Article Text


Molecular basis of virulence
  1. Division of Infection and Immunity
  2. IBLS, Joseph Black Building
  3. University of Glasgow
  4. Glasgow G12 8QQ
    1. ADAM FINN
    1. Sheffield Institute for Vaccine Studies
    2. Department of Paediatrics
    3. Sheffield Children’s Hospital
    4. Western Bank
    5. Sheffield S10 2TH

      Statistics from

      Advances in molecular biology, cell biology, and several other areas of science have changed the way we understand the mechanisms in which microbial pathogens interact with their hosts. This trend is set to continue with the advent of microbial genome sequencing, in vivo gene expression analysis, and other related techniques. The availability of these techniques and advances in other areas such as protein expression and crystallography has allowed the understanding of host pathogen interaction at the molecular and even atomic level. However, despite these powerful approaches the basic concept advanced many years ago by Smith that pathogenicity or virulence is a multifactorial property that consists of five basic steps is still valid today.1 The molecular basis of virulence can still be considered under these five headings: (1) attachment to the host (via mucous membranes); (2) entry into the host (usually); (3) multiplication within the host; (4) Interference with host defence systems; (5) damage to the host.

      These five stages are not mutually exclusive. The factors produced by pathogens that mediate these steps are termed the determinants of microbial pathogenicity. The molecular basis of these steps will be considered and specific examples will be given to demonstrate the basic principles. Finally it should be emphasised that expression of determinants of pathogenicity is usually regulated and systems exist for environmental sensing and quorum sensing to allow appropriate expression of virulence factors.

      (1) Attachment to the host

      Attachment of bacteria to host cells is mediated by adhesins, which have been identified for many bacterial species.2-5Most adhesins are proteins which usually bind to carbohydrate receptors on the host cell surface. Perhaps the most studied adhesins are the fimbrial adhesins of Escherichia coli. These can be divided into two families. The K88, K99, CFA/I, and CFA/II adhesins mediate attachment to gut epithelium while type 1, P, and S fimbriae are associated with infections of the urinary tract. Fimbriae tend to be composed of a helical structure of major protein subunits which act as a support for the minor protein subunits. The minor protein subunits determine the receptor specificity for type 1, P, and S fimbriae whereas receptor specificity is determined by the major protein subunits in the attachment to gut epithelium. Fimbriae genes have been cloned and sequenced and the individual amino acids responsible for the interaction with the host cell receptors have in some cases been determined, for example lysine and arginine residues at positions 116 and 118 appear to be important for the Sfa minor subunit of S fimbriae.6 Host cell receptors for attachment of bacteria have also been characterised at the molecular level, for example β-D-Gal groups are the ligand for K88 fimbriae. It should be pointed out that interaction between a pathogen and host is not mediated by a single adhesin/receptor interaction. A good example of this fact is the interaction of the respiratory pathogen Streptococcus pneumoniae (the pneumococcus) with host cells. At least four types of receptor have been proposed for the pneumococcus and the interaction with these receptors depends on the phenotype of the organism.7 Pneumococci present at least three interchangeable variants which can be distinguished by their colony morphology into opaque, semitransparent, and transparent. The molecular basis of the phase variation is unknown but these phenotypes differ in their ability to colonise the nasopharynx.8 Nasopharyngeal cells bear a receptor, GlcNAcβ1-3Gal, which is recognised by transparent phase variants. Lung cells in the resting state bear two types of receptor, GalNcβ1-4Gal and GalNcβ1-3Gal, which are recognised by both opaque and transparent types. Cytokine activated lung cells also express the resting receptors but also present platelet activating factor (PAF) receptors. Transparent phase variants are able to adhere to PAF receptors whereas opaque variants are not. Thus, adhesion to epithelial cells should be considered as a dynamic situation in which both bacteria and host cell receptors differ according to site of isolation or activation state.

      (2) Entry into the host

      Entry involves either direct invasion of the epithelial cells of a mucus membrane or passage between them followed by invasion of the deeper tissues. Not all pathogens invade; Vibrio choleraecauses disease by toxin production from within the intestinal lumen.

      Invasion of epithelial cells by the gut pathogen Shigella flexneri has been investigated in some detailed and elegant studies.9 Entry of shigella into cells is triggered by three gene products IpaB (invasion plasmid antigen B), IpaC and IpaD. IpaB is a haemolysin responsible for the release of shigella from the vacuole into the cytoplasm where the organisms then move by polymerisation of host actin under the influence of the intracellular spread gene icsA. This motility results in the generation of protrusions from one cell into the next. The protrusion is ‘clipped-off’ to form a vacuole within the next cell. The double membranes of this vacuole are then lysed by the product of the icsB gene, the organisms are released into the cytoplasm and the cycle begins again. A note of caution is warranted here with regard to the use of tissue culture cells. Although the mechanism described above applies to HeLa cells, shigella cannot invade across a brush border. Further studies using animal models indicated that invasion was more complex. Shigella enter the colonic mucosa through M-cells and then infect macrophages causing apoptosis programmed cell death). Apoptosis is mediated by the product of the ipaB gene. Inflammation resulting from the release of cytokines results in infiltration of phagocytes which damage the basal membrane and disrupt the epithelium. The brush border is disrupted and shigella invade the cells. This example highlights the need to use several approaches to understand the events occurring during pathogenesis.

      (3) Multiplication in the host

      Once in its specific niche the pathogen must multiply. The ability to multiply is a characteristic of all living organisms and the success of a pathogen depends on the degree to which it can multiply upon reaching its specific niche and secure its potential transmission to a new host. The speed of multiplication will also affect the type of disease caused. Rapid multiplication leads to acute disease whereas slow multiplication may be advantageous in chronic disease. Little is known of the molecular basis of multiplication within the host. The environmental factors and nutrients that determine growth rate in tissues remain largely unknown. Most detailed studies on factors controlling growth in vivo have concerned molecular mechanisms for overcoming iron restriction. Pathogens attempt to obtain iron by one of several mechanisms including (a) from haemoglobin as hemin or heme, for example S pneumoniae,10 (b) directly from ferrated transferrin or lactoferrin, for example Neisseria gonorrhoeae,11 (c) indirectly from iron binding proteins by the production of siderophores, for example E coli,12 (d) from intracellular iron stores, for example Mycobacterium tuberculosis.13

      (4) Interference with host defence systems

      To survive within the host the pathogen must either prevent the immune response or circumvent its action. There are many virulence factors associated with interference of host defence, including polysaccharide capsules, protein toxins and lipopolysaccharides.

      Polysaccharide capsules interfere with the processes of phagocytosis and complement mediated bacterial killing. The importance of the capsule in the virulence of some organisms has been demonstrated at the molecular level. A non-capsular mutant of the pneumococcus generated by transposon mutagenesis was one million times less virulent than its capsular parent.14 The molecular mechanisms of activity of capsular polysaccharide still remain to be defined. It is known that sialyl groups on capsular polysaccharides of E coli K1 and group B streptococci may prevent the activation of the complement pathway by these organisms.4 15 Antibodies to capsular polysaccharide usually confer protective immunity through opsonisation of bacteria promoting phagocytosis. Encapsulated pathogens are still major causes of human disease due to the production of diverse capsular types (for example pneumococcus) or to the capsule being non-immunogenic (for example E coli K1 and Neisseria meningitidis group B). The non-immunogenicity of these organisms may be due to the molecular structure of the polysaccharide subunits being similar to sugars found on host cells.

      Pathogens may also produce toxins that interfere with the immune response. The pneumococcus, for example, produces pneumolysin.16 At high concentrations this membrane damaging toxin lyses all eukaryotic cells. At sublytic concentrations it has a range of effects on the cells and soluble molecules of the immune system. Pneumolysin inhibits the respiratory burst of phagocytes and also inhibits random migration and chemotaxis by these cells.17 The toxin inhibits antibody production by lymphocytes and blocks mitogen induced proliferation of B-cells. In vivo the toxin induces a large inflammatory response and this may be due to its ability to stimulate the production of inflammatory cytokines (interleukin-1β and tumour necrosis factor-α )18 and to activate the classical complement pathway. Activation of the classical complement pathway is due, at least in part, to the ability of the toxin to bind to the Fc portion of IgG.19 Molecular analysis has allowed the regions of the toxin responsible for these activities to be identified and modified. The role of the activities of the toxin have been investigated in the context of the whole bacterium20 by using gene replacement techniques to construct versions of the pneumococcus expressing altered versions of the toxin.21 These studies demonstrate the power of molecular analysis when used in combination with other techniques such as animal models in allowing the role of individual proteins, protein domains or even amino acids in the pathogenic process to be investigated.

      Studies on the molecular basis of the activity of lipopolysaccharides (endotoxin) are not as advanced as studies with protein virulence factors. It is known that long O side chains are required for serum resistance in E coli and salmonella.

      (5) Damage to the host

      Damage to the host can be mediated either directly by production of toxins or indirectly by the induction of gross inflammation and immunopathologic reactions.

      There is a wealth of information in the literature concerning the molecular action of bacterial protein toxins (reviewed in Alouf and Freer22 ). In some cases the contribution of the toxin to disease process is obvious (for example in cholera or tetanus) and the action of the toxins concerned is understood at the molecular level. Cholera toxin, for example, is an example of an ADP ribosylating toxin. The enzymatic action of the toxin results in the ADP ribosylation of a regulatory G protein of the adenylate cyclase complex in enterocytes. This results in increased levels of cAMP which in turn alters ion transport across the epithelium and leads to the diarrhoea which is the key feature of the disease. The heat labile toxin of E coli works in a similar manner to cholera toxin. Other examples of ADP ribosylating toxins include pertussis toxin, diphtheria toxin and toxin A of Pseudomonas aeruginosa. The effects of these toxins vary according to the cellular target of ADP ribosylation. Diphtheria and toxin A inhibit protein synthesis while pertussis toxin uncouples signal transduction. Other toxins are proteases. Tetanus toxin is a zinc protease that cleaves synaptobrevin, a protein involved in neurotransmitter release. Another large group of toxins are known as the membrane damaging toxins. These include pore forming proteins such as Staphylococcus aureus alpha toxin and thiol-activated toxins including pneumolysin. The molecular mechanisms of action of some activities of these toxins have been elucidated, but others and contribution to disease of the process of pore formation is still unclear.

      Stimulation of inflammation may occur due to the inappropriate or excessive production of cytokines or activation of the complement pathway. Inappropriate cytokine production or activation of the complement cascade may be triggered by bacterial toxins such as pneumolysin. Bacterial cell wall components also stimulate cytokine production. Stimulation of inflammatory cytokine production by lipopolysaccharide of Gram negative bacterial cell walls mediates endotoxic shock. Release of inflammatory cell wall fragments from the pneumococcus during autolysis mediates inflammation and data suggest that this inflammation is the major contributing factor to pathology in pneumococcal meningitis. Treatment of pneumococcal meningitis with cell wall active antibiotics such as penicillin may have the short term effect of promoting inflammation and it has been suggested that these antibiotics should be used in conjunction with anti-inflammatory agents.23 The molecular mechanisms of induction of cytokines by lipopolysaccharide and Gram positive cell walls are now beginning to be understood at the molecular level and have been reviewed.7 24

      Immunopathologic reactions may also lead to damage to the host. These reactions occur due to bacterial antigens including the production of antibodies that are cross reactive to human structures as is seen for example in endocarditis after infection with group A streptococci. Glomerulonephritis can also occur after streptococcal infection and is due to the deposition of immune complexes in the kidney. The molecular determinants of some of these reactions have been established. The M-protein of streptococci, for example, has been shown to share epitopes with antigens expressed in heart tissue.25Ankylosing spondylitis may be caused by antibodies to klebsiella reacting with antigens expressed by lymphocytes from individuals with the HLA B27.26

      Regulation of virulence

      When it encounters a host, a pathogen must adapt to changing environments and express appropriate virulence factors. Virulence genes may be regulated in response to a range of environmental stimuli including pH, temperature, oxygen tension, and inorganic metal ion concentration. Knowledge about the molecular basis of virulence gene regulation is rapidly increasing. The commonest mechanism involves two component regulatory systems in which one component (the sensor) detects the environmental stimulus while the other (the reponder, usually a DNA binding protein) is responsible for altering gene expression. The sensor protein is usually a membrane spanning kinase which autophosphorylates on stimulation. The phosphate is then transferred to the response regulator proteins which then affect gene expression. Examples of this type of system are regulation of permeability in E coli in response to osmotic stimuli (EnvZ/OmpR) and regulation of motor control in E colichemotaxis (CheA/CheY, CheB). These types of system, which have been comprehensively reviewed,27 have common features but differ in their exact mechanisms.

      Bacterial pathogens produce a range of molecules that allow them to cause disease via the five stages described above. It should be emphasised that these five stages are not discrete steps and the interaction of the pathogen with its host is dynamic. The pathogen continually monitors its environment and produces virulence factors according to the signals it receives. There is also a host contribution to the process such that genetic differences between individuals will make one host environment different to another and affect the interactions and signals that occur between the pathogen and its environment. An understanding of the events that occur at the molecular level both in the action of individual virulence factors and in the coordinate regulation of virulence as a whole is a continuing aim that will be aided by the new trend of total genome sequencing of pathogens. An understanding of the molecular events involved in the disease process will allow us to generate new weapons to use in the continuing battle against infectious diseases.



      The issues discussed by Professor Mitchell raise some important questions about the characteristics of the many human diseases that occur as a consequence of infection.

      Micro-organisms have evolved alongside humans and, indeed, are able to evolve much more rapidly due to their short generation times. The nature of their relationships with us vary from, at one extreme, imperceptible parasitism (for example commensal colonisation of the skin and mouth) or even full blown symbiosis (for example the mitochondria in all our cells—thought to derive from exogenous microbes) to overwhelming illness and death at the other (for example septicaemia due to Neisseria meningitidis or infection with HIV), with examples of infections of every conceivable site, rate, and severity in between.

      Upon first consideration, it seems surprising that “we” and “they”, driven by the same fundamental aims to survive and multiply, should have thrown up such a myriad of diverse relationships. However, the existence of so many possible distinct mechanisms of transmission and sites of colonisation explains the large number of distinct niches that exist for micro-organisms.

      More puzzling is why so many infections are virulent (that is cause disease) at all. Sometimes it is fairly obvious how induction of certain symptoms of disease contributes to the survival and spread of invading micro-organisms (for example diarrhoea in cholera and sneezing in rhinovirus infection). However, a “purpose” for disease features is often less clear. For example it is hard to see how the survival ofStreptococcus pneumoniae is enhanced by causing countless episodes of acute ear pain in preschool children and occasional episodes of meningitis. Some well known infectious diseases (for example tetanus and human infections with Pseudomonas spp) would appear to be the results of extraordinary and largely irrelevant excursions into the human for microbes whose modus vivendiis inhabiting the soil and stagnant water.

      Of course, as doctors, we tend to focus our attention on disease-producing micro-organisms rather than the many others which have taken an evolutionary stance not unlike that of sheep—achieving enormous success in terms of sheer numbers by being useful, keeping quiet, not causing problems, and putting up with rough treatment at times. Most of us have learned to adopt what could be termed the “microbe centred” view of infectious diseases, namely that these are examples of “invaders” penetrating and colonising the human host and doing damage along the way. As Mitchell’s article points out, the manifestations of disease are as often as not due to the host response, or over-response, to the situation.

      Clearly eukaryotic organisms evolved a fundamental principle of survival—manifest as the immune response—very early on, namely: “if it’s inside me but isn’t me, kill it before it kills me”. This process remains the central dogma of our immunity because, despite important exceptions such as mitochondria and long term colonisation with herpes viruses and mycobacterium tuberculosis which (usually) don’t do any harm, owners of immune systems which give invading microbes the benefit of the doubt don’t survive well.

      Our immune responses set clear boundaries to the hoards of microbes that surround us (outside and inside). Our secreted protein and phagocyte defences endeavour to inflict instant death on the invader and our specific immunity acts as a slower but highly organised search and destroy machinery for viruses or larger microbes which succeed in penetrating our initial defences.

      Thus, infectious diseases, when they occur, are the result of the interaction between microbe and host. In other words, infection and immunity are two sides of the same coin and cannot be studied legitimately in isolation from each other.

      Experience with vaccines has already shown the potential power of inducing host responses in the prevention of infectious diseases. As knowledge of the molecular details of our interactions with micro-organisms continues to grow, this will certainly also lead to new interventions designed specifically to treat disease by modifying the host response or by interfering with microbe-host interactions, as well as to additional conventional agents designed to inhibit or kill the microbe.

      View Abstract

      Request permissions

      If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.