial effector proteins that affect host mitochondria. We found that P. aeruginosa significantly decreased Samm50 (p ≤ 0.001) expression, compared with uninfected cells ( Figure 6C). Previous studies have found that Samm50 depletion protects cells from caspase-3induced cell death [34]. Expression of immune regulators. IL-6 (A) and IL-1β (C) were detected in cell lysates (control or P. aeruginosa-infected (2.6 × 10 4 CFU/mL)). IL-6 (B) and IL-1β (D) were detected in EVs (control or P. aeruginosa-infected (2.6 × 10 4 CFU/mL)). Experiments were performed using four or five independent experiments. The results are presented as mean ± SEM values.
The pro-inflammatory cytokine, IL-1β, has critical roles in the activation of many inflammatory pathways and adaptive immunity. IL-1β levels can modulate in response to bacterial infection [31]. IL-1β levels in cell lysates were slightly increased after P. aeruginosa infection, compared with the control group ( Figure 5C). The results of another study indicated there is a statistically significant difference in intensity of expression (p ≤ 0.001) of IL-1β in control compared with P. aeruginosa groups of cell lysates [14]. In this study, we found IL-1β within the packaging of the control and P. aeruginosa-infected BV-2-derived EVs. Taken together, these results suggest that EVs act as cargo carriers for cytokines produced by BV-2 cells ( Figure 5D).

Pseudomonas Aeruginosa Decreases Samm50 Expression in Cell Lysates
Caspases are modulators of cell death in response to stimuli (e.g., infections, stress) [32]. We evaluated cleaved caspase-3 expression in cell lysates and EVs. The presence of cleaved caspase-3 indicates the manifestation of apoptotic pathways within the cell as it responds to the bacterial infection [33]. Cleaved caspase-3 was expressed in cell lysates and slightly decreased as a result of P. aeruginosa infection ( Figure 6A). However, in EVs there was relatively no change between the control and the P. aeruginosa-infected groups ( Figure 6B). Samm50 is a component of the sorting and assembly machinery complex (SAM). It is in the outer membranes of mitochondria and has an essential role in the maintenance of mitochondrial cristae structure [34,35]. Samm50 targets contaminants and bacterial effector proteins that affect host mitochondria. We found that P. aeruginosa significantly decreased Samm50 (p ≤ 0.001) expression, compared with uninfected cells ( Figure 6C). Previous studies have found that Samm50 depletion protects cells from caspase-3-induced cell death [34].

EV Treatment Decreases Cell Viability
The viability of BV-2 cells after EV treatment decreased by 72 h. We speculated that the cell viability decreases observed in Figure 1 could partially be mediated by an unknown EV-associated mechanism. We observed the effects on the BV-2 cells after EV treatment with BV-2-derived EVs (control) and EVs derived from P. aeruginosa infection (infected). By 72 h, infected BV-2-derived EVs decreased BV-2 cell viability at both concentrations (10 µg and 40 µg) by approximately 0.23-0.24 fold, compared with the control cells (Figure 7).

EV Treatment Decreases Cell Viability
The viability of BV-2 cells after EV treatment decreased by 72 h. We speculated that the cell viability decreases observed in Figure 1 could partially be mediated by an unknown EV-associated mechanism. We observed the effects on the BV-2 cells after EV treatment with BV-2-derived EVs (control) and EVs derived from P. aeruginosa infection (infected). By 72 h, infected BV-2-derived EVs decreased BV-2 cell viability at both concentrations (10 µg and 40 µg) by approximately 0.23-0.24 fold, compared with the control cells (Figure 7).

EV Treatment Decreases Cell Viability
The viability of BV-2 cells after EV treatment decreased by 72 h. We speculated that the cell viability decreases observed in Figure 1 could partially be mediated by an unknown EV-associated mechanism. We observed the effects on the BV-2 cells after EV treatment with BV-2-derived EVs (control) and EVs derived from P. aeruginosa infection (infected). By 72 h, infected BV-2-derived EVs decreased BV-2 cell viability at both concentrations (10 µg and 40 µg) by approximately 0.23-0.24 fold, compared with the control cells (Figure 7).

Discussion
Gram-negative bacteria are associated with approximately two million nosocomial infections and 90,000 fatalities annually in the USA [36]. Due to high rates of P. aeruginosa infection, the increases in multi-drug resistant P. aeruginosa, and the evolution of the bacterium against antibiotics (e.g., cephalosporins, carbapenems, and aminoglycosides), it is responsible for fatal health-related threats worldwide [37]. In the USA, the number of infections is increasing and the number of effective treatments is decreasing. Thus, new biological information regarding pathogenesis is needed to develop effective treatments against pathogenic microbes.
In our study, we performed an evaluation of EV biogenesis and composition to understand the effect of P. aeruginosa on the microglia cell line, BV-2. Microglia are the resident brain macrophages that maintain tissue homeostasis. Microglia function is one of the primary defenses against bacterial infection and brain injury, and stimulate tissue repair [38]. Microglia recognize many Gram-negative pathogenic bacteria that can colonize the central nervous system [39]. In addition to P. aeruginosa, Escherichia coli and Vibrio vulnificus have effects on microglial cells [40][41][42]. Hoogland et al. developed a live E. coli mouse model that mimics the delirium-associated infection [41]. They found that compared with control mice, mice infected with live E. coli have microglial activation by 72 h post-inoculation. Using flow cytometry analysis, they also found that by 72 h, microglia from E. coli-infected mice had increases in mean cell size and CD45 expression, compared with the controls [41]. Mayer et al.'s results suggested that V. vulnificus MO6-24/O lipopolysaccharide activates rat microglia in vitro and stimulates the release of superoxide anion in vivo [42]. Other species of Gram-negative bacteria likely enter and replicate in the central nervous system [43]. However, few to no study results have been published on this topic.
Disease-associated proteins (e.g., β-amyloid [44], superoxide dismutase [45], and α-synuclein [46]) are released from cells during the EV response. The secreted proteins likely participate in the pathogenic response via interaction with recipient cells [47]. Alpha-synuclein can be found in the plasma and cerebrospinal fluid of humans, which further supports the hypothesis that EVs can be used as biomarkers for disease [47]. The results of infection biology studies indicate that EVs are released by mammalian cells infected with bacteria, infectious prion proteins, and viruses [48].
Using an in vitro model for human physiology, we observed the release of EVs in BV-2 cell culture media in response to P. aeruginosa infection. Pseudomonas aeruginosa cannot be isolated from the body to distinguish which EVs were produced by which organism. Using NTA, we found a significant decrease in EV particle number/mL after P. aeruginosa infection. Decreased EV numbers correlated to a significant decrease in cell viability after P. aeruginosa (2.6 × 10 4 CFU/mL) infection (Figure 2). At 72 h post-inoculation, the EV particle/cell number was 706 EVs/per control cell versus 205 EVs/per infected cell. This result indicated that in response to infection, EVs were being released at slower rates, which correlated with the decreasing cell survival rate. When used in combination, the qPCR, dot blot analysis, and ELISA allowed us to examine mRNA, EV markers (e.g., LAMP-1, TSG101), and proteins (e.g., HSPs, ILs, caspases, etc.) that are found internally and externally on EVs (Figures 3-6).
EVs have a vital role in cell-to-cell communication by transferring mRNA through an endocytic process. While we tested miRNAs and mRNA for expression within BV-2-derived EVs (Supplementary Figure S3), the most significant findings were those related to the chemokine ligand family. The chemokine ligands, CCL4 and CCL5, have active roles in the inflammatory response in tissues. Our results indicated that chemokine ligands were in BV-2-derived EVs; specifically, P. aeruginosa infection caused a significant decrease in CCL4 packaged within EVs (Figure 3). We found that these two chemokines were produced in BV-2-derived EVs. To our knowledge, this finding is the first in the context of P. aeruginosa, chemokine ligands, and EVs. The chemoattractant, CCL4, is a major chemokine for lymphocytes, naïve dendritic cells, monocytes, and natural killer cells [49]. The downregulation of CCL4 production by microglial-derived EVs in response to P. aeruginosa suggested that microglial cells do not have an initial role in immune response activation. These findings were not consistent with the findings of Sun et al., which suggest that P. aeruginosa infection causes an increase in CCL4 production and subsequently initiates the host immune response [49]. The inconsistent findings could be due impartially to Sun et al. using mast cells rather than microglial cells in their study.
HSP70 is packaged within EVs released from astrocytes under stressful conditions, such as heat or oxidative stresses and subsequently has a pro-survival effect on neurons [50]. HSP70 s protection of brain cells against other stressors could be due to how it functions during the prevention of damaging pro-inflammatory responses [51]. Our results indicated that HSP70 was significantly expressed in BV-2-derived EVs at the 72-h time point. Therefore, it likely has a critical role in cytoprotection ( Figure 4C). Toll-like receptors (TLRs) use HSP70 as a ligand on immune cells (i.e., microglia) [52]. In our previous study, we did not find TLR2 [14], but that does not exclude participation of other TLRs. Other TLRs would need to be tested to determine which TLR contributes to this pathway. HSP secretion also occurs independently, or in the lack of cell death, which suggests that secretion could be occurring via active mechanisms such as EVs.
HSP90β levels of expression significantly increased within the BV-2 EVs derived after P. aeruginosa infection ( Figure 4D). The effects of HSP90β inhibition in the brain are not well-studied. However, HSP90β does regulate HSP70 induction [53]. This study found that P. aeruginosa also generally released EVs. However, we hypothesize that the predominant EV populations are derived from the BV-2 cell line. Our results suggested that both HSP70 and 90β are upregulated in BV-2-derived-EVs in the presence of P. aeruginosa infection. We also found HSP70 and HSP90β in cell lysates. HSPs protect against an array of stress-related factors, including bacterial infection, hyperthermia, oxygen radicals, heavy metals, and ethanol [20].
Pathogens also use the host's HSPs to their specific advantage. Henderson et al. found that bacteria express numerous molecular chaperones on their cell surfaces. They can secrete these into the extracellular environment to act as virulence signals [54]. Hence, HSP70 is connected to microbial pathogenesis, immune responses, and apoptosis [7]. The stress response is vital for bacteria during adaptation to fluctuations in their biological state. HSP70 proteins and their cooperating chaperones create complex systems of protein-folding machines in prokaryotes and in eukaryotic cellular compartments and are expressed during stress conditions. Molecular chaperones regulate many tactics used by bacterial pathogens, including entry into the host, and replication and survival after entry [55]. A recent methodology to counteract bacterial infections is to exploit HSP70 and its co-chaperones as possible therapeutic targets. This approach allows prokaryotes to proceed through stress-induced changes caused by host responses or antibiotics [56]. Therapeutic treatment or a vaccine vehicle targeting these proteins may be an effective approach to limit bacterial pathogenesis.
To evaluate the role of immune regulators and P. aeruginosa infection, we examined IL-6 and IL-1β in BV-2 P. aeruginosa-derived EVs and cell lysates. Interleukins have a key part in systemic immune protection against infections, particularly bacterial infections [57]. Interleukins protect by activating the adaptive immune response that occurs after apoptotic reactions. The multifunctional cytokine, IL-6, has a central role in host immunity and it was present in EVs and cell lysates of both BV-2 groups, the control and P. aeruginosa-infected groups. IL-1β is a potent stimulator of IL-6 production in central nervous system astrocytes [58]. Microglial cells release IL-6 and IL-1β, which are both upregulated [29]. Our study found that interleukins were shuttled into the endocytic EVs from the cell in response to the Gram-negative pathogen, P. aeruginosa. We also tested and found immunomodulatory agents (IL-6 and IL-1β) and inflammatory response provoker, TNFα, in BV2-derived EVs. This result further supports the presence of immune regulators as part of the contents of EVs, in response to infection [14]. TNFα participates in biological effects including cell growth, cell differentiation, and cytotoxicity and, most importantly, in our study, anti-infection [59,60]. Finally, in a previous study, when we compared cell lysates of control to P. aeruginosa-infected cells, we found a statistically significant increase in intensity of P. aeruginosa-infected cells during evaluation of NLR family pyrin domain containing 3 (NLRP3) and B-cell lymphoma 2 (BCL-2) [14]. A list of proteins that we tested for presence in the BV-2-derived EVs is presented in Supplementary Figure S2.
Fundamentally, EV functionality and characteristics (e.g., cell-to-cell communication) suggest that they have one or more roles in the pathogenesis of bacterial infections, whether it is host protection or assisting during bacterial invasion, or both. Here, we found that BV-2 viable cell number decreased after treatment with EVs derived after P. aeruginosa infection (Figure 7). This finding directly correlated with the cell viability results (Figure 1). Although P. aeruginosa infection causes cell viability to decrease over time, our results ( Figure 7) suggested that in the presence of EVs, viable cells were also affected by molecular constituents packaged in the EVs. This area of EV research is not well-understood and requires further study.

Conclusions
Bacteria, particularly Gram-negative bacteria, can release EVs for intercellular communication in prokaryotic and eukaryotic cells [7]. Intercellular communication can be mediated through direct cell-to-cell communication or transfer of secreted molecular constituents [47]. During bacterial infections, EVs can have different roles during the pathogen's life cycle. EVs activate immune responses that combat the bacterium or assist the bacterium in the spread of the infection by spreading its virulence factors, or both [7]. Our findings suggested that EVs have an important role in infection biology. These vesicles may be directly pathogen-derived or released from microglial cells. In both cases, understanding their complex roles in cell-to-cell communication and immune modulation is important. A comprehensive investigation of EVs derived from non-infected and infected hosts is still needed to explain mechanisms related to pathogenesis and innovative therapeutics. These findings suggest that the potential to use EVs as a biomarker may guide the advancement of novel therapeutics for challenging and defiant pathogens. Furthermore, these studies shed light on the impact of Gram-negative bacteria on the central nervous system and EV biogenesis. In future studies, we will compare the effects of Gram-negative bacteria on EV biogenesis in vivo.

Supplementary Materials:
The following are available online at http://www.mdpi.com/2076-0817/8/4/297/s1, Figure S1: Dot blot analysis of cell lysates. Representative dot blot analysis of BV-2-derived cell lysates at 0.8 µg were tested with IL-6. The quantitative results can be found in Figure 5A. Figure S2: Additional proteins tested for expression in BV-2-derived exosomes and cell lysates. Figure S3: Additional miRNA and mRNA tested for presence in BV-2-derived exosomes and cell lysates.