Harrison, A. McNally, A. Diggle, unpublished data. Because we included likelihood of pleiotropy as a moderator variable in our analysis, we should be able to partition out between-study variance caused by the use of mutants that risk a bias toward positive results. If there is bias, we could find an overrepresentation of noisier experiments reporting higher magnitude results. We searched the literature for papers featuring infections of whole live host organisms with P.
Experiments further differed in the way infections were established and in the organs targeted. Each experiment compared infections with a control P. Twenty-eight experiments used mutant strains with clean deletions or transposon Tn5 insertions in genes encoding the pyoverdine biosynthesis pathway.
In these cases, pleiotropic effects are expected to be relatively low—i. The other 53 experiments used mutants where pleiotropic effects were likely or even certain. For example, some mutant strains carried mutations in pvdS , which encodes the main regulator of pyoverdine synthesis that also regulates the production of toxins and proteases Ochsner et al. Others carried mutations in pvdQ , encoding an enzyme known to degrade quorum-sensing molecules in addition to its role in pyoverdine biosynthesis Nadal Jimenez et al.
We combined data from the set of experiments described above in a meta-analysis to determine the extent to which pyoverdine's effect on virulence varied across four moderator variables: i host taxa, ii tissue types, iii pathogen wildtype background, and iv pyoverdine-mutation type. To obtain a comparable measure of virulence across experiments, we extracted in each instance the number of cases where a given infection type did or did not have a virulent outcome i.
We then took as our effect size the log-odds-ratio, i. Consistent with the theoretical prediction that host-pathogen interactions and host ecology are important modulators of virulence, we found considerable variation in the effect sizes across experiments and subgroups of all moderators Figure 1. Pyoverdine-deficient mutants showed substantially reduced virulence in invertebrate and mammalian hosts, whereas there was little evidence for such an effect in plants Figure 1A.
Overall, evidence for pyoverdine being an important virulence factor was weak for taxa with a low number of experiments i. We found that pyoverdine-deficient mutants exhibited reduced virulence in all organs and tissues tested, with the exception of plants Figure 1B. Comparing the effect sizes across wildtype strain backgrounds, we see that pyoverdine deficiency reduced virulence in experiments featuring the well-characterized PA14 and PAO1 strains Figure 1C whereas the reduction was less pronounced in experiments with less well-characterized wildtype strains.
This could be due to sampling error only a few experiments used these strains or it may be that these strains really behave differently from PA14 and PAO1. Finally, we observed that the nature of the pyoverdine-deficiency mutation matters Figure 1D. Infections with strains carrying well-defined mutations known to exclusively or at least primarily affect pyoverdine production showed a relatively consistent reduction in virulence. Conversely, where mutants were poorly-defined, or carried mutations likely to affect other traits beyond pyoverdine, here the virulence pattern was much more variable, with both reduced and increased virulence relative to wildtype infections Figure 1D.
We posit that at least some of the differences in observed virulence between these mutants and their wildtype counterparts was likely due to pleiotropic differences in phenotypes unrelated to pyoverdine.
Much of the variation we observe is probably due to other factors beyond those explored in Figure 1. The issue is that a we do not know what all these additional factors might be, and b the probably patchy distribution of experiments across the levels and ranges of these other factors would leave us with limited power to test for their effects. This leaves us with a core dataset comprising only those experiments where animal host models were infected with strains from well-defined PA14 or PA01 wildtype background, and survival vs.
Using this restricted dataset, we performed a series of meta-regression models to test for significant differences between subgroups of our moderator factors, and we also estimated the share of total variance in effect sizes that is explained by each moderator variable Figure 2.
These models revealed that infection type is the variable that explains the largest share of total variance For instance, in systemic infection models the pyoverdine-defective mutants showed strongly reduced virulence compared to the wild-type, whereas this difference was less pronounced in gut infections.
Host taxon explained only 8. This was interesting, because we predicted a priori that mutations with pleiotropic effects on other virulence factors could introduce within-study bias toward a greater effect of siderophore loss on virulence. Figure 2. Test for differences between subgroups of moderator variables with regard to the effect sizes for pyoverdine as a virulence factor in P.
Our baseline condition for all comparisons is the following: gut infections in invertebrate hosts, using the P. The effect size for this baseline scenario is set to zero. All other scenarios had more extreme negative effect sizes, and are therefore scaled relative to this baseline condition. Comparisons reveal that virulence in pyoverdine-deficient strains was significantly more reduced in systemic compared to gut infections, and that most effect size variation is explained by the infection type.
There were no significant effect size differences between any of the other subgroups. Values given in brackets indicate percentage of effect size heterogeneity explained by a specific moderator. In any field, there is a risk that studies with negative or unanticipated results may be less likely to get published e. Especially when negative or unanticipated results are obtained from experiments featuring low sample sizes and thus high uncertainty , the scientists responsible may be less inclined to trust their results, and consequently opt not to publish them.
This pattern could result in a publication bias, and an overestimation of the effect size. To test whether such a publication bias exists in our dataset, we plotted the effect size of each experiment against its inverted standard error Figure 3. If there is no publication bias, we would expect to see an inverted funnel, with effect sizes more or less evenly distributed around the mean effect size, irrespective of the uncertainty associated with each estimate i.
Instead, we observed a bias in our dataset, with many lower-certainty experiments that show strongly negative effect sizes i. Figure 3. Association between effect sizes and their standard errors across 81 experiments examining the role of pyoverdine production for virulence in P.
In the absence of bias, we should see an inverted funnel-shaped cloud of points, more or less symmetrically distributed around the mean effect size vertical dotted line. Instead, we see an over-representation of low-certainty experiments associated with strong negative effect sizes.
This suggests a significant publication bias: experiments with low-certainty and weak or contrary effects presumably do exist, but are under-represented here note the absence of data points in the cross-hatched triangle.
Effect sizes are given as log-odds-ratio. Each symbol represents a single experiment. Symbol colors and shapes stand for different host organisms red circles, invertebrates; blue squares, mammals; green diamonds, plants. Large symbols denote the experiments included in the core dataset. Note that due to the stronger weights accorded to high certainty experiments i. Our meta-analysis reveals that pyoverdine-deficient strains of the opportunistic pathogen P.
This confirms that iron limitation is a unifying characteristic of the host environment, making siderophores an important factor for pathogen establishment and growth within the host Parrow et al. However, we also saw that the extent to which pyoverdine deficiency reduced virulence varied considerably, and was quite modest in many instances. Nonetheless, these mutants were typically still able to establish a successful infection, and, in many cases, could still kill their host Romanowski et al.
These results support ecological theory predicting that the effect of a certain phenotype i. Our findings have direct consequences for any therapeutic approaches targeting this particular virulence factor. Because pyoverdine seems to be generally involved with virulence, treatments inhibiting pyoverdine production could have wide applicability and be effective against different types of infections across a wide host range.
However, given the variation observed and pyoverdine's generally modest effect on virulence, the clinical impact of such treatments would likely vary across infection contexts, and be limited to attenuating rather than curing the infection. This would mean that for P. Certainly promising is that pyoverdine seems to have a more consistent Figure 1A and more prominent Figure 2 although not significant effect in mammalian compared to invertebrate hosts.
From this observation, one could infer that pyoverdine may have potential as a target for infection control in humans. Our work demonstrates how meta-analyses can be used to quantitatively synthesize data from different experiments carried out at different times by different researchers using different designs. Such an analytical approach goes beyond a classical review, where patterns are typically summarized in a qualitative manner. For instance, a recent study proposed that three different virulence factors pyocyanin, protease, swarming of P.
Here we use a meta-analytic approach to quantitatively derive estimates of the overall virulence potential of a given bacterial trait and investigate variables that affect infection outcomes. We assert that such quantitative comparisons are essential to identify those virulence factors that hold greatest promise as targets for effective broad-spectrum anti-virulence therapies.
Our finding that effect sizes vary considerably across our assembled experiments provides a different perspective compared to that which one would obtain from a cursory reading of the literature. For instance, the first study investigating pyoverdine in the context of an experimental infection model Meyer et al. Although this experiment and its message have been widely cited including by ourselves , it may no longer be the strongest representative of the accumulated body of research on this topic.
As we see in Figure 1 , the effect size it reports is associated with a high uncertainty due to a comparatively low sample size. Moreover, the observed effect cannot unambiguously be attributed to pyoverdine because an undefined UV-mutagenized mutant was used. We highlight this example not to criticize it, but rather because it serves to demonstrate why drawing inferences from appropriately weighted aggregations of all available evidence is preferable to focusing solely on the results of a single study.
Our meta-analytic approach not only provides information on the overall importance of pyoverdine for P. For example, let us consider which types of studies were conspicuously absent from our dataset.
First, most experiments in our dataset employed acute infection models, even though P. This raises the question to what extent insights on the roles of virulence factors important in acute infections can be transferred to chronic infections. In the case of pyoverdine, we know that in chronically-infected cystic fibrosis airways, pyoverdine production is often selected against Wiehlmann et al.
Second, our comparative work shows that experiments were predominantly carried out with the well-characterized strains PAO1 and PA While these strains were initially isolated from clinical settings, they have subsequently undergone evolution in the laboratory environment Bragonzi et al. Therefore, while we found no overall differences between the lab strains used in our data set, we argue that it would still be useful to carry out additional studies on a range of clinical isolates to be able to make firm conclusions on the general role of pyoverdine as a virulence factor.
Finally, our data analysis revealed that low-certainty studies showing no or small effects of pyoverdine on virulence were under-represented in our data set, which points toward a systematic publication bias. It remains to be seen whether such biases are common with regard to research on virulence factors, and whether they result in a general overestimation of the effect these factors have on host survival or tissue damage.
With regard to pyoverdine, further studies are clearly needed to obtain a more accurate estimate of the true effect size. In addition to these issues of data availability, all meta-analyses unavoidably involve intrinsic assumptions and subjective decisions that can further influence the resulting outputs. For instance, although we have used the standard log-odds-ratio as our common metric, related metrics like risk ratios, while typically highly correlated, can sometimes produce different patterns in a meta-analysis.
In the present case, however, using risk ratios instead does not qualitatively affect the patterns we observed nor alter our conclusions. Furthermore, to facilitate calculation of a log-odds ratio in cases where zero counts appear as denominators, we have in this study adopted the common, yet ultimately arbitrary, convention of replacing these zero denominators with 0.
Had we used a different value as a substitute, say 0. A more fundamental issue is that when estimating population-level statistics across a collection of experiments, we typically assume that we are comparing like with like. Here, we have intentionally brought together a very diverse set of studies, and even though we have translated their individual effect sizes into a common metric and also stratified them by some of their major defining characteristics, their individual effect sizes nonetheless remain highly heterogeneous.
In effect, we are knowingly combining apples and pears, because we think that the resulting fruit salad is still something that is worth taking a look at. Alternative or additional ways of slicing up the data could yield models with lower residual heterogeneity, but then poor data coverage in combinations of subcategories could limit the accuracy of any parameter estimates we want to extract from such models.
In light of these issues, we advise readers to focus on the overall patterns our models reveal, rather than the specific values of the estimates they generate. While our study demonstrates the strength of quantitative comparative approaches, it is important to realize that extracting effect sizes is one of the biggest challenges in any meta-analysis.
This challenge was particularly evident for the experiments we found, which profoundly varied in the way data was collected and reported. As a consequence, we had to exclude many studies because they used measures of virulence that were only reported by a minority of studies, or because their reporting of results was unclear for a selected list of examples, see Table S3 in the Supplemental Material.
To amend this issue for future studies, we would like to first highlight the problems we encountered and then provide general guidelines of how data reporting could be improved and standardized.
One main problem we experienced was incomplete data reporting i. Another important issue was that different studies measured virulence using very different metrics. Some measured virulence at the tissue level i. Others focused on the dynamics of the bacteria themselves, taking this as a proxy for the eventual damage to the host.
There were both quantitative measures e. Survival data was sometimes presented as a timecourse, sometimes as an endpoint; sometimes as raw counts, sometimes as proportions. In most cases, the time scales over which survival was assessed were fairly arbitrary. Compiling such diverse measures of virulence is not simply time consuming, but it also generates extra sources of heterogeneity in the dataset, which might interfere with the basic assumptions of meta-analytical models Borenstein et al.
How can these problems be prevented in future studies? We propose the following. Finally, d studies leading to unexpected or negative results e. This will make comparisons across studies much easier and will provide a useful resource for future meta-analytic studies.
Currently, bacterial traits are subject to a binary categorisation whereby some are labeled as virulence factors while others are not. We demonstrate that traits' effects on virulence are anything but binary. Rather, they strongly depend on the infection context. Our study affirms meta-analysis as a powerful tool to quantitatively estimate the overall effect of a specific virulence factor and to compare its general importance in infections across different bacterial strains, hosts, and host organs.
Such quantitative comparisons provide us with a more complete picture on the relative importance of specific virulence factors. Such knowledge is especially valuable for opportunistic pathogens, which have a wide range of virulence factors at their disposal, and infect a broad range of host organisms Kurz et al.
Meta-analytical comparisons could thus inform us on which traits would be best suited as targets for anti-virulence therapies. Ideal traits would be those with high effect sizes and general importance across pathogen and host organisms. University of Warwick FH. The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Andersen, S. Long-term social dynamics drive loss of function in pathogenic bacteria. Attila, C. Pseudomonas aeruginosa PAO1 virulence factors and poplar tree response in the rhizosphere.
Banin, E. The potential of desferrioxamine-gallium as an anti-pseudomonas therapeutic agent. Beare, P.
Siderophore-mediated cell signalling in Pseudomonas aeruginosa : divergent pathways regulate virulence factor production and siderophore receptor synthesis.
Becker, K. Metal limitation and toxicity at the interface between host and pathogen. FEMS Microbiol. Begg, C. Operating characteristics of a rank correlation test for publication bias.
Biometrics 50, — Borenstein, M. Introduction to Meta-Analysis. West Sussex: John Wiley. A basic introduction to fixed-effect and random-effects models for meta-analysis. Methods 1, 97— Bragonzi, A. Pseudomonas aeruginosa microevolution during cystic fibrosis lung infection establishes clones with adapted virulence. Care Med. Braud, A. The Pseudomonas aeruginosa pyochelin-iron uptake pathway and its metal specificity.
Calderone, R. Virulence factors of Candida albicans. Trends Microbiol. Carter, M. A subtype of a Pseudomonas aeruginosa cystic fibrosis epidemic strain exhibits enhanced virulence in a murine model of acute respiratory infection. Clatworthy, A.
Targeting virulence: a new paradigm for antimicrobial therapy. Cox, D. The continuity correction. Biometrika 57, — Dubern, J. Integrated whole genome screening for Pseudomonas aeruginosa virulence genes using multiple disease models reveals that pathogenicity is host specific.
Dwan, K. Systematic review of the empirical evidence of study publication bias and outcome reporting bias. Egger, M. Bias in meta-analysis detected by a simple, graphical test. BMJ — Feinbaum, R. Genome-wide identification of Pseudomonas aeruginosa virulence-related genes using a Caenorhabditis elegans infection model. PLoS Pathog. Frydenlund Michelsen, C. Evolution of metabolic divergence in Pseudomonas aeruginosa during long-term infection facilitates a proto-cooperative interspecies interaction.
ISME J. Gallagher, L. Gaynes, R. Overview of nosocomial infections caused by gram-negative bacilli. Harrison, F. Cooperation and virulence in acute Pseudomonas aeruginosa infections. BMC Biol. Hauser, A. Severe Infections Caused by Pseudomonas aeruginosa. New York, NY: Springer. Google Scholar. He, J. The broad host range pathogen Pseudomonas aeruginosa strain PA14 carries two pathogenicity islands harboring plant and animal virulence genes.
Higgins, J. Quantifying heterogeneity in a meta-analysis. Hirsch, E. Impact of multidrug-resistant Pseudomonas aeruginosa infection on patient outcomes.
Expert Rev. Outcomes Res. Hohnadel, D. Mapping of mutations affecting pyoverdine production in Pseudomonas aeruginosa. Imperi, F. Repurposing the antimycotic drug flucytosine for suppression of Pseudomonas aeruginosa pathogenicity. Jeukens, J. Comparative genomics of isolates of a Pseudomonas aeruginosa epidemic strain associated with chronic lung infections of cystic fibrosis patients. Nadal Jimenez, P. Role of PvdQ in Pseudomonas aeruginosa virulence under iron-limiting conditions.
Microbiology , 49— Jimenez, P. The multiple signaling systems regulating virulence in Pseudomonas aeruginosa. Jiricny, N. Loss of social behaviours in populations of Pseudomonas aeruginosa infecting lungs of patients with cystic fibrosis. Josenhans, C. The role of motility as a virulence factor in bacteria. Kaneko, Y. The transition metal gallium disrupts Pseudomonas aeruginosa iron metabolism and has antimicrobial and antibiofilm activity.
Kazmierczak, B. Cross-regulation of Pseudomonas motility systems: the intimate relationship between flagella, pili and virulence. Recall that an adhesin is a protein or glycoprotein found on the surface of a pathogen that attaches to receptors on the host cell.
Adhesins are found on bacterial, viral, fungal, and protozoan pathogens. One example of a bacterial adhesin is type 1 fimbrial adhesin , a molecule found on the tips of fimbriae of enterotoxigenic E.
Recall that fimbriae are hairlike protein bristles on the cell surface. Type 1 fimbrial adhesin allows the fimbriae of ETEC cells to attach to the mannose glycans expressed on intestinal epithelial cells. Table 1 lists common adhesins found in some of the pathogens we have discussed or will be seeing later in this chapter. There is no indication that the bacteria entered the blood through an injury.
Instead, it appears the portal of entry was the gastrointestinal route. Listeria monocytogenes , the facultative intracellular pathogen that causes listeriosis, is a common contaminant in ready-to-eat foods such as lunch meats and dairy products.
Once ingested, these bacteria invade intestinal epithelial cells and translocate to the liver, where they grow inside hepatic cells. Listeriosis is fatal in about one in five normal healthy people, and mortality rates are slightly higher in patients with pre-existing conditions that weaken the immune response.
A cluster of virulence genes encoded on a pathogenicity island is responsible for the pathogenicity of L. These genes are regulated by a transcriptional factor known as peptide chain release factor 1 PrfA. One of the genes regulated by PrfA is hyl , which encodes a toxin known as listeriolysin O LLO , which allows the bacterium to escape vacuoles upon entry into a host cell.
A second gene regulated by PrfA is actA, which encodes for a surface protein known as actin assembly-inducing protein ActA. ActA is expressed on the surface of Listeria and polymerizes host actin. He is now experiencing a stiff neck and hemiparesis weakness of one side of the body. Concerned that the infection is spreading, the physician decides to conduct additional tests to determine what is causing these new symptoms.
After exposure and adhesion, the next step in pathogenesis is invasion , which can involve enzymes and toxins. Many pathogens achieve invasion by entering the bloodstream, an effective means of dissemination because blood vessels pass close to every cell in the body.
The downside of this mechanism of dispersal is that the blood also includes numerous elements of the immune system. Various terms ending in —emia are used to describe the presence of pathogens in the bloodstream.
The presence of bacteria in blood is called bacteremia. Bacteremia involving pyogens pus-forming bacteria is called pyemia. When viruses are found in the blood, it is called viremia. The term toxemia describes the condition when toxins are found in the blood.
If bacteria are both present and multiplying in the blood, this condition is called septicemia. Figure 1. This patient has edema in the tissue of the right hand. Such swelling can occur when bacteria cause the release of pro-inflammatory molecules from immune cells and these molecules cause an increased permeability of blood vessels, allowing fluid to escape the bloodstream and enter tissue. Some bacteria can cause shock through the release of toxins virulence factors that can cause tissue damage and lead to low blood pressure.
Gram-negative bacteria are engulfed by immune system phagocytes, which then release tumor necrosis factor , a molecule involved in inflammation and fever. Tumor necrosis factor binds to blood capillaries to increase their permeability, allowing fluids to pass out of blood vessels and into tissues, causing swelling, or edema Figure 1. With high concentrations of tumor necrosis factor, the inflammatory reaction is severe and enough fluid is lost from the circulatory system that blood pressure decreases to dangerously low levels.
This can have dire consequences because the heart, lungs, and kidneys rely on normal blood pressure for proper function; thus, multi-organ failure, shock, and death can occur. Some pathogens produce extracellular enzymes, or exoenzyme s , that enable them to invade host cells and deeper tissues. Exoenzymes have a wide variety of targets. Some general classes of exoenzymes and associated pathogens are listed in Table 2. Each of these exoenzymes functions in the context of a particular tissue structure to facilitate invasion or support its own growth and defend against the immune system.
For example, hyaluronidase S, an enzyme produced by pathogens like Staphylococcus aureus , Streptococcus pyogenes , and Clostridium perfringens , degrades the glycoside hylauronan hyaluronic acid , which acts as an intercellular cement between adjacent cells in connective tissue Figure 2. This allows the pathogen to pass through the tissue layers at the portal of entry and disseminate elsewhere in the body Figure 2.
Figure 2. Pathogen-produced nucleases, such as DNAse produced by S. As bacterial and host cells die at the site of infection, they lyse and release their intracellular contents. The DNA chromosome is the largest of the intracellular molecules, and masses of extracellular DNA can trap bacteria and prevent their spread.
This strategy is also used by S. Enzymes that degrade the phospholipids of cell membranes are called phospholipases. Their actions are specific in regard to the type of phospholipids they act upon and where they enzymatically cleave the molecules.
The pathogen responsible for anthrax , B. When B. Phospholipases can also target the membrane that encloses the phagosome within phagocytic cells.
As described earlier in this chapter, this is the mechanism used by intracellular pathogens such as L. The role of phospholipases in bacterial virulence is not restricted to phagosomal escape.
Many pathogens produce phospholipases that act to degrade cell membranes and cause lysis of target cells. These phospholipases are involved in lysis of red blood cells, white blood cells, and tissue cells. Bacterial pathogens also produce various protein-digesting enzymes, or proteases. Proteases can be classified according to their substrate target e.
One example of a protease that contains a metal ion is the exoenzyme collagenase. Collagenase digests collagen, the dominant protein in connective tissue.
Collagen can be found in the extracellular matrix, especially near mucosal membranes, blood vessels, nerves, and in the layers of the skin. Similar to hyaluronidase, collagenase allows the pathogen to penetrate and spread through the host tissue by digesting this connective tissue protein. The collagenase produced by the gram-positive bacterium Clostridium perfringens , for example, allows the bacterium to make its way through the tissue layers and subsequently enter and multiply in the blood septicemia.
Once the host cells have died, the bacterium produces gas by fermenting the muscle carbohydrates. The widespread necrosis of tissue and accompanying gas are characteristic of the condition known as gas gangrene Figure 3. Figure 3. The illustration depicts a blood vessel with a single layer of endothelial cells surrounding the lumen and dense connective tissue shown in red surrounding the endothelial cell layer. Collagenase produced by C. In addition to exoenzymes, certain pathogens are able to produce toxin s , biological poisons that assist in their ability to invade and cause damage to tissues.
The ability of a pathogen to produce toxins to cause damage to host cells is called toxigenicity. Toxins can be categorized as endotoxins or exotoxins. The lipopolysaccharide LPS found on the outer membrane of gram-negative bacteria is called endotoxin Figure 4. During infection and disease, gram-negative bacterial pathogens release endotoxin either when the cell dies, resulting in the disintegration of the membrane, or when the bacterium undergoes binary fission.
The lipid component of endotoxin, lipid A , is responsible for the toxic properties of the LPS molecule. Lipid A is relatively conserved across different genera of gram-negative bacteria; therefore, the toxic properties of lipid A are similar regardless of the gram-negative pathogen.
If the concentration of endotoxin in the body is low, the inflammatory response may provide the host an effective defense against infection; on the other hand, high concentrations of endotoxin in the blood can cause an excessive inflammatory response, leading to a severe drop in blood pressure, multi-organ failure, and death.
Figure 4. Lipopolysaccharide is composed of lipid A, a core glycolipid, and an O-specific polysaccharide side chain. Lipid A is the toxic component that promotes inflammation and fever. A classic method of detecting endotoxin is by using the Limulus amebocyte lysate LAL test. The amebocytes will react to the presence of any endotoxin. This reaction can be observed either chromogenically color or by looking for coagulation clotting reaction to occur within the serum.
An alternative method that has been used is an enzyme-linked immunosorbent assay ELISA that uses antibodies to detect the presence of endotoxin. Unlike the toxic lipid A of endotoxin, exotoxin s are protein molecules that are produced by a wide variety of living pathogenic bacteria.
Although some gram-negative pathogens produce exotoxins, the majority are produced by gram-positive pathogens. Exotoxins differ from endotoxin in several other key characteristics, summarized in Table 4.
In contrast to endotoxin, which stimulates a general systemic inflammatory response when released, exotoxins are much more specific in their action and the cells they interact with. Each exotoxin targets specific receptors on specific cells and damages those cells through unique molecular mechanisms. As discussed earlier, endotoxin can stimulate a lethal inflammatory response at very high concentrations and has a measured LD 50 of 0.
By contrast, very small concentrations of exotoxins can be lethal. For example, botulinum toxin , which causes botulism , has an LD 50 of 0. The exotoxins can be grouped into three categories based on their target: intracellular targeting, membrane disrupting, and superantigens. Table 5 provides examples of well-characterized toxins within each of these three categories. The intracellular targeting toxin s comprise two components: A for activity and B for binding.
Thus, these types of toxins are known as A-B exotoxins Figure 5. The B component is responsible for the cellular specificity of the toxin and mediates the initial attachment of the toxin to specific cell surface receptors. Once the A-B toxin binds to the host cell, it is brought into the cell by endocytosis and entrapped in a vacuole. The A and B subunits separate as the vacuole acidifies. The A subunit then enters the cell cytoplasm and interferes with the specific internal cellular function that it targets.
Figure 5. Figure 6. The mechanism of the diphtheria toxin inhibiting protein synthesis. The A subunit inactivates elongation factor 2 by transferring an ADP-ribose. This stops protein elongation, inhibiting protein synthesis and killing the cell. Four unique examples of A-B toxins are the diphtheria, cholera, botulinum, and tetanus toxins.
The diphtheria toxin is produced by the gram-positive bacterium Corynebacterium diphtheriae , the causative agent of nasopharyngeal and cutaneous diphtheria. After the A subunit of the diphtheria toxin separates and gains access to the cytoplasm, it facilitates the transfer of adenosine diphosphate ADP -ribose onto an elongation-factor protein EF-2 that is needed for protein synthesis. Hence, diphtheria toxin inhibits protein synthesis in the host cell, ultimately killing the cell Figure 6.
Cholera toxin is an enterotoxin produced by the gram-negative bacterium Vibrio cholerae and is composed of one A subunit and five B subunits. The mechanism of action of the cholera toxin is complex. The B subunits bind to receptors on the intestinal epithelial cell of the small intestine. After gaining entry into the cytoplasm of the epithelial cell, the A subunit activates an intracellular G protein. The activated G protein, in turn, leads to the activation of the enzyme adenyl cyclase, which begins to produce an increase in the concentration of cyclic AMP a secondary messenger molecule.
Botulinum toxin also known as botox is a neurotoxin produced by the gram-positive bacterium Clostridium botulinum.
It is the most acutely toxic substance known to date. The toxin is composed of a light A subunit and heavy protein chain B subunit. The B subunit binds to neurons to allow botulinum toxin to enter the neurons at the neuromuscular junction.
Normally, neurons release acetylcholine to induce muscle fiber contractions. This has the potential to stop breathing and cause death. Because of its action, low concentrations of botox are used for cosmetic and medical procedures, including the removal of wrinkles and treatment of overactive bladder.
Another neurotoxin is tetanus toxin , which is produced by the gram-positive bacterium Clostridium tetani. This toxin also has a light A subunit and heavy protein chain B subunit.
Unlike botulinum toxin, tetanus toxin binds to inhibitory interneurons, which are responsible for release of the inhibitory neurotransmitters glycine and gamma-aminobutyric acid GABA. Normally, these neurotransmitters bind to neurons at the neuromuscular junction, resulting in the inhibition of acetylcholine release. Tetanus toxin inhibits the release of glycine and GABA from the interneuron, resulting in permanent muscle contraction.
The first symptom is typically stiffness of the jaw lockjaw. Violent muscle spasms in other parts of the body follow, typically culminating with respiratory failure and death. Figure 7 shows the actions of both botulinum and tetanus toxins. Figure 7. Mechanisms of botulinum and tetanus toxins. Membrane-disrupting toxins affect cell membrane function either by forming pores or by disrupting the phospholipid bilayer in host cell membranes.
Two types of membrane-disrupting exotoxins are hemolysin s and leukocidins , which form pores in cell membranes, causing leakage of the cytoplasmic contents and cell lysis.
These toxins were originally thought to target red blood cells erythrocytes and white blood cells leukocytes , respectively, but we now know they can affect other cells as well. The gram-positive bacterium Streptococcus pyogenes produces streptolysins , water-soluble hemolysins that bind to the cholesterol moieties in the host cell membrane to form a pore. The two types of streptolysins, O and S, are categorized by their ability to cause hemolysis in erythrocytes in the absence or presence of oxygen.
Streptolysin O is not active in the presence of oxygen, whereas streptolysin S is active in the presence of oxygen. Other important pore-forming membrane-disrupting toxins include alpha toxin of Staphylococcus aureus and pneumolysin of Streptococcus pneumoniae.
Bacterial phospholipases are membrane-disrupting toxin s that degrade the phospholipid bilayer of cell membranes rather than forming pores. We have already discussed the phospholipases associated with B. These same phospholipases are also hemolysins. Other phospholipases that function as hemolysins include the alpha toxin of Clostridium perfringens , phospholipase C of P.
Some strains of S. PVL consists of two subunits, S and F. The S component acts like the B subunit of an A-B exotoxin in that it binds to glycolipids on the outer plasma membrane of animal cells.
The F-component acts like the A subunit of an A-B exotoxin and carries the enzymatic activity.
0コメント