atlantica coincidien-

that is, single cells give rise to colonies with low frequency.

2.1.4 Dry weight and protein

The most direct way to measure growth is to quantify the dry weight of cells in a culture. The cells are harvested by either centrifugation or fi ltration, dried to a constant weight, and care-fully weighed. Since protein increases parallel with growth, one can also measure growth by doing protein measurements of the cells. The cells are harvested by either centrifugation or fi ltration or (more usually) precipitated fi rst with acid or alcohol, then recovered by centrif-ugation or fi ltration. A simple colorimetric test for protein is then performed.

2.2 Growth Physiology

The vast topic of growth physiology includes the regulation of rates of synthesis of

macro-molecules, the regulation of the timing of DNA synthesis and cell division, and such adaptive physiological responses to nutrient availabil-ity as changes in gene expression, homeostasis adaptations to the external environment, and the coupling of the rates of biosynthetic path-ways to the rates of utilization of products for growth (metabolic regulation). From this point of view, most of this text is concerned with the physiology of growth, and we therefore return to the topic in subsequent chapters. What follows here is an introduction to growth physiology as it pertains to the synthesis of macromolecules during steady state growth, and very gen-eral adaptations of cells to nutrient depletion.

We begin with a discussion of the sequence of growth phases through which a population of bacteria progresses after being inoculated into a fl ask of fresh media.

2.2.1 Phases of population growth

When one measures the growth of populations of bacteria grown in batch culture, a progres-sion through a series of phases can be observed (Fig. 2.3). The fi rst phase is frequently a lag phase, in which no net growth occurs (i.e., no increase in cell mass). This is followed by a phase of exponential growth, in which cell mass increases exponentially with time. Following exponential growth, the culture enters the sta-tionary phase, a phase of no net growth. After a stationary period, cell death occurs in a fi nal stage called the death phase. Notice in Fig. 2.3 that, prior to the exponential phase of growth, when cell division occurs, the cells increase in

Fig. 2.3 Growth kinetics in batch culture: solid line, mass; dashed line, viable cells. Note that if we defi ne growth as an increase in mass, then only the solid line accurately refl ects the growth of the culture. The dashed line refl ects growth only when it is parallel to the solid line.

mass (solid line); that is, they grow larger. At the end of exponential growth, the cells con-tinue to divide after growth has ceased (dashed line), but they become smaller. It would seem to be an advantage to a population of bacteria to continue to divide after growth has ceased in the population, since in this way more cells can be produced for distribution to new sites, where conditions for continued growth may be better.

One consequence of the uncoupling of growth from cell division during the lag and late log phases is that the size of the cell varies during growth in batch culture. For the investigator, the noncoincidence of growth and cell division during stages of batch growth has a practical consequence: namely, cell counts are not always a valid measurement of growth.

Lag phase

When cells in the stationary phase of growth are transferred to fresh media, a lag phase often occurs. In this situation, the lag phase is due basically to the time required for the physi-ological adaptation of stationary phase cells in preparation for growth. Usually the longer the cells are kept in the stationary phase, the lon-ger is the lag phase when they are transferred to fresh media. Some of the physiological changes that occur in the stationary phase are described in Section 2.2.2.

There are several possible reasons for the lag phase. In some cases the lag phase accommo-dates the requirement for time for the cells to recover from toxic products of metabolism, such as acids, bases, alcohols, or solvents, that may accumulate in the external medium. Sometimes, new enzymes or coenzymes must be synthesized before growth can resume. Such synthesis will be necessary if, for example, the fresh medium is dif-ferent from the inoculum medium and requires a change in the enzyme composition of the cells.

If signifi cant cell death occurs in the stationary phase, an apparent lag phase will be measured because the inoculum includes dead cells that contribute to the turbidity. The lag phase can be avoided if the inoculum is taken from the expo-nential phase of growth and transferred to fresh medium of the same composition.

Stationary phase

Cells stop growing and enter the stationary phase for various reasons. Among these are exhaustion of nutrients, limitation of oxygen,

and accumulation of toxic products (e.g., alco-hols, solvents, bases, acids). The accumulation of toxic products is frequently a problem for fermenting cells because instead of being con-verted to cell material, most of the nutrient is excreted as waste products. The excretion of fermentation end products is discussed in Chapter 15. For a review of the physiology of stationary phase cells, read ref. 1.

Death

Cell death can result from several factors.

Common causes of death include the depletion of cellular energy and the activity of autolytic (self-destructive) enzymes. Some bacteria begin to die within hours of entering the stationary phase. However, many bacteria remain viable for longer periods. For example, some bacte-ria sporulate or form cysts when exponential growth ceases. The spores and cysts are resting cells that remain viable and will germinate in fresh media. As discussed next, even nonsporu-lating bacteria can adapt to nutrient depletion and remain viable for long periods in stationary phase.

2.2.2 Adaptive responses to nutrient limitation

In the natural environment bacteria are fre-quently faced with starvation conditions and enter intermittent periods of no growth or very slow growth. In some environments the gen-eration time may be many days or even months because the nutrient levels are so dilute. When a bacterial culture faces the depletion of an essen-tial nutrient, the culture may keep growing by inducing specifi c uptake systems to scavenge the environment for the nutrient or a source thereof, and/or it may induce the synthesis of enzymes that can use an alternative source of the nutrient. For example, bacteria starved for inorganic phosphate may induce the synthesis of a high-affi nity inorganic phosphate uptake system as well as the synthesis of enzymes capa-ble of degrading organic phosphates to release inorganic phosphate (phosphatases). (See the discussion of the Ntr and PHO regulons in Sections 19.4 and 19.5, respectively.) If the bacteria cannot bring into the cells the essential nutrient, then the population will stop growing and dividing. Under these circumstances, the population enters stationary phase or, in the

from self-digestion of cell material, including the cell envelope. Some gram-negative bacteria (e.g., Pseudomonas) bleb off outer membrane vesicles as they become smaller. Frequently, the decrease in size can be accompanied by a change from rod-shaped cells to coccoid-shaped cells, as discussed next.

Morphological changes

Along with a reduction in size, some bacteria (e.g., Arthrobacter) change from rod-shaped cells to coccoid cells in stationary phase. Similar morphological changes from rods to small coccoid shapes can occur upon starvation in several other bacteria, including Klebsiella, Escherichia, Vibrio, and Pseudomonas.

Changes in surface properties

When certain marine bacteria are starved, the cell surface becomes hydrophobic and the cells are more adhesive.⁸ Vibrio synthesizes surface fi brils and forms cell aggregates when starved for a long time. These changes in the surfaces of starved cells make them more adhesive.

Presumably, it is advantageous for nutrient-limited cells to adhere to particles that have adsorbed nutrient on the surfaces.

Changes in membrane phospholipids

In E. coli all the unsaturated fatty acids in the membrane phospholipids become converted to the cyclopropyl derivatives by means of the methylation of the double bonds. The advan-tage to the cyclopropane fatty acids is not known, since it does not appear that mutants unable to synthesize cyclopropane fatty acids are at a survival disadvantage when faced with environmental stresses such as starvation and high or low oxygen tension.

Changes in metabolic activity

When bacteria enter stationary phase, their overall metabolic rate slows. In addition, during starvation many bacteria experience a signifi -cant increase in the turnover (metabolic break-down and resynthesis) of protein and RNA.

Presumably, in starved cells the protein and ribosomal RNA can serve as an energy source to maintain viability and crucial cell functions.

The latter would include solute transport sys-tems, an energized membrane, and ATP pool levels.

case of certain bacteria, the cells may sporulate or encyst. (See Section 23.3 for a description of sporulation in B. subtilis.)

Lately, increased attention is being given to physiological changes that occur in bacteria that enter the stationary phase when they are experimentally subjected to starvation.^2–5 These bacteria undergo physiological changes that result in metabolically less active cells that are more resistant to environmental hazards. This property has been associated for a long time with bacteria that sporulate or form desicca-tion-resistant cysts upon nutrient deprivation.

The spores and cysts represent metabolically inactive or less active stages of the life cycle of the organism. When nutrients become avail-able once more, the spores and cysts germi-nate into vegetative cells that grow and divide.

More recently, it has been discovered that when faced with nutrient deprivation, even some bacteria that do not form spores (e.g., E. coli, Salmonella, Vibrio, Pseudomonas) undergo adaptive changes and thereupon enter station-ary phase. Some of these changes also result in resistant, metabolically less active cells, and such responses may be common in most if not all bacteria. However, not all the effects can be rationalized in terms of survivability, and their physiological role is not yet known. Some of the changes that occur in cells that are starved are described next. For a review of starvation in bacteria, see ref. 6.

Changes in cell size

As discussed earlier, cells that enter stationary phase upon carbon exhaustion generally become smaller because they undergo reductive divi-sion; that is, they keep dividing for 1 to 2 h after growth has ceased. Reductive division results in the production of more cells, which may be advantageous for dispersing the population. In some cases there may be several cell divisions in the absence of growth so that the cell size differs radically from what is found in the growing cell.

Some bacteria decrease in length from several micrometers (e.g. 5–10 μm) to approximately 1 to 2 μm or even less. (See note 7.) As discussed in ref. 1, in addition to undergoing size reduc-tion due to continued division in the absence of growth, bacteria can become smaller during starvation after reductive division has been com-pleted. The additional decrease in size results

starvation sigma factor. It is also known as σ³⁸ because it has a molecular weight of 38 kDa.

Sometimes called the “master regulator of the stationary phase response,” RpoS has been found in other gram-negative bacteria in addi-tion to E. coli, but not in gram-positive bacte-ria.¹² Gram-positive bacteria use a different sigma factor (σ^B) to regulate the transcription of genes homologous to some of the σ^s -dependent genes in E. coli.¹³

Proteins whose synthesis depends upon σ^s include a catalase made during stationary phase (HP II), which presumably accounts for resistance to H₂O₂, an exonuclease III that can repair DNA damage due to H₂O₂ and near-UV radiation, and an acid phosphatase. Because wild-type rpoS mutants do not have station-ary phase resistance properties (measured as percentage survival) to heat, high salt (osmotic shock), near-UV, H₂O₂, or prolonged starva-tion (e.g., several days of starvastarva-tion for carbon or nitrogen), it has been concluded that the products of rpoS-dependent genes are necessary for survival under stressful conditions during this phase.⁸ The transcription of the rpoS gene (measured by using a rpoS–lacZ fusion plas-mid) increases dramatically in a nutrient-rich medium as cells approach and enter stationary phase and during starvation for specifi c nutri-ents such as phosphate.^14,15 (See note 16 for a description of lacZ gene fusions.) Homologues to rpoS exist in other gram-negative bacteria, including pathogens. Mutations in rpoS in the intestinal pathogen Salmonella result in the attenuation of virulence in mice.^17,18 One might suppose that RpoS aids pathogens to survive stress such as oxidative stress, which they might encounter inside host cells, or pH stress (e.g., in the stomach).

2. RpoS is a global regulator that is important for stress responses in exponentially growing cells as well as in stationary phase

The RpoS subunit of RNA polymerase should be viewed as a global regulator that also functions during responses to stress during exponential growth.¹⁹ Increased RpoS levels during expo-nential growth can be caused by slow growth, temperature upshift from 30 °C to 42 °C, and high osmolarity. For example, during osmotic upshifts in exponentially growing cultures RpoS levels increase. Such cultures become not Changes in protein composition

Bacteria may synthesize 50 to 70 or more new proteins under conditions of carbon, nitrogen, or phosphate starvation. Many of the proteins serve specifi c functions related to the nutrient that is in low concentration. For example, phos-phate starvation induces the synthesis of PhoE porin, which is an outer membrane channel for anions, including phosphate. This helps the bacterium bring in more phosphate. Another example is the nitrogen fi xation genes that are induced when the cells are starved for nitrogen.

In addition to the proteins of known function, there are many proteins made under all condi-tions of starvation for which there is presently no known function, although many of the pro-teins are presumed to be involved in resistance to stress.

Changes in resistance to environmental stress

Cells entering stationary phase also become more resistant to environmental stresses such as high temperatures, osmotic stress, and cer-tain chemicals. For example, E. coli reportedly becomes more resistant in stationary phase to high temperature, hydrogen peroxide, high salt, ethanol, solvents such as acetone and toluene, and acidic or basic pH. These resistant prop-erties are due to the synthesis of a starvation sigma factor that has four names: RpoS, σ^s, σ³⁸, and KatF.

1. The rpoS gene in E. coli encodes a sigma factor required for transcription of genes expressed during stationary phase and starvation

The reasons for concluding that rpoS encodes a sigma factor are as follows: (1) the predicted amino acid sequence of the rpoS gene product, based upon nucleotide sequence data, suggests that it is a sigma factor, and (2) studies with purifi ed protein have confi rmed that the protein binds to RNA polymerase and directs transcrip-tion of rpoS-dependent genes. RpoS is required for the transcription of a regulon that includes at least 50 genes that encode proteins induced by carbon starvation when cells enter station-ary phase; it is also necessstation-ary for the transcrip-tion of several genes whose activities increase during phosphate and nitrogen starvation.^9–12 Accordingly, the protein is also called σ^s, for

limitations of carbon, phosphate, or nitrogen sources. How bacteria survive such nutritional stress is fundamental to an understanding of their physiology in natural ecosystems.

Th e control of ribosomal RNA synthesis There is much controversy surrounding the question of what regulates the rate of synthesis of ribosomes, although it is clear that ribosome synthesis is coupled to growth rates (Section 2.2.3).²⁰ As shown in Fig. 2.4, faster growing cells have more ribosomes (refl ected in cellular RNA) per unit cell mass. For example, the num-ber of ribosomes in an E. coli cell can vary from fewer than 20,000 to about 70,000 depending upon the growth rate. In addition to growth rate control, amino acid starvation results in the inhibition of ribosome synthesis. Probably sev-eral factors are involved in regulating rates of ribosome synthesis. For example, Fis, a DNA-binding protein whose synthesis is increased in rich media, positively regulates transcription of rRNA genes. Another DNA-binding protein, H-NS, antagonizes Fis stimulation (see later).

A very important regulator is guanosine tetra-phosphate (ppGpp), which increases when cells are subjected to amino acid starvation or carbon and energy limitation, and slows the synthesis of rRNA (and tRNA). The global regulator ppGpp is best known as the effector that inhib-its rRNA and tRNA synthesis during the strin-gent response, an effect originally discovered as the inhibition of rRNA and tRNA synthesis due to amino acid starvation.²¹ See note 22 for other physiological effects of ppGpp and ref. 23 for a recent review.

1. The stringent response

The stringent response in E. coli is a temporary inhibition in the synthesis of ribosomal RNA and transfer RNA when the cells are shifted to a medium in which they are starved for an amino acid. It works in the following way. When cells are shifted from a medium rich in amino acids to a minimal medium, “uncharged”

tRNA accumulates, causing an insuffi ciency of aminoacylated tRNA, which in turn results in

“idling” or “stalled” ribsomes. An enzyme on the “stalled” ribosomes called RelA [(p)ppGpp synthetase I, or PSI], which is the product of the relA gene, becomes activated by uncharged tRNA that binds to the A site. Activated RelA only osmotolerant but also thermotolerant and

resistant to H₂O₂.

3. Regulation of RpoS levels can be at the transcriptional, post-transcriptional, and post-translational levels

Interestingly, the reason for the increased lev-els of RpoS during stress responses is not sim-ply increased transcription of the rpoS gene.

The levels of RpoS are regulated not only at the transcriptional level but also at the post-tran-scriptional and post-translational levels. Thus, the regulation of RpoS levels is very complex.

For example, the rate of degradation of RpoS decreases when cells enter stationary phase.

The half-life of RpoS in exponential cultures of E. coli growing at 37 °C is less than 2 min.

However, the half-life is greater than 30 min when cells enter stationary phase. See Box 2.1 and the references therein for discussion of how the levels of RpoS are adjusted, including tran-scriptional regulation of the rpoS gene, regula-tion of translaregula-tion of the RpoS messenger RNA, and the role of proteases in the degradation of RpoS.

Other regulators for stationary phase gene expression

Although σ^s is certainly a major transcription factor for gene expression during station-ary phase, additional regulatory factors exist (reviewed in ref. 5). This conclusion is sug-gested partly because the variation in the tim-ing of expression durtim-ing stationary phase of σ^s-dependent genes, depending upon the gene in question, clearly points to regulatory fac-tors in addition to σ^s. In fact, there are several

Although σ^s is certainly a major transcription factor for gene expression during station-ary phase, additional regulatory factors exist (reviewed in ref. 5). This conclusion is sug-gested partly because the variation in the tim-ing of expression durtim-ing stationary phase of σ^s-dependent genes, depending upon the gene in question, clearly points to regulatory fac-tors in addition to σ^s. In fact, there are several