The type II secretion system as an underappreciated and understudied mediator of interbacterial antagonism

ABSTRACT

Interbacterial antagonism involves all major phyla, occurs across the full range of ecological niches, and has great significance for the environment, clinical arena, and agricultural and industrial sectors. Though the earliest insight into interbacterial antagonism traces back to the discovery of antibiotics, a paradigm shift happened when it was learned that protein secretion systems (e.g., types VI and IV secretion systems) deliver toxic “effectors” against competitors. However, a link between interbacterial antagonism and the Gram-negative type II secretion system (T2SS), which exists in many pathogens and environmental species, is not evident in prior reviews on bacterial competition or T2SS function. A current examination of the literature revealed four examples of a T2SS or one of its known substrates having a bactericidal activity against a Gram-positive target or another Gram-negative. When further studied, the T2SS effectors proved to be peptidases that target the peptidoglycan of the competitor. There are also reports of various bacteriolytic enzymes occurring in the culture supernatants of some other Gram-negative species, and a link between these bactericidal activities and T2SS is suggested. Thus, a T2SS can be a mediator of interbacterial antagonism, and it is possible that many T2SSs have antibacterial outputs. Yet, at present, the T2SS remains relatively understudied for its role in interbacterial competition. Arguably, there is a need to analyze the T2SSs of a broader range of species for their role in interbacterial antagonism. Such investigation offers, among other things, a possible pathway toward developing new antimicrobials for treating disease.

INTERBACTERIAL ANTAGONISM AND THE EMERGENT ROLE OF PROTEIN SECRETION SYSTEMS

Recently, there has been a renewed and expanded interest in interbacterial antagonism, that is, when one bacterium compromises another’s viability (1 – 6). Interbacterial antagonism occurs across the full range of ecological niches and includes all major phyla. Thus, understanding its mechanisms has much significance for clinical, environmental, agricultural, and industrial arenas. The earliest insight into interbacterial antagonism traces back to the time when antibiotics were revealed, and since then, various metabolites, peptidic bacteriocins, colicins, and perforin-like proteins have demonstrated antibacterial activity (1 – 3, 7 – 13). In these cases, which encompass Gram-negative bacteria and Gram-positive bacteria, the antibacterial factors are released from the producer by cell lysis or via the action of the Sec-translocon, ABC-type transporters, efflux pumps, or outer membrane (OM) vesicles and thereafter diffuse toward the target bacterium (1, 2, 14 – 16). However, some of these antibacterial proteins, for example, WapA of Bacillus subtilis, are exported by the Sec-translocon not to the extracellular milieu but to the cognate bacterial surface and once there mediates a form of contact-dependent growth inhibition (3, 16 – 18). As a variation on this theme, SitA is transferred from the surface of Myxococcus xanthus to the target’s OM in a process called OM exchange (2, 19, 20).

A paradigm shift occurred when it was realized that protein secretion systems, which are multi-component, membrane-spanning apparatuses, can also mediate the delivery of toxic “effectors” into competitors (1 – 3, 14, 21, 22). The most widely studied of this type of secretion system is the type VI secretion system (T6SS). Present in various Gram-negative bacteria, the T6SS is a spear-like apparatus (i.e., a repurposed phage fiber) that directly contacts the competing bacteria and then injects lipases, nucleases, peptidoglycan hydrolases, and other effectors (23 – 30). We and others have shown that type IV secretion systems (T4SSs), which are present in a subset of Gram-negative bacteria and evolutionarily related to conjugation systems, can also promote contact-dependent, interbacterial killing by delivering multiple protein effectors (22, 31 – 40). Some conjugative T4SSs mediate interbacterial antagonism independently of protein or DNA-based cargo (5). Contact-dependent interbacterial killing has also been linked to the type VII secretion system of some Gram-positive bacteria (41 – 45) and a subclass of the Gram-negative type V secretion system known as contact-dependent inhibition (CDI) (14, 46 – 50). In yet another example, variants of the type I secretion system of some Gram-negative bacteria secrete bacteriocins into the extracellular milieu or deliver other toxic proteins to the producer’s surface for cell-to-cell antagonism (3, 21, 51, 52). Finally, a version of the Gram-negative type III secretion system helps M. xanthus degrade bacterial prey (19, 53), and there is speculation that effectors of the type IX secretion system of Bacteroidota hinder competitors (54, 55). Not surprisingly, many bacteria use multiple methods for antagonizing competitors, including contact-dependent and contact-independent mechanisms and the utilization of more than one secretion apparatus (2, 3, 22, 56). Yet, there is another type of protein secretion system, the type II secretion system (T2SS), which, though well studied for other reasons, has been largely overlooked for its role in interbacterial competition.

THE T2SS

Evolutionarily related to the type IV pilus apparatus, T2SSs mediate a multi-step form of protein secretion (57 – 65). Proteins to be secreted by this system (substrates) are first transported across the inner membrane by the Sec or Tat translocon. Once in the periplasm, the substrates assume their tertiary conformation and, in some cases, oligomerize. Finally, the folded substrates are transited across the OM by the T2SS apparatus. In this last step, the T2SS “pseudopilus” behaves like a piston or Archimedes screw to propel the substrates through the T2SS’s OM secretin and deliver them into the extracellular space. The T2SS apparatus is typically composed of 12 core proteins, although there are instances of some bacteria having fewer constituent parts (62). Finally, in some cases, additional chaperones aid with the stabilizing and secreting of the substrates (66, 67). What ultimately causes a substrate to be recognized by the T2SS apparatus is not clear but likely involves the protein’s tertiary structure (61, 68).

Although, at one time, referred to as the main terminal branch of the general secretory pathway (57, 69, 70), the T2SS is not universal in Gram-negative bacteria (71). Indeed, in its canonical form, the T2SS is mainly present in the Proteobacteria and, even there, is not 100% conserved (62, 72, 73). Hence, the T2SS is rightly considered as a specialized system that (only) a subset of Gram-negative bacteria has evolved for growth in the environment and/or infection of host organisms (62, 74). However, many human and animal pathogens are known to express a T2SS, including, among others, Acinetobacter baumannii, Burkholderia cenocepacia, Escherichia coli, Klebsiella pneumoniae, Legionella pneumophila, Pseudomonas aeruginosa, Stenotrophomonas maltophilia, Vibrio cholerae, and Yersinia enterocolitica (Fig. 1) (62, 73, 75 – 85). T2SS-expressing plant pathogens include Dickeya dadantii, Erwinia amylovora, Pectobacterium carotovorum, Xanthomonas campestris, and Xylella fastidiosa, among others (Fig. 1) (62, 86 – 93). Just as the number of pathogenic species that have been shown to express a functional T2SS has increased in recent years, as reflected in Fig. 1, the number of processes ascribed to T2SSs during infection has steadily increased and currently encompasses tissue degradation in a range of body sites, plant cell wall degradation, and subversion of host defense factors, including complement, neutrophils, reactive oxygen species and mucus layers, suppression of immune signaling and cytokine destruction, adherence to surfaces, biofilm formation, invasion of and growth within host cells, host cell death and lysis, alterations in ion flux, reductions in blood coagulation, and nutrient assimilation after the breakdown of proteins, lipids, and carbohydrates (62, 94 – 113). T2SSs are also active in many environmental, non-pathogenic species (Fig. 1), expediting an expanding list of metabolic processes and symbioses, for example, iron, manganese, and iodate reduction, hydrocarbon degradation, and nutrient trafficking (62, 114 – 127). For many of the genera in Fig. 1, there are additional species in the genus that also carry the genes for a T2SS but functional analyses have not yet been reported (66, 73, 76, 128 – 144). In addition to the pathogenic and non-pathogenic genera listed in Fig. 1, there are many other genera within the Proteobacteria that encode the genes for a T2SS (that are distinguishable from the genes for a type IV pilus) and likely express T2SS-dependent proteins (Fig. 2) (73, 135, 145 – 156). The genes for T2SSs are typically present within the bacterial chromosome; however, there are examples of the system being encoded within a plasmid (108, 146, 157).

Fig 1.

Species in which a secreted protein/activity or phenotype is linked to the T2SS. Species belonging to the α-Proteobacteria are shaded in purple, β-Proteobacteria in orange, γ-Proteobacteria in blue, δ-Proteobacteria in yellow, and non-Proteobacteria in green. This is not necessarily an exhaustive list.

Fig 2.

Additional genera within the Proteobacteria that carry genes for a T2SS. This is not necessarily an exhaustive list.

The output of a T2SS can range from one to dozens of proteins, encompassing a diverse array of peptidases, proteases, phosphatases, carbohydrate-degrading enzymes (e.g., cellulases, chitinases, and mucinases), lipolytic enzymes (lipases and phospholipases), nucleases (DNase and RNase), reductases, pore-forming proteins, ADP-ribosylating toxins, and novel proteins (62, 99, 102, 158 – 168). Although most T2SS substrates ultimately exist (only) in the extracellular milieu, some also locate to the surface of the expressing cell (62, 107, 111, 169 – 174). Some bacteria encode two or three T2SSs that might mediate the release of different sets of substrates (62, 153, 175). Despite the vast amount of work done on T2SSs, a significant role for these systems in interbacterial antagonism has not been described or posited in the many reviews on the T2SS that extend from 1990 to the present (2, 3, 14, 18, 21, 22, 57, 58, 60 – 62, 69 – 71, 73, 79, 114, 174, 176 – 222). Consequently, the impression has been that T2SSs are not important for interbacterial competition but are devoted to virulence or nutrient assimilation.

CONNECTIONS BETWEEN T2SSs AND INTERBACTERIAL ANTAGONISM

From the mid-1960s to the present, bacteriolytic enzymes have been detected in the culture supernatants of different environmental, Gram-negative bacteria (223 – 228). Such enzymes that have been characterized include the peptidoglycan-targeting α-lytic proteases, β-lytic proteases, and CwhA amidases of Achromobacter lyticus and Lysobacter sp. and the lysozyme-like enzymes and lipases from M. xanthus and other myxobacteria (229 – 240). The documented presence of a signal sequence in the N-terminus of many of these enzymes suggests that at least some of them are substrates of the T2SS. Compatible with such a scenario, the T2SS apparatus genes are upregulated at the time when Lysobacter capsici produces its bacteriolytic proteases and M. xanthus preys on other bacteria (125, 241). However, a formal linkage to the T2SS, for example, documenting the loss of the secreted protein in a T2SS mutant’s supernatant, has not occurred yet. The first clear connection of a secreted, bacteriolytic enzyme to a T2SS began in 1993 when the LasA elastase of P. aeruginosa was shown to be equivalent to a previously defined staphylolytic enzyme in P. aeruginosa supernatants (242 – 247). In 1998, LasA was confirmed as being a substrate of the P. aeruginosa T2SS, when it proved to be undetected in supernatants of an xcp T2SS mutant (248 – 250). A member of the M23 family of peptidases (251, 252), LasA lyses Staphylococcus aureus by cleaving the pentaglycine within the peptidoglycan of that target cell (242, 247). Despite these data, the antibacterial function of LasA was not featured in the many reviews on T2SSs and interbacterial antagonism that later appeared (as noted above), although attention was frequently directed toward the role of the protein’s elastase activity in infection. Incidentally, another staphylolytic enzyme dependent on the P. aeruginosa T2SS has been suggested, but the identity of that factor remains undefined (253 – 255). The next connection between an antibacterial activity and a T2SS occurred in 2020 and involved a marine species of Pseudoalteromonas. Specifically, an M23-peptidase known as pseudoalterin was found to be secreted via the T2SS and to promote the killing of S. aureus and various other marine Gram-positive bacteria (122, 256). As is the case for LasA, pseudoalterin acts on the peptide chain within the peptidoglycan of its target bacterium (122). The third link between antibacterial activity and a T2SS came in 2021, when the NlpC/P60 endopeptidase (PnpA) secreted via the T2SS of Photobacterium damselae was shown to degrade in vitro purified Vibrio peptidoglycan (257 – 259). Yet, an outstanding question from this study is how PnpA naturally bypasses the Gram-negative target’s OM in order to reach the peptidoglycan. One possibility is that a T2SS-dependent lipase or an effect of another secretion system first disrupts the lipid bilayer creating a pathway for PnpA to access the periplasm. A final study linking a T2SS to interbacterial antagonism occurred in 2022, when a T2SS mutant of Plesiomonas shigelloides was found to be impaired for killing E. coli upon co-incubation on solid media (81). The secreted bactericidal protein(s) of P. shigelloides remains unknown, however. Based on these data, the T2SS can, in fact, be a mediator of interbacterial antagonism, and it is conceivable that many T2SSs have antibacterial output. Yet, the T2SS still remains understudied for its role in interbacterial competition, especially when compared to other protein secretion systems.

CONCLUDING THOUGHTS AND FUTURE QUESTIONS

Despite what has been the prevailing impression, it is logical that T2SSs would be another means for interbacterial antagonism. For example, the different proteases/peptidases, lipases, and carbohydrate-degrading enzymes that are secreted by T2SSs could theoretically alter many moieties on the surface or in the envelope of a competitor leading to a loss of function or cell death (while not necessarily harming the producer). Based on the examples above, peptidoglycan appears to be a common target for antibacterial T2SSs. When the competitor is a Gram-positive bacterium, an enzyme acting on peptidoglycan might alone suffice. But, when the competitor is another Gram-negative bacterium, enzymes that act on the target’s OM would seem to be also necessary for effective competition. On the other hand, some T2SS substrates might act indirectly, for example, by processing foodstuffs in the extracellular milieu in a way that makes them less accessible or useful to competitors. Since some T2SS substrates (also) reside on the producer’s surface, T2SSs might even facilitate a novel form of contact-dependent killing. Finally, it is possible that some T2SS substrates potentiate the action of another antibacterial secretion system, just as some T2SSs act to enhance the effects of those other systems on eukaryotic hosts (222, 260). Overall, T2SSs likely contribute to a multi-pronged strategy of interbacterial antagonism, especially for those Gram-negative bacteria that do not have one or more of the other systems. Aside from these types of mechanistic questions, it will be beneficial for future investigations to discern what other T2SS-encoding bacteria (Fig. 1 and 2) use their T2SS for antibacterial antagonism, to what degree, and with what types of effectors. Current Basic Local Alignment Search Tool (BLASTP) searches indicated that proteins with significant amino acid sequence similarity to LasA, pseudoalterin, or PnpA are encoded within the genomes of many of these other species (Fig. 3), further suggesting that these organisms might similarly employ their T2SS for interbacterial competition. Yet, given the ecological diversity of the >100 genera in Fig. 1 and 2, it is likely that new types of effectors and new forms of competition will also be revealed. For such an endeavor, it will be valuable to assess the role of the T2SS in models that simulate natural niches, whether that be an aquatic or terrestrial habitat, the rhizosphere, or infection of an animal or human host. Another interesting question will be if any known or yet-to-be-defined T2SS substrates that target bacteria also confer activity against fungi or protists. Further investigation of T2SS substrates as agents of antibacterial antagonism also offers a possible pathway toward identifying new antimicrobials that could be used to treat infectious diseases. Along those lines, LasA has been used as a treatment for experimental staphylococcal eye infections (261 – 263). In sum, an expanded appreciation for T2SSs is likely to yield important new insight into the mechanisms of interbacterial antagonism, pathogenesis and potential disease therapies, and diverse ecological niches.

Fig 3.

Known antibacterial effectors of T2SSs and some of their homologs encoded within the genomes of other T2SS-encoding species. (a) BLASTP results using the LasA sequence from P. aeruginosa strain PA01 (accession no. NP_250562) as the query. Top hits were proteins from other species of Pseudomonas; however, they were not presented in order to focus on related proteins that occur in other genera. The five examples given are proteins that show some of the greatest levels of similarity to LasA and are from diverse species that are known to encode a T2SS. (b) BLASTP results using the pseudoalterin sequence from Pseudoalteromonas sp. strain CF6-2 (accession no. WP_237115101) as the query. The examples listed are five that show some of the greatest levels of similarity to pseudoalterin and are from a range of non-Pseudoalteromonas species that are known to encode a T2SS. (c) BLASTP results using the PnpA sequence from P. damselae strain MT1415 (accession no. 6SQX_B) as the query. The proteins listed are five that showed some of the greatest levels of similarity to PnpA and are from non-Photobacterium species that are known to encode a T2SS.