Discovery of coordinately regulated pathways that provide innate protection against interbacterial antagonism

Abstract

Bacterial survival is fraught with antagonism, including that deriving from viruses and competing bacterial cells. It is now appreciated that bacteria mount complex antiviral responses; however, whether a coordinated defense against bacterial threats is undertaken is not well understood. Previously, we showed that Pseudomonas aeruginosa possess a danger-sensing pathway that is a critical fitness determinant during competition against other bacteria. Here, we conducted genome-wide screens in P. aeruginosa that reveal three conserved and widespread interbacterial antagonism resistance clusters (arc1-3). We find that although arc1-3 are coordinately activated by the Gac/Rsm danger-sensing system, they function independently and provide idiosyncratic defense capabilities, distinguishing them from general stress response pathways. Our findings demonstrate that Arc3 family proteins provide specific protection against phospholipase toxins by preventing the accumulation of lysophospholipids in a manner distinct from previously characterized membrane repair systems. These findings liken the response of P. aeruginosa to bacterial threats to that of eukaryotic innate immunity, wherein threat detection leads to the activation of specialized defense systems.

Keywords: Pseudomonas aeruginosa; Tn-seq; infectious disease; innate immunity; interbacterial antagonism; microbiology; phospholipase; toxin.

Figure 1.. Multiple pathways under Gac/Rsm control contribute to P. aeruginosa defense against antagonism.

(A) Transposon library sequencing-based comparison of the fitness contribution of individual P. aeruginosa genes during growth competition with wild-type B. thai versus B. thai ∆T6S. Genes under Gac/Rsm control (blue) and those encoding core Gac/Rsm regulatory factors (purple, labeled) are indicated. (B) Overview of the Gac/Rsm pathway (SD, Shine–Dalgarno). The rsmY and rsmZ genes encode small RNA molecules that sequester the translational regulator RsmA. Hybrid sensor kinase PA1611 is here renamed RetS-regulating sensor kinase A (RskA). (C) Left: P. aeruginosa gene clusters hit (greater than threefold in replicate screens; Figure 1—source data 1) in this study. Numbers below genes indicate transposon insertion ratio (B. thai ∆T6S/wild-type) for screen in (A) followed by the replicate screen (Figure 1—figure supplement 1); light toned genes were hit in only one replicate. Asterisk indicates genes for which insertion was undetected in libraries obtained from B. thai wild-type competition. Right: zoom-in of boxed region of (A) with genes colored corresponding to clusters at left. (D, E) Recovery of P. aeruginosa cells with the indicated genotypes following growth competition against B. thai wild-type (D) or ∆T6S (E). For interbacterial competition assays, the mean ± SD of biological duplicates and associated technical duplicates is shown: *p<0.05 using unpaired two-tailed Student’s t-test.

Figure 1—figure supplement 1.. Multiple pathways under Gac/Rsm control contribute to P. aeruginosa defense against antagonism.

(A) Growth of wild-type P. aeruginosa in co-culture with the indicated B. thai strains. (B) Transposon library sequencing-based comparison of the fitness contribution of individual P. aeruginosa genes during growth competition with wild-type B. thai versus B. thai ∆T6S. Genes under Gac/Rsm control (blue) and those encoding core Gac/Rsm regulatory factors (purple, labeled) are indicated. This experiment is a biological replicate of that in Figure 1A. (C) P. aeruginosa gene clusters hit (Figure 1—source data 1) in this study. (D) Zoom-in of boxed region of (B) with genes colored corresponding to clusters at left. The interbacterial competition assay represents two biological replicates with two associated technical replicates.

Figure 1—figure supplement 2.. Antagonism resistance clusters possess functionally related genes and are in diverse bacteria.

(A) Schematic depicting the arc1 gene cluster of P. aeruginosa. Predicted enzymatic activity of Arc1A, B, D, F is indicated (right) and placed onto a phospholipid biosynthesis pathway (bottom). (B) Depiction of arc2 genes from the indicated species. Genes encoding the predicted MoxR-like ATPase and von Willebrand factor (VWF) domain protein are indicated below the respective genes. Orthologs are colored to highlight synteny. (C) Depiction of arc3 genes from the indicated species. S. dynata exemplifies an instance of translational fusion of arc3A,B. Domains of unknown function constituting the two proteins are indicated below.

Figure 1—figure supplement 3.. Interbacterial competition assays measure the contribution of arc1-3 to P. aeruginosa fitness during antagonism by B. thai.

(A) Recovery of P. aeruginosa strains bearing the indicated single or double deletions within arc1-3 following growth competition against wild-type B. thai. (B) Recovery of P. aeruginosa cells with the indicated genotypes following growth competition against wild-type B. thai. The mean ± SD of biological duplicates and associated technical duplicates is shown: n.s. corresponds to p>0.05 by unpaired two-tailed Student’s t-test. The fold difference in growth yield between mutants and the parental strain is indicated above.

Figure 2.. Arc pathways can provide toxin-specific antagonism defense that eclipses that of other cellular factors.

(A) Recovery of B. thai cells following growth competition against a relative abundance of P. aeruginosa containing the indicated gene deletions, colored according to Figure 1C. (B) Relative survival of P. aeruginosa Arc1-3-inactivated mutants following growth in competition with an excess of the indicated B. thai strains (mutant CFU/parental CFU × 100 for each B. thai competitor). The loss of ColA or Tle3 activity in B. thai increases the relative survival of P. aeruginosa lacking Arc2 or Arc3 activity, respectively. (C, D) Transposon library sequencing-based comparison of the fitness contribution of individual P. aeruginosa genes during growth competition with wild-type B. thai versus B. thai lacking Tle3 (C) or ColA (D) activity, colored according to Figure 1C. Values in parentheses correspond to the number of genes hit (greater than threefold change in transposon insertion frequency) within each cluster compared to the number within each cluster hit in the initial screens (B. thai wild-type versus B. thai ∆T6S) followed by the average transposon insertion frequency ratio (B. thai mutant/B. thai wild-type) of the genes in the depicted screen. (E) Fitness contribution of each P. aeruginosa Arc pathway and the H1-T6SS to defense against Tle3 versus ColA. For interbacterial competition assays, the mean ± SD of biological duplicates and two technical replicates is shown: *p<0.05 using unpaired two-tailed Student’s t-test.

Figure 2—figure supplement 1.. Arc1-3 do not contribute to the survival of P. aeruginosa exposed to common environmental stresses.

Survival of the indicated P. aeruginosa strains during exposure to hydrogen peroxide (50 mM, 30 min) (A), high salinity (3 M NaCl, 20 hr) (B), high temperature (55°C, 30 min) (C), or detergent (0.5% [w/v] SDS) (D). The mean ± SD of a representative biological replicate with three technical replicates is shown.

Figure 2—figure supplement 2.. Arc1-3 are subject to tight regulation by the Gac/Rsm signaling pathway.

Anti-VSV-G immunoblot analysis was used to probe the levels of the indicated VSV–G-tagged (–V) Arc proteins encoded at their respective native genomic loci in P. aeruginosa wild-type, ∆retS, or ∆gacS backgrounds. The asterisk denotes a nonspecific VSV-G antibody-reactive band. The arrow highlights the position of Arc2G–V. Blot images are representative of at least two biological replicates. See Figure 2—figure supplement 2—source data 1–6.

Figure 2—figure supplement 3.. The loci encoding Tle1 and Tle3 are duplicated in B. thai, and Tle3 induces potent self-intoxication.

(A) Schematic depicting the tle1/tli1 (top) and tle3/tli3 (bottom) loci in the genome B. thai E264 strain. Predicted catalytic serine residues mutagenized in this study are indicated by red asterisks. (B) Recovery of B. thai strain sensitized to Tle3 intoxication (∆tle3∆tli3) following growth competition against B. thai wild-type or a strain in which both orthologs of Tle3 contain an inactivating point mutation. The mean ± SD of two biological replicates with two technical replicates is shown.

Figure 3.. Arc3B is a predicted massive polytopic membrane protein with functionally complementing family members that occur widely in Gram-negative and -positive bacterial phyla.

(A) Schematic depiction of Arc3A and Arc3B based on bioinformatic predictions. (B) Phylogeny of bacterial phyla with representative arc genes depicted and colored as in (A). Membrane-associated regions (grayed) and the number of predicted transmembrane segments are indicated. (C) Recovery of P. aeruginosa strains bearing the indicated mutations and containing a control chromosomal insertion (–) or insertion constructs expressing Arc3B proteins derived from assorted bacteria (P. aer, P. aeruginosa; P. pro, P. protegens; B. cen, B. cenocepacia; P. syr, P. syringae; C. rod, Citrobacter rodentium) following growth competition against B. thai wild-type. (D) Recovery of the indicated P. protegens strains following growth competition against B. thai wild-type or a strain lacking Tle3 activity. For interbacterial competition assays, the mean ± SD of biological duplicates and two or three associated technical replicates is shown: *p<0.05 using unpaired two-tailed Student’s t-test.

Figure 3—figure supplement 1.. EstA and Arc3B inactivation does not significantly impact P. aeruginosa fitness in pairwise growth competition experiments with B. thai.

Recovery of the indicated P. aeruginosa strains following growth in competition with B. thai wild-type or a strain lacking Tle3 activity (tle3^S264A). Means ± SD of biological duplicates with three technical replicates s shown.

Figure 4.. Arc3 prohibits Tle3-catalyzed lysophospholipid accumulation by a mechanism distinct from known pathways.

(A–D) Mass spectrometric (A) and radiographic thin layer chromatography (TLC) (B) analysis of major phospholipid and lysophospholipid species within lipid extracts derived from the indicated mixtures of P. aeruginosa and B. thai strains. P. aeruginosa were grown in ³²PO₄^2- prior to incubation with B. thai. Strains were allowed to interact for 1 hr before phospholipids were harvested. Radiolabeled molecules of interest within biological triplicate experiments resolved by TLC were quantified by densitometry (C, D) (E) Mass spectrometric analysis of major phospholipid and lysophospholipid species within lipid extracts derived from mixtures containing B. thai lacking Tle3-specific immunity factors competing with the indicated B. thai strains. (F) Recovery of the indicated P. aeruginosa strains following growth competition against B. thai wild-type or a strain lacking Tle3 activity. (G) Radiographic TLC analysis of products extracted after incubation of P. aeruginosa-derived spheroplasts with purified radiolabeled lysophosphatidylethanolamine (LPE). For interbacterial competition assays, the mean ± SD of biological replicates and associated technical replicates is shown: *p<0.05 using unpaired two-tailed Student’s t-test. TLC images shown are representative of at least two biological replicates. See also Figure 4—source data 1–3.

Figure 4—figure supplement 1.. Tle3 intoxication does not impact free fatty acids nor intact parent phospholipids in P. aeruginosa.

Mass spectrometric analysis of free fatty acid (A) and parent phospholipid species (B) within lipid extracts derived from the indicated mixtures. Data represent the mean ± SD of biological duplicates with two technical replicates.

Figure 4—figure supplement 2.. Arc3B genes are encoded adjacent to aas genes in diverse bacteria.

Schematic depicting arc3B and aas loci in the indicated bacterial species.

Figure 4—figure supplement 3.. Overexpression of aas fails to complement the competitive defect associated with Arc3A inactivation.

Recovery of the indicated P. aeruginosa strains carrying empty pPSV39 (control) or pPSV39-aas grown in competition with an excess of B. thai with Isopropyl β-D-1-thiogalactopyranoside (IPTG) for induction of expression from the plasmid. Data represent mean and standard deviation from technical duplicates from one biological replicate representative of two conducted.

Author response image 1.. Inactivation of Arc3A does not detectably impact P. aeruginosa growth in competition with assorted phospholipase producing species.

(A) Recovery of P. aeruginosa with the indicated genotypes following growth in competition with an excess of wild-type or Tle1-inactivated B. thai. (B and C) Recovery of P. aeruginosa (B) or competitive index obtained (C; final donor:recipicient ratio divided by initial donor:recipienct ratio) following growth in competition with the indicated organisms.