Diversion of the Metabolic Flux from Pyruvate Dehydrogenase to Pyruvate Oxidase Decreases Oxidative Stress during Glucose Metabolism in Nongrowing Escherichia coli Cells Incubated under Aerobic, Phosphate Starvation Conditions

Abstract

Ongoing aerobic metabolism in nongrowing cells may generate oxidative stress. It is shown here that the levels of thiobarbituric acid-reactive substances (TBARSs), which measure fragmentation products of oxidized molecules, increased strongly at the onset of starvation for phosphate (P_i). This increase in TBARS levels required the activity of the histone-like nucleoid-structuring (H-NS) protein. TBARS levels weakly increased further in ΔahpCF mutants deficient in alkyl hydroperoxide reductase (AHP) activity during prolonged metabolism of glucose to acetate. Inactivation of pyruvate oxidase (PoxB) activity decreased the production of acetate by half and significantly increased the production of TBARS. Overall, these data suggest that during incubation under aerobic, P_i starvation conditions, metabolic flux is diverted from the pyruvate dehydrogenase (PDH) complex (NAD dependent) to PoxB (NAD independent). This shift may decrease the production of NADH and in turn the adventitious production of H₂O₂ by NADH dehydrogenase in the respiratory chain. The residual low levels of H₂O₂ produced during prolonged incubation can be scavenged efficiently by AHP. However, high levels of H₂O₂ may be reached transiently at the onset of stationary phase, primarily because H-NS may delay the metabolic shift from PDH to PoxB.

Escherichia coli possesses numerous mechanisms aimed at protecting and repairing cell constituents exposed to toxic agents. Defense mechanisms in growing cells are generally transient responses that allow rapid adaptation to stress. However, in nature, bacteria may alternate between growing and nongrowing states because of nutritional starvation. Survival of starved cells poses a specific problem because of the low capacity of stationary-phase cells to synthesize proteins. Starved cells can at least in part resolve this problem by accumulating RpoS (σ^s) at the entry into stationary phase. The induction of the numerous genes of the RpoS regulon helps to protect nongrowing cells against various stresses (14, 16).

Moreover, it has been suggested that metabolic fluxes may be redistributed in starved cells, which may decrease the endogenous production of reactive oxygen species (ROS). For instance, cells incubated under aerobic, glucose starvation conditions shift from aerobic to fermentative metabolism in an ArcA-dependent manner (e.g., pyruvate dehydrogenase synthesis is decreased, whereas pyruvate formate-lyase synthesis is increased) (24). Such a shift may decrease the production of NADH and the respiratory chain-mediated generation of ROS (24, 28). In the aerobic respiratory chain, NADH dehydrogenase II produces superoxide and hydrogen peroxide (H₂O₂) by autooxidation of its reduced flavin adenine dinucleotide cofactor (20). H₂O₂ is primarily detoxified by the alkyl hydroperoxide reductase (AHP) complex (AhpCF), which efficiently scavenges low levels of H₂O₂ but is inefficient against high levels of H₂O₂ (27).

In contrast to cells incubated under aerobic, glucose starvation conditions which shift from aerobic to anaerobic metabolism, cells incubated under aerobic, P_i starvation conditions may continue to metabolize glucose through aerobic metabolism, which may generate ROS (11). The cellular levels of ROS, which are not diluted by growth any more, may hence steadily increase in P_i-starved cells. This idea is supported by the finding that the colony-forming ability of ahp mutants decreases after 2 days of incubation under aerobic, P_i starvation conditions (22). However, these data might indicate that P_i-starved cells are primarily exposed to ROS at the exit of rather than during stationary phase (5). It has been shown that ahp mutants accumulate DNA damage and fragmentation products of oxidized molecules, measured as thiobarbituric acid-reactive substances (TBARSs), when they exit stationary phase and enter the lag phase in fresh Luria broth (LB) medium (12). Therefore, TBARS measurements were used in this study to determine whether oxidative stress occurs in cells when they are incubated under aerobic, P_i starvation conditions.

MATERIALS AND METHODS

Bacterial strains.

The Escherichia coli K-12 strains are described in Table 1. Mutations were introduced into strains by P1 transduction (21). For strains carrying the leuB and argE3(Oc) mutations, cultures were routinely checked for the absence of Leu⁺ contaminants and Arg⁺ suppressors, which can grow in old cultures by consuming nutrients excreted into the medium (11).

TABLE 1.

Bacterial strains

Strain	Genotype	Source or reference
ENZ1203	F−thr-1(Am) leuB6 Δ(argF-lac)U169 glnV44 sulA211 hisG4(Oc) rpsL31 argE3(Oc) thi-1 rpoS+	22
ENZ1257	ENZ1203 sulA::lacZ	22
ENZ1399	ENZ1257 ΔahpCF::kan	22
ENZ1643	ENZ1257 poxB176::lacZ Cmr	P1(YYC912) × ENZ1257
ENZ1645	ENZ1203 poxB176::lacZ Cmr	P1(YYC912) × ENZ1203
ENZ1646	ENZ1203 ΔahpCF::kan	P1(ENZ1399) × ENZ1203
ENZ1667	ENZ1203 hns::neo	P1(MC4100hns) × ENZ1203
MC4100hns	hns::neo	30
YYC912	poxB176::lacZ Cmr	7

Media and culture conditions.

The minimal medium used for liquid cultures was essentially the morpholinepropanesulfonic acid (MOPS) medium described by Neidhardt et al. (23), which notably contained 40 mM MOPS, 86 mM NaCl, 9.5 mM NH₄Cl, 5 mM K₂HPO₄ and 20 mM glucose, supplemented with five vitamins (0.02 mM thiamine, 0.02 mM calcium pantothenate, 0.02 mM p-aminobenzoic acid, 0.02 mM p-hydroxybenzoic acid, and 0.02 mM 2,3-dihydroxybenzoic acid) and six amino acids (0.8 mM leucine, 0.4 mM threonine, 0.2 mM histidine, 0.4 mM arginine, 0.4 mM isoleucine, and 0.6 mM valine) (pH 7.2). The P_i- and glucose-limiting media contained 0.1 mM K₂HPO₄ plus 9.8 mM KCl and 3 mM glucose, respectively (11). Strains were grown in MOPS medium for 24 h, diluted 1:200 into MOPS medium at time zero, and incubated further with aeration in a water bath rotary shaker (Aquatron Infors HT, 150 rpm) at 37°C, unless otherwise indicated. Culture optical density at 600 nm (OD₆₀₀) was measured with a Jasco V-530 spectrophotometer in cells of 1-cm path length. The pH of the media was determined at ≈25°C.

Levels of glucose and by-products.

The concentrations of glucose (glucose HK assay; Sigma), pyruvate (Roche), and glutamate, formate, acetate, and d-and l-lactate (Boehringer Mannheim/R-Biopharm) were assayed with enzymatic tests according to the instructions of the manufacturers.

Determination of TBARS concentration.

Cells were centrifuged at 12,000 × g for 15 min, suspended in phosphate buffer (pH 7.0) containing 0.1 mM butylated hydroxytoluene (Sigma) as an antioxidant, and disrupted by one passage through a French pressure cell at 10⁸ Pa. The extract was clarified by centrifugation at 27,000 × g for 20 min, and the supernatant was centrifuged for 90 min at 230,000 × g (k = 96, rotor type 70.1Ti; Beckman) at 4°C. The pellet fraction, which contained cytoplasmic membrane vesicles, was suspended in phosphate buffer containing 0.1 mM butylated hydroxytoluene, precipitated with 10% (wt/vol) trichloroacetic acid, incubated with 4.6 mM thiobarbituric acid (Sigma) in 50% (vol/vol) glacial acetic acid plus 0.1 mM butylated hydroxytoluene for 1 h at 95°C, mixed with 1 ml of butanol, and centrifuged at 11,000 × g for 3 min. The absorbance of the butanol phase was determined at 532 nm. The concentration of TBARS (ɛ₅₃₂ = 156 mM⁻¹ cm⁻¹) was expressed in nanomoles per milligram of protein (6, 13). Protein content was determined by the Bradford dye-binding assay (Bio-Rad) with bovine serum albumin as the standard.

RESULTS AND DISCUSSION

TBARS levels increase significantly in cells incubated under P_i starvation conditions.

Cells incubated for 4 days under glucose and P_i starvation conditions exhibited TBARS levels 2.5- and 8-fold higher than those of exponentially growing cells, respectively (Table 2, Fig. 1). These data suggest that ROS levels may eventually increase more strongly in P_i- than in glucose-starved cells.

TABLE 2.

TBARS levels increase in cells incubated for 4 days under aerobic, P_i starvation conditions^a

Medium	Growth phase	OD600	Mean TBARS concn, nmol/mg, ± SD (fold increase)
MOPS	Exponential	0.7	0.06 ± 0.01
Glucose limiting	Stationary	0.7	0.15 ± 0.02 (2.5)
Pi limiting	Stationary	0.8	0.48 ± 0.04 (8)

^aStrain ENZ1257 was inoculated into 100 ml of MOPS (5 mM P_i plus 20 mM glucose), glucose-limiting (5 mM P_i plus 3 mM glucose), or P_i-limiting (0.1 mM P_i plus 20 mM glucose) medium at time zero and incubated further with aeration in 500-ml Erlenmeyer flasks. When cells were either in exponential growth phase or in stationary phase, i.e., on day 4 of incubation under glucose or P_i starvation conditions (starved cells entered stationary phase after ≈10 h of incubation), the OD₆₀₀ of the cultures and TBARS concentrations were determined. TBARS values are the means of two experiments. The values relative to those in exponentially growing cells are indicated in parentheses.

FIG. 1.

TBARS levels as a function of time of incubation under aerobic, P_i starvation conditions in ahp⁺ and ahp mutant cells. Strains ENZ1257 (ahp⁺) (○) and ENZ1399 (ENZ1257 ΔahpCF) (•) were inoculated into 100 ml of P_i-limiting medium (0.1 mM P_i plus 20 mM glucose) (time zero) and incubated under aerobic conditions in 500-ml Erlenmeyer flasks. TBARS concentrations were determined between days 1 and 4 of incubation (OD₆₀₀ ≈ 0.7). The values for exponentially growing cells were obtained from cells grown for 6.5 h (OD₆₀₀ ≈ 0.7) in MOPS medium (5 mM P_i plus 20 mM glucose). The data are from a representative experiment from three separate trials. Variations between experiments were ≤20%. The inset shows the growth curves of ahp⁺ and ahp mutant cells in P_i-limiting medium.

TBARS levels depend on the amount of glucose metabolized in ahp cells.

Inactivation of the ahpCF genes increased by ≈1.9-fold the levels of TBARS in cells incubated for 4 days in P_i-limiting medium initially containing 20 mM glucose (from 0.48 nmol/mg in ahp⁺ cells to 0.90 nmol/mg in ahp mutants; Tables 2 and 3). These data suggest that one role of AHP is to protect cell constituents against ROS produced during P_i starvation.

TABLE 3.

TBARS levels increase as a function of the amount of glucose consumed in ahp mutants starved of P_i for 4 days^a

Initial glucose concn (mM)	No. of expt	OD600	Glucose (mM)	Mean TBARS concn (nmol/mg) ± SD
5	2	0.6	≤0.4	0.52 ± 0.04
10	2	0.7	≤0.4	0.66 ± 0.03
20	4	0.7	≤0.4	0.90 ± 0.07
30	4	0.7	4.2 ± 0.1	1.09 ± 0.10
40	2	0.7	14.3 ± 0.1	0.93 ± 0.10

^aStrain ENZ1399 (ENZ1257 ΔahpCF) was inoculated into 100 ml of P_i-limiting medium (0.1 mM P_i) containing increasing concentrations of glucose, and incubated further under aerobic conditions in 500-ml Erlenmeyer flasks. On day 4 of incubation, the OD₆₀₀ of the cultures, glucose concentrations in the spent medium, and TBARS levels were determined.

TBARS levels measured on day 4 of incubation in ahp mutants increased from ≈0.5 to ≈1.0 nmol/mg as a function of the concentration of glucose added in the range from 5 to 30 mM (Table 3). TBARS levels remained steady at ≈0.5 nmol/mg in ahp⁺ cells incubated under the same conditions (data not shown). Therefore, AHP may help scavenge ROS produced during prolonged incubation in P_i-limiting medium in the presence of excess glucose (≈3 mM glucose was consumed at the entry into stationary phase by ≈10 h of incubation; data not shown).

However, when ahp mutants were incubated for 4 days in P_i-limiting medium initially containing 30 or 40 mM glucose, similar levels of TBARS were produced and similar amounts of glucose were consumed (≈26 mM) (Table 3). Therefore, the ROS generated during incubation in P_i-limiting medium may primarily be produced endogenously through glucose metabolism rather than by autooxidation of glucose present in the medium (27).

Overall, these data suggest that metabolism of glucose in P_i-starved cells may generate ROS. However, consumption of glucose had little effect on biomass formation (OD₆₀₀) (Table 3), which suggests that metabolism may primarily be achieved by overflow metabolism, i.e., by excretion of by-products.

AHP prevents a late increase in TBARS levels.

TBARS levels were measured in ahp⁺ and ahp mutant strains as a function of the time of incubation in P_i-limiting medium containing 20 mM glucose (Fig. 1). TBARS levels increased in roughly two steps. First, a strong increase (≈8-fold) occurred during the transition from exponential growth to stationary phase (day 1 of incubation) in both ahp⁺ and ahp mutant cells. Second, a weaker increase (≈2-fold) preferentially occurred between day 1 and day 4 of incubation in ahp mutants.

The differential roles of AHP suggest that the levels of H₂O₂ reached at the onset of P_i starvation may be too high to be scavenged efficiently by AHP (27). However, between days 1 and 4 of incubation, when the rate of production of H₂O₂ may be reduced, AHP can scavenge the low levels of H₂O₂ accumulated in P_i-starved cells as a result of glucose metabolism.

TBARS levels increase strongly in poxB mutants.

In P_i-starved cells, glucose was primarily metabolized to acetate, which was excreted into the incubation medium. On day 5 of incubation in P_i-limiting medium initially containing 40 mM glucose, acetate levels reached 29.4 mM and the pH of the medium decreased from ≈7.2 to ≈4.8 (Table 4); the concentrations of pyruvate, lactate, aspartate, glutamate and formate in the incubation medium were lower than 0.4 mM (Table 4 and data not shown).

TABLE 4.

Metabolic pattern in cells starved of P_i for 5 days^a

Strain	OD600	Final pH	Mean concn (mM) ± SD
			Glc	Ace	Pyr
ENZ1257	0.7	4.8	11.6 ± 0.2	29.4 ± 1.8	0.31 ± 0.01
ahp	0.7	5.1	13.2 ± 0.6	24.0 ± 0.1	0.08 ± 0.01
poxB	0.8	4.7	14.3 ± 0.6	13.0 ± 0.1	3.9 ± 0.1

^aStrains ENZ1257, ENZ1399 (ENZ1257 ΔahpCF), and ENZ1643 (ENZ1257 poxB) were inoculated into 50 ml of P_i-limiting medium (0.1 mM P_i plus 40 mM glucose) and incubated further under aerobic conditions in 500-ml Erlenmeyer flasks. On day 5 of incubation, the OD₆₀₀ and pH of the cultures and the concentrations in the culture supernatants (adjusted to pH 7) of glucose (Glc), acetate (Ace), and pyruvate (Pyr) were determined. Data are the means of two experiments.

The low concentration of formate in the spent medium indicates that pyruvate formate-lyase played no significant role in the metabolism of pyruvate in P_i-starved cells. Therefore, the metabolism of pyruvate should occur through aerobic enzymes: the pyruvate dehydrogenase (PDH) complex (AceEF Lpd) or pyruvate oxidase (PoxB) or both (10).

The PDH complex is primarily used during aerobic growth to catalyze the oxidative decarboxylation of pyruvate to acetyl coenzyme A (acetyl-CoA) with the concomitant reduction of NAD. Since the rate of synthesis of the AceF subunit of the PDH complex decreases by approximately twofold at the onset of P_i starvation (29), a decrease in PDH activity might account at least in part for the apparent decrease in the production of H₂O₂ during prolonged incubation under P_i starvation conditions.

PoxB can catalyze the oxidative decarboxylation of pyruvate directly to acetate with the concomitant reduction of flavin adenine dinucleotide and ubiquinone, bypassing NADH oxidation by the respiratory chain (10). It has been suggested that PoxB, which is strictly regulated by RpoS, could play a beneficial role at very low growth rates and at the entry into stationary phase under microaerobic conditions, when neither PDH nor pyruvate formate-lyase can function efficiently (1, 7).

Inactivation of poxB caused significant changes in the metabolic pattern of cells incubated for 5 days in P_i-limiting medium initially containing 40 mM glucose (Table 4). Whereas the total amount of glucose consumed by poxB mutants (25.7 mM) was only slightly lower than that consumed by poxB⁺ cells (28.4 mM), the amount of acetate excreted into the culture medium by poxB mutants (13.0 mM) was 2.2-fold lower than that excreted by poxB⁺ cells (29.4 mM) (the concentrations of acetate remained steady between days 4 and 6 in poxB mutants, whereas they continued to increase for 6 days up to ≈33 mM in poxB⁺ cells; data not shown). In contrast, the amount of pyruvate excreted by poxB mutants was 12.5-fold higher than that excreted by poxB⁺ cells (Table 4).

These data suggest that (i) at least half of the pyruvate that is metabolized to acetate in P_i-starved cells is metabolized through PoxB rather than through the PDH complex-phosphotransacetylase-acetate kinase pathway and (ii) part of the pyruvate that is not metabolized in poxB mutants is excreted directly into the medium.

Inactivation of poxB caused a dramatic increase in TBARS levels during prolonged incubation in P_i-limiting medium initially containing 30 mM glucose (Table 5). Whereas poxB⁺ and poxB mutant cells exhibited similar levels of TBARS during the exponential growth phase (0.06 nmol/mg) and on day 1 of incubation (≤0.45 nmol/mg), TBARS levels increased sharply to 1.79 nmol/mg on day 4 of incubation in poxB mutants, while TBARS levels barely increased in poxB⁺ cells (0.46 nmol/mg) (Table 5). Under the same conditions (on day 4 of incubation in P_i-limiting medium containing 30 mM glucose), TBARS levels increased to 1.09 nmol/mg in ahp mutants (Table 3).

TABLE 5.

TBARS levels in poxB mutants^a

Strain	Mean TBARS concn (nmol/mg) ± SD
	Exponential	Day 1	Day 4
ENZ1203	0.06 ± 0.01	0.41 ± 0.01	0.46 ± 0.01
poxB	0.06 ± 0.01	0.45 ± 0.03	1.79 ± 0.13

^aStrains ENZ1203 and ENZ1645 (ENZ1203 poxB) were inoculated into 100 ml of MOPS containing 30 mM glucose and 5 mM P_i (exponential) or 0.1 mM P_i (day 1 and day 4). Data are the means of two experiments.

Overall, these data suggest that in poxB mutants, although part of the pyruvate that is not metabolized through PoxB is excreted, part may be metabolized through the PDH complex in addition to the amount normally metabolized in poxB⁺ cells, which may increase the production of NADH and eventually the production of H₂O₂ by the respiratory chain. Therefore, by diverting part of the metabolic flux from the PDH complex, PoxB may protect P_i-starved cells against oxidative stress.

H-NS modulates the production of TBARS.

During the exponential growth phase, synthesis of the PDH complex is decreased, whereas synthesis of PoxB is increased in hns mutants compared to the hns⁺ parents (2, 15, 19). Therefore, a role in growing cells of the histone-like nucleoid-structuring protein H-NS may be to increase PDH activity (which gives rise to NADH and acetyl-CoA) and to decrease PoxB activity (which gives rise directly to acetate and eventually to acetyl-CoA through acetate kinase and acetyl-CoA synthetase, both activities requiring ATP) (1, 18). This specific role of H-NS, which may result primarily from its capacity to decrease RpoS levels in exponential phase (4), may help growing cells to sustain an efficient energetic metabolism and a rapid growth rate. In fact, the generation time of hns mutants was 3.5-fold higher than that of hns⁺ cells in MOPS medium (4) (Table 6). However, it was rather surprising that TBARS levels were threefold higher in hns mutants than in hns⁺ cells during the exponential growth phase (Table 6). This result may be explained at least in part by the recent finding that overexpression of poxB in PDH⁻ mutants, which mimic the behavior of hns mutants, increases by ≈2-fold the specific rate of consumption of oxygen during growth (1). This is probably required to compensate for the higher consumption of ATP needed to metabolize further PoxB-derived acetate to acetyl-CoA (1). Higher activity of the tricarboxylic acid cycle enzymes and of the aerobic respiratory chain (8, 10) may eventually increase the production of ROS during growth in hns mutants.

TABLE 6.

H-NS modulates the production of TBARS^a

Strain	Pi (mM)	Growth phase (h of growth)	Glucose consumed (mM)	Mean TBARS concn, nmol/mg, ± SD (fold increase)
ENZ1203	5	Exponential (7)	3.3	0.05 ± 0.01
ENZ1203	0.1	Stationary (24)	8.1	0.41 ± 0.02 (8.2)
hns	5	Exponential (22)	2.1	0.15 ± 0.02 (3)
hns	0.1	Stationary (47)	9.8	0.20 ± 0.03 (4)

^aStrains ENZ1203 and ENZ1667 (ENZ1203 hns::neo) were inoculated (125 μl from overnight cultures in LB medium) into 100 ml of MOPS (5 mM P_i plus 20 mM glucose) or P_i-limiting medium (0.1 mM P_i plus 20 mM glucose) at time zero. Exponentially growing cells in MOPS medium were taken at an OD₆₀₀ of 0.7 (hns⁺; time = 7 h; generation time ≈ 52 min) or at an OD₆₀₀ of 0.4 (hns mutants; time = 22 h; generation time ≈ 3 h). TBARS values are the means of two experiments. The values relative to those in hns⁺ cells in the exponential growth phase are indicated in parentheses.

Whereas H-NS activity decreased TBARS levels during growth, H-NS activity increased TBARS levels at the onset of P_i starvation. TBARS levels increased strongly at the entry into stationary phase in hns⁺ cells (Table 6), while they barely increased in hns mutants (Table 6), although similar amounts of glucose were eventually consumed by both strains. These data may be explained simply by the hypothesis that at the entry into stationary phase, pyruvate metabolism occurs through the PDH complex in hns⁺ cells, whereas part of the metabolism may have already occurred through PoxB in hns mutants.

Overall, these data suggest that at the onset of P_i starvation, H-NS delays the shift from PDH to PoxB, which prolongs the production of NADH by the PDH complex and the production of ROS by NADH dehydrogenase II. The ≈4-fold increase in the rate of synthesis of H-NS at the onset of P_i starvation (29) may thus help to increase oxidative stress in P_i-starved cells.

Conclusion.

The data presented here provide the outlines to an understanding of how oxidative stress occurs in P_i-starved cells during the metabolism of glucose to acetate (Fig. 2). At the onset of stationary phase, H-NS may help to sustain a high metabolic flux through the PDH complex, which generates NADH and eventually ROS as a result of the activity of NADH dehydrogenase II in the respiratory chain. However, the rate of production of NADH may decrease rapidly, primarily because of diversion of the metabolic flux from PDH to PoxB. The residual low levels of H₂O₂ generated during prolonged metabolism of glucose may be efficiently scavenged by AHP.

FIG. 2.

Diagram showing the proposed pathways for the metabolism of pyruvate in P_i-starved cells. Abbreviations: Q, ubiquinone; G6P, glucose 6-phosphate; cyt, cytochrome quinol oxidase; FAD, flavin adenine dinucleotide; PEP, phosphoenolpyruvate; PDHc, PDH complex; Pta, phosphotransacetylase; AckA, acetate kinase.

There is evidence that the PoxB shunt-AHP antioxidant activity integrated system may be specific to P_i-starved cells. First, even though glucose-starved cells accumulate RpoS and induce the expression of poxB (although at levels approximately twice lower than in P_i-starved cells; unpublished data), metabolism of glucose could occur primarily through an aerobic phosphoenolpyruvate-glyoxylate cycle (9), which may generate ROS (Table 2) (22), and eventually fermentative pathways, which may prevent the generation of ROS (17, 24). However, the increase in oxidative protein damage in glucose-starved cells results primarily from the synthesis of aberrant proteins rather than from an increase in ROS levels (3). Second, nitrogen-starved cells accumulate low levels of RpoS and poorly induce poxB (unpublished data). However, nitrogen-starved cells could prevent the generation of ROS by diverting glucose metabolism towards products such as glycogen (26). Finally, sulfur-starved cells may be exposed to oxidative stress as a result of the reduction of Fe³⁺ to Fe²⁺ by cystine, which increases the production of hydroxyl radicals through the Fenton reaction (25). Therefore, it appears that each type of metabolic perturbation may generate oxidative stress through different mechanisms, and cells may respond to these perturbations by specific strategies which are just beginning to be revealed.