Salmonella, Shigella, and Escherichia, specifically E. coli, are the best known members of the Enterobacteriaceae, a large bacterial family that includes many enteric pathogens. Every year in the United States, according to the CDC, close to 400,000 cases of bacterial gastroenteritis are reported that are attributed to these three bacteria [1-3]. By CDC estimates, there could be at least twenty times more cases that go unreported [1,2].

Typical treatment of these infections is with bactericidal antibiotics. Depending on the severity of the infection, patients are prescribed either a β-lactam or a fluoroquinolone antibiotic [1,2]. However, in recent years, many strains have acquired resistance to commonly used antibiotics [4]. Many recent studies have documented how the use of antibiotics has resulted in the spread of resistance genes and other virulence factors through horizontal gene transfer [4-6]. Some state that antibiotics can no longer be used because of how quickly the populations are mutating and how widespread the changes are [7].

An alternative approach sparking considerable interest is to arrest the spread of pathogenicity rather than to kill the infecting pathogen [8]. Virulence genes, also known as pathogenic traits, are carried on pathogenicity islands that are frequently transferred across species boundaries by conjugation [9,10]. One important step in the transfer process is the signaling that occurs between donors and recipients [11]. A possible target of new drug therapy would be interruption of this signaling process and prevention of the spread of pathogenicity.

Pathogenicity islands are a specific type of mobile genetic element [4]. Mobile genetic elements are considered parasitic in nature because, unlike chromosomes which carry essential genes for life and reproduction, these mobile genetic elements carry accessory genes for non-essential processes [8]. Essential genes are those that encode for functions such as DNA replication and maintenance, transcription, translation and regulatory mechanisms. Non-essential processes include pathways for the biosynthesis of secondary metabolites and degenerate catabolic pathways [8]. Pathogenicity islands are metabolically costly for a host cell to maintain. Production of the various proteins they encode along with their separate replication requires resources and consumes energy, causing an increase in generation time compared to plasmid-free strains. Therefore, in order to balance their cost, pathogenicity islands have to confer some benefit onto the host [9,12].

Virulence genes encode for a variety of gene products, including proteins for attachment, entry, and toxins. Genera within the Enterobacteriaceae use a Type III Secretion system as an entry mechanism [10,13-15]. Type III secretion systems work by puncturing a pore in the host cell and injecting proteins into the host, like a needle. Both Shigella and Salmonella use them to start an infection by inducing uptake into host cells [14,16]. Another virulence factor shared by the members is the Shiga toxin, found in some strains of both Shigella spp. and E. coli. The Shiga toxins of E. coli, especially enterohemorrhagic strains that include E. coli O157:H7, are known to have been transferred via a pathogenicity island [17]. Though not virulent towards the targeted host, genes encoding bactericidal compounds also are advantageous to the pathogen, allowing them to kill off their competition in order to free up resources for colonization within their environment [18]. Some pathogenicity islands also carry antibiotic resistance genes [7].

Pathogenicity islands are very diverse, both in location and what genes are present in their genetic maps [4,10]. (Figure 1) They are either found inserted in the chromosome or as plasmids [9,10,21]. Pathogenicity islands such as the incompatibility group IncP-1 plasmids carried by Enterobacteriaceae have been found to be plasmids with a conserved backbone of genes that encode for a number of essential traits, and a variable set of interchangeable cassettes for other traits [22]. The backbone contains all the genes required for the replication, inheritance, and control of the island. A mechanism used by the pathogenicity islands of Enterobacteriaceae to maintain inheritance is “addiction” [22]. Addiction results from the presence of toxin and antitoxin-encoding genes in the island. These genes encode a stable toxin and labile antitoxin which degrades faster than the toxin. In order for the cell to survive, it must constantly transcribe and translate the genes for the antitoxin. Therefore, it must maintain the island [22]. Another mechanism is inheritance control by “active partitioning systems”. This mechanism ensures that all daughter cells receive the island, often by post-segregational sorting. On the other part of the island, the interchangeable cassettes include the virulence factors themselves [22]. For example, the IncP-1 plasmids are known to switch in and out cassettes for different antimicrobial resistance [22]. Similarly, an island found in Salmonella typhimurium has an insertion sequence, which when present, changes its phenotypic expression of flagella. In fact, different combinations of genes in a pathogenicity island have been found within the other members of Enterobacteriaceae, but none of them have the same combination. Nor do any of them use all of the possible open reading frames known to be in the island [16].

Figure 1 A generic pathogenicity island. Here the pink shaded area is the conserved backbone consisting of genes that regulate replication, transfer, and maintenance of the island. The other areas are the variable interchangeable cassettes. Here, the green could be a gene for entry and attachment, the blue a gene for toxin production, and the purple for antibiotic resistance. These islands can vary in their makeup. Some include all resistance genes, like the IncP-1 plasmids of Enterobacteriaceae [19], while others can include genes for a specific function like the acid survival island of E. coli [20].

The function of most interest in this review of pathogenicity islands is their ability to be transferred [8]. Movement of genetic information between parent and offspring is known as vertical gene transfer [12]. In comparison, mobile genetic elements are involved in horizontal gene transfer, the movement of genetic information from one cell to another non-offspring cell [4]. More often than not, traits acquired through horizontal gene transfer will be transmitted to the next generation. Of the three types of horizontal gene transfer, pathogenicity islands are commonly involved in conjugation.

Conjugation involves either direct cell-to-cell contact or a bridge, commonly a pilus, which allows DNA to be transferred from a donor to a recipient [9]. Conjugation can occur not only within members of the same species, but between members of different species, even those that are widely divergent [21]. Recently, the rate of transfer has been found to be the same, whether in one species or in mixed cultures [23]. This adds evidence to the assumption that not only can conjugation occur easily between species, but also that reservoir species that contain pathogenicity islands exist. In fact, bacteria-laden soil and water treatment plants have been found to be environments that favor conjugation, especially those with biofilms [4,6,21]. In order for donor cells to begin conjugation, they require some sort of signal that indicates the presence of recipients in the environment [11]. This is to make sure the cell is not constantly using resources in order to maintain the conjugation machinery indefinitely [12]. It is also a protective measure because pili and other proteins involved in transfer are targets for phages [12].

Conjugation signaling is separate from quorum sensing, a well-known mechanism of communication amongst bacteria [11]. Quorum sensing is a mechanism in which members of the same species secrete signaling molecules into their environment to determine the number of individuals present in the vicinity by the concentration of signal that accumulates in the extracellular environment [24]. It is used to determine if sufficient members are present in order to create and maintain a biofilm, to produce bioluminescent molecules, to sporulate, or to cause disease [19].

Once bound by the receptor, CSM activate a downstream signal transduction pathway that facilitates the assembly of an initiation complex that activates transcription of genes encoding transfer proteins [11,21]. Those transfer proteins are then assembled into both a conjugative pore, through which the DNA will pass, and a pilus, which stabilizes the mating pair during the transfer [21,23,25]. In addition to the role played by the CSM in transfer protein assembly, it has also been found that stimulation by CSM makes conjugation at a high frequency possible [26].

The CSM used by Enterobacteriaceae are actually the same signaling molecules used for quorum sensing, acyl homoserine lactones [12,20]. Previous studies have assumed that these molecules were only used in high density situations, like those required for biofilm production, but a recent study found that acyl homoserine lactones can be used as communication between two cells, even those with a large distance between them [24]. However, they are not produced by either Salmonella or E. coli [20,27]. Therefore, for either of those two species to begin conjugation, they must be activated by some other bacteria in the environment, like Yersinia or Pseudomonas [20]. In fact, work done in rabbit and cow gut has found that signaling between Yersinia and Salmonella actually intensifies the virulence of the Salmonella infection [20].

Receptors for acyl homoserine lactones are of one family, LuxR receptors [25]. The LuxR receptors are common among non-enteric Gram negative bacteria, like Chromobacterium and Pseudomonas [25]. On the other hand, members of Enterobacteriaceae produce a homolog of the LuxR receptor known as SdiA [20]. The SdiA receptor of Salmonella and its related genes are all acquired horizontally and allow Salmonella to detect a wide variety of acyl homoserine lactones. Detection of CSM allows Salmonella to upregulate genes and become more virulent within the host. On the other hand, in E. coli, detection of acyl homoserine lactones activates an acid survival island, which allows E. coli to survive in the rumen of cows and pass through the upper gastrointestinal tract, arriving intact at the recto-anal junction, which it prefers to colonize [20].

In recent years, the use of antibiotics has been shown to select for the presence of pathogenicity islands [4,12]. Antibiotics are used either to kill bacteria or to prevent their growth. Therefore, in order to survive, bacteria have acquired antibiotic resistance genes, some of which can be found in pathogenicity islands [9]. Through this process, the islands themselves are also acquired and spread. Some genes in the pathogenicity islands allow the bacteria to be more virulent, by changing their spreading mechanism. Some genes change the mode of entry used by pathogens, allowing them to mask themselves more efficiently [7,28,29]. In general, the use of antibiotics has been found to increase the rate at which bacteria are acquiring new pathogenicity islands [6].

Conjugation signaling presents a unique target for drug therapy to prevent the spread of pathogenicity islands [30]. A drug that targets this signaling could be a modified CSM that acts as an inhibitor or a synthesized molecule capable of binding to the receptor and modifying its activity [30]. In both cases, the goal is to disrupt downward in-cell signaling from the receptor so no conjugation occurs [30]. An example of a modified CSM has been worked on using Streptococcus. Streptococci use short peptide chains as CSM. In vitro a modified version was created by making a single amino acid substitution in the receptor binding region. This modified CSM not only prevented gene transfer, but it also inhibited other mechanisms of virulence, preventing colonization of the lungs of mice. This type of inhibition is competitive, and increased concentrations of the modified CSM were shown to have increased inhibition [31]. (Figure 2) Similarly, a possible mechanism to stop transfer of pathogenicity islands among Enterobacteriaceae are synthesized CSM antagonists. These differ from normal CSM in two structural locations, but are still able to bind the LuxR receptors. One antagonist prevents a transcription factor from binding DNA and activating transcription, while the other reduces or eliminates transcriptional activation by attenuating the interaction of the factor with RNA polymerase. These were shown in vivo to inhibit production of a pigment that is used by Chromobacterium violaecum to kill the nematode host, insuring survival of the host [25].

Figure 2 (a) Graphic of conjugation signaling molecules present in a donor cell’s environment. An internal signaling cascade starts from the membrane-bound receptor and travels to the pathogenicity island where it activates the transfer genes. These transfer genes begin producing the proteins needed to create and support a pilus for DNA transfer. (b) Graphic of conjugation signaling molecules and inhibitors present in the environment of a donor cell. The inhibitors out-compete the conjugation signaling molecule and as a result, do not allow for activation of the transfer genes, preventing pilus formation and conjugation. (c) Adapted from Zhu L, Lau GW, PLos Pathogens. 7(9):e1002241. The transformants per ml (T/ml) in the presence of three different conjugation signaling molecules (normal, modified signal 1 and modified signal 2) are 1,000 T/ml, 10 T/ml and 1 T/ml, respectively. These data show that the modified signaling molecules drop the number of transformants by a significant degree.

Advantages to using such inhibitors are blocking spread of the island, controlling the pathogen population, and also preventing virulence within the host [25,30,31]. However, since the CSM is the same molecule used for quorum sensing, it is hard to pinpoint exactly which processes are affected, and whether or not they are processes encoded by the pathogenicity island or the host. Further work into the downstream in-cell signaling is worth undertaking.

Mechanisms with other targets to stop the spread of pathogenicity and pathogenesis itself have also been presented in recent studies. One method focuses on using salicylidene acylhydrazides to block the type III secretion systems [32]. The other method attempts to cure cells of pathogenicity islands by inserting new post-segregational systems [33]. The former method holds some promise as it has been shown to inhibit E. coli virulence, but the latter has not been tested in vivo and therefore remains inconclusive.

Enteric pathogens cause disease using the traits in their pathogenicity islands. Now that antibiotic resistance is so common, alternate therapies are becoming attractive in order to treat these infections.

As an alternative treatment, drugs that target signaling have theoretical advantages over antibiotics [7,8]. Unlike antibiotics that kill, these drugs are only targeting the spread of the pathogenicity islands [33]. Since pathogenicity islands do not carry essential genes, this will not kill the bacterium, nor prevent its growth [33]. That diminishes the selective pressure impinged upon bacteria by antibiotics. It is postulated that without this selective pressure, many bacteria will lose their pathogenicity islands due to their cost of maintenance [5,24]. That is the ultimate goal for this type of therapy; that instead of killing the bacteria, allow them to live in their niche as long as they are not pathogenic and detrimental [24,33].

To date, most work has been done in non-enteric pathogens, focused on competitive inhibitors. Working with members of Enterobacteriaceae and studying non-competitive inhibitors are two more avenues worth following to further this therapy idea.