Functional and Structural Diversity of Bacterial Contact-Dependent Growth Inhibition Effectors

Abstract

Bacteria live in complex communities and environments, competing for space and nutrients. Within their niche habitats, bacteria have developed various inter-bacterial mechanisms to compete and communicate. One such mechanism is contact-dependent growth inhibition (CDI). CDI is found in many Gram-negative bacteria, including several pathogens. These CDI⁺ bacteria encode a CdiB/CdiA two-partner secretion system that delivers inhibitory toxins into neighboring cells upon contact. Toxin translocation results in the growth inhibition of closely related strains and provides a competitive advantage to the CDI⁺ bacteria. CdiB, an outer-membrane protein, secretes CdiA onto the surface of the CDI⁺ bacteria. When CdiA interacts with specific target-cell receptors, CdiA delivers its C-terminal toxin region (CdiA-CT) into the target-cell. CdiA-CT toxin proteins display a diverse range of toxic functions, such as DNase, RNase, or pore-forming toxin activity. CDI⁺ bacteria also encode an immunity protein, CdiI, that specifically binds and neutralizes its cognate CdiA-CT, protecting the CDI⁺ bacteria from auto-inhibition. In Gram-negative bacteria, toxin/immunity (CdiA-CT/CdiI) pairs have highly variable sequences and functions, with over 130 predicted divergent toxin/immunity complex families. In this review, we will discuss biochemical and structural advances made in the characterization of CDI. This review will focus on the diverse array of CDI toxin/immunity complex structures together with their distinct toxin functions. Additionally, we will discuss the most recent studies on target-cell recognition and toxin entry, along with the discovery of a new member of the CDI loci. Finally, we will offer insights into how these diverse toxin/immunity complexes could be harnessed to fight human diseases.

Keywords: contact-dependent growth inhibition; structure; toxin function; toxin/immunity complex; type V secretion system.

FIGURE 1

Introduction to CDI proteins and a working model of CDI toxin translocation. (A) Domain architecture of CdiA and organization of the cdi locus. CdiA is colored by domain, including the secretion signal (SS), two partner secretion (TPS) domain, FHA-1, receptor-binding domain (RBD), tyrosine-proline rich domain (YP), FHA-2, pretoxin (PT) domain, VENN motif, and CdiA-CT entry domain and CdiA-CT cytotoxic (Toxic) domain. (B) Schematic of CDI. CdiB (green) presents CdiA (colored by domain as in A) onto the surface of the inhibitor cell. When the CdiA-RBD (orange) recognizes its target-cell receptor (yellow), CdiA secretion continues, and FHA-2 (dark green) translocates CdiA-CT (red) into the target-cell. Once in the target-cell periplasm, the CdiA-CT is cleaved after the PT-motif, the entry domain recognizes a specific inner membrane receptor (purple) and the cytotoxic domain is translocated across the inner membrane into the cytosol. (C) Structure of CdiB. CdiB is an OMP β-barrel. Key components like α-helix H1 (blue) and loop L6 (pink) are highlighted [PDB ID 6WIL (Guerin et al., 2020)].

FIGURE 2

The diverse range of structures for BECR family CdiA-CT toxins and their CdiA-CT/CdiI complexes. (A) As representatives of BECR structural folds we show and box ColD (PDB ID: 5ZNM (Chang et al., 2018)) and ColE3 (PDB ID: 2B5U). ColD and ColE3 are colored as in the left panels described below. CDI BECR toxin and toxin/immunity complex structures are shown from: (B) K. pneumoniae 342 (Kp342) [PDB ID: 6CP9 (Gucinski et al., 2019)], (C) E. coli NC101 [PDB ID: 5I4Q (Jones et al., 2017)], (D) E. coli STEC_O31 [PDB ID: 5HKQ (Michalska et al., 2018)], (E) P. aeruginosa PABL017 [PDB ID: 6D7Y (Allen et al., 2020)], (F) E. cloacae ATCC 13047 (ECL) [PDB ID: 4NTQ (Beck et al., 2014)], and (G) E. coli 3006 (CdiA-CT/CdiI^EC3006) [PDB ID: 6CP8 (Gucinski et al., 2019)]. For each pair, the left panel displays the toxin alone in cartoon representation with its BECR core structure colored by secondary structure with β-strands in red, α-helices in green, and loops in wheat, and the remainder of the secondary elements are colored in light pink. Active site residues are shown as yellow spheres. In the right panel is the CdiA-CT/CdiI complex with CdiI colored by secondary structure: β-strands in magenta, α-helices in cyan, and loops in salmon. CdiA-CT has a semi-transparent molecular surface with white cartoon representation and with active site residues shown as yellow spheres.

FIGURE 3

Representative structures of the PD-(D/E)XK family CdiA-CT toxins and CdiA-CT/CdiI complexes from E. coli TA271 (CdiA-CT/CdiI^TA271) [PDB ID: 4G6U (Morse et al., 2012)], B. pseudomallei 1026b (CdiA-CT/CdiI^1026b) [PDB ID: 4G6V (Morse et al., 2012)], and B. pseudomallei E479 (CdiA-CT/CdiI^E479) [PDB ID: 5J4A (Johnson et al., 2016b)]. (A) CdiA-CT in cartoon representation with the core PD-(D/E)XK structure colored by secondary structure with β-strands in red, α-helices in green, and loops in wheat, and the remainder of the secondary elements colored in light blue. Notably, the structure of TA271 includes additional structure at the N-terminus (gray) domain, a Zn²⁺ ion (green sphere), and the TA271 core is interrupted by a protruding β-hairpin (dark blue). (B) Structures of PD-(D/E)XK family CdiA-CT/CdiI complexes. CdiI is colored by secondary structure: β-strands in magenta, α-helices in cyan, and loops in salmon. CdiA-CT is in white cartoon representation with a semi-transparent molecular surface and active site residues shown as yellow spheres. In the TA271 complex the bound metal ion and the TA271 β-hairpin are colored as in the toxin alone. The extra N-terminal region of TA271 is omitted.

FIGURE 4

CdiA-CT toxins with unique characteristics. (A) Y. kristensenii (Ykris) [PDB ID: 5E3E (Batot et al., 2017)] has some structural homology to BECR family members, but most closely resembles RNase A [PDB ID: 4B36 (Thiyagarajan and Acharya, 2013)]. i) RNase A is shown in cartoon depiction with β-strands in magenta, α-helices in cyan, and loops in salmon, and with conserved disulfide bonds shown as orange spheres. ii) CdiA-CT^Ykris in cartoon representation with its BECR core structure colored by as in Figure 2 left panels. iii) The CdiA-CT/CdiI^Ykris complex with CdiI colored as in Figure 2 right panels. In (B,C) the accessory protein (CysK or EF-Tu) is shown in beige, while on the left the toxin and immunity protein are colored by secondary structure where CdiA-CT has β-strands in yellow, α-helices in red, and loops in green; CdiI has α-helices in cyan and loops in pink. Active site residues for CdiA-CT are shown as yellow spheres. In the complex on the right, CdiA-CT is colored red, CdiI cyan and the accessory protein is in beige. (B) In UPEC 536, CdiA-CT forms a complex with CdiI and CysK [PDB ID: 5J5V (Johnson et al., 2016a)]. The C-terminal residue of CdiA-CT (green spheres) that inserts into the CysK active site is highlighted. Shown on the right is the overall oligomerization of the CdiA-CT/CdiI/CysK complex, where CysK dimerization results in a dimer of heterotrimers. (C) In E. coli NC101, CdiA-CT interacts directly with CdiI and domain 2 (DII) of EF-Tu, while CdiI interacts with DII and domain 3 (DIII) of EF-Tu (PDB ID: 5I4R; (Jones et al., 2017). On the right, the CdiA-CT/CdiI/EF-Tu complex is shown, where CdiI forms the center of the dimer of heterotrimers.

FIGURE 5

Potential avenues toward utilizing CDI toxins to treat human disease. (A) The toxin/immunity complex can be disrupted by designing small molecules that bind the immunity protein and free the toxin to initiate cell death (Williams and Hergenrother, 2012; Unterholzner et al., 2013; Morse et al., 2015). (B) A fusion toxin/immunity protein is generated that encodes an onco-specific POM (osPOM) at the immunity N-terminus that signals for oncoprotein E6 polyubiquitination in cervical cancer cells (Preston et al., 2016). Ubiquitination initiates the degradation of the immunity protein, freeing the toxin to initiate cell death. (C) Taking advantage of the upregulation of specific proteases in cancer and virus-infected cells, a fusion toxin/immunity protein can be designed with a linker that is selectively cleaved by an upregulated protease (Chan et al., 2015; Yeo et al., 2016). Protease cleavage frees the toxin and results in cell death. (D) A DNA sequence complementary to an onco-miRNA is placed upstream to the immunity gene, and functions as an osPOM. When the toxin and immunity genes are transcribed, the oncomiRNA binds the complementary sequence and initiates degradation of the immunity mRNA. Thus, in cancer cells, transcription of the immunity gene is silenced while toxin transcription continues, leading to toxin-induced cell death (Turnbull et al., 2019; Houri et al., 2020).