Early Acidification of Phagosomes Containing Brucella suis Is Essential for Intracellular Survival in Murine Macrophages

Reagents.

N-Hydroxysuccinimidyl ester 5- (and -6)-carboxyfluorescein (NHS-CF), N-hydroxysuccinimidyl ester 5- (and -6)-carboxytetramethylrhodamine (NHS-Rho), and Oregon Green 488 carboxylic acid succinimidyl ester (Oregon Green 488) were purchased from Molecular Probes (Eugene, Oreg.); bafilomycin A₁ and monensin were purchased from Sigma.

Bacterial strains and growth conditions.

The strain used throughout the experiments was B. suis 1330 (ATCC 23444). Bacteria were grown in tryptic soy (TS) broth (Difco Laboratories) at 37°C for 24 h to stationary phase. The bacteria were harvested by centrifugation, washed twice in phosphate-buffered saline (PBS), and used immediately. Killed organisms were obtained either by incubation at 65°C for 30 min or by incubation at 37°C for 1 h with 300 μg of gentamicin per ml.

Preparation of opsonized bacteria.

B. suis were opsonized with polyclonal murine anti-Brucella antibodies for 30 min at 37°C and washed once in PBS.

Labeling of bacteria with fluorescent probes.

Log-phase cultures (10⁹ bacteria/ml) were washed twice in PBS and resuspended in 1 ml of PBS containing 0.05% Tween 80. After addition of the mix of fluorescent probes NHS-CF or Oregon Green 488 and NHS-Rho (10 μl of each at 10 mg/ml), the suspension was vortexed. The bacteria were incubated at 4°C for 30 min in the dark. The labeled bacteria were centrifuged, and the reaction was stopped by addition of Tris-HCl (pH 8.3) at a final concentration of 100 mM. The bacteria were again incubated at 4°C for 15 min in the dark, washed twice in PBS, and used immediately for infection. Under these conditions 100% of the bacteria were labeled, as controlled by fluorescence microscopy.

Cell culture.

J774.A1 cells (ATCC TIB 67), a murine macrophage-like cell line, were grown in RPMI 1640 medium (Gibco/BRL) containing 10% heat-inactivated fetal calf serum and 5 mM glutamine (complete medium) at 37°C and 5% CO₂. Cells were resuspended at 10⁵ cells/ml and cultured for 1 day.

Infection and intracellular viability assay of B. suis in J774 cells.

Experiments were performed as described previously (8). Briefly, J774 cells were infected at a multiplicity of infection (MOI) of 20 bacteria per cell with stationary-phase B. suis 1330 in RPMI 1640 for 30 min. Cells were washed three times with PBS and reincubated in RPMI 1640 with 10% fetal calf serum and gentamicin at 30 μg/ml for at least 1 h (1 h postinfection). At different time points, cells were washed once with PBS and lysed in 0.2% Triton X-100. Bacterial counts (CFU) were determined by plating serial dilutions on TS agar and incubation at 37°C for 3 days. The relative intracellular survival has been defined as 100% at 1 h postinfection.

Addition of vacuolar-pH-neutralizing reagents to infected cells.

At 1 h postinfection, one of the following reagents was added to J774 cells: 30 mM NH₄Cl, 100 nM bafilomycin, or 50 μM monensin. The effect of these substances was studied by determining the number of viable bacteria 3 and 5 h after addition of the reagents. Infected but untreated cells were analyzed in parallel. All experiments were performed in quadruplicate. A possible toxic effect of the substances on B. suis was tested by exposure in TS broth for 5 h followed by viability assays. Macrophage viability was verified by trypan blue dye exclusion assays.

In another set of experiments, NH₄Cl (30 mM, final concentration) was added to J774 cells at 90 min and at 7 h after the beginning of infection with B. suis as described above. Infected but untreated cells were analyzed in parallel. The intracellular survival was determined at 1.5, 7, 24, and 48 h. All experiments were performed in triplicate.

Measurement of phagosomal pH by fluorescence microscopy.

J774 cells were cultured in Lab-Tek chambered coverglass (Nunc) with 400 μl per well of a suspension at 10⁵ cells/ml. After 1 day of incubation, cells were infected for 45 min by adding carboxyfluorescein (CF)-Rho or Oregon Green 488-rhodamine (Rho)-labeled bacteria in serum-free RPMI 1640 medium (200 μl per well) at a ratio of 100 bacteria per cell. Cells were then washed twice in PBS and incubated in complete medium containing gentamicin at 30 μg/ml to eliminate extracellular bacteria. For pH neutralization, bafilomycin A₁ was added at a concentration of 100 nM together with gentamicin at the end of the infection period. At various time points postinfection, the medium was removed and replaced by PBS for observation. For each point, several pairs of images (generally four) were acquired at CF- and Rho-specific wavelengths. Acquisitions were made by videofluorescence microscopy with a Cool View camera (Photonic Science, East Sussex, England) and an image processor linked to an inverted Leica DM IRB microscope (Leica, Rueil-Malmaison, France). Filters used to detect CF or Oregon Green 488 fluorescence consisted of an excitation bandpass filter (450 to 490 nm), a dichroic mirror (510 nm), and an emission bandpass filter (515 to 560 nm). Filters used to detect Rho fluorescence consisted of an excitation bandpass filter (515 to 560 nm), a dichroic mirror (580 nm), and a longpass emission filter (>590 nm). Microscope settings were performed under Rho fluorescence to avoid extensive CF photobleaching, and CF or Oregon Green 488 fluorescence was acquired rapidly. After acquisition, images were analyzed by the imaging system VISIOLAB 1000 (Biocom, Les Ulis, France). For each image, background was defined as the grey-level value corresponding to the majority of pixels. Fluorescent objects were selected by the software on the Rho image (the number of objects per image generally varied from 20 to 50), and the selection was reported on the CF or Oregon Green 488 image. The average grey-level value was calculated for each selection, the background was subtracted from the corresponding value, and the CF/Rho or Oregon Green 488/Rho ratio was calculated for each pair of image. Standard deviations were obtained from the mean of four CF/Rho or Oregon Green 488/Rho ratio values.

An in situ calibration curve of the CF/Rho or Oregon Green 488/Rho emission ratio versus pH was made at the end of each experiment with the same data acquisition parameters (39). Infected J774 were incubated for 1 h in a buffer of defined pH containing 10 μM K⁺/H⁺ ionophore nigericin to collapse pH gradients across membranes, 130 mM KCl–30 mM sodium acetate-acetic acid (for pH 3.6, 4.0, 4.5, or 5.0) or 10 μM nigericin–130 mM KCl–30 mM Na₂HPO₄-NaH₂PO₄ (for pH 6.0, 7.0, 8.0, or 9.0). Pairs of CF-Rho or Oregon Green 488-Rho images were acquired and processed as described above, and a calibration curve was obtained.

Labeling of B. suis with fluorescent probes does not affect bacterial viability and intracellular replication of the bacteria.

The method that we used to determine the pH in Brucella-containing vacuoles was first described by Oh and Straubinger (32) and is based on the fact that CF emission intensity varies with pH whereas Rho fluorescence intensity is independent of pH and is used as a reference signal. The ratio of CF and Rho fluorescence therefore allows calculation of the pH. The fluorescent products NHS-CF and NHS-Rho had to be covalently bound to the surface of the bacteria prior to infection. Compared to control labelings performed with Escherichia coli in preliminary experiments, the efficiency of CF and Rho labeling of B. suis was always lower, regardless of the labeling conditions with respect to concentrations of probes, temperature, or time of incubation (not shown). With heat-killed B. suis, the labeling efficiency was significantly higher, possibly due to alterations of the bacterial cell surface structure. However, the method was validated for the labeling of live B. suis as well, allowing measurement of the pH of phagosomes containing live or nonviable B. suis.

To examine a possible effect of bacterial labeling with the fluorescent marker mixture on viability of B. suis, we compared the viability of dye-labeled bacteria with that of nonlabeled bacteria. There was no significant difference (data not shown). To compare the intracellular behavior within J774 macrophages, we performed infection experiments and monitored the multiplication of labeled and unlabeled opsonized bacteria over a period of 48 h (Fig. 1). The rates of bacterial uptake, as determined by the ratio of CFU at 1 h postinfection to CFU at the end of the 30-min infection period, were identical for unlabeled and labeled bacteria, and both intracellular growth curves showed an identical behavior of the bacteria with respect to survival and replication (Fig. 1), leading to the assumption that binding of the pH probes does not affect the course of the events taking place during phagocytosis and within the Brucella-containing phagosome.

Quantitation of pH in individual phagosomes containing B. suis.

For in situ calibration allowing the calculation of pH from the CF/Rho emission ratio, the ionophore nigericin was used to artificially adjust the intracellular pH to the values defined for the incubation buffers, i.e., pH 3.6, 4.0, 5.0, 6.0, 7.0, 8.0, and 9.0. A representative calibration curve is shown in Fig. 2A. For each experiment, an independent calibration curve was established because of variability in labeling intensity and in data acquisition parameters.

Unopsonized B. suis were taken up poorly by J774 cells (1 to 2% infected cells). The rate of phagocytosis was markedly increased (90 to 100% infected cells), however, when the bacteria were opsonized with polyclonal anti-Brucella antibodies. As a consequence, in all experiments performed for the measurement of intraphagosomal pH which required rapid acquisition of a large number of images, B. suis was opsonized before infection. Video-fluorescence microscopy analysis showed that the pH in the phagosomes containing live B. suis decreased to values of pH 4.0 ± 0.5 (Fig. 2B). The rate of acidification was high: at 1 h postinfection, the phagosome was already acidic and remained at this level for at least 5 h, the duration of the experiment. pH values measured at the various time points were not significantly different from one another.

Quantitation of pH in individual phagosomes containing live B. suis, using Oregon Green 488 as reporter of pH.

Fluorescein derivatives are the most widely used pH-sensitive probes, with a pK_a of around 6.4, which therefore allows accurate pH measurements in the range 5 to 8. Nevertheless, by using NHS-CF in our experiments, we could determine phagosomal pH values in the range 4 to 8 (Fig. 2A), as also reported previously by Oh and Straubinger (32) and Montcourrier et al. (30). Recently, a new probe, Oregon Green 488, was described as a more appropriate reporter of acid pH, since its pK_a of around 4.7 allows a more accurate measurement of pH in the range 3.5 to 6.0 (41) than CF. Oregon Green 488 has the same maximum excitation and emission wavelengths as CF. Using the method described above, we performed experiments with this probe. A representative calibration curve, obtained with Oregon Green 488-Rho-double-labeled B. suis, is shown in Fig. 3A. The pH dependence was optimal between pH 3.6 and 5.0. Our results revealed that the pH inside phagosomes containing live bacteria varied between 4.2 and 4.5 and remained at this level until at least 5 h postinfection (Fig. 3B). These results confirmed those obtained with CF.

Effect of the specific vacuolar proton-ATPases inhibitor bafilomycin A₁ on phagosome acidification.

Vacuolar proton-ATPases participate in the variety of mechanisms that have been found to be implicated in the acidification of vacuoles (27). To investigate the nature of the mechanism responsible for acidification of the compartment containing B. suis, we studied the effect of the macrolide antibiotic bafilomycin A₁ on phagosome acidification. BafilomycinA₁ has been described as a specific inhibitor of vacuolar ATPases (38), but it has no toxic effect on B. suis (our control experiments [data not shown]) or on other gram-negative bacteria (6). Acidification of the compartment containing live B. suis was inhibited by bafilomycin A₁ compared to the vacuoles in untreated cells (Fig. 4A). One hour of treatment with the inhibitor was sufficient to obtain a shift of the pH to about 6.

Similar experiments were conducted with gentamicin-killed B. suis (Fig. 4B). Phagosomes containing killed bacteria are usually routed to lysosomes. The pH of the vacuoles containing gentamicin-killed B. suis was rapidly acidified to an average value of 4.5 (1 h postinfection), and again, the pH values measured over the 5 h-period were not significantly different from one another. After treatment with bafilomycin A₁, the intraphagosomal pH fluctuated between 6.5 and 7. Thus, the major mechanism of acidification of Brucella-containing vacuoles appeared to be proton pumping via vacuolar proton-ATPases sensitive to bafilomycin.

Effect of bafilomycin A₁ and other vacuolar-pH-neutralizing reagents on short-term survival of B. suis in macrophages.

The results described above led us to investigate the importance of low phagosomal pH in the survival and replication of B. suis within J774 cells. In addition to bafilomycin, two other reagents raising the pH of acidic compartments by different mechanisms were used: NH₄Cl as a lysosomotrophic substance, and monensin as a cationic ionophore. One hour postinfection, macrophages were treated with the vacuolar-pH-neutralizing reagents, and viable intracellular bacteria were enumerated following recovery 3 and 5 h later, i.e., 4 and 6 h postinfection. Infected but untreated cells were analyzed in parallel.

Experiments were performed on bacteria opsonized with specific antibodies as already described for the pH quantitation experiments (Fig. 5A) and on nonopsonized bacteria (Fig. 5B). In both cases, the three reagents significantly decreased the viability of intracellular B. suis. After a 3-h treatment, the number of viable B. suis in J774 cells was clearly reduced; the survival rate was 7.6 to 22.6%, compared to 49.5% for the untreated control. After a 5-h treatment, only 2.8 to 3.6% of the bacteria were recovered, as opposed to 29% for the control (Fig. 5A). With nonopsonized bacteria, the survival of intracellular B. suis varied from 20.8 to 23.9% (untreated control, 54.4%) 3 h after the beginning of the treatments, and only 1.5 to 8.8% survived (untreated control, 51.2%) after 5 h of treatment (Fig. 5B). These data indicated that the observed enhanced decrease in survival and replication of B. suis could be linked to the loss of acidification in the compartment containing the bacteria. Differences in survival observed for all treated samples, compared to the respective untreated controls, were statistically significant (P < 0.01; Student t test). We excluded a direct toxic effect of bafilomycin, NH₄Cl, and monensin on the bacteria by incubation for 5 h in TS broth containing the same concentrations of the three reagents as in the J774 infection assays described above. The viability of B. suis was not affected under these conditions (data not shown). To mimic intraphagosomal conditions, we incubated B. suis in TS broth at pH 4 and 6 in the absence and presence of bafilomycin (100 nM) over a period of 24 h and then performed viability assays by serial dilutions and plating at 0, 3, 6, and 24 h of incubation. Viability was not affected in the presence of bafilomycin (data not shown).

Early but not late neutralization of vacuolar pH by ammonium chloride inhibits survival of B. suis in J774 macrophages.

The strong reduction of early intracellular survival of opsonized and nonopsonized B. suis observed in the presence of the three vacuolar-pH-neutralizing agents described above led us to investigate the fate of the intracellular bacteria over a longer period of infection, until 48 h. Our study had to be limited to the effects of NH₄Cl, as monensin and bafilomycin were cytotoxic to J774 cells starting at 12 and 24 h postinfection, respectively, whereas 30 mM NH₄Cl had no toxic effect until the end of the experiments (not shown). Early neutralization of intracellular compartments at 90 min after the beginning of infection resulted in a constant decline of the bacteria until their complete eradication at 48 h postinfection, as opposed to infection of untreated cells (Fig. 6). In contrast, the addition of NH₄Cl to J774 cells at 7 h had no effect on normal intracellular growth of B. suis: after an initial reduction in intracellular viability of 30 to 50%, the bacteria multiplied 800- to 1,000-fold (Fig. 6). These results indicated that for intracellular multiplication of B. suis, an acidic intraphagosomal pH is essential only in the early phase of infection. In our system, neutralization at 6 to 7 h after the first contact of B. suis with the macrophage, i.e., when bacteria began to multiply, did not alter the outcome of infection.