Chlamydia infection does not alter the polarization state of non-polarized (M0) macrophages

Based on reports of intracellular pathogens being able to induce an M2 profile in macrophages [17,19], we initially hypothesized that Chlamydia would do the same. We assessed the state of macrophage polarization using RT-PCR as this remains the gold standard method. Our positive controls, macrophages polarized to M1 or M2, showed expected expression profiles of key M1 or M2 markers (Figure 1A). Non-polarized bone marrow derived macrophages (BMDM) showed very little difference in the expression of key M1 or M2 genes upon infection compared to the large differences shown in polarized macrophages (Figure 1B). Chlamydia 16sRNA was highly expressed in infected macrophages, but was not present in uninfected macrophages.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 1. Chlamydia muridarum does not induce a gene expression profile of polarization in naïve macrophages.

Gene expression profiles determined by RT-PCR of BMDM. (A) Positive controls of polarized BMDM show differential expression in key M1- and M2-associated genes relative to expression seen in M0 macrophages. (∆) denotes detectable Retnla expression in M2 macrophages but no detection in M0 or M1 macrophages. (B) M0 macrophages were infected for 24 hr with 1 MOI Chlamydia. 16sRNA detected in infected macrophages by RT-PCR, but not uninfected indicative of bacterial presence. Expression of M1-associated genes and expression of M2-associated genes show no significant alteration with infection as determined by one-sample t-test against hypothetical value of 1.0. (∆∆) denotes no Retnla detected in uninfected or uninfected macrophages. Data are averages ±SEM from three independent experiments.

https://doi.org/10.1371/journal.pone.0069421.g001

Polarized macrophages are not infected equally

As it is known that M1 macrophages have an enhanced ability to control intracellular bacteria, we examined how pre-polarized primary macrophages may affect the outcome of infection with Chlamydia. We infected pre-polarized macrophages with Chlamydia over a time course of 6 to 48 hr. At the designated time points we fixed and directly stained the macrophages for Chlamydia (Figure 2). At early time points most macrophages contained a number of small inclusions staining for Chlamydia. The number of inclusions per cell or number of cells infected appeared independent of macrophage phenotype, although this was not possible to quantify owing to the dispersed nature of the early inclusions. Interestingly, we observed these inclusions to be perinuclear in nature, consistent with previous reports on Chlamydia inclusion trafficking to the microtubule organizing centre (MTOC) [20]. At later time points (>24 hr), large inclusions comparable in morphology to those seen during Chlamydia growth in fibroblasts, were visible only in M0 and M2 infected macrophages but not M1 macrophages. Small inclusions were still seen in all macrophage phenotypes at these time points, albeit at a lower frequency than at earlier time points. Secondary infections were evident at 48 hr, visible as a cluster of infected cells (data not shown).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 2. Polarized macrophages are not infected equally.

BMDM plated at 2.5 x10⁵ cells/well in 24 well plate and pre-polarized prior to infection for 6–48 hr. Cells fixed and stained with anti-LPS antibody. Representative images of duplicate wells at 6 hr and 24 hr p.i. By 24 hr p.i. large, visible inclusions (white arrows) where seen in M0 and M2 macrophages, but not M1 macrophages. Scale bar is 50µm.

https://doi.org/10.1371/journal.pone.0069421.g002

Enumeration assays were used to quantify the effect of macrophage polarization on Chlamydia growth. As mature Chlamydia EB escape from infected cells into the supernatant, we harvested the supernatant and cell lysate for enumeration. This however presented a subsequent problem as macrophages release copious cytokines, some of which could affect Chlamydia growth during enumeration in fibroblasts. For this reason we performed ultracentrifugation to separate EB from cell debris and supernatants for each BMDM sample prior to enumeration. Using this protocol, we showed a 20-40 fold decrease in inclusion forming units (IFU) after growth in M1 macrophages versus M0 or M2 macrophages respectively (24 hr, p<0.001; 48 hr p<0.001) (Figure 3A). Low levels of Chlamydia at 6 and 12 hr post-infection (p.i.) represent residual EB left in each well after washing, since during these time points intracellular Chlamydia are expected to be in the non-infectious RB form.

We repeated experiments at 24 hr p.i. with an increased number of individual experiments (Figure 3B) to assess whether the trend of higher IFU in M2 versus M0 macrophages seen in Figure 3A was significant. For each individual experiment we normalized IFU seen in M1 and M2 macrophages to IFU in M0 macrophages to account for inter-experiment variation. This analysis showed M2 macrophages to harbor a significantly higher Chlamydia load than M0 macrophages, suggesting enhanced susceptibility to infection.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 3. M1 macrophages demonstrate lower Chlamydia load than M0 and M2 macrophages.

Polarized BMDM were infected for the indicated time points. (A) Chlamydia growth seen at 24 and 48 hr p.i. was significantly higher in M0 and M2 macrophages versus M1 macrophages as determined by repeated measures two-way ANOVA with Boneferroni post-test. Data is mean ±SEM from three independent experiments. (B) At 24 hr p.i. M2 macrophages contain significantly more IFU than M0 macrophages whilst M1 macrophages contain less. As inter-experimental IFU differed, IFU in M1 and M2 normalized to that in M0 prior to analysis. Data is mean ±SEM from six independent experiments and analyzed by one-sample t-test against hypothetical value of 1.0.

https://doi.org/10.1371/journal.pone.0069421.g003

RT-PCR demonstrates control of Chlamydia by M1 macrophages involves persistence

As measuring Chlamydia IFU provides information on the number of viable infectious Chlamydia, we examined the state of intracellular Chlamydia through RT-PCR. For these experiments any free EB were excluded by removing media and washing cells once with PBS prior to RNA extraction. We initially examined C. muridarum 16sRNA expression as an indicator of the number of viable intracellular Chlamydia. Relative to that observed in M0 macrophages, 16sRNA levels were significantly lower in M1 macrophages from 6 to 48 hr p.i. (Figure 4A), whereas 16sRNA levels remained at similar levels in M2 macrophages compared to M0 macrophages.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 4. RT-PCR of Chlamydia 16sRNA demonstrates that 16sRNA remains stable in M1 macrophages, but increases in M0 and M2.

Polarized BMDM infected with Chlamydia were harvested at times indicated. (A) Pfaffl calculated fold change in 16sRNA in polarized macrophages (M1/M2) versus M0 demonstrates significantly lower Chlamydia load in M1 macrophages at all timepoints. Treatments and timepoints analyzed separately by one-sample t-test versus hypothetical mean of 1.0 (B) Difference in ∆Ct between HPRT and 16sRNA for all samples demonstrates that decreased fold change in 16sRNA in M1 macrophages is due to increased 16sRNA expression in M0 macrophages rather than a decrease in M1. Data analyzed by repeated measures two-way ANOVA and Bonferroni’s post test. All data is mean ±SEM of three separate experiments.

https://doi.org/10.1371/journal.pone.0069421.g004

As fold change does not distinguish whether the observed change is from increased expression in the control (M0) or reduced expression in the treatment (M1), we re-analyzed this data as delta cycle threshold (∆Ct) (Figure 4B). To minimize error, concentrations of cDNA template were standardized and we kept the Ct constant for both HPRT and 16sRNA for all samples. The results show that 16sRNA, which was expressed at higher levels than HPRT in all samples, was significantly lower in M1 than M0, but no different in M2 than M0. In M1 macrophages ∆Ct 16sRNA did not change significantly from 6hr timepoint as assessed by repeated measures one-way ANOVA with Dunnett’s post test. This indicates that M0 and M2 macrophages are permissive to intracellular Chlamydia growth, whereas M1 macrophages appear to control Chlamydia.

To further examine the nature of Chlamydia control in M1 macrophages we measured bacterial gene expression. To date, persistence in C. muridarum has not been assessed through gene expression, so we selected candidate genes based on genes differentially expressed in other species of Chlamydia during persistence. Using this approach we screened genes reported to be up-regulated -Euo, IncA, GroEL2 and GroEL3 [21,22] - or down-regulated - omp1, omcB, FtsK and Tal [23-25]- in M1 and M0 macrophages after 24 hr growth (data not shown). In preliminary experiments Euo was consistently up-regulated and Tal and Omp1 were consistently down-regulated so were selected for further experiments. Analysis of the expression of these genes showed significant differential expression during growth in M1 versus M2 macrophages at 6 to 24 hr post infection (Figure 5). At 48 hr p.i. the significance was lost, most likely due to secondary infections seen in M0 macrophages leading to asynchronized infection.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 5. Control of Chlamydia by M1 macrophages involves the induction of persistence as demonstrated by RT-PCR.

Polarized macrophages were infected for the indicated timepoints. Chlamydia gene expression is not altered by growth in M2 macrophages versus growth in M0 macrophages. In M1 macrophages, a gene expression profile characteristic of persistence was seen from 6 hr p.i. infection to 24 hr p.i. whereby Euo was up-regulated and Omp1 and Tal were down-regulated. Data is mean ±SEM of three independent experiments and was analyzed using one-sample t-test against hypothetical mean of 1.0.

https://doi.org/10.1371/journal.pone.0069421.g005

Electron microscopy (EM) confirms that replication and persistence depend critically on macrophage polarization

Classically, Chlamydia persistence was identified through the observation of inclusions with enlarged RB of aberrant morphology. Immunofluorescence staining of Chlamydia detects Chlamydia antigen, but not live intact Chlamydia. For this reason we sought to confirm the profiles of replicative and persistent growth of Chlamydia with EM. We examined uninfected macrophages (not shown) and infected macrophages of each phenotype (Figure 6). As previously reported [26], and as observed with direct staining, inclusions within macrophages typically do not fuse to form a large inclusion, rather they tend to remain small with a low number of Chlamydia particles. As macrophages are rich with phagosomes and lysosomes, the detection of average smaller inclusions was difficult, however fully mature, enlarged inclusions were seen in M0 and M2 macrophages at a frequency similar to those observed with immunofluorescence and as previously reported [12,27]. Despite extensive scanning of infected M1 macrophage EM sections, such inclusions were not seen, however smaller intracellular vesicles with 1-3 amorphous particles of up to 2 µm in length were observed in most infected, but not uninfected M1 macrophages. We interpret these to be aberrant particles are Chlamydia in the persistence state, thus supporting gene expression profiles seen as mentioned above.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 6. Electron microscopy of infected polarized macrophages demonstrates productive growth in M0 and M2 macrophages but induction of Chlamydia persistence in M1 macrophages.

Polarized BMDM infected for 24 hr prior to processing for electron microscopy. Mature Chlamydia inclusions were seen in M0 and M2 macrophages (top panels). Bottom left panel shows representative uninfected M1 macrophage. Mature inclusions were not seen in infected M1 macrophages. Small inclusions with enlarged amorphous particles, consistent with Chlamydia in the persistent state, were seen frequently in infected M1 macrophages (arrows, bottom panels).

https://doi.org/10.1371/journal.pone.0069421.g006

Ethics statement

All experiments were approved by the Animal Resource Centre of the University Health Network (UHN), Toronto (Animal Use Protocol #1072). UHN maintains an animal care and use program certified by the Canadian Council on Animal Care (CCAC) and all procedures are conducted in accordance with guidelines in the Province of Ontario’s Animals for Research Act.

Chlamydia muridarum

All cells were incubated at 37° C, 5% CO₂. McCoy B fibroblasts (ATCC) were passaged in αMEM (Invitrogen) supplemented with 10% FCS (Invitrogen), 10 µg/ml gentamicin (Sigma) and 25 µg/ml vancomycin (Sigma). C. muridarum (ATCC) was cultured in McCoy B cells in this media further supplemented with 1 µg/ml cycloheximide (Sigma) and 8.8% d-glucose (Sigma) as previously reported [42]. Chlamydia EB were isolated by density ultracentrifugation using a 50%-20% Gastrografin gradient (Roche) and were stored in sucrose phosphate-glutamic acid buffer (8.5 mM Na₂HPO₄, 4 mM NaH₂PO₄, 220 mM sucrose, 0.5 mM l-glutamic acid, pH 7.4 [Sigma]) at -80° C prior to use. IFU of purified EB was determined through an enumeration assay as described below.

Animals and Bone Marrow Isolation

L-cells (a gift from Dr. F. Tsui, originally from ATCC) were plated at 5 x10⁵ cells/75 cm² flask (Corning) in 25 ml RPMI 1640 (Invitrogen) with 10 µg/ml gentamicin, 25 µg/ml vancomycin, 50 µM mercaptoethanol and cultured for one week after which L-cell-conditioned media was harvested, filtered and frozen at -20° C. Bone marrow was harvested from femurs and tibias of 8-12 week old male BALB/c mice. Mice were obtained from Jackson Laboratories and housed in a 12 hr light/dark cycle under SPF conditions as mandated by the Canadian Council on Animal Care.

Bone marrow was differentiated into macrophages using supplemented RPMI and 20% L-cell conditioned media as previously described [43]. Macrophage purity was routinely >95% as assessed by flow cytometry for CD11b and F480 (data not shown).

Bone marrow derived macrophage (BMDM) polarization, infection and repolarization

BMDM were plated in 24-well plates (Corning) in bone marrow media without L-cell conditioned media, at a density of 5 x10⁵ cells/well unless indicated. These cells were incubated for 10-12 hr prior to infection without treatment for non-polarized (M0) or BMDM were polarized to either M1 macrophages with 20 ng/ml IFNγ (RnD Systems) and 100 ng/ml LPS (Sigma) or to M2 macrophages with 20 ng/ml IL-4 (RnD Systems) as reported [44]. Polarization of macrophages in our hands matched that reported in the literature [15,45] as assessed by supernatant nitrite, arginase activity, cytokine release (Figure S1 in Supporting Information S1) and gene expression (Figure 1A). We unexpectedly found IL-10 to be upregulated in M1 macrophages but not M2 macrophages, however this has been reported in the literature [46]. For infection with Chlamydia, media was removed and 1 multiplicity of infection (MOI) was added. The plates were spun at 600 g for 20 min at 37° C to synchronize infection. The supernatant was replaced to remove free Chlamydia. As macrophage polarization is believed to be a plastic process, cytokines were replaced after infection and left in media for the duration of experiments. LPS was not replaced as prolonged exposure (>16 hr) can induce M2 polarization of macrophages due to endotoxin tolerance [47].

Enumeration assay

For enumeration of intracellular and extracellular viable infectious Chlamydia, infected wells were first scraped to lyse BMDM. Cell lysate in media was harvested and frozen at -80° C. McCoy B cells were plated at 2.5 x10⁵ cells/well in 24-well plates overnight to generate confluent monolayers. Lysate was thawed at 37° C and sonicated for 15 sec. The lysate was spun at 600 g for 5 min to remove cellular debris, prior to sterile ultracentrifugation at 19,500 g (L80-70m, Beckman Coulter) for 30 min at 4C to separate Chlamydia from conditioned media. The Chlamydia pellet was serially diluted and McCoy B cells were infected by centrifugation as above. Media was not removed and McCoy B cells were incubated for 24 hr prior to counting of infected cells.

Immunofluorescent microscopy

For direct staining of Chlamydia in BMDM or McCoy B cells, FITC-conjugated, anti-LPS antibody was used as per the manufacturer’s recommendations (Biorad). Stained samples were viewed with Nikon ECLipse TE2000U. ImageJ (NIH) was used to merge phase contrast and fluorescent images. For enumeration assays, the number of inclusions per well were counted whereby one infected cell was considered to be one inclusion forming unit (IFU).

RT-PCR

RNA was extracted from cells using TRIzol (Invitrogen) and was reverse transcribed using random primers and superscript II (Invitrogen). RT-PCR was performed with an ABI 7900HT (Applied Biosystems) system using power SYBR green (Applied Biosystems). RT-PCR data was analyzed with SDS 2.4.2 (Applied Biosystems). Primers for RT-PCR were designed with Primer Express (Applied Biosystems), tested for intra- and inter-species cross-reactivity using primerBLAST (NCBI) and ordered from ACGT corp (Figure S2 in Supporting Information S1). Primer efficiencies were assessed through construction of serial dilution standard curves and were between 90 and 105%. Multiple housekeeper genes were initially assessed, and HPRT was selected as it proved to be unaffected by polarization or Chlamydia infection (data not shown). 16sRNA is a commonly used housekeeper for Chlamydia [21] and does not appear to be differentially expressed during stress [48]. Pfaffl equation was used to calculate fold-change gene of interest in comparing treatments to controls [49].

Electron microscopy

For electron microscopy, BMDM were plated at 2 x10⁶ cells/well in 6 well plates (Corning) and were subsequently polarized and infected as above. For harvesting, macrophages were washed once with PBS (Invitrogen) and treated for 10 min with accutase (Sigma). Cells were flushed prior to fixation to dislodge from plate and processed as previously described [50]. A Hitachi H7000 electron microscope with XR-60 camera (AMT Co.) was used to examine specimens.

Statistics

GraphPad Prism 5 (TreeStar) was used for the generation of all graphs and statistical analysis. All data was expressed as mean with standard error of mean (SEM) of at least three independent experiments. To calculate statistical significance of fold changes, a one-sample t-test was used versus a hypothetical mean of 1.0, with this value representing no change in treatment relative to control. For time course analyses, repeated measures two-way ANOVA was used with Bonferroni’s post test. For all graphs: * p = 0.01–0.05, ** p = 0.001–0.01, *** p < 0.001.