In Search of the Mechanism of Persister Formation

Identification of the mechanism of persister formation presents a formidable challenge due to an apparent redundancy of persister genes. Thus, attempts to identify persister genes by screening transposon insertion libraries for either increased or decreased survival to antibiotics were not successful (Hu and Coates 2005; Spoering et al. 2006). In a recent report, phoU was identified as a putative persister gene using a similar approach. However, the phoU mutant had a decreased MIC to a number of antibiotics. This suggests that phoU::Tn is a pleiotropic mutation. It is interesting to compare this experience with identifying genes controlling another function that produces dormant cells: sporulation. It is easy to obtain specific spo mutants specifically lacking the ability to make spores from a knockout library (*), and this is indeed how most spo genes were identified. Genes controlling tolerance resemble in this regard those coding for multidrug resistance pumps (MDRs). In P. aeruginosa, for example, there are at least 13 RND family MDRs, but knockout out of most of them does not produce a discernible phenotype (*). At the same time, overexpressing any single MDR produces multidrug resistance. We therefore reasoned that persister genes may be identified by screening or selecting a library cloned into an expression vector for gain of function. In this case, even a weak contributor to a multigene function can be identified when overexpressed. However, this approach is problematic as well, since overproduction of many proteins leads to misfolded toxic products that can stop cell growth and will create an artifact emulating a dormant state, as discussed above. It appears that standard approaches of molecular genetics are poorly suited to search for persister genes, which probably explains the slow pace of discovery in this area.

Another barrier to discovery has been a lack of approaches to isolate persister cells. The first method to isolate persisters was recently reported, based on simply sedimenting surviving cells from a culture lysed by ampicillin (Keren et al. 2004b). This method has its limitations: it requires a rapidly growing culture for ampicillin to lyse it, and the fraction of persisters in such a population is small, approximately 10"5. In E. coli, this necessitated the use of a hipA7 strain overproducing persisters. In addition, these persisters are exposed to an antibiotic. These limitations notwithstanding, enough cells were collected to obtain a gene expression profile. The profile showed downregulation of proteins involved in energy production and nonessential functions such as flagellar synthesis, suggesting that persisters are dormant cells. This is consistent with the finding that persisters formed by a hipA7 (high persistence) strain of E. coli are nongrowing (or slow growing) cells (Balaban et al. 2004). The profile also pointed to proteins that may be responsible for dormancy: RMF, a stationary state inhibitor of translation (Yoshida et al. 2002), SulA, an inhibitor of septation (Walker 1996), and toxin-antitoxin (TA) module elements RelBE, DinJ, and MazEF (Christensen and Gerdes 2003; Christensen et al. 2003). Homologs of TA modules are found on plasmids where they constitute a maintenance mechanism (Hayes 2003). Typically, the toxin is a protein that inhibits an important cellular function such as translation or replication, and forms an inactive complex with the antitoxin. The toxin is stable, while the antitoxin is degradable. If a daughter cell does not receive a plasmid after segregation, the antitoxin level decreases due to proteolysis, leaving a toxin that either kills the cell or inhibits propagation. TA modules are also commonly found on bacterial chromosomes, but their role is largely unknown. MazEF was proposed to serve as a programmed cell death mechanism (Sat et al. 2001). However, it was reported recently that MazF and an unrelated toxin RelE do not actually kill cells, but induce stasis by inhibiting translation, a condition that can be reversed by expression of corresponding antitoxins (Pedersen et al. 2002; Christensen et al. 2003).

Expression of RelE, a toxin that causes reversible stasis by inhibiting cleaving mRNA and inhibiting translation, strongly increased tolerance to antibiotics (Keren et al. 2004b). Expression of a toxin HipA increased tolerance as well (Falla and Chopra 1998; Correia et al. 2006; Korch and Hill 2006; Vazquez-Laslop et al. 2006). Interestingly, a bioinformatics analysis indicates that HipA is a member of the Tor family of kinases, which have been extensively studied in eukaryotes (Schmelzle and Hall 2000), but have not been previously identified in bacteria. HipA is indeed a kinase, it autophosphorylates on ser150, and site-directed mutagenesis replacing it, or other conserved amino acids in the catalytic and Mg2+-binding sites abolishes its ability to stop cell growth and confer drug tolerance (Correia et al. 2006). Knowing that HipA is a kinase provides an additional tool to search for the target, which is yet to be identified.

Deletion of potential candidates of persister genes noted above does not produce a discernible phenotype affecting persister production, possibly due to the high degree of redundancy of these elements. In E. coli, there are at least ten toxin-antitoxin (TA) modules, and more than 60 in Mycobacterium tuberculosis (Gerdes et al. 2005).

Several independent lines of evidence point to persister dormancy: lack of growth in the presence of antibiotics (by contrast to resistant mutants), downregu-lation of biosynthetic pathways, and an elegant demonstration of slow growth or no growth in persisters formed by the E. coli hipA7 strain (Balaban et al. 2004). In the latter study, cells were placed in troughs of a multichannel chip that restricts mobility and makes it possible to simultaneously videotape growth and division of many individual cells in the channels. The device also made it possible to flush the medium, and application of ampicillin caused lysis of cells. However, cells that did not lyse were those that had little growth preceding the application of ampicillin.

Based on these data, we reasoned that dormancy may be used to physically sort naïve persister cells from a wild type population (Shah et al. 2006). Dormancy implies low levels of translation, which can then enable differential sorting based on expression of a detectable protein. In E. coli ASV, a degradable GFP is inserted into the chromosome in the X attachment site and expressed from the ribosomal rrnBP1 promoter, the activity of which is proportional to the rate of cell growth (Fig. 4). The half-life of degradable GFP is less than 1 h, and it should be effectively cleared from dormant cells. This would then enable sorting of dim persister cells. A logarithmically growing population of E. coli ASV was sorted with a high-speed cell-sorter using forward light scatter, which detects particles based on size. This enabled detection of cells irrespective of their level of fluorescence. Sorting by fluorescence showed that the population consisted of two strikingly different types of cells: a bright majority and a small subpopulation of cells with no detectable fluorescence (Fig. 4). Fluorescent microscopy confirmed that the sorted bright cells were indeed bright green, while the dim ones had no detectable fluorescence. The dim cells were also smaller than the fluorescent cells, and in this regard resembled stationary state cells. Sorted dim cells were exposed to a high level of ofloxacin that rapidly kills both growing and nongrowing normal cells, but has no effect on persisters. The majority of this subpopulation survived, as compared to a drastic drop in viability of the sorted bright cells. This experiment showed that the sorted dim cells are dormant persisters.

The sorting method provides a general approach to obtaining naïve persisters from a wild type population of any species. These cells can then be used to obtain an expression profile and to study a variety of functions by biochemical methods. However, there is still room for improvement: sorting is relatively slow, costly, and results in a limited amount of material that precludes proteome analysis, for example. A rapid method for obtaining large quantities of persisters has yet to be developed.

Knowing that persisters are dormant cells bolsters the case for TA module involvement in persister formation. Indeed, TAs seem to be ideally suited for the task. Reversible action of toxins such as RelE and MazF, inhibition of important cellular functions by toxins capable of creating a dormant state, and the presence of TA modules in the chromosomes of all known free-living bacteria makes them attractive candidates for persister genes.



64 128 192 Forward Scatter

Healthy Chemistry For Optimal Health

Healthy Chemistry For Optimal Health

Thousands Have Used Chemicals To Improve Their Medical Condition. This Book Is one Of The Most Valuable Resources In The World When It Comes To Chemicals. Not All Chemicals Are Harmful For Your Body – Find Out Those That Helps To Maintain Your Health.

Get My Free Ebook

Post a comment