CpG ODN enhances uptake of bacteria by mouse macrophages

Abstract

Unmethylated CpG motif in synthetic oligodeoxynucleotide (CpG ODN) or bacterial DNA is well recognized for its role in innate immunity, including enhancing production of NO and cytokines by macrophages. In the present study, we demonstrated the effect of CpG ODN on the phagocytic uptake of bacteria by macrophages. Flow cytometric analysis of mouse macrophages (RAW 264·7) incubated with fluorescein isothiocyanate (FITC)-labelled Burkholderia pseudomallei, Salmonella enterica serovar Typhi or Escherichia coli showed that CpG ODN increased the uptake of these bacteria by mouse macrophages. The enhancement of bacterial uptake by CpG ODN was concentration-dependent. The increase of bacterial uptake by CpG ODN-activated macrophages shown above is consistent with the result of bacteria internalization study using a standard antibiotic protection assay. There was also an increase in the rate and degree of multi-nucleated giant cell formation, phenomena which have been shown previously to be unique when the cells were infected with B. pseudomallei. These observations may provide significant insights for future investigation into host cell–pathogen interaction.

Introduction

Bacterial genomic DNA or synthetic oligonucleotide containing unmethylated cytosine followed by guanine (CpG ODN) is now recognized as a potent immunostimulator. This family of compounds can induce the production of a variety of proinflammatory cytokines (both in vitro and in vivo) including TNF-α, IL-12 and IFN-α [1–4]. Unmethylated CpG oligodeoxynucleotides in the context of particular flanking sequences (CpG motif) are required for specific activation [5]. Although the mechanism for its action is still not understood fully, current data indicate that the CpG DNA requires complexing with Toll-Like Receptor 9 (TLR 9) [6–10]. The interference of endosomal acidification by chloroquine abolishes cytokine production by suppressing MAPK activation, thus indicating that endocytosis is required for TLR 9 signalling [7,11].

The specific CpG ODN 1826 (5′ TCCATGACGTTCCT GACGTT 3′) used in the present study is known to be an excellent immunostimulator in the murine model, both in vitro and in vivo [12]. In addition to the cytokine production mentioned above, in the present study we demonstrated that CpG ODN also increased bacterial uptake by macrophages, using Burkholderia pseudomallei as a model. B. pseudomallei is a causative agent of melioidosis, a potentially fatal disease in tropical regions, e.g. northern Australia and south-east Asian countries. This facultative intracellular Gram-negative bacillus can survive and multiply inside phagocytic and non-phagocytic cells [13]. After internalization, B. pseudomallei can escape from the membrane-bound phagosome into the cytoplasm [13]. Although the mechanism for intracellular survival of B. pseudomallei in the macrophages is not well understood, recent data suggest that this bacterium can invade macrophages without triggering a substantial quantity of inducible nitric oxide synthase (iNOS), which plays a major role in bacterial intracellular killing [14]. This unique interaction may be related, at least in part, to its unusual lippolysaccharide (LPS) structure, shown previously to possess a relatively weak macrophage activating activity [15,16]. Moreover, unusual morphological changes of the cells infected with this bacterium have been reported. B. psueodmallei is able to induce cell-to-cell fusion, leading to a multi-nucleated giant cell formation (MNGC) in both phagocytic and non-phagocytic cells [17,18]. Staining of the infected cells with rhodamine-conjugated phalloidine indicated the actin rearrangement into a comet-tail formation occurs which most probably facilitates the spreading of B. pseudomallei to neighbouring cells [17].

Materials And Methods

Reagents and cell culture

Mouse macrophage cell line (RAW 264·7) was obtained from American Type Culture Collection (ATCC, Rockville, MD, USA). If not indicated otherwise, the cells were cultured in Dulbeccco's modified Eagles’ medium (DMEM) (Gibco Laboratories, Grand Island, NY, USA) supplemented with 10% fetal bovine serum (FBS) (HyClone, Logan, UT, USA) at 37°C under a 5% CO₂ atmosphere.

B. pseudomallei strain 844 (arabinose-negative strain) was isolated from a patient admitted to Srinagarind Hospital in the melioidosis endemic Khon Kaen province of Thailand. The bacterium was identified originally as B. pseudomallei based on its biochemical characteristics, colonial morphology on selective media, antibiotic sensitivity profiles and reaction with polyclonal antibody [15, 19, 20]. Salmonella enterica serovar Typhi (S. typhi) and Escherichia coli used for comparison throughout these experiments was maintained at Ramathibodi Hospital (Mahidol University, Bangkok, Thailand) and kept as stock culture in our laboratory.

Nuclease-resistant phosphorothioate ODNs were kindly made available by A. M. Krieg (Coley Pharmaceutical Group, Wellesley, MA, USA). The sequences of the ODNs used were the immunostimulatory CpG ODN 1826 (5′ TCCATGACGTTCCT GACGTT 3′) and non-CpG containing ODN 1982 (5′ TCCAG GACTTCTCTCAGGTT 3′) served as ODN control. The possible contamination with a trace amount of lipopolysaccharide (LPS) in these samples was determined by a Limulus amoebocyte lysate assay (LAL assay, BioWhittaker, Walkersville, MD, USA) and expressed as endotoxin units (EU) per ml. The lower limit for the detection of LPS in our laboratory was 0·03 EU/ml. The Salmonella typhimurium LPS (Sigma, St Louis, MO, USA) used as reference had an activity of 0·230 EU/ng. Both ODNs used exhibited endotoxic activity less than 7·5 × 10⁻⁸ EU/ng.

Labelling of bacteria with fluorescein isothiocyanate (FITC)

B. pseudomallei, S. typhi and E. coli were cultured in tryptic soy broth (TSB) at 37°C until a concentration of approximately 1 × 10⁹ colony-forming units (CFU)/ml was attained. The bacterial suspension was incubated with an equal volume of 0·1 m carbonate buffer pH 9·5 containing 1 mg of FITC. After 30 min of incubation at 37°C, the FITC-conjugated bacteria were washed thoroughly with phosphate-buffered saline (PBS). The viability of labelled bacteria was determined by colony count and the final concentration was adjusted with PBS to approximately 2·5 × 10⁸ CFU/ml.

Uptake of labelled bacteria by murine macrophages

Macrophages (1 × 10⁶) were seeded in a 24-well plate and incubated overnight at 37°C. The cells were washed three times with PBS and then exposed to FITC-conjugated bacteria at a multiplicity of infection (MOI) of 10 : 1. The plates were centrifuged at 250 g for 10 min to facilitate contact between bacteria and macrophages before incubating at 37°C. After 1 h of incubation, the cells were washed three times with PBS and then fixed with 1% paraformaldehyde in PBS for 30 min before analysing by flow cytometry using FACStar^PLUS. A total of 10 000 events were acquired in the list mode using lysys ii software. Debris was excluded by light scatter gating. The gated population was analysed for FL-1 intensity. The labelled cells counted represented those associated with bacteria on the surface and/or inside the cells.

Intracellular survival and multiplication of bacteria in mouse macrophages

To determine intracellular survival and multiplication of the bacteria, a standard antibiotic protection assay was performed as described previously [14]. The macrophages (1 × 10⁶) were cultured overnight in a 24-well plate and then exposed to bacteria at MOI 2 : 1. After 1 h of incubation, extracellular bacteria were removed and the cells were washed three times with 2 ml of PBS. Residual bacteria that adhered to the cell surface were killed by incubating in DMEM containing 250 µg/ml kanamycin (Gibco Laboratories) for 1 h. The cells were washed three times with PBS and the intracellular bacteria were liberated by lysing the macrophages with 0·1% Triton X-100 and plated the released bacteria on tryptic soy agar. The number of intracellular bacteria was determined by bacterial colony counting.

Quantification of multi-nucleated giant cell (MNGC) formation from macrophages infected with B. pseudomallei

In order to quantify the degree of MNGC, the macrophages (1 × 10⁶) were cultured first overnight on a coverslip, as described previously [17]. After 1 h of incubation with B. pseudomallei at MOI 2 : 1, the macrophages were washed three times with PBS and then incubated in DMEM containing 250 µg/ml kanamycin (Gibco Laboratories) for 2 h to kill residual extracellular bacteria. The macrophages were incubated further in DMEM containing 20 µg/ml kanamycin before being visualized for MNGC formation by light microscope at a magnification of × 400. At different time intervals, the coverslips were washed with PBS, fixed for 15 min with 1% paraformaldehyde and then washed sequentially with 50% and 90% ethanol for 5 min each. The coverslips were air-dried before staining with Giemsa [17]. For enumeration of MNGC formation, at least 1000 nuclei per coverlsip were counted and the percentage of multi-nucleated cells was calculated. The MNGC was defined as the cell possessing more than one nuclei within the same cell boundary.

Results

In order to investigate the effect of CpG ODN (1826) on the ability of macrophages to take up bacteria, the cells were treated overnight with various concentrations of CpG ODN (1826) and non-CpG ODN (1982) before being infected with B. pseudomallei, S. typhi or E. coli conjugated with FITC at 37°C, as described in Materials and methods. As shown in Fig. 1, CpG ODN (1826) was able to increase the percentage of cells associated with all three bacteria. The enhancement of bacterial uptake by CpG ODN (1826) was also found to be dose-dependent. However, when the cells were preactivated with non-CpG containing ODN (1982), the percentage of macrophages associated with bacteria was not enhanced, even when tested at a concentration as high as 1 µg/ml. Similar experiments were carried out at 4°C in order to ensure that the signals originated from the internalized bacteria and not from the bacteria adhered onto the cell surface. The results showed that at 4°C a very low percentage of the cells associated with bacteria was observed, even when the concentration of CpG ODN as high as 1 µg/ml was used (Fig. 1, insert). These observations indicated that the increase in the percentage of macrophage cells associated with bacteria at 37°C was due to internalization and not adherence of FITC-conjugated bacteria.

Fig. 1.

CpG ODN (1826) enhances percentage of bacteria-associated cells. Mouse macrophages (RAW 264·7) were exposed to live FITC-conjugated B. pseudomallei (a), FITC-conjugated S. typhi (b) or FITC-conjugated E. coli (c) at MOI 10 : 1 as described in Materials and methods. After 1 h of incubation at 37°C, the association of FITC-labelled bacteria with macrophages was analysed by flow cytometry. Macrophages preincubated overnight with CpG ODN (1826) (•) but not with non-CpG ODN (1982) (○) exhibited increased percentage of bacteria-associated cells in a dose-dependent manner. It should be noted that when the experiments were carried out at 4°C instead of 37°C, the only a low level of association could be observed (insert), suggesting that at 37°C, CpG ODN (1826) enhanced bacterial internalization but not adherence. These results are representative of three separate experiments.

In order to determine the effect of CpG ODN on intracellular survival and multiplication of bacteria, the macrophages were treated overnight with 1 µg/ml of either CpG ODN, non-CpG ODN or LPS before being infected with B. pseudomallei or S. typhi at MOI 2 : 1. After 1 h of infection, the macrophages were washed three times and incubated further for 1 h at 37°C in the medium containing 250 µg/ml of kanamycin to kill extracellular bacteria that might remained adhering to the cell surface. The intracellular bacteria were liberated, plated on tryptic soy agar and the colonies were counted as described. It should be mentioned that under this condition, kanamycin was effective in killing more than 99% of both B. pseudomallei and S. typhi (data not shown). In the presence of CpG ODN the number of B. pseudomallei and S. typhi inside the macrophages was increased significantly while no such increase was observed in the cells treated with non-CpG ODN or LPS (Fig. 2). However, when the CpG ODN-activated macrophages were pretreated with cytochalasin D, a chemical that inhibits polymerization of microfilament, the number of intracellular bacteria was decreased markedly (Table 1), confirming that the CpG ODN enhanced the bacterial uptake of macrophages. These results are also consistent with the data presented in Fig. 3 showing the CpG ODN enhancement of MNGC formation. We have observed previously that B. pseudomallei was able to induce MNGC formation in both phagocytic and non-phagocytic cells [17]. In this experiment, the degree of MNGC formation in B. pseudomallei-infected macrophages culture was qunatitated after Giemsa staining. The results presented in Fig. 3 showed that when the macrophages preactivated with CpG ODN (1826), as much as 20% of the infected culture exhibited MNGC formation with 4 h. At 8 h after infection, more than 80% of the infected cells were found to be in a stage of MNGC formation. In contrast, in the absence of CpG ODN, no MNGC could be detected at 4 h, but became detectable at 6 h (10%) and increased gradually to 30% at 8 h after infection. Similar results were observed in the macrophages treated with non-CpG. It should be noted that, in the uninfected cells, less than 1% of MNGC was observed in both CpG ODN or non-CpG ODN treated cells. These results suggest that CpG ODN not only accelerated the rate of MNGC but also increased the degree of MNGC formation in the macrophages infected with B. pseudomallei.

Fig. 2.

CpG ODN increases internalization of bacteria. Mouse macrophages were pretreated overnight with 1 µg/ml of CpG ODN (1826), non-CpG containing ODN (1982) or 0·1 µg/ml of LPS before incubating for 1 h with B. pseudomallei or S. typhi at MOI 2 : 1. Extracellular bacteria were removed and killed, then intracellular bacteria released after lysis were determined by standard antibiotic protection assay as described in Materials and methods. Data represent mean ± s.d. of three separate experiments, each carried out in duplicate. *P < 0·05 by Student's t-test. ▪, B. pseudomallei; , S. typhi.

Table 1.

Inhibition of CpG ODN-enhanced bacterial uptake by cytochalasin D

Treatment	No of bacteria(mean±s.d.)
	Without cytochalasin D	With cytochalasin D (2.5 µg/ml)
P. pseudomallei	3.80 × 105 ± 0.28	1.43 × 103 ± 0·03
B. pseudomallei+ 1826	7.41 × 105 ± 0.50	1.10 × 103 ± 0·07
B. pseudomallei+ 1982	3.55 × 105 ± 0.41	1.19 × 103 ± 0·10
S. typhi	1.02 × 105 ± 0.14	1.28 × 103 ± 0·20
S. typhi+ 1826	6.69 × 105 ± 1.47	1.05 × 103 ± 0.02
S. typhi+ 1982	1.04 × 105 ± 0.09	1.01 × 103 ± 0.16

Fig. 3.

CpG ODN enhances the rate and extent of multi-nucleated giant cell (MNGC) formation of B. pseudomallei-infected macrophages. Mouse macrophages were pretreated overnight with 1 µg/ml of CpG ODN (1826) (•) or non-CpG containing ODN (1982) (○) before being infected with B. pseudomallei at MOI 2 : 1. Untreated cells served as control (□). At 4, 6 and 8 h postinfection, the cells were fixed, stained with Giemsa, and the number of MNGC was determined by microscopic examination (400 ×). Data represent mean ± s.d. of three separate experiments.

Chloroquine is an endosomal maturation/acidification inhibitor shown previously to abolish CpG ODN-mediated cytokine production and cell activation [7,11]. In order to investigate the possible inhibitory effect of chloroquine on the bacterial uptake of macrophages, the macrophages were preincubated first with or without 2 µg/ml of chloroquine for 2 h before being activated with 1 µg/ml of CpG ODN (1826). After overnight incubation, the cells were infected with B. pseudomallei or S. typhi at MOI 2 : 1 for 1 h. The extracellular bacteria were removed and the adhering bacteria on the macrophage surface were killed with kanamycin, as described above. After 1 h, the intracellular bacteria were determined by a standard antibiotic protection assay as described. The results presented in Fig. 4 showed that while the chloroquine could not inhibit the internalization of either B. pseudomallei or S. typhi, it was able to reduce the number of intracellular bacteria from the CpG ODN (1826)-activated macrophages. Similarly, the enhancement of MNGC formation induced by CpG ODN was also decreased significantly when the infected cells were pretreated with chloroquine (Fig. 5), thus indicating that the increase of MNGC formation by CpG ODN (1826) also required endosomal maturation/acidification (Fig. 5). The circumstantial evidence presented in these experiments support the contention that the CpG ODN enhancement of bacterial uptake by macrophages required endosomal maturation/acidification.

Fig. 4.

Effect of chloroquine on bacterial internalization by macrophages preactivated with CpG ODN (1826). Mouse macrophages were pretreated with chloroquine (2 µg/ml) for 2 h before incubating overnight with 1 µg/ml of CpG ODN (1826). The cells were then exposed to either B. pseudomallei or S. typhi at MOI 2 : 1 for 1 h. Intracellular bacteria were determined as described. Data represent mean ± s.d. of three separate experiments, each carried out in duplicate. *P < 0·05 by Students's t-test. ▪, B. pseudomallei; , S. typhi.

Fig. 5.

Chloroquine reduces multinucleated cell formation in CpG ODN-activated macrophages. Mouse macrophages were cultured on coverslips in the presence or absence of 2 µg/ml chloroquine for 2 h before being activated overnight with 1 µg/ml of CpG ODN (1826). The cell cultures were then exposed to B. pseudomallei at MOI 2 : 1. After 8 h of exposure, the cells were fixed and stained with Giemsa, and the number of MNGC was determined microscopically (400×). Data represent mean ± s.d. of three separate experiments, each carried out in duplicate *P < 0·05 by Student's t-test.

Discussion

Bacterial genomic DNA containing unmethylated cytosine followed by guanine (CpG DNA) is a well-recognized immunomodulator. It appears to function as one of the ‘danger signals’ to trigger innate immunity against microbial infection, as well as modulating adaptive immune response [1]. Unmethylated CpG oligodeoxynucleotides (CpG ODNs) in the context of particular flanking sequences (CpG motif) are a required structure for this specific recognition [1,5]. Subsequently, it was shown that the short synthetic oligodeoxynucleotide-containing CpG motif could also serve as a danger signal for the vertebrate immune system [21]. It has been shown that CpG ODN is a potent immunomodulator that induces significant protective immunity in a number of infections caused by bacteria, virus and protozoa. For instance, mice injected with CpG ODN have been shown to respond with a rapid production of IL-12 and IFN-γ leading to enhanced resistance to infection by Listeria monocytogenes or Leishmania major [22,23]. Activation of the innate immune cells, e.g. macrophages, by CpG DNA or synthetic CpG ODN results in the induction of an array of proinflammatory cytokines such as TNF-α, IL-12 and type I interferon both in vivo and in vitro [1–3]. Moreover, the synthetic CpG ODN, including the one used in this study (1826), can also stimulate the production of bactericidal effector substances such as nitric oxide [24]. The molecular mechanism of CpG DNA or CpG ODN in immunostimulation is almost identical to that of the more classical inducers of innate immune system, i.e. LPS [21]. In fact, both the CpG DNA and CpG ODN, like the LPS, can induce septic shock in mice [25]. To activate the immune cells, CpG DNA or CpG ODN requires TLR9 while LPS requires TLR4 [8,9]. In this study, we have demonstrated a novel role of CpG ODN in macrophage activation. The results from flow cytometry, together with those from intracellular bacterial survival and MNGC formation, suggested that CpG ODN (1826) was able to increase uptake of both intracellular (B. pseudomallei and S. typhi) and extracellular (E. coli) by mouse macrophages (Figs 1 and 2). The enhancement of phagocytic activity induced by CpG ODN (1826) was also concentration-dependent (Fig. 1). An increase in the macrophage ability to take up bacteria by CpG ODN (1826) was not observed in the cells activated with a classical immunostimulator such as LPS (data not shown).

To initiate signalling, the CpG DNA-stimulated cells recruit MyD88 to an endosome-like vesicular structure where it co-localizes with CpG DNA [1,5]. Inhibitors of endosomal maturation/acidification, e.g. chloroquine, can prevent cytokine production by inhibiting phosphorylation of p38 and JNK [11]. In the present study, we have demonstrated further that the increase of phagocytic activity and MNGC formation of CpG ODN-activated macrophages (Figs 3 and 4) was abolished when the cells were pretreated with chloroquine, thus suggesting that the enhancement of bacterial uptake by CpG ODN (1826) also requires endosomal maturation/acidification. These observations confirm and extend the data reported previously by other investigators [7,11].

Synthetic CpG containing ODN has been shown to activate the production of several proinflammatory cytokines, such as TNF-α and IL-12. The production of these cytokines by macrophages, in turn, stimulates NK cells to secrete IFN-γ [26]. The latter plays a critical role in immunity against bacterial infection, for instance, by stimulating NO production to interfere with intracellular bacterial survival and multiplication inside the macrophages and MNGC formation induced by B. pseudomallei [14]. The enhancement of bacterial uptake by CpG ODN-activated macrophages, together with the stimulation of IFN-γ secretion from NK cells, may eventually result in enhancement of bacterial clearance by the host. However, the exact molecular mechanism by which the CpG DNA or CpG ODN increase the phagocytic activity of macrophages is not understood fully and remains to be investigated.