In situ studies on AMP deaminase as a control system of the adenylate energy charge in yeasts

The role of AMP deaminase reaction in the stabilization of the adenylate energy charge was investigated using permeabilized yeast cells. The addition of P_i or Zn²⁺, which inhibits AMP deaminase, remarkably retarded the depletion of total adenylate pool and the recovery of the adenylate energy charge. Polyamine, an activator of the enzyme, decreased total adenylates, resulting in the enhanced recovery of the energy charge in situ. AMP deaminase can act as a regulatory enzyme in the system that stabilizes the adenylate energy charge in yeast cells under the conditions of severe metabolic stress.     

Regulation of cytosol 5'-nucleotidase by adenylate energy charge

In the physiological range of the adenylate energy charge in liver (0.7-0.9), th rate of AMP-hydrolysis catalysed by rat liver cytosol 5'-nucleotidase (5'-ribonucleotide phosphohydrolase, EC 3.1.3.5) increased sharply with decreasing energy charge. In addition, a decrease in the concentration of Pi caused marked acceleration of the AMP-hydrolysing activity over the physiological range of adenylate energy charge. These responses seem to serve to protect the cells against a metabolic stress which could result from sudden utilization of ATP by removal of AMP. The AMP-hydrolysing activity of this enzyme decreased sharply as the size of the adenine nucleotide pool decreased in the physiological range. This effect may be a self-limiting response to prevent excess depletion of the pool. IMP-hydrolysing activity of this enzyme increased with increasing adenylate energy charge. But no marked response to its variation within the physiological range was observed. On the basis of the data obtained in this study, the IMP-hydrolysing activity of the cytosol 5'-nucleotidase in rat liver cells seems to be comparable to that of AMP deaminase reaction, but the AMP-hydrolysing activity was estimated to be less than 10% of AMP deaminase reaction at energy charge value of about 0.7. This strongly suggests that the AMP leads to IMP leads to inosine pathway is more significant that the AMP leads to adenosine leads to inosine pathway in rat liver.     

Effects of sterol manipulation on microsomal desaturase activities in Tetrahymena: with regard to thermal acclimation

The role of NH+4 ion and AMP deaminase reaction in the activation of phosphofructokinase with respect to its response to the adenylate energy charge was investigated using permeabilized yeast cells. (a) Phosphofructokinase and AMP deaminase were activated by the decrease in the adenylate energy charge. The addition of NH+4 further stimulated the phosphofructokinase activity in the presence of intracellular level of K+, and the optimal energy charge value giving the maximal response of the enzyme was shifted from 0.3 to the value above 0.5. (b) The increase in NH+4 ion produced through the activation of AMP deaminase by spermine which shows no direct action on the phosphofructokinase activity can activate phosphofructokinase with shift of the optimal energy charge value of the enzyme to 0.5 in the presence of K+, whereas the optimal energy charge value for AMP deaminase reaction was not affected by the addition of spermine. Phosphofructokinase can be activated most effectively by the physiological decrease in the energy charge under the condition of increased NH+4 in the presence of K+. The possibility that the interaction of phosphofructokinase with AMP deaminase under hypoxic condition might be a contributing factor to the Pasteur effect is discussed.     

Adenosine triphosphate-activated adenylate deaminase from marine invertebrate animals. Properties of the enzyme from lugworm (Arenicola cristata) body-wall muscle.

Adenylate deaminase (AMP aminohydrolase, EC 3.5.4.6) from lugworm (Arenicola cristata) body-wall muscle was partially purified by extraction in KCl solutions and chromatography on phosphocellulose. Enzyme activity was eluted from the column at two salt concentrations. Both forms show co-operative binding of AMP (Hill coefficient, h, 2.85) with s0.5 values of 20 mM and 15.6 mM. ATP and ADP act as positive effectors lowering h to 1.07 and s0.5 to 2mM. The apparent Ka (activation) for ATP was 1.5mM. GTP is an inhibitor with an apparent Ki of 0.12 mM. In vivo the ATP-activated adenylate deaminase is in the active form and may be regulated by changes in GTP concentrations. Adenylate deaminase may act as a primary ammonia-forming enzyme in ammonotelic marine invertebrates with the purine nucleotide cycle.     

The regulatory role of spermine and fatty acid in the interaction of AMP deaminase with phosphofructokinase

The role of fatty acid and polyamine in the interaction of AMP deaminase (EC 3.5.4.6)-ammonium system with glycolysis was investigated using permeabilized yeast cells. (1) The addition of fatty acid inhibited the activity of AMP deaminase in situ, resulting in a decrease in the total adenylate pool depletion, and in the recovery of the adenylate energy charge. (2) The addition of fatty acid resulted in an indirect decrease in the activity of phosphofructokinase (EC 2.7.1.11) through a reduced level of ammonium ion; fatty acid itself did not inhibit phosphofructokinase activity in the presence of excess ammonium ion. (3) Spermine protected AMP deaminase from inhibition by fatty acid: the increased ammonium level enhanced phosphofructokinase activity, glycolytic flux and the recovery of the energy charge. In contrast, alkali metals, which are also activators of AMP deaminase had little effect on the inhibition of the enzyme. The inhibition of glycolysis by fatty acid and its reversal by polyamine can be accounted for by the changes in ammonium ion through the action of AMP deaminase-ammonium system, and the physiological relevance is discussed.     

Major determinants of purine excretion from human lymphoblasts

WI-L2 B lymphoblasts deficient in hypoxanthine-guanine phosphoribosyltransferase (HGPRT) excreted amounts of hypoxanthine two to three times larger than CEM T lymphoblasts deficient in HGPRT, despite similar growth rates. ATP consumption occurred at a higher rate in WI-L2 cells than in CEM cells when cultivated in a glucose-free buffer, because of higher RNA synthesis in WI-L2 cells. The introduction of actinomycin D and azaserine resulted in lower hypoxanthine excretion in WI-L2 cells than in CEM cells, not in parallel with changes of the adenylate pool size. When the energy charge was high, de novo purine synthesis was a major determinant for purine excretion. The adenylate pool ratio (AMP/ATP) change caused by the introduction of oligomycin was greater during ATP depletion and recovery in WI-L2 cells than in CEM cells. WI-L2 cells were observed to have AMP deaminase activity three to four times higher than CEM cells. The major component of AMP deaminase in these cells was liver type. The higher rate of RNA synthesis caused greater changes of (AMP/ATP) and required higher AMP deaminase activity for recovery. When the energy charge was low, AMP deaminase was a major determinant for purine excretion.     

Regulation of platelet AMP deaminase activity in situ.

The regulation of platelet AMP deaminase activity by ATP, GTP and phosphate was studied in human platelets in situ, and in vitro after partial purification. In intact platelets, a similar 50% decrease in cytosolic ATP was induced by either glucose starvation or treatment with H2O2. During starvation, AMP deaminase was in the inhibited state, as ATP consumption was mostly balanced by the accumulation of AMP. During H2O2 treatment, however, the enzyme was in the stimulated state, as the AMP formed was almost completely deaminated to IMP. Cytosolic GTP fell by 40-50% in both starvation and H2O2 treatment. In contrast, intracellular phosphate was 4-5-fold higher in starved than in H2O2-treated cells. These data point to phosphate as the main regulator of AMP deaminase activity in situ. This conclusion was verified by kinetic analysis of partially purified AMP deaminase. At near-physiological concentrations of MgATP, MgGTP and phosphate, the S0.5 (substrate half-saturation constant) for AMP was 0.35 mM. Half-maximal stimulation by MgATP occurred at a concn. between 2 and 3 mM. This stimulation was antagonized by the inhibitory effects of phosphate (IC50 = 2.0 mM) and MgGTP (IC50 = 0.2-0.3 mM), which acted in synergism (IC50 is the concentration causing 50% inhibition). We conclude that the difference in adenylate catabolism between starved and H2O2-treated platelets is due to the distinct phosphate concentrations. During starvation, refeeding and H2O2 treatment, the values of the adenylate charge and the phosphorylation potential were kept closely co-ordinated, which may be effected by AMP deaminase.     

AMP deaminase reaction as a control system of glycolysis in yeast. Role of ammonium ion in the interaction of phosphofructokinase and pyruvate kinase activity with the adenylate energy charge

The role of ammonium ion and AMP deaminase (EC 3.5.4.6) reaction in the activation of phosphofructokinase (EC 2.7.1.11) and pyruvate kinase (EC 2.7.1.40) by the decrease in the adenylate energy charge was investigated using permeabilized yeast cells. Response of AMP deaminase, phosphofructokinase, and pyruvate kinase to variation in the energy charge is typical of the ATP-regenerating enzymes: an activation with the decrease in the energy charge under the in situ conditions. The addition of polyamine activated AMP deaminase in situ, resulting in the subsequent increase in ammonium production, which can stimulate the phosphofructokinase activity with the increase in the optimal energy charge value giving maximal activity of the enzyme. The optimal energy charge value of phosphofructokinase was 0.2-0.25 in the absence of ammonium ion and was shifted to the value above 0.5 by the addition of ammonium ion, whereas Pi, an activator of the enzyme showed little effect on the increase in the optimal energy charge value. The optimal energy charge value of AMP deaminase and pyruvate kinase was not affected by the addition of their effectors. Modulation of the response to the energy charge of phosphofructokinase and pyruvate kinase was analyzed in terms of the "activation coefficient," which was defined as the ratio of the activity at the energy charge of 0.6 to that at the value of 0.9. Activation of phosphofructokinase by the physiological decrease in the energy charge (0.9 to 0.6) can be enhanced by the increase in ammonium ion specifically, although the coefficient of pyruvate kinase remained unaffected by ammonium ion. These results suggest that the AMP deaminase reaction as an ammonium-forming reaction can participate in a key role in the stimulation of phosphofructokinase or glycolytic flux in cells.     

Negative modulation of Escherichia coli NAD kinase by NADPH and NADH.

NAD kinase was purified 93-fold from Escherichia coli. The enzyme was found to have a pH optimum of 7.2 and an apparent Km for NAD+, ATP, and Mg2+ of 1.9, 2.1, and 4.1 mM, respectively. Several compounds including quinolinic acid, nicotinic acid, nicotinamide, nicotinamide mononucleotide, AMP, ADP, and NADP+ did not affect NAD kinase activity. The enzyme was not affected by changes in the adenylate energy charge. In contrast, both NADH and NADPH were potent negative modulators of the enzyme, since their presence at micromolar concentrations resulted in a pronounced sigmoidal NAD+ saturation curve. In addition, the presence of a range of concentrations of the reduced nucleotides resulted in an increase of the Hill slope (nH) to 1.7 to 2.0 with NADH and to 1.8 to 2.1 with NADPH, suggesting that NAD kinase is an allosteric enzyme. These results indicate that NAD kinase activity is regulated by the availability of ATP, NAD+, and Mg2+ and, more significantly, by changes in the NADP+/NADPH and NAD+/NADH ratios. Thus, NAD kinase probably plays a role in the regulation of NADP turnover and pool size in E. coli.     

AMP deaminase from yeast. Role in AMP degradation, large scale purification, and properties of the native and proteolyzed enzyme

Eukaryotes have been proposed to depend on AMP deaminase as a primary step in the regulation of intracellular adenine nucleotide pools. This report describes 1) the role of AMP deaminase in adenylate metabolism in yeast cell extracts, 2) a method for large scale purification of the enzyme, 3) the kinetic properties of native and proteolyzed enzymes, 4) the kinetic reaction mechanism, and 5) regulatory interactions with ATP, GTP, MgATP, ADP, and PO4. Allosteric regulation of yeast AMP deaminase is of physiological significance, since expression of the gene is constitutive (Meyer, S. L., Kvalnes-Krick, K. L., and Schramm, V. L. (1989) Biochemistry 28, 8734-8743). The metabolism of ATP in cell-free extracts of yeast demonstrates that AMP deaminase is the sole pathway of AMP catabolism in these extracts. Purification of the enzyme from bakers' yeast yields a proteolytically cleaved enzyme, Mr 86,000, which is missing 192 amino acids from the N-terminal region. Extracts of Escherichia coli containing a plasmid with the gene for yeast AMP deaminase contained only the unproteolyzed enzyme, Mr 100,000. The unproteolyzed enzyme is highly unstable during purification. Substrate saturation plots for proteolyzed AMP deaminase are sigmoidal. In the presence of ATP, the allosteric activator, the enzyme exhibits normal saturation kinetics. ATP activates the proteolyzed AMP deaminase by increasing the affinity for AMP from 1.3 to 0.2 mM without affecting VM. Activation by ATP is more efficient than MgATP, with half-maximum activation constants of 6 and 80 microM, respectively. The kinetic properties of the proteolyzed and unproteolyzed AMP deaminase are similar. Thus, the N-terminal region is not required for catalysis or allosteric activation. AMP deaminase is competitively inhibited by GTP and PO4 with respect to AMP. The inhibition constants for these inhibitors decrease in the presence of ATP. ATP, therefore, tightens the binding of GTP, PO4, and AMP. The products of the reaction, NH3 and IMP, are competitive inhibitors against substrate, consistent with a rapid equilibrium random kinetic mechanism. Kinetic dissociation constants are reported for the binary and ternary substrate and product complexes and the allosteric modulators.     

The role of adenosine monophosphate nucleosidase in the regulation of adenine nucleotide levels in Azotobacter vinelandii during aerobic-anaerobic transitions

When cultures of Azotobacter vinelandii are made anaerobic the adenylate pool size remains constant or increases slightly while the adenylate energy charge decreases. Under these conditions, cell growth stops but the cells remain viable for at least 5 h with the decreased energy charge. The changes in the adenylate pool during the aerobic-anaerobic transition include: the formation of adenylates as a result of RNA degradation; the degradation of a portion of the excess AMP to form hypoxanthine by the sequential actions of AMP nucleosidase and adenine deaminase; an increase in the total adenylate pool which is stabilized at approximately 1.5 times the level in growing cells; and stabilization of the adenylate energy charge at a value near 0.3. The degradation of AMP is regulated by AMP nucleosidase, an allosteric enzyme which is activated by MgATP^2? and inhibited by P_i. The in vivo activity of AMP nucleosidase was estimated by measuring the rate of hypoxanthine formation in the culture or by measuring the activity of purified enzyme at the concentrations of AMP, ATP, and P_i found in the cells. The maximum estimated in vivo rate of AMP degradation was less than 3% of the catalytic capacity of AMP nucleosidase. Thus ample activity is present for rapid adjustments of the AMP levels in these cells. Expression of AMP nucleosidase catalytic activity is tightly controlled since high constant concentrations of intracellular AMP can be maintained for extended time periods at low adenylate energy charge values. Under these conditions controlled degradation of AMP can occur to maintain a constant AMP concentration.     

Purification and properties of AMP-deaminase from human uterine smooth muscle. Regulation by adenylate energy charge and activated fatty acids

At pH 7.0 and physiological concentrations of the main regulatory ligands (ATP, ADP, orthophosphate), human uterine muscle AMP-deaminase follows a hyperbolic type of saturation kinetics with S0.5 parameter value about 2 mM. The enzyme is regulated by adenylate energy charge (AEC) variations, being the most active at the AEC value 0.5-0.6 or 0.5-0.7, depending on the size of the total adenine nucleotide pool. Long-chain acyl-CoA strongly inhibit activity of the enzyme, influencing mainly the maximum velocity of the reaction.     

AMP deaminase from baker's yeast. Kinetic and molecular properties.

The kinetic and molecular properties of AMP deaminase [AMP aminohydrolase, EC 3.5.4.6] purified from baker's yeast (saccharomyces cerevisiae) were investigated. The enzyme was activated by ATP and dATP, but inhibited by Pi and GTP in an allosteric manner. Alkali metal ions and alkaline earth metal ions activated the enzyme to various extent. Kinetic negative cooperativity was observed in the binding of nucleoside triphosphates. Kinetic analysis showed that the number of interaction sites for AMP (substrate) and Pi (inhibitor) is two each per enzyme molecule. The molecular weight of the native enzyme was estimated to be 360,000 by sedimentation equilibrium studies. On polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate, the enzyme gave a single polypeptide band with a molecular weight of 83,000, suggesting that the native enzyme has a tetrameric structure. Baker's yeast AMP deaminase was concluded to consist of two "promoter" units which each consist of two polypeptide chains with identical molecular weight.     

Regulatory properties of 14-day embryo and adult hen skeletal muscle AMP deaminase

Chromatography on phosphocellulose column revealed changes in the elution profile of 14 day-old chicken embryo and adult hen skeletal muscle AMP deaminase. In the presence of 5 mM potassium the enzyme from embryo muscle exhibited a sigmoid-shaped plot of the reaction rate versus substrate concentration. The increase of KCl concentration up to 100 mM diminished distinctly sigmoidicity of the plot. Micromolar concentrations of ADP or ATP activated, whereas GTP at the same concentrations inhibited the embryo and hen skeletal muscle AMP deaminase while 5 mM KCl was present in the incubation medium. 100 mM potassium concentration diminished the effect of ADP and ATP but not of GTP. Palmitoyl-CoA inhibited strongly the embryo skeletal muscle adenylate deaminase but had no effect on the activity of the hen enzyme. Alanine inhibited only the adult hen enzyme. The embryo and hen AMP deaminase differed also in the specificity to adenylate analogues and exhibited a different dAMP/AMP ratio. The data presented indicate that kinetic and regulatory properties of the two developmental forms of AMP deaminase are different.     

AMP deaminase as a control system of glycolysis in yeast. Mechanism of the inhibition of glycolysis by fatty acid and citrate

The role of fatty acid and citrate on the interaction of the AMP deaminase (EC 3.5.4.6) reaction with glycolysis was investigated using permeabilized yeast cells. (a) Linolenate and citrate inhibited glycolytic flux and the recovery of the adenylate energy charge; however, linolenate remarkably retarded the depletion of the total adenylate pool, which was not at all affected by the addition of citrate. (b) Linolenate inhibited AMP deaminase activity in situ, resulting in the subsequent decrease in ammonium production, which reduced the activity of 6-phosphofructokinase (EC 2.7.1.11), whereas linolenate itself had no ability to inhibit the phosphofructokinase activity in the presence of excess ammonium concentration. (c) Citrate inhibited the activity of phosphofructokinase in situ in the presence and absence of ammonium ion, followed by an inhibition of glycolysis; however, AMP deaminase activity was not inhibited by citrate. The inhibition of glycolysis by fatty acids can be accounted for by the lowered activity of phosphofructokinase as a result of the decreased level of ammonium ion through the inhibition of the AMP deaminase reaction by these ligands, whereas the effect of citrate on glycolysis is a direct inhibition of phosphofructokinase without affecting the activity of AMP deaminase. Fatty acid and citrate, a principal metabolic product of fatty acid oxidation, can be responsible for the control of glycolysis in two different manners.     

Interrelationships between the membrane-bound ATP-dependent energy systems of the gastric mucosa and the gastric cytoprotective effect of beta-carotene on the development of gastric mucosal damage produced by 0.6 M HCl in rats

The effects of different doses (0.01-0.1-1.0-10.0/mg/kg-1) of beta-carotene were studied on gastric secretory responses of 4 hr pylorus-ligated rats: development of gastric mucosal damage (as assessed by number and severity of lesions) produced by intragastric administration of 0.6 M HCl; tissue level of adenosine triphosphate (ATP), adenosine diphosphate (ADP), adenosine monophosphate (AMP), adenylate pool (ATP + ADP + AMP), ratio of ATP X ADP-1, "energy charge" (ATP + 0.5 ADP X X (ATP + ADP + AMP)-1) (during the development of gastric mucosal damage by 0.6 M HCl and of gastric cytoprotection by beta-carotene. It was found that beta-carotene did not decrease the gastric secretory responses of 4 hr pylorus-ligated rats; The development of gastric mucosal damage could be decreased dose-dependently by the administration of beta-carotene; the ATP transformation could be decreased by beta-carotene; the tissue levels of cAMP and AMP could be increased significantly and dose-dependently by beta-carotene; the ratio of ATP X ADP-1 could be increased significantly and dose-dependently by beta-carotene; the values of adenylate pool and "energy charge" remained unchanged.(ABSTRACT TRUNCATED AT 250 WORDS)     

Modification by liposomes of the adenosine triphosphate-activating effect on adenylate deaminase from pig heart.

Adenylate deaminase (AMP deaminase, EC 3.5.4.6) of a high substrate specificity was purified from pig heart by chromatography on cellulose phosphate. The enzyme shows a co-operative binding of AMP [h (Hill coefficient) 2.35, with SO.5 (half-saturating substrate concentration) 5mM]. ATP and ADP act as positive effectors, lowering h to 1.55 and SO.5 to 1 mM. The addition of liposomes (phospholipid bilayers) to ATP-activated or ADP-activated enzyme causes a further shift of the h value to 1.04 and SO.5 to 0.5 mM. For ATP-activated enzyme the addition of liposomes increases Vmax. by about 100%, and for ADP-activated enzyme by 50%. Liposomes have no effect on the kinetics of AMP deaminase in the absence of ATP and ADP, and neither do they influence the inhibitory effect of orthophosphate on heart muscle AMP deaminase. Metabolic implications of these findings are discussed.     

Purification and properties of AMP-deaminase from human kidney.

AMP-deaminase from human kidney (cortex and medulla) was purified and the physicochemical properties were characterized. The enzyme from both portions of the kidney exhibited identical kinetics and regulatory properties. At optimal pH (6.6), the AMP-deaminase studied exhibited a distinctly sigmoidal substrate saturation kinetics, with the half-saturation parameter (S0.5) as high as 10 mM. ATP at 1 mM strongly activated the enzyme, decreasing S0.5 nearly 10-fold. The activating effect of ADP was less strong. Orthophosphate inhibited the enzyme, but the inhibition observed was weak (Ki approximately 16 mM) and had a pure competitive character. At pH 7.2, physiological for the kidney cortex, orthophosphate inhibition became even weaker and became partially competitive. Variations in the adenylate energy charge had potent effects on the activity of AMP-deaminase, depending on the size of the total adenine nucleotide pool examined. The results of gel filtration and SDS-PAGE indicated that human kidney AMP-deaminase is an oligomeric enzyme composed of four, probably identical, subunits weighing about 37 kDa each.     

Relationship between ATP and energy charge during lethal metabolic stress of the marine isopod Cirolana borealis

The relationship between ATP and energy charge was studied in individuals of Cirolana borealis under heavy metabolic stress caused by anoxia or exposure to toluene. Prolonged anoxia led to a lowering of the ATP content to about 10% after 4 days, with a simultaneous decrease in energy charge to about 0.25. A lowering of the total adenylate pool reduced the fall in energy charge somewhat, but this effect was marked only in late anoxia when the individuals had become inactive. Exposure to 0.14 mM toluene for 8 days led to a similar decrease in ATP and energy charge. Exposure to 1.4 mM toluene for 24 h led to only slight changes in the adenine nucleotide pool, although the individuals became narcotized within a few hours. The energy charge associated with moribund individuals thus varied much. The mechanism of energy charge stabilization through reduction of the adenine nucleotide pool as ATP declined seemed to be of little significance for the survival of the individuals.     

Energy Metabolism of Rickettsia typhi: Pools of Adenine Nucleotides and Energy Charge in the Presence and Absence of Glutamate

The obligate intracellular bacterium Rickettsia typhi was examined for its ability to generate and maintain an adenylate energy charge in an extracellular environment. Freshly purified organisms were incubated, at 34 degrees C and pH 7.4, with or without glutamate and various other metabolites, and the levels of ATP, ADP, and AMP were determined. Of the metabolites tested, glutamate and glutamine were the most effective for the generation of ATP. In the presence of glutamate, there was a rapid increase in the level of ATP, followed by a moderate decrease during 150 min of incubation. The energy charge increased from a level of 0.2 to 0.5 to about 0.7 to 0.75, and then slowly declined to about 0.45 to 0.6. In the absence of glutamate, after an occasional initial surge in ATP level as the temperature was changed from 4 to 34 degrees C, there was a sharp decline in both ATP and energy charge (to 0.1 and sometimes to 0.01). The rickettsiae maintained their ability to regenerate their energy charge upon the addition of glutamate for about 30 min, but this ability declined with further incubation. In contrast to Escherichia coli, the decline in ATP in R. typhi was accompanied by a sharp increase in the level of AMP and the total adenylate pool. No adenine or adenosine was recovered from rickettsiae incubated with labeled AMP, ADP, or ATP. From these experiments and the demonstration reported elsewhere that rickettsiae transport the adenine nucleotides, it can be concluded that the adenylate energy charge in R. typhi is governed by the salvage of the adenine nucleotides rather than their unphosphorylated precursors. Thus, R. typhi undergoes greater shifts in energy charge than other bacteria, a phenomenon which may account for their instability in an extracellular environment. Under optimal conditions the adenylate energy charge of R. typhi approaches levels that border on those generally regarded as adequate for growth.