Genome characteristics
Pairwise DNA sequence comparisons were made to sequences deposited in the EzBioCloud database. The 16S rRNA gene sequence (1,436 bp) from our isolate, strain porpoise, showed >99% identity to six different Proteus species. To identify the bacterium more accurately, WGS was performed. Genome-based taxonomy determined strain porpoise belongs within the species faecis.
The genome of P. faecis porpoise consists of 3,802,708 bp. The number of contigs is 74. The contig L50 is 9 and the contig N50 is 135, 430. No plasmid replicons were detected. The features of all nine genomes used in this study are presented in Table 1.
Putative virulence factors
Table 2 lists the putative virulence factors found in the genomes of the nine different P. faecis strains used in this study. The production of cytotoxic haemolysins is very common in pathogens (Hacker and Hughes 1985). For example, P. mirabilis secrete haemolysins which insert themselves into the membrane of host cells resulting in pore formation, and cytotoxicity (Braun and Focareta 1991). In our study P. faecis porpoise exhibited beta-haemolysis when grown on blood agar. A putative haemolysin protein (Fig. S1) was found in all nine strains.
A type VI secretion system (T6SS) was found in all nine strains. This is not unexpected since T6SS gene clusters have been found in over 8000 genomes (Zhang et al. 2022). Twenty five percent of gram-negative pathogens have this apparatus (Bingle et al. 2008). A T6SS functions in a contact dependent manner. Neighboring cells are first punctured and then injected with multiple lethal effectors (Ho et al. 2014; Basler et al. 2015). The T6SS of Proteus mirabilis HI4320 is encoded by 17 highly conserved genes (Alteri and Mobley 2016). From our study, the amino acid sequences of a putative T6SS component TssG are shown in Fig. S2. We also found Phospholipase A1 which is an antibacterial T6SS effector. For delivery into a bacterial cell, this protein interacts with the VgrG cargo protein (Flaugnatti et al. 2016). The amino acid sequences of Phospholipase A1 are shown in Fig. S3.
The type III secretion system (T3SS) is present in numerous Gram-negative bacteria. Some are human pathogens capable of causing dysentery, plague, and typhoid fever (Hueck 1998; Abby and Rocha 2012). The T3SS machinery allows bacteria to interact with target cells (Hueck 1998; Galán et al. 2014) and disrupt cellular processes, such as signal transduction, vesicle transport and cytoskeletal dynamics (Büttner 2012). Based on sequence analysis, there were genes encoding a potential T3SS found on the genome of P. mirabilis HI4320 (Pearson et al. 2008). In our study, a putative T3SS gene cluster was found (Fig. 2). The amino acid sequences of a Type III secretion outer membrane pore forming protein are shown in Fig. S4. The T3SS machinery delivers virulence proteins called type III secreted effectors (T3SEs) into the cytoplasm of a host cell. T3SEs can cause changes which are detrimental to the host cell resulting in a disease (Galán et al. 2014; Jennings et al. 2017). However, only a few T3SEs have been found for bacteria with a known and functional T3SS (Hu et al. 2017). In this study two type III secretion proteins were identified (Fig. 2). Since these proteins are part of a T3SS gene cluster they could be T3SEs. The amino acid sequences are shown in Fig. S5a and b.
Biofilm is a highly structured microbial communities where bacterial cells attach to a biotic or abiotic surface and embedded in a matrix, blocking the killing from antimicrobial agents or host defense (Kostakioti et al., 2013; Lebeaux et al., 2014). Sometimes mineral crystals can also be found in the polysaccharide matrix (Donlan 2002). P. mirabilis is capable of biofilm formation. This can be seen in chronic wound (Rajpaul 2015) and in urinary tract infections (Armbruster et al. 2018). Interestingly, the bacterium can form a biofilm of a crystalline nature due to its ability to produce urease which can in some cases lead to encrustation (Jacoben et al. 2008). In this study a putative biofilm regulator and urease alpha subunit were found. The amino acid sequences are shown in Figs. S6 and S7, respectively.
In this study the ability of biofilm formation by P. faecis porpoise was investigated in DMEM medium, which is of low nutrition, simulating an in vivo environment. As shown in Fig. 3, P. faecis develops a biofilm. The ∆eseB strain and the ∆eseC strain were included as the negative control and positive control, respectively. The EseB filament-mediated bacterial cell-cell interaction prompts biofilm formation, however, EseC inhibits biofilm formation through sequestering EseE, a positive regulator for EseB (Gao et al. 2015; Liu et al., 2019). Since biofilms act as a primary barrier to protect bacteria from the detrimental effects of antibiotics, cytokines, and complement(Wakimoto et al. 2004), our results suggest P. faecis could evade host defenses and survive in vivo through biofilm formation.
Adherence to a biotic or abiotic surface is a critical first step in biofilm formation (Harriott et al. 2019). In Gram-negative bacteria, flagella and pili/fimbriae, often aid in the attachment (Berne et al., 2015). Fimbriae allow the adherence of P. mirabilis to medical devices and uroepithelial cells which can result in a urinary tract infection (Armbruster et al. 2018). From our study, the amino acid sequences from type IV pilus protein PilA are shown in Fig. S8.
Flagella are whip-like appendages which allow for the locomotion of bacteria (Haiko and Westerlund-Wikstrom 2013). Swarming motility is defined as the movement across a semisolid surface, which is operationally defined as multicellular, flagella-mediated surface migration of bacteria (Ha et al. 2014). It is well known that P. mirabilis is capable of swarming motility. In a mouse model it has been demonstrated that P. mirabilis swarmers were able to migration across urethral catheters and enter into the urinary tract (Li et al. 2002). It has also been shown that P. mirabilis swarmer cells are able to establish an ascending infection in a mouse kidney (Allison et al. 1994). Swarming is an important virulence factor since P. mirabilis strains that exhibit non-swarming or weakly swarming capabilities have a reduced ability to colonize the bladder and kidney of mice (Mobley et al. 1996). In this study the motility of the P. faecis porpoise was analysed on LB medium with 0.6% agar or on TSA medium with 0.3% agar. E. piscicida strains were included as the controls. It was observed that P. faecis porpoise can mobilize on the surface of soft agar. The WT strain (E. piscicida PPD130/91) can also move on a semisolid surface but failed to move with the depletion of FlhB (Fig. 4). This experiment indicates that P. faecis porpoise could have flagella, which allows for swarming. Some bacteria, such as Bacillus subtilis, swarm on a wide range of energy-rich media, whereas other bacteria, such as Salmonella enterica and Yersinia entercolitica, require the presence of particular supplements, such as glucose (Julkowska et al. 2005; Young et al. 1999; Harshey et al. 1994). Our results show that P. faecis porpoise can swarm on an agar surface without the need for special supplementation (Fig. 4). The amino acid sequences from flagellar cap protein FliD are shown in Fig. S9.
The twin-arginine translocation (Tat) is a type of protein transport pathway which can be found in both bacteria and archaea. This system is specific for precursor proteins which harbour a twin-arginine pair in the signal sequence. Unlike other translocase systems Tat translocases can translocate proteins which are completely folded across the phospholipid bilayer (Fröbel et al. 2012). Importantly, virulence factor proteins and virulence-associated proteins are Tat substrates (Pradel et al. 2003; Rossier Cianciotto 2005; Zhang et al. 2009). In P. mirabilis it was shown that the twin arginine translocation system is very important in motility, fitness and translocation of virulence factors within the bloodstream during an infection (Armbruster et al. 2019). Sequence information for the Twin-arginine translocation protein TatA, TatB and TatC in P. faecis strains are shown in Figs. S10, S11 and S12, respectively.
The caseinolytic protease (ClpP) is a serine protease which is conserved across several different kingdoms. In Mycobacterium tuberculosis a ClpP depletion resulted in a reduction in colony forming units (CFUs) as shown both in vitro and using a mouse model (Raju et al. 2012). In our study a putative ClpP was found. The amino acid sequences from the ATP-dependent Clp protease proteolytic subunit ClpP are shown in Fig. S13.
Iron is essential for all forms of life. Many proteins and enzymes require this metal in order to function. As a way of combating bacterial infections iron is sequestered by the host. In early studies it was thought that P. mirabilis lacked iron-scavenging systems (Miles and Khimji 1975; Evanylo et al. 1984). However, more recent studies have shown the bacterium is able to scavenge iron. For example, it is unable to make the siderophore enterobactin, but it can still use this molecule which was produced by other bacteria (Himpsl et al. 2010). In this study a ferric anguibactin-binding protein was found in all nine strains. The amino acid sequences are shown in Fig. S14. This siderophore has been documented as an important virulence factor in the fish pathogens Vibrio anguillarum (Wolf and Crosa 1986) and Photobacterium damselae subsp. piscicida (Osorio et al. 2015). In bacteria the ferric uptake regulator (Fur) protein tightly regulates the transport of siderophores in response to the availability of iron (Ernst et al., 1978; Köster 2001). The amino acid sequences of a putative Fur protein from P. faecis strains are shown in Fig. S15. In Escherichia coli, there are different importer systems. The FhuCDB system is used to import hydroxamate-based siderophores (Fecker and Braun 1983). In this study putative ferrichrome transport system permease protein FhuB (Fig. S16) and ferrichrome transport ATP-binding protein FhuC were found (Fig. S17). A putative iron transport gene cluster is shown in Fig. 5.
Antimicrobial resistance
P. mirabilis does not contain any beta-lactamases which are chromosomally encoded. As a result, wild-type strains are sensitive to all lactams (Girlich et al. 2020). However, since there have been several blaOXA carbapenemases, originating in Acinetobacter spp., found also in P. mirabilis strains (Girlich et al. 2020) the authors examined the antimicrobial resistance status of P. faecis porpoise. The bacterium was not resistant to the carbapenems tested in this study. It was resistant to ampicillin, but not other β-lactam antimicrobials (Table 3). The CARD database identified genes which could confer resistance to ampicillin, through mutations in the penicillin-binding proteins (D350N). The resistome of relatively few P. faecis isolates have been examined. No antimicrobial resistance genes were found in the genome of P. faecis CR112 isolated from a healthy fish (Paul et al. 2024). However, multidrug resistant (MDR) P. faecis FZP1097 was isolated from a wound secretion from a hospital located in Sichuan Province, China. The bacterium was resistant to ampicillin, aztreonam, ciprofloxacin, ceftriaxone and trimethoprim/sulfamethoxazole. It was believed there was an unknown resistance mechanism involved in the MDR phenotype (Li et al. 2002).