annurev.mi.47.100193

Annu. Rev. Microbiol. 1993.47:855-74 Copyright © 1993 by Annual Reviews Inc. All rights reserved THE STATIONARY PHASE OF THE BACTERIAL LIFE CYCLE Roberto Kolter Department of Microbiology and Molecular Genetics, Harvard Medical School, 200 Longwood Avenue, Boston, Massachusetts 021IS Deborah A. Siege Ie Department of Biology, Texas A & M University, College Station, Texas 77843 Antonio Torma Departamento de Bioquimica y Biologia Molecular, Facultad de Ciencias Quimicas, Universidad Complutense de Madrid, Madrid, Spain KEY WORDS: stationary phase, survival, starvation, gram-negative bacteria CONTENTS INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 856 Stages of the Escherichia coli Life Cycle . . . . . . . . . . . . . . . . . . . . . . . . . . 856 Properties of Starved Gram-Negative Bacteria ...................... 857 ENTRY INTO STATIONARY PHASE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 859 Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 859 Protein Synthesis During Entry into Stationary Phase . . . . . . . . . . . . . . . . . . 860 A Stationary Phase-Specijic Sigma Factor. . . . . . . . . . . . . . . . . . . . . . . . . 86 1 Other Rt'gulators of Stationary-Phase Gene Expression . . . . . . . . . . . . . . . . . 863 RESPONSES DURING PROLONGED STARVATION . . . . . . . . . . . . . . . . . . . 864 Protein Synthesis . . ..... . . . . . . . . . . . . . . . ... . . . . . . . . . . . . ... 864 Survival Kinetics ........................................ 865 Mutants with a Competitive Advantage in Stationary Phase .... . .. . ...... 865 REENTERING THE GROWTH CYCLE FROM STATIONARY PHASE . . . . . . . . 868 SUMMARY AND PERSPECTIVES 870 ABSTRACT In the natural environment bacteria seldom encounter conditions that permit periods of exponential growth. Rather, bacterial growth is characterized by long periods of nutritional deprivation punctuated by short periods that allow 855 0066-4227/93/1001-0855$02.00 A nn u. R ev . M ic ro bi ol . 1 99 3. 47 :8 55 -8 74 . D ow nl oa de d fro m w w w .an nu al re vi ew s.o rg by U ni ve rs ity o f S ci en ce & T ec hn ol og y of C hi na o n 09 /0 6/ 12 . F or p er so na l u se o nl y. 856 KOLTER, SIEGELE & TORMO fast growth, a feature that is commonly referred to as the feast-or-famine lifestyle. In this chapter we review the recent advances made in our understanding of the molecular events that allow some gram-negative bacteria to survive prolonged periods of starvation. After an introductory description of the properties of starved gram-negative bacteria, the review presents three aspects of stationary phase: entry into stationary phase, responses during prolonged starvation, and reentry into the growth cycle. INTRODUCTION In the natural environment bacteria seldom encounter conditions that permit continuous balanced growth. When nutrients are plentiful, bacteria can sustain relatively fast growth rates. But the very fact that bacterial populations can use nutrients efficiently to generate rapid increases in their biomass means that they are nutritionally starved most of the time. Still, these organisms can survive for extremely long periods in the absence of nutrients. Laboratory conditions do not exactly reflect what bacteria find in nature. However, one can simulate short periods of nutrient availability and prolonged periods of starvation by growing cultures in synthetic media. In most media, exponentially growing cells quickly use up the available nutrients and cease their exponential increase in biomass, thus entering a phase of the culture referred to as stationary phase. This part of the bacterial life cycle has always attracted the attention of investigators, and in recent years, through the application of modem genetic and biochemical approaches, exciting discov­ eries have been made with regards to the molecular mechanisms that bacteria utilize to survive during stationary phase. The ability of many bacteria to form dormant spores or multicellular aggregates in response to starvation has been extensively studied, and the reader is referred to several recent reviews on the subject (46, 52, 82, 83), two of which appear in this volume (17, 38). This review focuses on the responses that gram-negative organisms, in particular Escherichia coli, mount when confronted with starvation conditions in the laboratory. Several review articles that treat the same subject from different perspectives have appeared recently (32, 47, 54, 55, 80). Stages of the Escherichia coli Life Cycle During normal exponential growth, E. coli cells undergo cycles of cell growth and division in which daughter cells are virtually identical to the mother cell. Theoretically, the cessation of growth in response to starvation could result simply from the arrest of metabolic activity anywhere along this growth cycle. Growth could then be reinitiated by restarting the cycle from its point of arrest once nutrients were again available. Whereas this scheme might be possible, the sudden arrest of growth could halt key metabolic processes, DNA A nn u. R ev . M ic ro bi ol . 1 99 3. 47 :8 55 -8 74 . D ow nl oa de d fro m w w w .an nu al re vi ew s.o rg by U ni ve rs ity o f S ci en ce & T ec hn ol og y of C hi na o n 09 /0 6/ 12 . F or p er so na l u se o nl y. STATIONARY PHASE 857 The STATIONARY PHASE E. coli Life Cycle � ENTRY+- C,,--�) C,,--_) C, ___ :�) � C MAINTENANCE .... ... EXIT C ____ ) .. C __ �) EXPONENTIAL PHASE Figure 1 The three stages of stationary phase as part the E. coli life cycle. replication in particular, at stages at which severe and irreparable damage could occur. In order to insure their survival, bacteria should be able to make an orderly transition into stationary phase such that the cell cycle is not arrested randomly. In addition, bacteria must also be able to remain viable during prolonged periods of starvation and to exit stationary phase and return to the exponential cell cycle when starvation is relieved. Thus the physiology of starved bacteJia can be divided into three stages: entry into stationary phase, maintenance of viability, and exit from stationary phase (Figure I). The results reviewed here show that, upon being starved, gram-negative bacteria can enter a developmental program that results in metabolically less active and more resistant cells. However, in contrast to other microbial developmental programs such as sporulation, this starvation-induced differentiation does not appear to be ;m all-or-none process involving an irreversible commitment to a program. Rather, the differentiation appears more gradual: the slower the growth rate of the culture, the more growing cells resemble starved cells. Properties /Jf Starved Gram-Negative Bacteria Morphological changes that are brought about by starvation are apparent through both light and electron microscopic examination. The familiar rod shape of growing E. coli is lost in stationary phase because cells become A nn u. R ev . M ic ro bi ol . 1 99 3. 47 :8 55 -8 74 . D ow nl oa de d fro m w w w .an nu al re vi ew s.o rg by U ni ve rs ity o f S ci en ce & T ec hn ol og y of C hi na o n 09 /0 6/ 12 . F or p er so na l u se o nl y. 858 KOLTER, SIEGELE & TORMa much smaller and almost spherical as the result of several cell divisions without an increase in cell mass (39, 48). A number of marine bacteria greatly decrease in size during starvation and develop into forms termed ultra­ microcells that reduce their size by a combination of reductive divisions and endogenous metabolism (47). It has been proposed that such dramatic size reduction during starvation may improve a strain's survival by increasing cell numbers (58). Within the cell, the relative volume and disposition of the subcellular compartments also changes; the cytoplasm is condensed and the volume of the periplasm increases (72). Changes in the cell envelope that result from starvation reflect the need for protection and insulation from stressful environments. On their surface, starved cells are covered with more hydrophobic molecules that favor adhesion and aggregation (47). Membranes may become less fluid and less permeable as fatty acid composition changes. For example, in E. coli all unsaturated membrane fatty acids are converted to the cyclopropyl derivatives as cells enter stationary phase (19). E. coli, when starved at low temperatures, produce curli, a fibronectin-binding filament that may also be involved in aggregation (l0, 69). Fimbrae-like structures and cellular aggregates or clumps are also characteristic of starved Vibrio spp. (2). In addition, the cell wall undergoes structural changes when cells are starved; these alterations may be correlated with increased resistance to autolysis (67, 92). The chromosome undergoes topological changes consistent with the reduc­ tion in gene expression observed in starved cells. After several hours in stationary phase, changes in the negative superhelical density of plasmids become apparent (13) and the nucleoid condenses (12, 59). This could in part result from the production of large amounts of the histone-like proteins H-NS (84) and Dps (8) during starvation. Bacterial genera such as Escherichia, Salmonella, and Vibrio are not generally considered to form differentiated cells as a result of starvation. Clearly these species do not form classical spores, but the morphological changes described above make it evident that major structural changes do occur when they are starved. These changes, in combination with changes in metabolism and physiology, confer on starved gram-negative cells many of the properties of classical spores. The spores that result from starvation of many gram-positive bacteria are characterized by their extreme resistance to different environmental stresses. In nonsporulating gram-negative bacteria, starvation also induces the devel­ opment of a more resistant state (54). E. coli cells that have been starved are more resistant to heat shock, oxidative stress, and osmotic Challenge than exponential-phase cells (42, 43, 49). Although resistance to these stresses can be induced during growth by exposure to nonlethal levels of heat, H2D2, or salt, the resistance produced by starvation is even more protective. A nn u. R ev . M ic ro bi ol . 1 99 3. 47 :8 55 -8 74 . D ow nl oa de d fro m w w w .an nu al re vi ew s.o rg by U ni ve rs ity o f S ci en ce & T ec hn ol og y of C hi na o n 09 /0 6/ 12 . F or p er so na l u se o nl y. STATIONARY PHASE 859 Many of the properties identified as stationary-phase induced may also be important for growth under conditions of limiting or poor nutrient availability. Hence for some physiological processes, stationary phase may represent a maximally slow growth rate (18). Many functions induced during stationary phase are aliso induced when the cells are growing with a long doubling time; several promoters induced during stationary phase also show an inverse proportionality of expression as a function of growth rate (6, 7, 18). In some cases, this inverse proportionality results in the presence of a constant number of gene-product molecules per cell, which may be important for the organi­ zation of processes such as cell division (94). ENTRY INTO STATIONARY PHASE Definition Bacteria growing in batch culture will inevitably reach a point when the growth rate decreases, indicating the onset of stationary phase. But what is really meant by stationary phase? Growth might be prevented by the exhaustion of any one of several essential nutrients, but does stationary phase represent a homogeneous physiological state? No--stationary-phase cultures is only a descriptive term: cultures in which the number of bacterial cells ceases to increase are said to be in stationary phase. This description does not distinguish whether or not the cells are metabolically active or even if they are undergoing cell division or not. It simply refers to a culture that shows no further increase in the number of cells, which points out some of the difficulties with the term. The time of the onset of stationary phase will differ depending on what criterion is used to define entry into this phase. Because of the reduction in cell size upon entry into stationary phase, cells will stop increasing in size while they continue to increase in number. If cell growth is measured by optical density, the defined onset of stationary phase will be much earlier than if the cessation of growth is defined as the moment when cell numbers cease increasing. Therefore, the investigator must recognize that entry into station­ ary phase is a transition period beginning at the point in the exponential phase when all cellular parameters cease increasing at equal rates, i. e. DNA, protein, and total cell mass no longer increase together, and continuing until the time when no further increase in cell number is detected. Just as one must recognize that the entry into stationary phase is a transition period for the cell, one should understand that the cell physiology during this time will vary enormously depending on the composition of the medium in the which cells are growing and on the conditions in which starvation came about. Experimentally, a culture can enter stationary phase in many ways. The most dearly defined ones involve starvation for a single nutrient, for A nn u. R ev . M ic ro bi ol . 1 99 3. 47 :8 55 -8 74 . D ow nl oa de d fro m w w w .an nu al re vi ew s.o rg by U ni ve rs ity o f S ci en ce & T ec hn ol og y of C hi na o n 09 /0 6/ 12 . F or p er so na l u se o nl y. 860 KOLTER, SIEGELE & TORMO which two methods can be used. One can take a culture in midexponential phase, spin out the cells, and resuspend them in identical medium that lacks one specific nutrient, e.g. a source of carbon. Alternatively, cultures can be incubated in medium containing one nutrient at a concentration low enough such that it will be exhausted before all other nutrients. Although the resuspension method is useful in defining a quick transition between conditions that permit growth and those that do not, it suffers from the fact that it artificially places the cells in a new medium. If any extracellular signaling molecules accumulate during the entry into stationary phase, these will be lacking in resuspension experiments. Protein Synthesis During Entry into Stationary Phase The patterns of proteins synthesized during the entry into stationary phase have been analyzed extensively using two-dimensional SDS polyacrylamide gel electrophoresis and gene fusions (31, 49, 65, 66, 85-87). The results obtained have provided an initial picture of the molecular events that occur in the cell as it senses starvation. Each nutritional starvation condition that leads to the cessation of growth results in the induction of a characteristic set of proteins that accompanies the inevitable decrease in the overall rate of protein synthesis. While the proteins that are induced vary widely depending on the conditions of starvation, a core set of 15-30 proteins is always induced in E. coli and has been designated the Pex (postexponential) proteins (55). A kinetic analysis of the proteins induced during the onset of stationary phase has revealed that not all proteins are induced with the same kinetics. E. coli, Salmonella typhimurium, and Vibrio spp. have temporal classes of gene expression. The expression of some genes is induced very early while others are not induced for many hours. The length of time that their expression is on also varies, but little is known about the molecular mechanisms responsible for the different kinetics. Proteins synthesized by starved cell during entry into stationary phase are involved in maintaining viability during prolonged starvation. Inhibiting protein synthesis with chloramphenicol during the first few hours in stationary phase greatly increases the rate at which cultures lose viability, while there is only a small decrease in viability when protein synthesis is inhibited after cells have been in stationary phase for several hours (66, 72). Some of the functions needed for maintaining maximal viability may function specifically in stationary phase and be dispensable during growth. This idea is supported by the isolation of E. coli mutants that, while appearing normal during logarithmic growth, fail to survive during stationary phase (49, 56, 90). Given the importance of the proteins made during stationary phase, the regulation of their synthesis must be critical for the survival of the cell. A nn u. R ev . M ic ro bi ol . 1 99 3. 47 :8 55 -8 74 . D ow nl oa de d fro m w w w .an nu al re vi ew s.o rg by U ni ve rs ity o f S ci en ce & T ec hn ol og y of C hi na o n 09 /0 6/ 12 . F or p er so na l u se o nl y. STATIONARY PHASE 861 A Stationary Phase-Specijic Sigma Factor The starvation-induced expression of many genes is controlled by an alterna­ tive sigma factor known as (fs or (f38. Several investigators independently identified the gene encoding (fs, and thus it initially received several different designations. Researchers found several regulatory mutations, without realiz­ ing they were in the same locus, that mapped near 59 min and affected a variety of processes: near ultraviolet light resistance (nur) (93), acid phospha­ tase produ1ction (appR) (91), and HPII catalase production (katF) (50). The first indication that this locus encoded an alternative sigma factor came from the nucleotide sequence of katF (60). Subsequently, a search for carbon starvation-·inducible gene fusions identified csi2: :lacZ, an insertion at 59 min with a pleiotropic phenotype with respect to acid phosphatase, HPII catalase production, and overall stationary-phase gene expression (49). This finding led to the recognition that appR and katF were different names for the same locus. Based on the gene's nucleotide sequence and its role in activating and repressing the synthesis of many proteins at the onset of starvation, the gene designation rpoS was proposed (49). Subsequent studies further demonstrated the central role of rpoS in the development of increased resistance at the onset of starvation (56). More recently, the purified protein was shown to have sigma-factor function in vitro (89). Hence we refer to the gene as rpoS and its product as (fs. At least 30 proteins require rpoS for their expression during starvation (56). Over a dozen of these proteins have been identified and many of them are important for the development of the resistant state seen in stationary phase (32) . A partial list of rpoS-dependent genes and their products or function includes: katE, HPII catalase (50); xthA, exonuclease III (75); appA, acid phosphatase (91); mcc, microcin C7 (25); boLA, cell-shape detennination (16, 48); osmB, lipoprotein (45); treA, periplasmic trehalase (34); otsAB, trehalose synthesis (34); cyxAB, third cytochrome oxidase (20); gLgS, glycogen primer (33); dps, DNA protection (8); and csgA, curli fibronectin binding fibers (68). With all of these genes identified, one might expect that a consensus sequence for rpoS-dependent promoters would have been derived. Although a possible -10 and -35 consensus sequence that differed from the (f70 consensus was proposed (48), the consensus broke down as more rpoS-de­ pendent