Intraspecies Competition in Serratia marcescens Is Mediated by Type VI-Secreted Rhs Effectors and a Conserved Effector-Associated Accessory Protein

Abstract

The type VI secretion system (T6SS) is widespread in Gram-negative bacteria and can deliver toxic effector proteins into eukaryotic cells or competitor bacteria. Antibacterial T6SSs are increasingly recognized as key mediators of interbacterial competition and may contribute to the outcome of many polymicrobial infections. Multiple antibacterial effectors can be delivered by these systems, with diverse activities against target cells and distinct modes of secretion. Polymorphic toxins containing Rhs repeat domains represent a recently identified and as-yet poorly characterized class of T6SS-dependent effectors. Previous work had revealed that the potent antibacterial T6SS of the opportunistic pathogen Serratia marcescens promotes intraspecies as well as interspecies competition (S. L. Murdoch, K. Trunk, G. English, M. J. Fritsch, E. Pourkarimi, and S. J. Coulthurst, J Bacteriol 193:6057-6069, 2011, http://dx.doi.org/10.1128/JB.05671-11). In this study, two new Rhs family antibacterial effectors delivered by this T6SS have been identified. One of these was shown to act as a DNase toxin, while the other contains a novel, cytoplasmic-acting toxin domain. Importantly, using S. marcescens, it has been demonstrated for the first time that Rhs proteins, rather than other T6SS-secreted effectors, can be the primary determinant of intraspecies competition. Furthermore, a new family of accessory proteins associated with T6SS effectors has been identified, exemplified by S. marcescens EagR1, which is specifically required for deployment of its associated Rhs effector. Together, these findings provide new insight into how bacteria can use the T6SS to deploy Rhs-family effectors and mediate different types of interbacterial interactions.

Importance: Infectious diseases caused by bacterial pathogens represent a continuing threat to health and economic prosperity. To counter this threat, we must understand how such organisms survive and prosper. The type VI secretion system is a weapon that many pathogens deploy to compete against rival bacterial cells by injecting multiple antibacterial toxins into them. The ability to compete is vital considering that bacteria generally live in mixed communities. We aimed to identify new toxins and understand their deployment and role in interbacterial competition. We describe two new type VI secretion system-delivered toxins of the Rhs class, demonstrate that this class can play a primary role in competition between closely related bacteria, and identify a new accessory factor needed for their delivery.

FIG 1

Type VI secretion system-associated Rhs proteins in Serratia marcescens. (A) Comparison of T6SS gene clusters and distant loci encoding PAAR domain-containing Rhs proteins between S. marcescens Db10 and three other strains of S. marcescens: SM39, WW4, and BIDMC81. Conserved T6SS components are shown in blue, with core components TssA-M indicated by single letters and others indicated by common names (VgrG and Hcp are core components TssI and TssD). Tae4 family effectors are shown in purple, Tai4/4a family immunity proteins are in pink, uncharacterized genes conserved in S. marcescens are in gray, and strain-specific genes are white. Genes encoding Rhs proteins and EagR accessory proteins (DUF1795; asterisk) are shown in green; distinct C-terminal domains and putative cognate RhsI proteins are indicated with different colors. Sequence data were obtained from NCBI databases, and genomic identifiers are given for selected genes (a dash indicates a nonannotated open reading frame manually identified as likely encoding an RhsI or Tae4 protein). (B) The domain organization of Rhs proteins of S. marcescens Db10. The Rhs domains are as defined in reference , and positions of PAAR motifs, Rhs repeats, and a partial HNH endonuclease domain are indicated. Amino acid numbering is given below each protein.

FIG 2

Mutants of Serratia marcescens lacking the immunity determinant RhsI1 or RhsI2 are sensitive to the action of type VI secretion system-delivered toxin Rhs1 or Rhs2, respectively. (A) Number of target cells recovered following coculture of a target strain lacking rhsI1 (S. marcescens Db10 ΔrhsI1 ΔtssH mutant) with wild-type (WT) or mutant (ΔtssE, Δrhs1, Δrhs2, and Δrhs1 Δrhs2) strains of S. marcescens Db10 as the attacker. None indicates coculture of the target with sterile medium alone, and the ΔtssE mutant has a nonfunctional T6SS. (B) Recovery of the rhsI1 mutant carrying either a vector control plasmid (ΔrhsI1+VC; the ΔrhsI1 ΔtssH mutant with pBAD18-Kan) or complementing plasmid expressing rhsI1 in trans (ΔrhsI1+RhsI1; ΔrhsI1 ΔtssH mutant with pSC636) as the target strain following coculture with attacker strains carrying either a vector control plasmid (VC; pBAD18-Kan) or a complementing plasmid expressing rhs1 in trans (+Rhs1; pSC643) as indicated. (C) Recovery of a target strain lacking rhsI2 (Db10 Δrhs2 ΔrhsI2 mutant) following coculture with WT or mutant (ΔtssE, Δrhs1, Δrhs2, and Δrhs1 Δrhs2) strains of S. marcescens Db10 as the attacker. (D) Recovery of the rhsI2 mutant carrying either a vector control plasmid (ΔrhsI2+VC; Δrhs2 ΔrhsI2 mutant with pBAD18-Kan) or complementing plasmid expressing rhsI2 in trans (ΔrhsI2+RhsI2; Δrhs2 ΔrhsI2 mutant with pSC672) as the target strain following coculture with WT or ΔtssH attacker strains also carrying the vector control plasmid (VC; pBAD18-Kan). (A to D) Points show means ± standard errors of the means (SEM) (n ≥ 3). (E) Immunoblot detection of Hcp, Ssp1, and Ssp2 in cellular (cell) and secreted (sec) fractions of WT or mutant strains of S. marcescens Db10.

FIG 3

C-terminal domain of Rhs2 has DNase activity, and the C-terminal domain of Rhs1 is a cytoplasmic-acting antibacterial toxin. (A) Degradation of plasmid DNA is observed on expression of Rhs2-CT in Escherichia coli. Plasmid DNA recovered from Escherichia coli BL21(DE3) pLysS carrying plasmids expressing the C-terminal domain of Rhs2 (Rhs2-CT; pSC674) or Rhs2-CT with RhsI2 (Rhs-CT+RhsI2; pSC675), with (+) or without (−) IPTG induction. S, size standards. (B) Heterologous expression of Rhs1-CT in Escherichia coli is toxic and alleviated on coexpression of RhsI1. Serial dilutions of Escherichia coli MG1655 carrying the empty vector (VC; pBAD18-Kan) or plasmids expressing full-length Rhs1 (Rhs1; pSC643), the C-terminal domain of Rhs1 (Rhs1-CT; pSC635), or Rhs1-CT with RhsI1 (Rhs-CT+RhsI1; pSC637) were spotted onto rich medium (LB) or M9 minimal medium. Gene expression was repressed or induced by the inclusion of 0.2% d-glucose or l-arabinose, respectively, in the media. (C) Subcellular localization of RhsI1. Cells of S. marcescens Db10 expressing an RhsI1-HA fusion protein (from pSC640) were subjected to fractionation and immunoblotting using antibodies against the HA tag, RNAP (RNA polymerase β subunit; cytoplasmic control), MBP (maltose binding protein; periplasmic control), or TssJ (membrane control). WC, whole cell; PP, periplasm; CYT, cytoplasm; MEM, total membrane.

FIG 4

Contribution of Rhs proteins to type VI secretion system-mediated inter- and intraspecies antibacterial activity of Serratia marcescens Db10. (A) Recovery of target organisms P. fluorescens 55, Escherichia coli MC4100, and Enterobacter cloacae ATCC 13047 following coculture with wild-type (WT) or mutant (ΔtssE, Δrhs1, Δrhs2, and Δrhs1 Δrhs2) strains of S. marcescens Db10 as the attacker. None indicates coculture of the target with sterile medium alone. Points show means ± SEM (n ≥ 3). (B) Recovery of target organisms S. marcescens SM39 and S. marcescens ATCC 274 following coculture with the strains of S. marcescens Db10 described for panel A. Points show means ± SEM (n = 4). (C) Recovery of S. marcescens ATCC 274 carrying the empty vector (VC; pBAD18-Kan) or a plasmid expressing RhsI1 and RhsI2 (RhsI1,I2; pSC673) following coculture with the WT or Δrhs1 Δrhs2 strain of Db10 carrying an empty vector. Points show means ± SEM (n = 4).

FIG 5

DUF1795 family protein EagR1 is specifically required for Rhs1-mediated antibacterial activity. (A and B) Recovery of target organism S. marcescens Db10 ΔrhsI1 ΔtssH (ΔrhsI1) strain or S. marcescens Db10 Δssp4 Δsip4 (Δsip4) strain following coculture with wild-type (WT) or mutant (ΔtssE, ΔeagR1, and Δrhs1) strains of S. marcescens Db10 as the attacker. The ΔeagR1 mutant is an in-frame deletion mutant of SMDB11_2277. (C) Immunoblot detection of Hcp, Ssp1, and Ssp2 in cellular (cell) and secreted (sec) fractions of WT or mutant strains of S. marcescens Db10. (D) Recovery of target organism S. marcescens Db10 Δrhs2 ΔrhsI2 (ΔrhsI2) strain following coculture with WT or mutant (ΔtssE and ΔeagR1) strains of S. marcescens Db10 as the attacker. This is part of the same experiment as that shown Fig. 2C, and the data for the control strains are repeated from that figure. (E) Recovery of the rhsI1 mutant carrying a vector control plasmid (ΔrhsI1; ΔrhsI1 ΔtssH mutant with pBAD18-Kan) as the target strain following coculture with attacker strains carrying either a vector control plasmid (VC; pBAD18-Kan) or complementing plasmid expressing SMDB11_2277 in trans (+2277; pSC658). Points show means ± SEM throughout (n ≥ 4).