O-antigen-deficient Francisella tularensis Live Vaccine Strain mutants are ingested via an aberrant form of looping phagocytosis and show altered kinetics of intracellular trafficking in human macrophages - PubMed (original) (raw)
O-antigen-deficient Francisella tularensis Live Vaccine Strain mutants are ingested via an aberrant form of looping phagocytosis and show altered kinetics of intracellular trafficking in human macrophages
Daniel L Clemens et al. Infect Immun. 2012 Mar.
Abstract
We examined the uptake and intracellular trafficking of F. tularensis Live Vaccine Strain (LVS) and LVS with disruptions of wbtDEF and wbtI genes essential for synthesis of the O antigen of lipopolysaccharide. Unlike parental bacteria, O-antigen-deficient LVS is efficiently killed by serum with intact complement but not by serum lacking terminal complement components. Opsonization of O-antigen-deficient LVS in serum lacking terminal complement components allows efficient uptake of these live bacteria by macrophages. In the presence of complement, whereas parental F. tularensis LVS is internalized within spacious pseudopod loops, mutant LVS is internalized within tightly juxtaposed multiple onion-like layers of pseudopodia. Without complement, both parental and mutant LVSs are internalized within spacious pseudopod loops. Thus, molecules other than O antigen are important in triggering dramatic pseudopod extensions and uptake by spacious pseudopod loops. Following uptake, both parental and mutant LVSs enter compartments that show limited staining for the lysosomal membrane glycoprotein CD63 and little fusion with secondary lysosomes. Subsequently, both parental and mutant LVSs lose their CD63 staining. Whereas the majority of parental LVS escapes into the cytosol by 6 h after uptake, mutant LVS shows a marked lag but does escape by 1 day after uptake. Despite the altered kinetics of phagosome escape, both mutant and parental strains grow to high levels within human macrophages. Thus, the O antigen plays a role in the morphology of uptake in the presence of complement and the kinetics of intracellular growth but is not essential for escape, survival, altered membrane trafficking, or intramacrophage growth.
Figures
Fig 1
Generation and confirmation of the LVS Δ_wbtDEF_ mutant. (A) The 8.9-kb suicide plasmid pSMP22-RT1 contains the 3′ two-thirds of the wbtD gene with no stop codon, a frame-shifted wbtF gene missing both the start and stop codons, and a chloramphenicol resistance gene cassette (camR) between the mutated wbtD and wbtF genes. The locations of the primer pairs (F and R) used for amplification of the various regions on the plasmid vector or the chromosome and ampR, a gene cassette conferring resistance to ampicillin, are indicated. (B) Genetic organization of a portion (9 kb) of the F. tularensis LVS O-antigen gene cluster. H, HindIII. The location of the primer pair used to amplify the wbtE gene is indicated. (C) PCR analysis of the LVS Δ_wbtDEF_ mutant (lane 1), the parental LVS (lane 2), and the suicide plasmid pSMP22-RT1 (lane 3). Lane M, 1-kb plus DNA ladder. FR, flanking region. (D) Southern hybridization analysis of the genomic DNA of the LVS Δ_wbtDEF_ mutant (lane 1) and the parental LVS (lane 2). A 3-kb HindIII genomic DNA fragment was detected in the parental LVS with an oligonucleotide probe for wbtE but not in the Δ_wbtDEF_ mutant (left gel), whereas a 2.5-kb HindIII genomic DNA fragment was detected in the Δ_wbtDEF_ mutant with an oligonucleotide probe to the cat cassette but not in the Δ_wbtDEF_ mutant (right gel). M, biotin-labeled 1-kb plus DNA ladder.
Fig 2
Characterization of properties of F. tularensis LVS Δ_wbtDEF_ and WbtIG191V mutant strains. (A) O-antigen biosynthesis is disrupted in the Δ_wbtDEF_ and WbtIG191V mutants. Whole-cell lysates from the parental LVS (lane 1), the WbtIG191V (lane 2) and Δ_wbtDEF_ (lane 3) mutant strains, and the WbtIG191V mutant strain complemented with an intact copy of the wbtI gene (lane 4) were subjected to immunoblot analysis using monoclonal antibody FB11 specific for F. tularensis LVS or polyclonal antibodies to F. tularensis proteins KatG and Bfr. While the parental LVS lane exhibits the characteristic ladder pattern of LPS immunoreactivity, no O-antigen-immunoreactive material was detected in the WbtIG191V or Δ_wbtDEF_ lane despite comparable amounts of KatG and Bfr. The O-antigen ladder is restored in the WbtIG191V mutant strain by complementation with an intact copy of the wbtI gene (lane 4). The molecular masses of the protein standards are indicated. (B) O-antigen-deficient mutants are killed by serum with an intact complement (comp) pathway but not by heat-inactivated serum or C7-deficient AB serum. Parental LVS or O-antigen-deficient LVS (either WbtIG191V or Δ_wbtDEF_) was suspended in RPMI 1640 medium with or without either 10% human AB serum with intact complement, 10% heat-inactivated human AB serum, C7-deficient [C7(−)] AB serum, or C7-deficient serum supplemented with purified C7 (7 μg/ml) and incubated for 10 min at 37°C. The number of CFU surviving the 10-min incubation was determined by plating serial dilutions (the values shown are means and standard errors). Treatment of the WbtIG191V and the Δ_wbtDEF_ O-antigen-deficient strains with AB serum or C7-deficient serum supplemented with purified C7 reduced CFU by more than 3 log units compared with HI-ABS or C7-deficient-serum treatment (P < 0.001; one-way ANOVA). In contrast, the parental LVS and wbtI_-complemented strains were resistant to treatment with AB serum and C7-deficient serum supplemented with purified C7. (C and D) Uptake of O-antigen-deficient F. tularensis LVS is promoted by C7-deficient AB serum. O-antigen-deficient LVS mutants were suspended in DMEM containing either 10% HI-AB serum or 10% C7-deficient AB serum and incubated with monolayers of THP-1 cells for 90 min at 37°C at an MOI of 5:3 (bacteria/macrophages). The monolayers were washed 3 times to remove nonadherent bacteria. Plasma membranes were stained by incubation with Alexa Fluor 633-labeled WGA for 5 min at 18°C, and the monolayers were fixed with formaldehyde, permeabilized with saponin, and stained for F. tularensis with a red fluorescent antibody. (C) Numbers of intracellular bacteria per macrophage were enumerated by fluorescence microscopy. Opsonization of the bacteria with C7-deficient serum greatly enhances the uptake of all 4 strains (P < 0.01; one-way ANOVA). The data shown are means and standard errors. (D) A representative confocal image in the xz plane shows an intracellular WbtIG191V mutant (arrow, red) beneath the macrophage plasma membrane (arrowhead, pseudocolor green), and the macrophage nucleus (stained blue by DAPI). (E) O-antigen-deficient F. tularensis LVSs show a substantial delay in growth but ultimately are able to grow to high levels in monolayers of THP-1 cells, and complementation of the WbtIG191V mutant with an intact copy of the wbtI gene eliminates the lag and restores intracellular growth kinetics to those of the parental LVS. Opsonized or unopsonized parental F. tularensis LVS or O-antigen-deficient mutant LVSs were incubated with THP-1 cells in DMEM containing either C7-deficient AB serum or heat-inactivated AB serum, and uptake was synchronized by pelleting the bacteria onto the monolayers by centrifugation. The macrophage monolayers were washed to remove nonadherent bacteria and incubated at 37°C in DMEM containing 10% HI-FBS and gentamicin to prevent extracellular growth of the bacteria, and bacterial CFU in the monolayers were determined at sequential times thereafter. Under both the opsonized and unopsonized conditions, CFU for the WbtIG191V and Δ_wbtDEF O-antigen-deficient strains were significantly lower than for the complemented and parental strains at the 6-h to 24-h time points (P < 0.01; two-way ANOVA with Bonferroni posttests). The data shown are means and standard errors. The experiments were performed at least twice with similar results.
Fig 3
Complement-opsonized O-antigen-deficient F. tularensis LVSs are internalized by asymmetric and often redundant tightly adherent pseudopod loops. The uptake of C7-deficient serum-opsonized parental LVS (A and B) and O-antigen-deficient LVSs (Δ_wbtDEF_ [C, E, and G]; WbtIG191V [D and F]) by human monocyte-derived macrophages was examined by TEM. Whereas the parental LVS bacteria are engulfed within spacious, asymmetric pseudopod loops (A and B), the O-antigen-deficient mutants are internalized with more tightly adherent asymmetric loops (C and D) and within overlapping, onion-like arrays of adherent pseudopodia (E to G). (H) For comparison, uptake of E. coli by conventional phagocytosis is shown. The experiment was conducted three times with similar results. Bars, 1 μm.
Fig 4
Uptake morphology of O-antigen-deficient mutant F. tularensis LVS is influenced by complement opsonization. (A and B) C7-deficient serum-opsonized WbtIG191V O-antigen-deficient LVS is internalized by human monocyte-derived macrophages via tightly adherent and often redundant pseudopod loops. (C and D) In contrast, in the absence of complement opsonization, the WbtIG191V O-antigen-deficient LVS is internalized in spacious, asymmetric pseudopod loops. (E to H) As in the case of the parental F. tularensis LVS, the complemented mutant is internalized via spacious, asymmetric pseudopod loops both in the presence (E and F) and in the absence (G and H) of serum opsonization. The experiment was conducted twice with similar results. Bars, 0.5 μm (A to D, G, and H) and 1.0 μm (E and F).
Fig 5
Immunofluorescence evaluation of colocalization of CD63 and Texas Red-dextran with parental F. tularensis LVS, O-antigen-deficient LVS Δ_wbtDEF_, and latex beads. Uptake of fluorescent blue 1-μm latex beads (arrowheads) and either LVS-GFP (A, B, E, and F, arrows) or LVS Δ_wbtDEF_-GFP (C, D, G, and H, arrows) by human THP-1 cells was synchronized by centrifuging the bacteria and beads onto the monolayer at 4°C and warming the monolayers to 37°C for 30 min. The monolayers were washed to remove nonadherent beads and bacteria, incubated in fresh medium at 37°C, fixed at sequential times thereafter, permeabilized, and stained for CD63 with a Texas Red-conjugated antibody and for host and bacterial DNA with DAPI (blue fluorescence). (A, C, E, and G) Merged color images of the green-fluorescent bacteria, red CD63 fluorescence, and blue fluorescence of the beads and nuclei. (B, D, F, and H) CD63 red fluorescence shown in black and white. (A to D) At the 30-min time point, the LVS-GFP (A and B, arrows) and LVS Δ_wbtDEF_-GFP (C and D, arrows) are partially rimmed by CD63 staining, and the latex beads (arrowheads) are more uniformly rimmed with CD63 staining. (E to H) At 22 h, the LVS-GFP (E and F, arrows) and LVS Δ_wbtDEF_-GFP (G and H, arrows) have multiplied extensively, and the majority no longer colocalize with CD63 fluorescence. In contrast, the latex beads (arrowheads) in these heavily infected cells continue to colocalize strongly with CD63. (I) Colocalization of latex beads and LVS-GFP or Δ_wbtDEF_-GFP with CD63 was evaluated for at least 40 bacteria or beads for each time point from 30 min to 22 h postinfection. The LVS Δ_wbtDEF_-GFP and the parental strain had significantly less colocalization with CD63 than the latex beads at the 3-h to 22-h time points (P < 0.01; two-way ANOVA with Bonferroni posttests). The data represent means ± standard errors. The experiment was performed twice with similar results.
Fig 6
Immunofluorescence evaluation of colocalization of Texas Red-dextran with parental F. tularensis LVS, O-antigen-deficient Δ_wbtDEF_ LVS, and latex beads. Lysosomes of PMA-differentiated THP-1 macrophages were prelabeled by incubation of the macrophages with lysine-fixable Texas Red-dextran (10,000 MW) prior to synchronizing uptake of fluorescent blue 1-μm latex beads and LVS-GFP (A, B, E, and F) or LVS Δ_wbtDEF_-GFP (C, D, G, and H) as described above. The macrophages were washed, incubated at 37°C, fixed at sequential times thereafter, stained with DAPI, and evaluated by fluorescence microscopy as described above. (A, C, E, and G) Merged fluorescence images showing green-fluorescent bacteria, red-fluorescent Texas Red-dextran, and blue-fluorescent beads and nuclei. (B, D, F, and H) Texas Red-dextran fluorescence is shown in black and white. (A to H) LVS-GFP (A, B, E, and F, arrows) and LVS Δ_wbtDEF_-GFP (C, D, G, and H, arrows) bacteria have little or no colocalization with Texas Red-dextran at either 30 min (A to D) or 22 h (E to H) after infection. In contrast, the 1-μm fluorescent blue latex beads (arrowheads) colocalize intensely with Texas Red-dextran at both the 30-min and 22-h time points. (I) Colocalization of latex beads and LVS-GFP or Δ_wbtDEF_-GFP with Texas Red-dextran was evaluated for at least 40 bacteria or beads for each time point from 30 min to 22 h postinfection. LVS Δ_wbtDEF_-GFP and the parental strain had significantly less colocalization with Texas Red-dextran than the latex beads at all time points (P < 0.001; two-way ANOVA with Bonferroni posttests). The data represent means ± standard errors. The experiment was performed twice with similar results.
Fig 7
Ultrastructural analysis of the time course of phagosome escape in human THP-1 cells by parental F. tularensis LVS and O-antigen-deficient LVS in human THP-1 cells. Monolayers of PMA-differentiated human THP-1 cells were infected with LVS or LVS Δ_wbtDEF_ that had been opsonized with C7-deficient AB serum (A) or without complement opsonization (B) or infected with WbtIG191V O-antigen-deficient LVS or the complemented mutant in the presence or absence of opsonization with C7-deficient serum (C). The macrophages were fixed at 1 h to 21 h and processed for transmission electron microscopy. Bacteria were scored as having escaped into the cytosol if less than 50% of their circumference was surrounded by a membrane bilayer. The difference between the Δ_wbtDEF_ mutant and the parental LVS in mean percent escaped was statistically significant at both the 6- and 12-h time points for both the opsonized (A) and unopsonized (B) conditions (P < 0.01; two-way ANOVA with Bonferroni posttests). In panel C, the difference between the WbtIG191V O-antigen-deficient LVS and the complemented mutant in mean percent escaped at the 8-h time point was statistically significant (P < 0.01; two-way ANOVA with Bonferroni posttests) under both the opsonized and unopsonized conditions. The data represent means ± standard errors. The experiments were performed twice with similar results.
Fig 8
Ultrastructural analysis of LVS and O-antigen-deficient LVS in human THP-1 cells in the absence of complement opsonization. (A and B) In the first 6 h after uptake of LVS bacteria, even in the absence of complement opsonization, the majority of LVS bacteria (indicated by asterisks) are found within membrane-bound vacuoles. (C to G) However, by 12 h after uptake (C and D), the majority of the LVS bacteria (asterisks) have escaped into the cytosol, while in the absence of serum opsonization, the majority of LVS Δ_wbtDEF_ bacteria remain within membrane-bound vacuoles in the first 12 h after uptake (6 h [E] and12 h [F and G]). By 12 h after uptake, many of the LVS Δ_wbtDEF_ phagosomes are decorated with a characteristic electron-dense fibrillar coat (faint in panel F and more obvious in panel G), which we also observed on the parental LVS phagosomes at earlier time points. (H) Blebbing and disruption of the fibrillar coat appear to accompany escape of the LVS Δ_wbtDEF_ bacteria into the cytosol. (I and J) By 21 h after uptake, the majority of the LVS Δ_wbtDEF_ bacteria have escaped and are free in the cytosol. (K and L) We observed similar features for the WbtIG191V mutant LVS. Panel L shows WbtIG191V mutant LVS bacteria at 16 h postinfection that have escaped into the cytosol. The _wbtI_-complemented strain (K) exhibits morphological interactions with the host cell similar to those of the LVS, whereas the WbtIG191V mutant strain (L) lacks the electron-lucent zone that is seen in the parental LVS. The electron-lucent zone surrounding the bacteria (K, arrowheads) is restored in the complemented strain. The experiment was conducted twice with similar results. Bars, 1 μm (A to E and J to L) and 0.5 μm (F to I).
References
- Ben Nasr A, Klimpel GR. 2008. Subversion of complement activation at the bacterial surface promotes serum resistance and opsonophagocytosis of Francisella tularensis. J. Leukoc. Biol. 84:77–85 -PubMed
Publication types
MeSH terms
Substances
LinkOut - more resources
Full Text Sources
Miscellaneous