What we perceive as virulence factors are often the tools that bacteria use to establish their colonies within the human host. Bacteria are driven by the same requirements as people: food, shelter, and a place with enough room to raise the offspring. Each of these virulence "tools" is only useful in specific circumstances, and they require considerable energy to manufacture. For these reasons, bacteria must regulate the production of these "tools" and express them only when needed.
Like the settlers who moved to the Western United States, many bacteria are simply looking for a nice comfortable place with sufficient food and space for them to lay down roots and build a small colony. Individuals or small numbers of bacteria have limited metabolic needs and maintain an inconspicuous existence. While in small numbers, they keep a low profile to avoid provoking an inflamed response from the local inhabitants. As the numbers of individuals and their needs grow, there would be greater need for food and for space. Settlers would start to deforest and abuse the land, attack and push away other inhabitants, and build fortresses for protection. Similarly, bacteria in small numbers may establish small colonies or establish a mutual coexistence agreement with their hosts to live as normal flora. Many bacteria, even bacteria such as Staphylococcus aureus, can coexist with their human hosts without causing disease as individuals or in small numbers in specific environments such as the skin. In hostile environments, the bacteria may need to anchor their colonies to human tissue with adherence factors to avoid being washed away or seek shelter within the cells of the host to hide from detection. As the numbers of pathogenic bacteria increase and the colony requirements for food and space increase, the bacteria may become more aggressive and virulent. They may release tissue-destroying toxins or protect themselves with biofilms or wall themselves off from the host with coagulase-built walls of fibrin. These actions provide for the basic requirements of the colony but, like the settlers action, will damage the tissue environment. The damage caused by the bacteria produces pathology; the mechanism is termed virulence, the tools are virulence factors, and the result is disease.
Unlike the genes encoding proteins that are essential for growth of the bacteria (housekeeping genes), genes for virulence factors (categorized as accessory genes) are needed only under certain conditions. Because constant expression of virulence genes would be metabolically expensive, bacteria have evolved the means to turn-on and turn-off the synthesis of these processes when appropriate. Expression of virulence genes is controlled primarily at the level of mRNA production by regulating when and where the RNA polymerase will transcribe the DNA into mRNA. An interconnecting network of regulatory proteins and regulatory RNA promotes or prevents the binding of the RNA polymerase to the DNA and determine which mRNAs and proteins will be made.
As for the settlers described above, many factors will influence the actions that must be taken. Specific triggers may cause minimal changes, whereas survival considerations may require global changes in the expression of many genes from many parts of the bacterial chromosome. The purpose of this article was to describe the triggers and mechanisms by which pathogenic bacteria control the expression of genes for virulence factors and survival.
Like the settlers, bacteria have many ways to sense and then signal changes in the temperature, food supply, and even the numbers of members within the colony. There is also a hierarchy of importance for the signals and their consequences. For the bacteria and the settler, the most important consideration is having a food supply and avoiding starvation followed closely by adaptive responses to the climate and environment. The colonist's clothes, how much wood to chop, and other behaviors will be influenced by the temperature and the severity of the environs. Similarly, bacterial behavior is responsive to signals from specific protein sensors which detect changes in temperature, but also pH, osmolarity, oxygen tension, and specific ions indicative of their environment (Fig. 1). Metabolic changes, indicated by the nitrogen balance or concentrations of specific metabolites, or stressors such as heat, salt, or ethanol concentration are also major influences. Some of these signals indicate a transition from outside to inside the human body and activate production of the necessary molecular machinery to raise a colony. Other sensors monitor the growth phase of the bacteria (exponential, postexponential, or stationary) or the number of bacteria within the local colony (presence of a quorum). Such information is used to optimize the metabolic machinery within the cell, allow the bacteria to adjust to a new environment (inside vs outside the body), promote the production of appropriate virulence factors, or take drastic measures for survival, such as sporulation. The control mediated by these triggers may require a global change in the operations of the bacteria, activation of multiple virulence factors, or fine tweaking of an enzyme pathway.
The ability to distinguish between the outside of the body and the different environments within the body allows bacteria to activate the correct tools for parasitism. Production of fimbriae and other adherence mechanisms that attach Escherichia coli to the lining of the urethra or the bowel are good examples of structures and proteins that would only be useful when the bacteria are in the body and not sitting in a pile of manure. Some virulence factors are only needed or are produced in sufficient concentration to be effective when there are sufficient numbers of bacteria, a quorum. For example, the presence of a quorum of bacteria is required to produce sufficient polysaccharide to form a useful biofilm to protect the colony or enough toxin (eg, degradative enzymes) to kill tissue and release metabolites to feed the colony. The product of a single cell would be insufficient to have an effect and would be a waste of metabolic currency.
Turning It On and Turning It Off: Control of Gene Expression
Translating the molecular signals from the sensors into expression of protein virulence factors is a highly regulated and coordinated process. Coordination of gene transcription can be accomplished in several ways. Small molecules, proteins, or even RNA can interact with sequences of DNA to facilitate (promoters) the binding of the RNA polymerase or prevent (repressor) the binding or the progression of the polymerase along the DNA to promote or prevent the production of mRNA. Binding of the RNA polymerase can also be influenced by the tertiary structure of the newly synthesized mRNA. Grouping of the genes for a metabolic pathway or structure together into operons of adjacent genes facilitates the coordination of their control by a shared set of promoters and repressors. A specialized, relatively large operon of virulence genes is termed a pathogenicity island. Grouping of genes for a specific virulence factor, such as a type III secretory device, under the control of a cluster of promoters allows coordinate expression of the multiple proteins for this complex molecular device. For example, the Salmonella pathogenicity island 1 (SPI-1) of Salmonella typhimurium contains at least 14 genes organized into 8 different operons which are coordinately controlled by outside signals, and 2 regulatory proteins, HilA and InvF, also encoded within the pathogenicity island (discussed later).1
By using promoter or repressor sequences responsive to the same molecular code word, the bacteria can mediate global control of genes spread around the chromosome or even on plasmids. The concentration of a small molecular indicator of the metabolic condition of the cell, such as cyclic adenosine monophosphate (cAMP) or ppGpp, can be sensed by proteins which activate promoters or repressor proteins. Cyclic adenosine monophosphate is used as a signal that glucose levels are low and ppGpp as a signal for nutrient depletion2 and that alternative metabolic pathways are required. Activation of biofilm production in Pseudomonas aeruginosa and the turn-on of toxins and turn-off of cell wall-associated proteins in S. aureus are also responsive to small molecules. Expression of biofilm production and other virulence factors is triggered when sufficient numbers of bacteria are present to justify the effort. In this process, termed quorum sensing, each bacterium makes a quantity of a small molecule which can be sensed by a cellular protein. For P. aeruginosa, the small molecule is an N-acyl homoserine lactone, and for S. aureus, it is a cyclic peptide (discussed later). When sufficient concentration of the molecule is reached, it triggers the activation of genes in different parts of the chromosome.
Even the level of DNA supercoiling will control gene expression. DNA supercoiling is maintained by topoisomerases, and several of these enzymes are responsive to levels of ATP. The level of supercoiling of the DNA will affect the ability of the regulatory proteins to bind to their promoter and repressor sites and will even affect the RNA polymerase's ability to bind to the DNA.
The overriding approach to global control of gene expression is mediated by a change in the composition of the RNA polymerase through variation of the σ factor.3,4 Sigma factors are an accessory component of the RNA polymerase (core enzyme consisting of α2ββ′ subunits) which help the polymerase choose which DNA sequences to transcribe into mRNA. Different σ factors promote binding to different promoter DNA sequences preceding specific genes or operons. In E. coli, different σ factors are used to promote transcription of genes for groups of proteins needed for exponential growth, for flagellar components and chemotaxis, for nitrogen metabolism, or for proteins required during the stationary growth phase, or to deal with different stresses including heat shock, acid, lack of iron, and starvation. Sporulation, which requires a global change in gene expression in answer to starvation, is also controlled by a special σ factor.
As students, the classic example of an operon was the lac operon which is regulated by the sugars that are available to the bacteria. The lac operon encodes the transport protein and enzymes for utilization of lactose instead of glucose and is controlled by global and specific signals. Normally, the lac operon is turned off by a repressor protein which binds to a sequence of DNA to block the RNA polymerase from binding to the gene and thus prevents production of mRNA. When the bacteria run out of glucose, many metabolic changes must occur in the bacteria to utilize alternate energy sources such as galactose. These are coordinated by cAMP which provides a cellwide global signal. Cyclic adenosine monophosphate binds to a protein which binds to a promoter sequence of DNA to attract the RNA polymerase to the lac operon; however, the expression of these proteins is not necessary unless galactose is present. Galactose binding to the repressor protein releases it from the gene to allow the gene to be transcribed into mRNA for the production of the required enzymes and transport proteins.
"Staph-ing" a Colony: Aggressiveness Growth With the Development of a Quorum
As indicated above, S. aureus controls the expression of its virulence genes according to cues that it gets from the environment, such as pH and carbon dioxide concentration, as well as the growth phase of the cell and the size of its colony. These cues provide the signals necessary to change the molecular repertoire to support an individual settler or a colony-dwelling resident. As the numbers of bacteria increase, adhesion functions are minimized, and cell wall thickness is less important, but toxin, protease production, and a biofilm become more important to provide food, growing space, and protection for the growing colony. The size of the colonial quorum is sensed by the amount of a small cyclic peptide termed an autoinducing peptide (AIP) that is produced by each bacterium. The AIP is produced by the agr locus of S. aureus which also produces sensor proteins to detect the AIP. As the concentration of AIP increases, the production of AIP, AIP sensors, and regulators of virulence genes also increase. This "autoinduction" of the production of the sensor and the sensed peptide is how the protein and the gene locus, the accessory gene regulator (agr),5 get their names. Agr can also be nicknamed the "aggression regulator." Staphylococcus aureus uses these AIPs as a means for taking the census of their colony and whether the colony is in exponential, postexponential, or stationary phase of growth.
As the concentration of AIP increases, the bacteria increase the production of RNA III. RNA III is an RNA which encodes the δ-hemolysin protein, but the RNA itself forms a structure which either directly or indirectly activates the genes of aggression. RNA III can activate a transcriptional regulator (SarT, SarS) which will control the synthesis of different proteins, or RNA III can interact directly with the mRNA for a specific protein, such as the α toxin, to promote the translation of the protein.6 In such a manner, the RNA III provides the means for coordinating a change in bacterial behavior. As the numbers of bacteria increase, the RNA III that is produced inhibits the production of the repressor of toxin (rot) protein. At low bacterial numbers, the rot inhibits the expression of α toxin; β-, δ-, and γ-hemolysins; the Panton-Valentine leukocidin; TSST-1; enterotoxins; exfoliatins; V8 serine; and other proteases. The RNA III also turns off functions that are more relevant to individual bacterial survival such as adherence (eg, fibronectin-binding protein) and the cell wall-associated protein A.
Different strains of S. aureus make different AIPs which differ by single amino acids, and the different AIP and corresponding agr variants are associated with different strains and different diseases.7 For example, certain agr groups are strongly linked with certain clinical syndromes. Groups I and II are linked to endocarditis, and group III to production of TSST-1 and hence the toxic shock syndrome. Group IV AIPs are associated with exfoliatin production and hence scalded skin syndrome.
The growth phase of individual S. aureus bacteria regulates the expression of toxin production through another major genetic regulator, the staphylococcal accessory regulator (sarA). This protein can turn-on production of RNA III to reinforce the quorum activation by AIPs. sarA can also bind directly to promoters and turn-on many toxin genes and turn-off other genes.
Oxygen levels, as in a deep wound, or the metabolic changes that accompany the lack of oxygen,6 trigger production of a regulator which turns-off α toxin and other virulence genes. α-Toxin requires oxygen for function. An increase in oxygen may promote expression of TSST-1 and other virulence genes.
An interesting consequence of the coordination of the control of cell wall peptidoglycan synthesis and toxin production by the agr locus occurs in some S. aureus strains that develop resistance to vancomycin. These GISA (glycopeptide-intermediate S. aureus) strains have a mutation that inactivates the agr locus, primarily in group II agr strains of S. aureus. In the absence of agr, the bacteria will have a thicker cell wall with more binding sites for vancomycin and hence require more glycopeptide for antimicrobial activity. The lack of agr, however, will prevent the activation of toxin production. This will reduce the virulence of the bacteria but possibly increase its success as a parasite.7 This interconnected example illustrates the complexity of the different turn-ons and turn-offs and the consequences of regulation of S. aureus virulence factors.
Enteric bacteria make many adjustments to accommodate the many different environments and conditions that they encounter during infection of the human body. Many of the triggers for turning on virulence factors sense and can distinguish between being in the gut and sitting in stool. The bacteria distinguish inside from outside by the temperature, pH, osmolarity, oxygen conditions, magnesium and phosphate levels, quorum size, and even epinephrine and norepinephrine levels.8,9
Environmental cues are essential for regulating the expression of the type III secretion (T3SS) devices of enterohemorrhagic E. coli, Salmonella, and Shigella. T3SS devices act as molecular syringes to inject virulence factors into host cells to facilitate attachment to the intestinal wall and invasion of the bowel (discussed further in Bower and Rosenthal10). An excellent depiction of T3SS action of E. coli can be seen at the Howard Hughes Institute Web site (http://www.hhmi.org/biointeractive/). The genes for the many parts of the T3SS are encoded within pathogenicity islands which are coordinately controlled by environmental and other signals.
Salmonella has the ability to sense and adapt to several different environments. These bacteria progress from outside the body into and through the stomach and small intestine and then cross the epithelial lining through M cells or epithelial cells to invade the body. After being gobbled up by macrophages, Salmonella species set up housekeeping within phagosomes. In making this transition, the bacteria coordinately turn-on or turn-off the expression of 20% of their genes.11,12 The initial control signal is delivered by acid sensors which activate individual genes as well as a unique σ factor, RpoS (σ S), to facilitate tolerance to acid and oxidative stress (acid tolerance response [ATR]). This amounts to 50 to 60 genes. Upon sensing the low oxygen tension and high osmolarity of the small intestine, Salmonella species activate the genes within the SPI-1 which encodes a T3SS. The T3SS injects proteins across the host membrane to rearrange the actin cytoskeleton and also promote nuclear responses to facilitate the internalization of the bacteria but leads to the eventual production of diarrhea.1,13 The SPI-1 pathogenicity island for the T3SS contains approximately 40 genes. These genes include the different components of the device, proteins to be secreted through the device, and the 2 controller proteins, HilA and InvF, mentioned earlier. The expression of HilA and InvF is determined by the oxygen tension, osmolarity, and the growth phase of the bacteria and is also influenced by global regulatory proteins. These 2 proteins control the expression of the other T3SS proteins of SPI-1. InvF also controls molecules that are secreted by the type III device that are encoded outside the pathogenicity island.1 Ultimately, contact with the host cell is required to trigger the secretion by the T3SS and invasion of the intestinal lumen.
Once inside the body, the Salmonella species are likely to be phagocytized by macrophages. The combination of low pH, low Mg+2, and low phosphate activates the ATR genes which again will provide protection against the acid in the lysosomal environment, but genes within a second pathogenicity island, SPI-2, will also be "turned-on."14 The SPI-2 contains approximately 40 genes encoding a different T3SS and the proteins that it secretes. Salmonella pathogenicity island 2 is important for intracellular growth in phagocytes.15
Salmonella virulence genes are also responsive to the metabolic and growth state of the bacteria through global regulators. These include alternative σ factors produced in response to cell stress, starvation, iron levels, or the stationary phase of growth.
Yersinia also up-regulate virulence factor expression upon internalization. Outside the body, at 26°C, Yersinia are more motile, but at 37°C and at pH 6, the bacteria turn-on their invasion machinery.16-19 These environmental cues activate production of a T3SS, and the bacteria set up their home by invading the intestinal lining. As for the Salmonella T3SS, cell contact activates the secretion device.
Enterohemorrhagic E. coli, including E. coli O157:H7, turn-on its pathogenicity island for the type III secretion device in response to the presence of other gut bacteria producing an aromatic compound (autoinducer 3) or even epinephrine within the gut.8 These compounds bind to a sensor protein which activates a protein kinase cascade, which in turn activates the transcription of the genes for the type III secretion device and a Shiga-like toxin.
As the settlers continued to expand their colonies, their demands on the environment became greater and greater. More land was cleared for the growing population, and toxic pollutants were released to the detriment of the land and waterways. Similarly, as the bacterial colony grows, the toxins and enzymes produced as virulence factors and the acids and gases produced as waste products disrupt tissues and prevent their proper functions and result in disease. Fortunately, the increased numbers increase the visibility of the colony and attract responses from the host innate and immune responses. Unfortunately, this is oftentimes too late to be effective.
Just as the virulence factors are bacterial weapons, they are also a potential Achilles heel. The exquisite control mechanisms that regulate the expression of the aforementioned virulence factors can be disrupted with drugs designed to block the action of the small molecules, proteins, and RNA which mediate the regulation. For example, antagonists of the quorum or environmental sensors might disrupt the signals that promote production of biofilms, regulate toxin production, or survival within a phagosome decreasing the virulence of the microbe and promoting its elimination. Future therapies may be able to evict bacterial settlers by preventing their ability to sense and respond to their human environment.
1. Eichelberg K, Galan J. Differential regulation of Salmonella typhimurium
type III secreted proteins by pathogenicity island I (SPI-1)-encoded transcriptional activators InvF and HilA. Infect Immun
2. Magnusson JU, Farewell A, Nystrom T. ppGpp: a global regulator in Escherichia coli
. Trends Microbiol
3. Boor KJ. Bacterial stress responses: what doesn't kill them can make them stronger. PLoS Biol
4. Kazmierczak MJ, Wiedmann M, Boor KJ. Alternative sigma factors and their roses in Bacterial virulence. Microbiol Mol Biol Rev
5. George EA, Muir TW. Molecular mechanisms of agr quorum sensing in virulent staphylococci. Chem Biochem
6. Bronner S, Monteil H, Prevost G. Regulation of virulence determinants in Staphylococcus aureus
: complexity and applications. FEMS Microbiol Rev
7. Sakoulas G, Moellering RC Jr, Eliopoulos GM. Adaptation of methicillin-resistant Staphylococcus aureus
in the face of vancomycin therapy. Clin Infect Dis
8. Reading NC, Sperandio V. Quorum sensing: the many languages of bacteria. FEMS Microbiol Lett
9. Cotter PA, DiRita VJ. Bacterial virulence gene regulation: an evolutionary perspective. Annu Rev Microbiol
10. Bower S, Rosenthal KS. Bacterial cell walls: the armor, artillery and Achilles heel. Infect Dis Clin Pract
11. Eriksson S, Lucchini S, Thompson A, et al. Unraveling the biology of macrophage infection by gene expression profiling of intracellular Salmonella enterica
. Mol Microbiol
12. Rhen M, Dorman CJ. Hierarchical gene regulators adapt Salmonella enterica
to its host milieus. Int J Med Microbiol
13. Suarez M, Russmann H. Molecular mechanisms of Salmonella
invasion: the type III secretion system of the pathogenicity island 1. Int Microbiol
14. Deiwick J, Nikolaus T, Erdogan S, et al. Environmental regulation of Salmonella
pathogenicity island 2 gene expression. Mol Microbiol
15. Garcia-del Portilla F. Salmonella intracellular proliferation: where, when and how? Microbes and Infection
16. Straley SC, Perry RD. Environmental modulation of the gene expression and pathogenesis in Yersinia
. Trend Microbiol
17. Cornelis GR. Yersinia
pathogenicity factors. In: Hormaeche CE, Penn CW, Smyth CJ, et al, eds. Molecular Biology of Bacterial Infection: Current Status and Future Perspectives
. Cambridge, MA: Cambridge University Press; 1992.
18. Stojiljkovic I, Hantke K. Transport of haemin across the cytoplasmic membrane through a haemin-specific periplasmic binding-protein-dependent transport system in Yersinia enterocolitica
. Mol Microbiol
19. McDonough KA, Barnes AM, Quan TJ, et al. Mutation in the pla
gene of Yersinia pestis
alters the course of the plague bacillus-flea (Siphonaptera: Ceratophyllidae) interaction. J Med Entomol