Previous Article | Next Article ![]()
Infection and Immunity, January 2007, p. 30-34, Vol. 75, No. 1
0019-9567/07/$08.00+0 doi:10.1128/IAI.01117-06
Copyright © 2007, American Society for Microbiology. All Rights Reserved.
,
Center for Biomedical Microbiology, BioCentrum-DTU, Technical University of Denmark, Lyngby, Denmark,1 Department of Clinical Microbiology, Rigshospitalet, Copenhagen, Denmark2
Received 17 July 2006/ Returned for modification 24 August 2006/ Accepted 2 October 2006
|
|
|---|
|
|
|---|
Antigen 43 (Ag43) is a surface protein of E. coli that confers bacterial cell-cell aggregation via an intercellular Ag43-to-Ag43 handshake mechanism (8, 12, 17). Ag43-mediated aggregation can be visualized macroscopically as flocculation of cells from static liquid suspensions; hence, the name flu was originally coined for the corresponding genetic locus by Diderichsen (6). Ag43 is a member of the autotransporter protein family, which is defined by the fact that the proteins contain all information required for traverse of the bacterial membrane system and final routing to the bacterial cell surface (9). Ag43 is found in most E. coli strains, including many pathogenic strains (12, 17, 19). Ag43 is a multifunctional protein and promotes bacterial binding to some human cells as well as biofilm formation on various surfaces (12, 22).
We have previously demonstrated that Ag43-mediated aggregation enhances bacterial tolerance to bactericidal agents (21). Human blood contains multiple antibacterial agents, including PMNs. With a view to see if Ag43-mediated bacterial aggregation provided enhanced resistance toward antibacterial agents, we probed the ability of Ag43-expressing E. coli K-12 to survive exposure to human PMNs.
|
|
|---|
flu derivatives of the E. coli reference strain MG1655, except for experiments where the urinary tract infectious E. coli strain 83972 was used. MG1655 and 83972 strains used for fluorescence microscopy investigations were fluorescently tagged with gfp and yfp, respectively, OS56 (MG1655
flu attB::bla-PA1/04/03-gfp*-T0), and VR48 (83972 attB::bla-rrnBP1-yfp-T0) as described previously (7), and MS528 (MG1655
flu
fim) was used in experiments not requiring fluorescence tagging. Transformants were selected on ampicillin-containing plates and checked for their ability to fluoresce. Proper insertion of the fragment into the
attachment site was confirmed by PCR. Bacteria were grown at 37°C on solid or in liquid LB supplemented with appropriate antibiotics. |
View this table: [in a new window] |
TABLE 1. Escherichia coli strains and plasmids used in this study
|
flu
fim), OS56, and VR48. All strains carrying the Ag43-expressing plasmid showed heavy aggregation, and the whole batch was used in the experiments. The RGD motif at position 208 in Ag43 was mutated using the QuikChange site-directed mutagenesis kit (Stratagene), such that RGD was changed to KGE using primers 655 (5'-CCGGGCAGTTTGTTAAAGGGGAAGCCGTACGCACAACCATC-3') and 656 (5'-GATGGTTGTGCGTACGGCTTCCCCTTTAACAAACTGCCCGG-3'). The resulting plasmid, pSF2, with mutant flu was introduced into E. coli strains MS528 and OS56. Isolation of human PMNs. Human blood samples were obtained from healthy volunteers by venous puncture and collected in BD Vacutainers with 0.129 M sodium citrate (388335; Becton Dickinson). The blood was mixed with dextran (T-500) 1:5, and the erythrocytes were left to precipitate for 40 min. The supernatant was applied to Lymphoprep (Axis-Shield PoC) and centrifuged at 850 x g for 15 min at 5°C. The pellet was treated with 2 ml 0.2% NaCl to lyse remaining erythrocytes. Lysis was terminated by addition of 2 ml 1.6% NaCl and 6 ml RPMI 1640. The cells were centrifuged at 350 x g for 10 min at 5°C, the supernatant was discarded, and the PMNs were resuspended in RPMI 1640 buffer at 5 x 106 PMNs/ml.
Blood sensitivity. Bacteria were tested for sensitivity to human whole blood by incubating the individual E. coli strains with blood to a final blood concentration of 75%. Overnight cultures of bacteria grown in LB were standardized and suspended in phosphate-buffered saline (PBS) (to a density of 4 x 109 CFU/ml). Aliquots of cells (triplicates) were kept in microtiter plates at room temperature for 2 h to allow aggregation, followed by addition of fresh human blood or RMPI 1640 buffer (control). The mixtures were shaken at 200 rpm and 37°C for 15 min, and serial dilutions were plated.
Bacterial internalization by PMNs. Internalization of bacteria was examined by confocal scanning laser microscopy (CSLM). Overnight cultures of fluorescent bacteria (green fluorescent protein [GFP] or yellow fluorescent protein [YFP]) were standardized and suspended in PBS to a density of 4 x 109 CFU/ml. Aliquots of cells (50 µl) were kept in microtiter plates at room temperature for 2 h to allow aggregation, followed by the addition of 150 µl freshly purified PMNs or RPMI 1640 buffer for control. Bacterium-PMN interaction was visualized after incubation for 45 min at 37°C and 200 rpm. All CSLM images were taken using a Zeiss LSM 510 system (Carl Zeiss) equipped with an argon laser and a helium-neon laser for excitation of the fluorophores. Simulated fluorescence projections were generated using the IMARIS software package (Bitplane AG). Images were further processed for display using the PhotoShop software (Adobe).
Internalization of E. coli in PMNs was assessed by 45-min incubation of bacteria and PMNs prepared as above, followed by the addition of gentamicin (18 µg/ml) and another 30 min of incubation. Bacterial survival in PMNs was tested by preparing mixtures of PMNs and bacteria as described above, followed by incubation for 0, 30, 60, and 120 min at 37°C and 200 rpm. Bacterial viability was determined by plating serial dilutions. To block bacterial uptake by phagocytosis, the isolated PMNs were treated with 3 µM cytochalasin D (C8273; Sigma-Aldrich) for 1 h at 37°C before incubation together with bacteria followed by gentamicin treatment as described above.
Fluorescence-activated cell sorter analysis of phagocytosis. Phagocytosis of bacteria by PMNs was analyzed using FACSort (Becton Dickinson) equipped with a 15-mW argon ion laser turned to 488 nm for excitation. Light scatter and exponentially amplified fluorescence parameters from at least 10,000 events were recorded in list mode after gating on light scatter to avoid debris, cell aggregates, and bacteria on its own. The instrument was calibrated using Calibrite (Becton Dickinson). PMNs were identified according to their morphology and their content of DNA. GFP-expressing E. coli cells, both nonexpressing control cells and Ag43-expressing cells, were mixed with PMNs as above. External E. coli cells were removed by washing in cold PBS. The remaining PMNs were fixed in fluorescence-activated cell sorter lysing solution (Becton Dickinson) containing 150 µg/ml propidium iodide (P-4170; Sigma) for discrimination of eukaryotic cells according to their DNA content.
|
|
|---|
![]() View larger version (19K): [in a new window] |
FIG. 1. E. coli exposed to human blood and internalization of Ag43-expressing E. coli by PMNs. (A) E. coli strain MG1655 flu, harboring control plasmid (pACYC184) or Ag43-encoding plasmid (pPKL330), was incubated with blood or buffer for 15 min, and bacterial viability was determined. Results are presented as survival in blood compared with buffer and means from three independent experiments; error bars indicate standard deviations (values were 4.4 x 108 and 6.0 x 108 CFU/ml in buffer for control plasmid and Ag43, respectively). (B) Bacterial uptake by human PMNs monitored by flow cytometry to determine the percentage of phagocytic PMNs after bacterial incubation for 45 min at 37°C. Phagocytic green fluorescent PMNs were detected after incubation with GFP-expressing E. coli control (pACYC184) and Ag43-expressing (pPKL330) cells followed by a wash of external cells. Values are means of results from 10 samples, and error bars indicate standard deviations (unpaired t test, P < 0.0001). The experiment was performed in duplicate with separate cultures and blood donors. (C) Internalization of bacteria in PMNs was assessed by addition of gentamicin (18 µg/ml) to mixtures of PMNs with control or Ag43 E. coli; bacterial viability was determined after 30 min of incubation. Error bars indicate standard deviations of results from three independent experiments. (D) Internalization of Ag43-expressing bacteria in PMNs was blocked by treatment of PMNs with cytochalasin D, leading to lack of protection against gentamicin. Error bars indicate standard deviations of results from two separate experiments. (E) Survival of phagocytosed Ag43-expressing E. coli inside PMNs. Coincubation of bacteria and PMNs for 0 to 120 min at 37°C was followed by serial dilutions and plating. Bacterial viabilities are presented relative to survival in buffer and are means of results from three independent experiments, and error bars indicate standard deviations.
|
![]() View larger version (106K): [in a new window] |
FIG. 2. Bacterial adherence to human PMNs. Fluorescent E. coli cells, nonexpressing control (pACYC184), Ag43-expressing (pPKL330), and AIDA-expressing (pLH44) cells, were mixed with PMNs. (A to C) Representative fluorescence micrographs of E. coli MG1655 flu (GFP), demonstrating that the majority of Ag43-expressing cells were taken up by PMNs in an opsonin-independent manner while neither control cells nor AIDA-expressing cells were taken up by PMNs. (D) E. coli MG1655 flu expressing the mutant Ag43 (pSF2; with RGD motif changed to KGE) was also taken up by PMNs. (E and F) Fluorescence micrographs of E. coli 83972 (YFP), nonexpressing control and Ag43-expressing cells, illustrating Ag43 adherence and uptake to PMNs. All images were obtained after coincubation of E. coli cells and PMNs for 60 min at 37°C. Bars, 40 µm (A, B, C, E, and F), 20 µm (D).
|
Ag43-assisted bacterial uptake by PMNs does not require the RGD motif. Proteins containing an RGD motif have been implicated in binding and internalization of microorganisms by human integrin receptors (10). Ag43 contains an RGD motif at position 208, and it seemed conceivable that this motif played a role in PMN uptake of Ag43-expressing bacteria. To test this, we made a version of Ag43 with an RGD-to-KGE change by PCR-assisted site-directed mutagenesis of the corresponding gene. However, bacteria expressing the KGE version of Ag43 were taken up by PMNs with similar efficiency as bacteria that expressed the wild-type version of Ag43 (Fig. 2D). The RGD-to-KGE change does not affect the aggregation properties of Ag43.
The bacterial uptake is Ag43 specific rather than dependent on aggregation.
The lack of RGD motif-dependent uptake led us to speculate that aggregation per se could provoke PMN uptake. To investigate this, we used another autotransporter protein, AIDA, which shares
25% sequence identity with Ag43 and also causes bacterial aggregation (23). Meanwhile, aggregating bacteria expressing AIDA were not taken up by PMNs any better than the controls (Fig. 2). Taken together, the results indicate that Ag43-assisted bacterial uptake by PMNs does not require RGD motifs nor does bacterial aggregation by itself suffice to provide uptake.
Ag43-expressing internalized cells survive inside PMNs. PMN phagocytosed bacteria are usually killed rapidly (14, 18), and we therefore set out to investigate the fate of phagocytosed Ag43-expressing bacteria. Accordingly, the survival rate of such bacteria was monitored for 120 min (Fig. 1E). It transpired that, during this period of time, no reduction in bacterial viability was observed, indicating that the tight aggregates of Ag43-expressing bacteria tolerated antimicrobial mechanisms inside the PMNs for extended periods of time. In contrast, bacteria devoid of Ag43 were not internalized at all and did not survive as well in whole blood compared to Ag43-expressing bacteria (Fig. 1A and 2). Ag43-mediated cell aggregation significantly protects bacteria against exposure to hydrogen peroxide (21). Since PMNs use peroxides to eliminate bacteria, similar protective mechanisms could be at play inside the PMNs. It therefore seems that Ag43 not only confers efficient bacterial uptake in PMNs but also seems to provide a survival mechanism for the bacteria after phagocytosis.
In this study, we have described a possible bacterial evasion technique, "hiding-in-the worst-place," i.e., inside the killer cells of the host defense. The bacterial strategy for the uptake and survival of Ag43-expressing E. coli might be evasion, survival, and possibly transport. Since the PMNs play a central role in the host defense, it should follow that the ability to foil PMN-mediated killing should be a major advantage for an infecting microorganism. E. coli is the primary cause of urinary tract infections in humans (24). The antibacterial defense of the urinary tract is highly dependent on innate immunity where PMNs are crucial players (25). Arguably, Ag43-mediated uptake and survival in PMNs could play an important role in this type of infections. The sessile state of bacteria (biofilms, microcolonies, and aggregates) shows an increased tolerance to a variety of antimicrobial measures compared to its planktonic counterparts (5, 20). Ag43 has been shown to be expressed by E. coli in the urinary tract in the mouse model (1). By expressing Ag43, the bacteria become readily internalized, and we speculate that they resist elimination internally due to aggregation. Firmly inside the PMN, the bacteria are protected from the remaining host defenses, as evidenced by the improved survival observed in whole blood. However, it needs to be clarified whether it is the internalization or the clumping which protects the cells. Even killing by antibiotics in general may be evaded within the PMNs. It is interesting to speculate that, once internalized, the intruder may be favored even further by being transported around the body by macrophage scavengers, which in turn phagocytose the dead PMNs. In this way, the PMNs become Trojan horses, which might liberate their content of bacteria at times and places that will increase the probability of successful infection.
Published ahead of print on 9 October 2006. ![]()
Supplemental material for this article may be found at http://iai.asm.org/. ![]()
|
|
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»