Comparative signature-tagged mutagenesis identifies Pseudomonas factors conferring resistance to the pulmonary collectin SP-A

Abstract

The pulmonary collectin, surfactant protein A (SP-A), is a broad spectrum opsonin with microbicidal membrane permeabilization properties that plays a role in the innate immune response of the lung. However, the factors that govern SP-A's microbial specificity and the mechanisms by which it mediates membrane permeabilization and opsonization are not fully understood. In an effort to identify bacterial factors that confer susceptibility or resistance to SP-A, we used comparative signature-tagged mutagenesis to screen a library of 1,680 Pseudomonas aeruginosa mutants for evidence of differential pulmonary clearance in SP-A-sufficient (SP-A) and SP-A-deficient (SP-A) mice. Two SP-A-sensitive P. aeruginosa mutants harboring transposon insertions in genes required for salicylate biosynthesis (pch) and phosphoenolpyruvate-protein-phosphotransferase (ptsP) were recovered. The mutants were indistinguishable from the parental wild-type PA01 with regard to opsonization by SP-A, but they exhibited increased susceptibility to SP-A-mediated membrane permeabilization. These results suggest that bacterial gene functions that are required to maintain membrane integrity play crucial roles in resistance of P. aeruginosa to the permeabilizing effects of SP-A.

Figure 1. SP-A Deficient Mice Are More Susceptible to Infection by P. aeruginosa

(A) SP-A^+/+ and SP-A^−/− mice (group of six) were infected intranasally with 1 × 10⁷ P. aeruginosa PA01. Bacteria were recovered from homogenized lung tissue 16 h after inoculation. p < 0.01. (B) Representative lung histology sections (stained with hematoxylin and eosin) from SP-A^+/+ and SP-A^−/− mice 16 h post intranasal instillation of PA01. Original magnification: 10 ×.

Figure 2. PCR-Based Signature-Tagged Mutagenesis

(A) Schematic drawing of PCR-based STM. Pools of 72 uniquely tagged mutants were intranasally inoculated into SP-A^+/+ and SP-A^−/− mice; 16 h later, lungs were harvested, homogenized, and plated. Approximately 10,000 colonies were harvested from the plates for genomic DNA extraction. PCR-amplification of tags was performed on the genomic DNA to screen for the presence or absence of DNA tags of each of the 72 mutants. Mutants whose DNA tags were present in the input pool and in the SP-A^−/− pool, but absent in the SP-A^+/+ pool (see white arrows) were further screened for susceptibility to SP-A. (B) Agarose gel of PCR-based STM, identifying the pch mutant (left panel). Attenuation in the first SP-A^+/+ mouse (right panel, m1) was confirmed in two additional SP-A^+/+ mice (right panel, m2 and m3). (C) Genetic loci of P. aeruginosa, when mutated, conferred increased sensitivity to in vivo killing by SP-A. DNA regions flanking pUTmini-Tn5 transposon insertions were cloned and sequenced. Similarity BLAST searches were performed against P. aeruginosa PA01 genomic sequence on NCBI and on http://www.pseudomonas.com. Black arrows indicate the approximate insertion site within the mutated ORFS. (D) TLC analyses indicate that the pch STM mutant is unable to synthesize pyochelin (see arrows). Wild-type PA01 grown in LB and the pch mutant grown in LB supplemented with 1 mM salicylate produced green color pyochelin (see arrows). In contrast, no observable pyochelin was produced by the pch and PA06331 strains. Pure pyochelin was used as control. (E) Restriction fragment length polymorphism analysis between parental wild-type PA01 and STM mutants confirmed that mutations in pch and ptsP were caused by a single insertion.

Figure 3. SP-A-Sensitive P. aeruginosa STM Mutants Compete Effectively with Parental Wild-Type PA01 in SP-A^−/− , but Not in SP-A^+/+ Mice

(A) In vivo competition assays between the wild-type and individual STM mutants after intranasal pulmonary inoculation of 6-wk-old SP-A^+/+ and SP-A^−/− mice. The mean CFU recovered from four mice in each group is shown. The CI is defined as the output ratio of mutant to wild-type bacteria divided by the input ratio of mutant to wild-type bacteria. *p < 0.013, #p < 0.011. (B) Single respiratory tract infection of pch and ptsP mutants in SP-A^+/+ and SP-A^−/− mice was performed as described for Figure 1. Attenuation is defined as the log₁₀ difference in CFU between wild-type and mutant bacteria recovered from lung tissue 16 h after inoculation. The mean ± standard error of six mice is shown. *p < 0.019, #p < 0.045.

Figure 4. Bacterial Opsonization by SP-A Does Not Result in Differential In Vivo Phagocytosis and Killing of STM Mutants

(A) In vitro phagocytosis assays were performed using live, GFP-expressing PA01 and isogenic pch and ptsP bacteria, with alveolar macrophages isolated from SP-A^+/+ mice. Internalized bacteria were counted using a phase contrast fluorescence microscope. The means of three experiments are shown. Two hundred macrophages were counted for each mouse. p = 0.0058 (PA01 vs PA01 + SP-A), p = 0.075 (pch vs pch + SP-A), and p = 0.109 (ptsP vs ptsP + SP-A). (B) In vitro phagocytosis of PA01, pch, and ptsP by human neutrophils isolated from healthy volunteers. Phagocytosis experiments were performed as described for (A). p = 0.037 (PA01 vs PA01 + SP-A), p = 0.577 (pch vs pch + SP-A), and p = 0.919 (ptsP vs ptsP + SP-A). (C–E) SP-A did not affect uptake or killing of STM mutants. PA01 (C), but not the STM mutants pch (D) and ptsP (E) were killed in significant amounts by neutrophils, independent of SP-A. The mean of three separate experiments is shown. p < 0.001 when comparing PA01 versus PA01 + neutrophils, and when comparing PA01-SP-A versus PA01-SP-A + neutrophils.

Figure 5. The pch and ptsP Mutants Are Preferentially Permeabilized and Killed by SP-A

(A) Fluorescence of a cleavage-activated phosphatase substrate ELF97 was examined in the presence or absence of 50 μg hSP-A for a period of 60–90 min. Human SP-A (hSP-A) preferentially permeabilized LPS mutant wbpL, but not the parental wild-type PA01. The wbpL strain was used as positive control for membrane permeability. The difference in membrane permeabilization between wbpL mutant and PA01 was significant from the 28^th min onward. *p < 0.05. (B) A mutant strain STMG2A7 that is virulent in both SP-A^+/+ and SP-A^−/− mice was as resistant to SP-A-mediated membrane permeabilization as the parental wild-type PA01. (C–D) Membrane permeability of the STM mutant pch and PA06331, a P. aeruginosa strain which is deleted in all of the structural genes required for pyochelin biosynthesis (Table 1.). The difference in membrane permeabilization between the mutants against wild-type PA01 was significant from the 35^th and 18^th min onward for mutant strains pch and PA06331, respectively. *p < 0.05. (E–F) Membrane permeability of the STM mutant ptsP and the ΔptsP, a P. aeruginosa strain with a nonpolar, inframe deletion of the ptsP gene. The difference in membrane permeabilization between the mutants against wild-type PA01 was significant from the 29^th and 28^th min onward, for mutant strains ptsP and ΔptsP, respectively. *p < 0.05.

Figure 6. SP-A-Mediated Membrane Permeabilization Directly Kills the pch and ptsP Mutants

(A) Live/Dead staining was performed on E. coli K12, PA01, pch, and ptsP cells, following 1 h membrane permeabilization with 100 μg/ml SP-A. Green-stained cells are alive whereas red-stained cells are dead. (B) Enumeration of live or dead bacteria following a 1 h exposure to hSP-A. At least 600–1,000 bacterial cells were counted under a fluorescence microscope. The mean of two experiments is shown.

Figure 7. Complementation Studies of SP-A-Sensitive Mutants pch and ptsP

(A) Culturing pch bacteria in LB supplemented with sodium salicylate restored its resistance to SP-A mediated membrane permeabilization. (B) In control experiments, we demonstrated that the enzymatic activity of bacterial phosphatases in ptsP bacterial cells was not inhibited by 1 mM sodium salicylate. (C) Provision of the wild-type ptsP gene on the pUCP19 in trans (pUCP-ptsP) restores the ability of STM mutant ptsP to resist SP-A mediated membrane permeabilization. In contrast, ptsP mutant expressing GFP gene on the pUCP19 (pstP-gfp) is permeabilized to similar levels as ptsP.