Michael J. Mahan is an Associate Professor and David A. Low is a Professor in the Department of Molecular, Cellular, and Developmental Biology, University of California, Santa Barbara.
Article Links:
Links to Other ASM Pages:
DNA Methylation Regulates Bacterial Gene Expression and Virulence
When this key enzyme is altered in mutants, they are no longer virulent and can serve as vaccines
Michael J. Mahan and David A. Low
DNA adenine methylase (Dam) plays a pivotal role in bacteria such as Escherichia coli and Salmonella—acting as a global regulator of gene expression and affecting a wide range of critical cellular functions, including DNA replication, DNA repair, transposition, and segregation of chromosomal DNA.
This extraordinary versatility stems from the inherent biochemical activity of Dam. Thus, by adding methyl groups to various sites along the cellular DNA, Dam alters interactions of a variety of regulatory proteins with their designated gene targets and, in the process, effectively controls expression of those genes. In some cases, such changes modulate bacterial virulence and also serve to elicit protective immune responses in host organisms that Salmonella or other bacterial pathogens may infect.
Through Dam Action, DNA Methylation Patterns Regulate Gene Expression
DNA methylation can affect gene expression by altering the affinity of regulatory proteins for DNA either through direct steric hindrance or, indirectly, through changes in DNA structure that affect the configuration of Dam target sites. Conversely, regulatory proteins that bind to nonmethylated Dam target sites can modulate gene expression by protecting them from methylation. Such stable, nonmethylated DNAs can also form heritable DNA methylation patterns that can affect gene expression in subsequent generations—a phenomenon that also is observed in eukaryotes, where it encompasses genomic imprinting, gene silencing, development, and tumorigenesis. Because no changes occur within the DNA primary sequence, this type of heritable gene regulation is considered "epigenetic."
Some of our insights into Dam control over gene expression come from studying the E. coli pap operon, which encodes pili that enable these bacteria to adhere to the urogenital tract and establish kidney infections. Genes encoding Pap pili are reversibly switched between the unexpressed state and the expressed state, a process known as phase variation. This ON-OFF pilin switching leads the bacteria to attach and detach, enabling them to colonize first the bladder and then the kidney, causing cystitis and pyelonephritis, respectively.
Pap phase variation is principally controlled by the leucine-responsive regulatory protein (Lrp) and the state of methylation of two specific pap regulatory DNA sequences that reside within Lrp-binding sites (Fig. 1). During the OFF-phase, Lrp binds to pap promoter proximal sites, inhibiting methylation at GATCprox and blocking pap transcription. During the ON-phase, Lrp binds to pap promoter distal sites, inhibiting methylation at GATCdistand activating pap transcription. Thus, methylation of two specific GATC sites modulates Lrp binding, controlling pap transcription. These pap DNA methylation patterns are heritable, regulating gene expression in subsequent generations.
DNA methylation patterns regulate expression of genes within several pili operons in both E. coli and Salmonella. Although Dam methylates most of the approximately 18,000 GATC sites along the E. coli chromosome, the binding of regulatory proteins at or near about 50 to 100 of those sites protects them from being methylated, leading to specific and stable methylation patterns. Cloning experiments indicate that some of these sites reside in gene control regions. Growth conditions can affect which among these GATC sites remain unmethylated, presumably reflecting fluctuations in levels of regulatory proteins that bind Dam target sites and alter gene expression in response to environmental conditions.
Dam Regulates Virulence in Salmonella
We recently examined the role of Dam in regulating gene expression in Salmonella, a pathogen that causes food and blood poisoning as well as typhoid fever in humans. Many genetically determined virulence functions of this pathogen are induced during infections but are not expressed, or expression is relatively low, when the cells are growing in vitro—"ON" in vivo, "OFF" in vitro. Because this response involves the coordinated expression of many genes when cells are moved from host tissues to synthetic media, it suggests that the cells depend on global regulatory factors, which we designated in vivo induced (ivi) genes.
Reasoning that mutations in this regulatory mechanism would likely impair the capacity of Salmonella to cause disease, we devised a means for screening mutations that permit ivi genes to be expressed when cells are grown in vitro on synthetic medium—"ON" in vivo and "ON" in vitro. This screening method yielded cells with mutations that resulted in the loss of Dam activity. In turn, those changes derepress the activity of more than 35 distinct ivi genes in Salmonella, indicating that Dam is a global regulator of Salmonella gene expression.
Because Dam affects expression of so many Salmonella genes that are preferentially expressed during infections, we pursued our studies of its role in pathogenesis. For instance, Salmonella strains that either lack Dam or overproduce it (DamOP) have severe virulence defects when tested in mice for their ability to cause typhoid fever. Thus, mice survive and show no overt signs of disease when inoculated with 10,000-fold more Dam- or DamOP cells than are needed of the wild-type (Dam+) strain to kill 50% of the animals (LD50). Put more simply, Dam plays an essential role in Salmonella pathogenesis.
Experiments Provide Insights about Salmonella Virulence and Vaccine Design
Immunologists and other researchers interested in developing vaccines often use Salmonella spp to construct experimental live bacterial vaccines—in part, to take advantage of their direct interactions with gut-associated lymphoid tissue (GALT) that can produce strong humoral responses. Moreover, because Salmonella can also invade and proliferate within host cells, these bacteria also are capable of eliciting strong cell-mediated immune responses.
With that work on bacterial vaccines in mind, we tested whether either one of our highly attenuated sets of Dam mutants could serve as a live vaccine to protect mice against infections by the wild type or other virulent strains of Salmonella. In one set of experiments, inoculating mice with the Dam–mutants protected them when they were subsequently exposed to the virulent wild-type strain, which only poorly colonized systemic tissue sites and produced no overt signs of disease among the Dam–-vaccinated mice.
Although this high degree of protection suggests that Salmonella Dam– mutants represent promising vaccine candidates, the effectiveness of these attenuated mutants is somewhat surprising since other highly attenuated mutants typically are cleared too rapidly from animals to elicit protective responses. However, Dam– Salmonella are less cytotoxic to M cells in gut mucosal tissue and survive within Peyer's patches of the mouse small intestine for an extended period, providing an opportunity to elicit host immune responses, according to our findings and those of Josep Casadesus of the University of Seville in Spain. Moreover, Dam–mutants only ineffectively invade or survive within systemic tissues—another factor contributing to their inability to cause disease.
Perhaps more importantly, when Dam– and DamOP strains are grown in broth, they express a series of specific proteins that differ from one another and from those produced by wild-type Salmonella strains. Thus, when Dam activity is abnormal, many other genes in such cells are expressed abnormally. We speculate that these changes, which affect expression among sets of genes, could lead to ectopic production of key antigens, better enabling them to elicit protective immunity when mice are inoculated with the mutant cells (Fig. 2).
To be effective against certain pathogens, vaccines need to protect individuals against a variety of different isolates. In the case of the more than 2,500 identified Salmonella strains that infect humans and animals, designing a broadly effective vaccine has proved a formidable challenge. However, by using live attenuated organisms that express the immunogens that are produced by many slightly varied strains, it might be possible to meet this challenge.
Because Salmonella Dam mutant strains aberrantly express several potential immunogens, we decided to test whether Dam mutant vac cines could elicit cross-protective immunity to distinct pathogenic strains—specifically, S. enterica serovar Typhimurium and S. enterica serovar Enteritidis, which are the two leading causes of Salmonella infection in humans, mainly from consuming contaminated chickens and eggs, and S. enterica serovar Dublin, which causes disease in calves. For example, mice were protected when they were vaccinated with Dam- mutants of Salmonella serovar Typhimurium and then challenged with either Salmonella serovar Dublin or Salmonella serovar Enteritidis. Reciprocal treatment with Dam–Salmonella serovar Enteritidis provides comparable cross-protection against subsequent challenge with either of the two other Salmonella pathogens.
Moreover, the cross-protection does not depend on the continuing presence of, or retreatment with, the vaccine strain. This finding is important because other investigators report that live vaccines sometimes elicit short-term protective responses that decline after those vaccines are cleared.
Dam Activity Affects Virulence in Several Other Bacterial Pathogens
Mutations in Dam and other adenine methyltransferases from an even wider variety of microbial pathogens can attenuate virulence, suggesting that enzymes which methylate DNA are pivotal and might prove essential in bacterial pathogenesis (see table).
For instance, Dam activity affects virulence of several other pathogenic bacteria, including Vibrio cholerae and Yersinia pseudotuberculosis, the causative agents of cholera and blood poisoning in mice, respectively. In contrast to E. coli and Salmonella, intact Dam is essential for viability in V. cholerae and Y. pseudotuberculosis (see table). However, mutants of these two species that overproduce Dam remain viable, and are attenuated to a significant degree.
These results suggest that Dam plays a role closely resembling that played by the cell cycle-regulated adenine methyltransferase (CcrM) in other microorganisms, including Caulobacter crescentus, Rhizobium meliloti, and Brucella abortus. Moreover, in B. abortus, mutants that overproduce CcrM are similarly no longer virulent, according to Lucy Shapiro and her collaborators at Stanford University in Stanford, Calif. Additionally, DNA methylation appears to play a role among some bacterial pathogens that infect plants, according to Noel Keene and Ching-Hong Yang at the University of California, Riverside. They find that Dam–mutants of the plant pathogen Erwinia chrysanthemi no longer cause disease in African violets and lettuce.
Based on our findings with Dam Salmonella mutants that serve as candidate vaccines for protecting animals against infections by pathogens of that species, we tested attenuated, Dam-overproducing mutants of Y. pseudotuberculosis for their effectiveness in protecting mice against virulent strains of this pathogen. Indeed, immunized mice were highly protected against a virulent and otherwise deadly wild-type challenge. Moreover, the virulent strain only poorly colonized the vaccinated mice, and induced no overt signs of disease.
These Dam-overproducing mutants of Y. pseudotuberculosis secrete several virulence-associated proteins [Yops] that, in wild-type strains, are not secreted under certain conditions. This altered, or ectopic, distribution of these proteins may contribute to the loss in virulence and capacity to elicit protective immune responses of these mutants. Because these results so closely parallel what we observed in salmonellae, we suspect that disrupting Dam activity may provide a general strategy for generating vaccines against still other bacterial pathogens.
Dam Affects Gene Expression, Virulence, and Protective Immune Responses
In affecting virulence and eliciting protective immune responses, Dam appears to act as a global regulator of gene expression within certain bacterial pathogens. By methylating DNA, Dam alters the affinity of regulatory proteins for various sites along the chromosome and thereby changes how many genes are expressed. Conversely, regulatory proteins sometimes bind to unmethylated, would-be Dam target sites, protecting them from becoming methylated.
Thus, competition between Dam and regulatory proteins for specific GATC sites regulates methylation patterns as well as expression of various genes. Environmental factors can influence this competitive equilibrium, yielding heritable changes in gene expression. These changes can arise, for example, when the levels or the binding affinities of regulatory proteins for their Dam target sites become altered. Moreover, when Dam activity in a pathogen is disrupted, the resulting mutant strain often becomes avirulent, making it a candidate to serve as a vaccine for inducing protective immunity. This potential is accentuated because such mutants may produce an expanded repertoire of potential antigens ectopically (Fig. 2).
Because DNA methylation patterns are inherited and may affect gene expression over many subsequent generations, the environment that parent cells experience can influence the behavior of daughter cells. And, because this form of "cellular memory" appears critical for coordinating bacterial gene expression throughout infection cycle, its activity and influence may explain, in part, Dam's critical role in determining virulence among such a widely diverse group of bacterial pathogens.
ACKNOWLEDGMENTS
We thank Doug Heithoff and Steve Julio for critical reading of the manuscript.
This work was supported by private donations from Jim and Deanna Dehlsen, University of California Biotech Program, the Santa Barbara Cottage Hospital Research Program, USDA grant 2000-02539 (to M.J.M), and National Institutes of Health (NIH) grant AI23348 (to D.A.L.).
SUGGESTED READING
Braaten, B. A., X. Nou, L. S. Kaltenbach, and D. A. Low. 1994. Methylation patterns in pap regulatory DNA control pyelonephritis-associated pili phase variation in E. coli. Cell 76:577-588.
Garcia del Portillo, F., M. G. Pucciarelli, and J. Casadesus. 1999. DNA adenine methylase mutants of Salmonella typhimurium show defects in protein secretion, cell invasion, and M cell cytotoxicity. Proc. Natl. Acad. Sci. USA 96:11578-11583.
Heithoff, D. M., R. L. Sinsheimer, D. A. Low, and M. J. Mahan. 1999. An essential role for DNA adenine methylation in bacterial virulence. Science 284:967-970.
Hendrich, B., and A. Bird. 2000. Mammalian methyltransferases and methyl-CpG-binding domains: proteins involved in DNA methylation. Curr. Topics Microbiol. Immunol. 249:55-74. Mahan, M. J., D. M. Heithoff, R. L. Sinsheimer, and D. A. Low. 2000. Assessment of bacterial pathogenesis by analysis of gene expression in the host. Annu. Rev. Genet. 34:139-164.
Marinus, M. G. 1996. Methylation of DNA, p. 782-791. In F. C. Neidhardt, R. Curtiss III, J. L. Ingraham, E. C. C. Lin, K. B. Low, B. Magasanik, W. S. Reznikoff, M. Riley, M. Schaechter, and H. E. Umbarger (ed.), Escherichia coli and Salmonella. Cellular and Molecular Biology, 2nd ed. ASM Press, Washington, D.C.
Reisenauer, A., L. S. Kahng, S. McCollum, and L. Shapiro. 1999. Bacterial DNA methylation: a cell cycle regulator? J. Bacteriol. 181:5135-5139.
Robertson, G. T., A. Reisenauer, R. Wright, R. B. Jensen, A. Jensen, L. Shapiro, and R. M. Roop RM 2nd. 2000. The Brucella abortus CcrM DNA methyltransferase is essential for viability, and its overexpression attenuates intracellular replication in murine macrophages. J. Bacteriol. 182:3482-3489. van der Woude, M., B. Braaten, and D. Low. 1996. Epigenetic phase variation of the pap operon in Escherichia coli. Trends Microbiol. 4:5-9.
Email: webmaster@asmusa.org