Macrolides and β-Lactam Antibiotics Enhance C3b Deposition on the Surface of Multidrug-Resistant Streptococcus pneumoniae Strains by a LytA Autolysin-Dependent Mechanism

Abstract

The emergence of Streptococcus pneumoniae strains displaying high levels of multidrug resistance is of great concern worldwide and a serious threat for the outcome of the infection. Modifications of the bacterial envelope by antibiotics may assist the recognition and clearance of the pathogen by the host immune system. Recognition of S. pneumoniae resistant strains by the complement component C3b was increased in the presence of specific anti-pneumococcal antibodies and subinhibitory concentrations of different macrolides and β-lactam antibiotics for all the strains investigated. However, C3b levels were unchanged in the presence of serum containing specific antibodies and sub-MICs of levofloxacin. To investigate whether LytA, the main cell wall hydrolase of S. pneumoniae, might be involved in this process, lytA-deficient mutants were constructed. In the presence of antibiotics, loss of LytA was not associated with enhanced C3b deposition on the pneumococcal surface, which confirms the importance of LytA in this interaction. The results of this study offer new insights into the development of novel therapeutic strategies using certain antibiotics by increasing the efficacy of the host immune response to efficiently recognize pneumococcal resistant strains.

INTRODUCTION

Streptococcus pneumoniae, also termed pneumococcus, is the leading bacterial cause of acute otitis media in children, community-acquired pneumonia, and nonepidemic meningitis and a frequent cause of bacteremia (15). Despite appropriate antibiotic treatment, invasive pneumococcal disease (IPD) is a very common infection associated with high rates of morbidity and mortality worldwide. The widespread use of antibiotics has been one of the major reasons for the emergence of clinical isolates that exhibit resistance to multiple antibiotics. A major threat to fighting IPD is the appearance of strains harboring high levels of antibiotic resistance, as has been recently reported in Europe (19). The development of resistance to a wide variety of antimicrobial drugs has allowed S. pneumoniae to attain the status of a so-called “superbug” (11). The complement system is one of the first lines of defense against invading pathogens, such as S. pneumoniae, with an essential role in both innate and adaptive immunity (23). Activation of complement cascades leads to the formation of the key component C3b, which is crucial in host defense against pneumococcus by coating the microorganism and stimulating phagocytosis (23). In cases where the invading pathogen displays multidrug resistance, antimicrobial concentrations in serum may be insufficient, and therefore, the outcome of the infection largely depends on the interaction between bacterial virulence factors and host immune mechanisms. Using mice infected with a pneumococcal resistant strain and treated with subtherapeutic doses of β-lactam antibiotics, we previously found that bacterial clearance and survival were increased in mice immunized against S. pneumoniae, suggesting a synergistic effect between antimicrobial chemotherapy and the acquired immune response (5, 28). Moreover, we have recently demonstrated that the efficiency of either human or mouse neutrophils in phagocytosing resistant strains of S. pneumoniae is markedly improved in the presence of serum containing specific antibodies and subinhibitory concentrations of cephalosporins, demonstrating cooperative roles of the immune system and antibiotics (4). However, whether this synergistic effect is partially due to increased recognition of S. pneumoniae by the complement system and whether it might be extended to other groups of antibiotics are unknown. The main goal of this study was to investigate the activation of C3b against three multiresistant pneumococcal strains in the presence of specific antibodies and subinhibitory concentrations of levofloxacin (LVX), erythromycin (ERY), azithromycin (AZM), midecamycin (MDM), amoxicillin (AMX), cefotaxime (CTX), and cefditoren (CDN).

MATERIALS AND METHODS

Bacterial strains and culture conditions.

The pneumococcal isolates used for this study were strain 1515/97 (serotype 6B), strain 69 (serotype 19F), and strain 48 (serotype 23F). Bacterial strains were grown at 37°C, 5% CO₂ in Todd-Hewitt medium supplemented with 0.5% yeast extract to an optical density at 580 nm (OD₅₈₀) of 0.4 to 0.5, and small aliquots were stored at −70°C in 10% glycerol as single-use aliquots.

Isogenic lytA mutants of strains 69 and 1515/97 were constructed by genetic transformation using DNA from strain P095 (D39 lytA::aphIII), which is a kanamycin-resistant derivative of strain D39 obtained after transformation with DNA from strain R924 (R6 lytA::aphIII) (16). lytA mutants had MICs identical to those of wild-type strains. The accuracy of the constructs was confirmed by PCR. Repeated attempts to construct a lytA mutant of strain 48 were unsuccessful.

Antibiotics used and susceptibility analysis.

The antibiotics used for this study were levofloxacin (LVX), erythromycin (ERY), azithromycin (AZM), midecamycin (MDM), amoxicillin (AMX), cefotaxime (CTX), and cefditoren (CDN). Antibiotics were purchased from Sigma-Aldrich Chemical Co., St. Louis, MO, except CDN, AZM, and MDM, which were supplied by Tedec-Meiji Pharma SA, Farma-Sierra, and Menarini, respectively. Susceptibility tests were assessed three times by the agar dilution technique (9) according to the criteria of the Clinical and Laboratory Standards Institute (CLSI).

Hyperimmune serum.

Sera containing specific antibodies against S. pneumoniae were obtained by immunizing groups of 5 BALB/c mice (up to 5 weeks old) with a heat-inactivated suspension of the different strains, as previously described (4, 5). The titers of specific IgG antibodies against strains 48, 69, and 1515/97 were 251 mg ml⁻¹, 371 mg ml⁻¹, and 1,056 mg ml⁻¹, respectively (4).

C3b binding assays.

Deposition of the key complement component C3b on the surfaces of the different strains in the presence or absence of subinhibitory concentrations of each antibiotic was detected by a flow cytometry assay as previously described (2, 29). Briefly, binding to C3b was analyzed by incubating 5 × 10⁶ CFU of the different S. pneumoniae strains in 10 μl of the corresponding hyperimmune mouse serum (diluted to 20% in PBS) for 2 h with or without supplementation of 0.5 MIC and 0.25 MIC of each antibiotic. After two washes in PBS-Tween 20 (0.01%), bacteria were incubated with 50 μl of a fluorescein isothiocyanate (FITC)-conjugated polyclonal goat anti-mouse C3b antibody (ICN-Cappel) diluted 1/300 in PBS for 30 min on ice. Bacteria were fixed in 3% paraformaldehyde and analyzed on a FACS Calibur flow cytometer (BD Biosciences) using forward and side scatter parameters to gate on at least 25,000 bacteria. The results were expressed as a relative percent fluorescence index that measures not only the proportion of fluorescent bacteria positive for C3b, but also the intensity of fluorescence that quantifies the C3b bound (18, 29).

Microscopy assays.

Formation of bacterial chains in the presence of antibiotics was measured by phase-contrast microscopy. In brief, 20 μl of a bacterial suspension containing 5 × 10⁶ CFU of the different strains was incubated for 2 h in phosphate-buffered saline (PBS) or hyperimmune serum or in the presence of 0.5 MIC of each antibiotic. Samples were analyzed using a Leica microscope with a 100× phase-contrast objective (DM4000B). All images were obtained using a Leica camera (DFC360-FX) and analyzed with the Leica software AF6000-DFC.

Statistical analysis.

The results of C3b deposition assays in the presence of hyperimmune serum and sub-MICs of antibiotics were compared to results obtained with the corresponding hyperimmune serum in the absence of each antibiotic by using a two-tailed Student's t test (for two groups), whereas analysis of variance (ANOVA) with a post hoc Dunnett test was chosen for multiple comparisons. The data presented are representative of results obtained from three repeated independent experiments with at least four replicas, and each data point represents the mean and standard deviation. GraphPad InStat version 3.0 (GraphPad Software, San Diego, CA) was used for statistical analysis.

RESULTS

Susceptibility studies of the pneumococcal clinical isolates confirmed that the three strains were resistant to all the β-lactam antibiotics evaluated, with strains of serotype 23F displaying the highest levels of resistance (Table 1). All the strains were susceptible to LVX as a representative of fluoroquinolones, whereas the three strains were resistant to all the macrolides evaluated, with strains 48 and 1515/97 showing the highest MICs (Table 1). Isogenic lytA-defective strains had susceptibility patterns to all the antibiotics evaluated that were identical to those of the corresponding clinical strains (data not shown).

Table 1.

Susceptibility profiles of clinical isolates to antibiotics expressed as MICs

Antibiotic	MIC (mg/liter) for strain (serotype):
	48 (23F)	69 (19F)	1515/97 (6B)
Amoxicillin	16	2	2
Cefditoren	4	2	1
Cefotaxime	8	4	2
Levofloxacin	1	1	1
Erythromycin	1,024	16	1,024
Azithromycin	1,024	8	1,024
Midecamycin	128	2	128
Tetracycline	128	64	8
Chloramphenicol	4	4	32

The interaction of the complement system with pneumococcal clinical isolates exposed to subinhibitory concentrations of antibiotics was first investigated using LVX, ERY, and CDN. Bacterial strains were opsonized with the corresponding hyperimmune mouse serum or in PBS or heat-inactivated (65°C for 30 min) serum (HIS) as negative controls in the presence or absence of subinhibitory concentrations of LVX, ERY, and CDN (Fig. 1). Deposition of the key complement component C3b was measured by flow cytometry (Fig. 1). Opsonization of the different strains with hyperimmune serum induced C3b deposition in comparison to bacteria incubated in either PBS or HIS, confirming the fidelity of the assay for detecting functional C3b (Fig. 1). In the presence of sub-MICs of LVX, C3b binding rates for all the strains were similar to those obtained by the hyperimmune serum alone, suggesting that the antibiotic does not affect the recognition of S. pneumoniae by the complement system (Fig. 1). In contrast, incubation of bacteria with sub-MICs of either ERY or CDN in the presence of serum containing antibodies significantly increased C3b levels on the bacterial surface compared to opsonization with serum alone, showing that ERY or CDN increased complement-mediated immunity against S. pneumoniae (Fig. 1).

Fig 1.

C3b deposition on the surfaces of pneumococcal strains using HIS or hyperimmune serum (S) containing specific antibodies in the presence or absence of 0.5 MIC and 0.25 MIC of LVX, ERY, or CDN. (A to C) Deposition of C3b on the surfaces of serotype 23F, 19F, and 6B strains, respectively. (D to F) Examples of flow cytometry histograms for C3b binding on the serotype 6B strain in the presence of LVX, ERY, and CDN, respectively. The results are expressed as relative percent fluorescence indices relative to the results for sera in the absence of antibiotics. The error bars represent standard deviations, and the asterisks mark results that are statistically significant compared to those for sera in the absence of antibiotics (two-tailed Student's t test; ***, P < 0.001 for all the strains, and ***, P < 0.001 for the overall comparison of C3b binding for ERY and CDN (one-way ANOVA with a post hoc Dunnet test).

To extend our findings to other macrolides and β-lactam antibiotics, C3b deposition on the bacterial surface was explored after exposure to AZM and MDM as representatives of macrolides or AMX and CTX as representatives of the β-lactam family (Fig. 2). Incubation of the different strains with serum containing specific antibodies and subinhibitory concentrations of the antibiotics mentioned above enhanced the deposition of C3b in comparison to bacteria incubated with serum alone. Overall, results for C3b binding were consistently increased for all three strains opsonized with hyperimmune serum and sub-MIC levels of each of the macrolides and β-lactams investigated (P < 0.001; one-way ANOVA); confirming that the two groups of antibiotics trigger the activation of the complement system in the presence of specific antibodies.

Fig 2.

C3b deposition on the surfaces of pneumococcal strains using hyperimmune serum (S) containing specific antibodies in the presence or absence of 0.5 MIC and 0.25 MIC of AMX, CTX, AZM, or MDM. (A to C) Deposition of C3b on the surface of the serotype 23F strain. (D to F) Deposition of C3b on the surface of the serotype 19F strain. (G to I) Deposition of C3b on the surface of the serotype 6B strain. The results are expressed as relative percent fluorescence indices relative to the results for sera in the absence of antibiotics. The error bars represent standard deviations, and the asterisks mark results that are statistically significant compared to those for sera in the absence of antibiotics (two-tailed Student's t test; ***, P < 0.001 for all the strains, and ***, P < 0.001 for the overall comparison of C3b binding (one-way ANOVA with a post hoc Dunnet test).

The pneumococcal protein LytA is the major autolysin of S. pneumoniae; it has an important function in pathogenesis by releasing pneumolysin and plays a fundamental biological role in bacterial lysis after exposure to certain antibiotics (14, 22). To evaluate the role of LytA in the improved complement immunity mediated by antibiotics, deposition of C3b in the presence or absence of either ERY or CDN was investigated using lytA mutants of strains 69 and 1515/97. Notably, LytA-deficient mutants showed a similar pattern of C3b binding independently of the addition of subinhibitory concentrations of ERY or CDN. These results suggest that the presence of LytA, the main autolytic enzyme of the bacterium, is involved in the cooperative effect between complement immunity and these antibiotics (Fig. 3).

Fig 3.

C3b deposition on the surfaces of isogenic lytA mutant strains using hyperimmune serum (S) containing specific antibodies in the presence or absence of 0.5 MIC and 0.25 MIC of ERY or CDN. (A) Deposition of C3b on the surface of strain 69 lytA. (B) Example of a flow cytometry histogram for C3b binding on strain 69 lytA. (C) Deposition of C3b on the surface of strain 1515/97 lytA. (D) Example of a flow cytometry histogram for C3b binding on strain 1515/97 lytA. The results are expressed as relative percent fluorescence indices relative to the results for sera in the absence of antibiotics.

Finally, chain formation was assessed by phase-contrast microscopy after incubation of the different pneumococcal isolates with PBS or 0.5 MIC of AMX, CDN, ERY, and AZM. In the absence of subinhibitory concentrations of antibiotics (Fig. 4A and B, G and H, and M and N), S. pneumoniae showed a diplococcus morphology, although it also included the formation of short chains that were more abundant for the serotype 19F strain (Fig. 4G). Treatment with β-lactam antibiotics, such as AMX or CDN, did not increase chain formation compared to bacteria incubated in PBS or serum (Fig. 4). However, bacteria incubated with 0.5 MIC of ERY increased chain formation of the 23F serotype strain, whereas for the other two strains, bacterial morphology remained unchanged regardless of the presence of ERY (Fig. 4E, K, and Q). Exposure to a different macrolide antibiotic, such as AZM, did not modify the bacterial chains of the different strains compared to bacteria incubated in PBS. Overall, chain formation mediated by the antibiotics of the study played a minor role in the morphology of the pneumococcal isolates investigated.

Fig 4.

Phase-contrast microscopy images of pneumococcal strains exposed to 0.5 MIC of AMX, CDN, ERY, and AZM in comparison to incubation in the absence of antibiotics (PBS). (A to F) Serotype 23F strain. (G to L) Serotype 19F strain. (M to R) Serotype 6B strain. Bars, 7.5 μm.

DISCUSSION

The emergence of pneumococcal strains with high levels of antibiotic resistance may clearly imply a significant challenge for empirical antibiotic treatment, supporting the need for a better understanding of the host immune response against this important human pathogen (19). The combined effects of β-lactam antibiotics with specific antibodies have been shown to improve bacterial sepsis against pneumococcal resistant strains (4, 5). Humoral and cellular immunities are essential for host defense during IPD, and clearance of pneumococci from the bloodstream is greatly dependent on opsonization by complement and phagocytosis (23). Using human and mouse neutrophils, we have recently demonstrated a synergistic effect in phagocytosis between CDN or ceftriaxone and serum from immunized mice (4). The results of the present study confirm that C3b deposition on the bacterial surface was enhanced in the presence of specific antibodies and subinhibitory concentrations of either β-lactams, such as AMX, CDN, and CTX, or macrolides, such as ERY, AZM, and MDM. These results support previous findings using AMX, CTX, or CDN, demonstrating that immunization against S. pneumoniae increases the efficacy of the antibiotic treatment with β-lactams by enhancing the ability of the host immune system to efficiently recognize and destroy pneumococcal resistant strains (4, 5, 26, 27). However, sub-MIC levels of LVX did not affect opsonization by C3b in any of the investigated strains, explaining why the treatment with subtherapeutic doses of LVX in mice previously immunized against a highly encapsulated serotype 6B S. pneumoniae strain did not increase survival rates (20).

Modulation of the host immune response by macrolides has been found by other authors for either Gram-negative or Gram-positive bacteria (13). The antimicrobial activity of macrolides was markedly improved against serum-resistant strains of Escherichia coli and Staphylococcus aureus when the clinical isolates were exposed to subinhibitory concentrations of ERY or AZM in the presence of human serum (17). Our results confirmed that pneumococcal resistant strains exposed to sub-MICs of different macrolides increased C3b deposition on the bacterial surface. The findings of this study also showed that loss of the major autolytic enzyme LytA was not associated with increased C3b mediated by exposure to β-lactams or macrolides, suggesting that LytA-mediated activity might play an important role in the enhanced recognition by the complement system.

Cell wall hydrolases (CWHs) are endogenous enzymes that specifically cleave covalent bonds of the bacterial cell wall. Most bacterial species contain one or more CWHs, and it has been proposed that these enzymes are partially responsible for the irreversible effects of certain antibiotics (14). LytA is the main peptidoglycan hydrolase of S. pneumoniae and is an important virulence factor because it releases pneumolysin from the cytoplasm and proinflammatory mediators (14). Antimicrobial drugs at subinhibitory concentrations may cause subtle modifications to the bacterial envelope that create greater exposure of specific bacterial ligands that usually might be at least partially hidden. This enhanced display might allow greater recognition by the complement system and phagocytic cells. The use of cephalosporins has been associated with increased bactericidal serum activity against Pseudomonas aeruginosa, whereas ERY treatment has been linked to small breakage points in the cell wall causing disruption of the bacterial envelope of Legionella pneumophila (7, 8). In addition, it has been shown that different macrolides at sub-MICs can modify the cell surface structure and destabilize the bacterial cell outer membrane, increasing permeability and killing of P. aeruginosa (3, 12, 21). An alternative mechanism involved in the enhanced C3b deposition of S. pneumoniae might be due to increased bacterial chain length after exposure to certain antibiotics, such as ERY, that may sensitize the pathogen to complement immunity (6). Phase-contrast microscopy images after bacterial incubation with subinhibitory concentrations of different antibiotics showed that the morphology of the three pneumococcal strains exposed to β-lactams remained unchanged, displaying a typical form of diplococci or short-chain morphology. In the presence of ERY, there was occasional chaining for the 23F serotype strain, whereas the other two strains adopted a normal morphology. Incubation with AZM did not modify chain length for any of the strains investigated, suggesting that under our experimental conditions, chain formation played a minor role in the enhanced C3b deposition. All macrolide antibiotics impair bacterial protein synthesis by acting on the 50S bacterial ribosomal subunit. Macrolides at subinhibitory concentrations, but not other groups of antibiotics, inhibit the production of pneumolysin, which is an important virulence factor of S. pneumoniae involved in evasion of C3b (1, 10, 24, 25). In addition, it has been shown that macrolide compounds, such as ERY, AZM, and clarithromycin, but not roxithromycin, significantly inhibited pneumolysin and PspA. This is important from the antimicrobial chemotherapy and the immunological perspectives because the combination of both PspA and pneumolysin is additive and highly effective in inhibiting complement activation (25).

Our results demonstrate that macrolides and β-lactam antibiotics increase the recognition of pneumococcal strains by C3b in the presence of acquired immunity and support the hypothesis that the major autolysin of the bacterium is involved in this process. Current vaccine strategies to improve antibody circulating levels to efficiently activate complement immunity, together with antibiotic treatment with macrolides or β-lactams, may offer a novel strategy to overcome the risk of treatment failure for pneumococcal strains displaying high levels of antibiotic resistance.