Adenylate EnergyCharge Escherichia CR341T28 Properties Heat-Sensitive … · E. COLIWITHHEAT-SENSITIVE ADENYLATE KINASE 1375 avalueofabout0.90(3-5),andproteinsynthesis appears to

JouMAwL OF BACTERIOLOGY, Mar. 1981, p. 1374-13850021-9193/81/031374-12$02.00/0

Vol. 145, No. 3

Adenylate Energy Charge in Escherichia coli CR341T28 andProperties of Heat-Sensitive Adenylate Kinase

CHRISTOPHER C. GLEMBOTSKI,t ASTRID G. CHAPMAN,4 AND DANIEL E. ATKINSON*Biochemistry Division, Department of Chemistry, University of California, Los Angeles, California, 90024

Received 19 May 1980/Accepted 5 December 1980

Escherichia coli strain CR341T28 will not grow at temperatures above 340C inliquid medium, and the adenylate kinase of this strain is heat sensitive. When aculture was shifted from a permissive (3000) to a nonpermissive (3600) temper-ature, the adenylate energy charge fell from 0.9 to 0.2, with a concurrent decreasein the number of viable cells and in the specific activity of adenylate kinase. Whencultures of the temperature-sensitive strain were grown at temperatures above300C, the adenylate energy charge, the specific activity of adenylate kinase, andthe growth rate were lower than the corresponding parameters for the parentalstrain. By isotopic labeling of the adenine nucleotides in vivo, it was determinedthat increasing growth temperatures between 30 and 340C for the heat-sensitivestrain resulted in a decrease in the adenosine triphosphate-to-adenosine mono-phosphate and adenosine triphosphate-to-adenosine diphosphate ratios. Between26 and 30'C the adenosine triphosphate-to-adenosine diphosphate ratio wasessentially normal in the temperature-sensitive strain, but the adenosine triphos-phate-to-adenosine diphosphate ratio was decreased. The adenylate ratios in theparental strain did not change between 30 and 340C. The adenylate kinase massaction ratio for each strain was essentially constant under all growth conditions.When assayed at 300C, the affinities of the enzyme from the mutant strain weresomewhat lower than those of the parent adenylate kinase. The mutant enzymealso did not exhibit the substrate inhibition that was observed at high adenosinemonophosphate concentrations with the parental enzyme. An increase in theassay temperature from 300 to 400C had little or no effect on the Km valuesdetermined for the parental adenylate kinase, but caused the Km values deter-mined for the mutant adenylate kinase to increase by a factor of two or more.

Adenylate kinase catalyzes interconversion ofthe adenine nucleotides: AMP + ATP = 2 ADP.This reaction provides the main route for recon-version ofmetabolically generatedAMP to ADP(and subsequently to ATP). Adenylate kinasethus is an essential enzyme, and is present athigh levels in energetically active tissues (20).Its activity is probably high enough in mostcases to hold the reaction components near equi-librium. If so, the three-component AMP/ADP/ATP system is reduced to one degree of freedomwith respect to mole fractions, relative concen-trations, or concentration ratios, and the speci-fication ofthe energy charge [(ATP + 0.5 ADP)/(ATP + ADP + AMP)] or of any two-compo-nent ratio fixes all mole fractions and concentra-tion ratios. A severe decrease in the activity ofadenylate kinase would restore a second degree

t Present address: Department of Physiology, University ofColorado Center for Health Sciences, 4200 East Ninth Avenue,Denver, Colorado 80262.

t Present address: Institute of Psychiatry, University ofLondon, London SE5 8AF.

of freedom and partially uncouple the two-com-ponent ratios from each other or from the energycharge.The activities of major regulatory enzymes in

vitro as a function of energy charge have beenreviewed (1, 2). Regulatory enzymes in anabolic,or ATP-utilizing (U-type), sequences show anincrease in activity as the energy charge in-creases from 0 to 1. Regulatory enzymes in cat-abolic, or ATP-regenerating (R-type), sequencesshow a decrease in activity as the energy chargeincreases. The midpoints and the steepest re-gions of the response curves of these enzymes tochanges in energy charge in vitro occur at energycharge values above 0.80. These sensitive andoppositely directed responses appear to play amajor role in the kinetic partitioning of metab-olites between catabolic and anabolic sequences.They should also hold the value of the energycharge within a narrow range. This expectationhas been bome out by analyses showing that inintact, growing, and actively metabolizing cellsof many types the energy charge is stabilized at

1374

E. COLI WITH HEAT-SENSITIVE ADENYLATE KINASE 1375

a value of about 0.90 (3-5), and protein synthesisappears to be depressed when the charge fallsbelow this value (25, 26).

Escherichia coli CR341T28 is a heat-sensitivestrain with a thermolabile adenylate kinase (6-8, 13) and a thermolabile glycerol-3-phosphateacyltransferase (10); both labilities apparentlyresult from a single mutation (9, 10, 13). Theadenylate kinase defect is the more thoroughlystudied of the two. Adenylate kinase of thisstrain is inactivated at elevated temperatures(above 36 to 37°C) both in vivo and in vitro (6,8, 13). The inactivation in vivo is associated witha rapid decline in the cellular ATP concentrationand concomitant cell death (7, 13).An objective of this study was to exploit the

temperature sensitivity of adenylate kinase inEscherichia coli CR341T28 in an attempt todetermine the consequences of limiting activityof this enzyme on growth and regulation. Wehoped that growth of this strain at temperaturesat which the activity of adenylate kinase wasgrowth limitig would allow for some degree ofuncoupling of the ATP/AMP and ATP/ADPratios from each other and from the energycharge. It might then be possible to evaluate therelative impacts of these ratios, which might beconsidered in a sense to be component factors ofthe energy charge, on the growth rate. Thesimplest prediction was that the ADP/AMP andATP/AMP ratios would decrease with increas-ing temperature. Changes, if any, in the ATP/ADP ratio and the energy charge could not bepredicted.We also wished to establish whether the re-

duced adenylate kinase activity in the mutantstrain might be at least in part due to alteredkinetic properties of the mutant enzyme, ratherthan merely to inactivation. Since the activity ofthis enzyme in the mutant cells appears to de-crease with a rise in culture temperature be-tween 30 and 34°C, it was also of interest toestablish whether the kinetic properties of themutant enzyme are affected by changes in tem-perature. We therefore purified adenylate kinasefrom the temperature-sensitive and parentalstrains and compared their kinetic propertiesover the relevant temperature range. The en-

zyme has been partially purified from E. coli byHolmes and Singer (16) and by Cousin et al. (8).

(Portions of these results have been presentedpreviously [A. G. Chapman, C. C. Glembotski,and D. E. Atkinson, Abstr. 11th FEBS Meet.1977, C-9, p. 816].)

MATERIALS AND METHODSOrganism. Escherichia coli strains CR341T28 f

thr leu Bl thy lacY met and CR34 (19) were derivedfrom E. coli K-12 (6, 8, 18). Strain CR341T28 is a

temperature-sensitive mutant containing a thermola-bile adenylate kinase (EC 2.7.4.3). This strain also hasa thermolabile mutant form of glycerol-3-phosphateacyltransferase (EC 2.3.1.15) (13), the first enzyme inthe phospholipid biosynthetic pathway, apparently asa consequence of the same mutation that causes ade-nylate kinase to be temperature sensitive (9). StrainCR34 (called the parent in this paper) is isogenic tostrain CR341T28 except for the mutation in adenylatekinase and for the absence of a methionine require-ment. Strain CR341 was not available for use duringthis study. Strain CR34 has been used in the past forcomparison to strain CR341T28 (19).Growth and sampling. Cells were grown aerobi-

cally at 30°C unless otherwise stated in medium con-taining 5 g of KH2PO4, 17 g of K2HPO43H20, 2 g of(NH4)2SO4, 0.2 g of MgCl2, 5 g of glucose, 10 mg ofthiamine-Cl, 200 mg each of thymidine, L-methionine,L-isoleucine, L-leucine, L-threonine, and L-vahine, and1.0 g of Casamino Acids (Difco Laboratories, Detroit,Mich.) per liter. Guanine (12.8 mg/liter) was addedwhen [3H]adenosine was to be incorporated into theadenine nucleotides. Starter cultures (50 ml) weregrown at 28°C for 12 to 14 h in Erlenmeyer flasks withrotary shaking. Experimental cultures were incubatedin gas dispersion flasks in a circulating water bath,with moist air forced through fritted-glass disks at arate of 150 ml/min. Cell growth was followed by mea-suring the optical absorbance at 540 nm (Afwo). Growthwas followed by measuring the number of particles perml with an Electrozone Celloscope (Particle Data, Inc.,Ehmhurst, Ill.) and by plating to determine the numberof viable cells. A 0.1- to 0.5-ml sample of a 1:106dilution of the culture was mixed with 10 ml of asolution containing 0.5% agar and 1% nutrient broth,and the resulting colonies were counted. For enzymepurification, a 15-liter culture of E. coli CR341T28 wasgrown in a glass carboy with moist air forced througha fritted-glass tube at a rate of 150 ml/min; E. coliCR34 (parent) was grown in six 2,000-ml Erlenmeyerflasks each containing 1 liter of medium and aeratedby rotary shaking. Cells were harvested by centrifu-gation in late exponential phase (Am0 = 1.0 for strainCR341T28 and Aw0 = 2.0 for strain CR34), and washedtwice with 50 mM Tris-chloride buffer, pH 7.4. Forexperiments involving a shift-up of the culture tem-perature, the mutant cells were grown in a water-jacketed gas dispersion flask equipped with a magneticstirrer and aerated with moist air at the rate of 150ml/min.

Perchloric acid extraction of adenine nucleo-tides. For the determination of adenine nucleotidesby the luciferase method, a 1.0-ml sample of the bac-terial culture was removed through a sampling port inthe incubation flask and pipetted into 0.2 ml of cold17% HC1O4. After 30 min at 4°C, the sample wascentrifuged at 7,000 x g for 15 min. A 0.8-ml portionof the supernatant was removed and neutralized with0.14 ml of 0.58 M KHCO3 in 2.6 M KOH. After 15 minthe sample was centrifuged at 7,000 x g for 15 min.The supernatant was then removed and stored at-70°C (5, 26). For the determination of adenine nu-cleotides by the radioisotope method, the same ex-traction procedure was used but the volumes werescaled down by a factor of ten.

VOL. 145, 1981

1376 GLEMBOTSKI, CHAPMAN, AND ATKINSON

Luciferase assay. ATP was assayed by the lucif-erase reaction (24) with a Luminescence Biometer (DuPont Co., Wilmington, Del.). ADP was determined bydifference after converting it to ATP with pyruvatekinase and assaying for ATP. AMP was determinedby difference after conversion to ADP with adenylatekinase and conversion of ADP to ATP with pyruvatekinase. These procedures have been previously de-scribed (26).Determination ofintracellular adenine nucleo-

tides. The intracellular nucleotide levels were deter-mined by subtracting the values for adenine nucleo-tides in the medium after the cells had been removedby filtration from values for the total culture. Thiscorrection is important in the CR341T28 cultures be-cause up to 15% of the total adenine nucleotides are inthe medium. The filtering apparatus was a 5-ml sy-ringe to which a 25-mm (biometer samples) or 13-mm(radioisotope samples) filter adapter (Millipore Corp.,Bedford, Mass.) had been attached. Millipore filterswith a pore size of 0.22 ,um were used to remove cells.

Incorporation of [3Hladenosine into adeninenucleotides. Cultures were grown to a density ofabout 7 x 108 cells per ml either in an Erlenmeyerflask on a shaker or in a sparger flask. A 0.27-mlsample was then placed in a small tube that contained30 pCi (35 to 50 Ci/mmol) of [2,8-3H]adenosine in 0.03ml of water at the same temperature as the cultureflask. This procedure is similar to one described pre-viously (13, 15). The incorporation was allowed toproceed for 90 s, at which time the adenine nucleotidepool was considered to be uniformly labeled. After thepulse-labeling, 0.20 ml of culture was removed andfiltered through a 0.22-pm pore Millipore filter 13 mmin diameter. A 0.1-ml aliquot of the filtrate was ex-tracted with 0.02 ml of 17% HC104, as was 0.1 ml ofthe culture that had not been filtered. The sampleswere then treated as previously described (see extrac-tion of the adenine nucleotides above.).

PEI-cellulose TLC separation of adenine nu-cleotides. A 5-pl volume of each sample was spotted1.8 cm from the short side of a poly(ethylene)iminecellulose thin-layer chromatography plate (10 cm by20 cm) that was impregnated with a fluor to facilitateviewing UV-absorbing compounds. Unlabeled ATP,ADP, and AMP were spotted as markers. The platewas soaked in anhydrous methanol for 10 min andthoroughly dried under a stream of cool air. The platewas then soaked in deionized water for 10 min andagain dried. The methanol and water washes removedsalts that otherwise interfere with the chromatogra-phy. The plate was developed in deionized water untilthe solvent front reached the top. This step requiredabout 60 min. After thorough drying, the plate wasplaced in aqueous 0.5 M LiCl and developed until thesolvent front was 12 cm above the origin (22). Thisstep required 30 min. The plate was again dried.

Elution and counting of adenine nucleotides.The spots corresponding to the marker nucleotideswere located with the aid of a UV lamp, cut out, andplaced in scintillation vials. A 0.7-ml volume of 4 MHCI was added to elute the nucleotides. After 30 min,15 ml of scintillation fluid (57% toluene, 43% anhy-drous ethanol containing 0.46 g of PPO (2,5-dipheny-loxazole) and 0.6 g of POPOP [bis(2-[5-phenyloxa-

zolyl])benzene) per liter) was placed in the vials, whichwere counted in a Tri-Carb Scintillation Spectrometer(model 2002; Packard Instrument Co., Inc., Rockville,Md.).Enzyme assays. For estimation of the specific

activity of adenylate kinase in vivo, cell samples weretaken directly from the liquid culture and frozen at-70°C. The samples were later thawed, and the cellswere broken by sonication. After centrifugation at27,000 x g for 50 min, the supernatant was dialyzedagainst 4 liters of 50mM Tris-chloride (pH 7.4) at 4°Cfor 15 to 20 h.

Adenylate kinase assays were performed, at 30°Cunless otherwise indicated, in 1.0-ml volumes by usinga temperature-controlled recording spectrophotome-ter. For estimations of activity in vivo and duringenzyme purification, the reverse reaction (productionof ATP and AMP from ADP) was routinely used. Thereaction was coupled to the hexokinase and glucose-6-phosphate dehydrogenase reactions, and the produc-tion of NADPH was observed as a function of time(21). Each assay contained 100 mM Tris-Cl (pH 7.4),100 mM KCI, 5 mM MgCl2, 10 mM glucose, 1 mMNADP+, 4 mM ADP unless otherwise stated, and 50pg of dialyzed hexokinase-glucose-6-PO4 dehydrogen-ase (1:1 by weight). All components except the ade-nylate kinase were incubated together in a cuvette for5 min at 30°C before the reaction was initiated byaddition of adenylate kinase. To monitor the produc-tion of ADP from ATP and AMP (forward reaction),the adenylate kinase reaction was coupled to the py-ruvate kinase and lactate dehydrogenase reactions,and the oxidation of NADH was followed spectropho-tometrically. Each assay contained 50 mM Tris-chlo-ride (pH 7.4), 5 mM MgCl2, 2.5 mM phosphoenolpyr-uvate, 0.26mM NaDH, the indicated amounts ofAMPand ATP, and 30Lg of dialyzed lactate dehydrogenase-pyruvate kinase mixture. Again, a stable backgroundrate was obtained before the reaction was initiated bythe addition of adenylate kinase. The specific activityof adenylate kinase was expressed as micromoles ofproduct formed per minute per milligram of protein.Affinity constants and Hill coefficients were calculatedfrom Hill plots {log(v/(V.. - v)] versus log S) afterthe maximum velocity had been estimated from Line-weaver-Burk plots.

Protein determination. The total protein was de-termined by the amido black method (23). Cells werecollected by centrifuging a 0.6-ml sample at 12,000 xg for 5 min at 4°C, and the supernatant was discarded.The pellet was suspended in 0.81 ml of water, and 0.09ml of Tris-chloride (pH 7.5) with 0.1% sodium dodecylsulfate was added; then 0.18 ml of 60% trichloroaceticacid was added and a portion of the resulting precipi-tate was spotted on a sheet of type GS Millipore filterpaper and assayed as published (23).

Materials. Poly(ethylene)imine cellulose thin-layer chromatography plates on plastic sheets wereobtained from E. Merck, AG, Darmstadt, Germany.The nucleotides were obtained from Sigma ChemicalCo., St. Louis, Mo. Hexokinase, glucose-6-phosphatedehydrogenase, myokinase, and pyruvate kinase wereobtained from Boehringer Mannheim Corp., NewYork, N.Y., and [2,8_3H]adenosine was obtained fromNew England Nuclear Corp., Boston, Mass.

J. BACTrERIOL.


Purification of adenylate kinase. Table 1 sum-marizes the purification of adenylate kinase from E.coli CR341T28. The enzyme from the parental strainwas purified according to the same procedure, exceptthat only 5 g of packed cells was used, and all volumeswere scaled down by a factor of seven. All steps of thepurification were carried out at 0 to 40C. Packed E.coli CR341T28 cells (35 g) were suspended in 350 mlof 50 mM Tris-chloride (pH 7.4), and sonicated in aBranson Sonifier, model LS75, at an intensity of 7 anda power of 7 mA for 45 s (three pulses of 15 s each).Further sonication led to partial inactivation of thetemperature-sensitive adenylate kinase. Cell debriswas removed by centrifugation at 20,000 x g for 15min. The supernatant solution was dialyzed againstthe extraction buffer (Tris-chloride, pH 7.4) for 12 to14 h. Streptomycin-S04 was added slowly at the rateof 1 pl of 25% streptomycin sulfate per ml of cell-freeextract per A unit at 260 nm. The suspension wasstirred for an additional 60 min, and the resultingprecipitate was removed by centrifugation for 30 minat 20,000 x g. Solid ammonium sulfate (291 mg/ml)was added to the supernatant solution slowly withrapid stirring to give a 50% saturated solution. Afteran additional 30 min of stirring, the precipitate wasremoved by centrifugation at 20,000 x g for 30 min.To the supernatant fraction was added 194 mg of solidammonium sulfate per ml to give an 80% saturatedsolution. The suspension was centrifuged, and theprecipitate was dissolved in 35 ml of 50 mM Tris-chloride (pH 7.4). The 50 to 80% ammonium sulfatefraction was applied directly to a Sephadex G-75 col-umn (2.6 by 92 cm) equilibrated with 50 mM Tris-chloride (pH 7.4). The column was eluted with thesame buffer at a flow rate of 0.7 ml/min. The adenylatekinase activity was collected between 110 and 160 mlof eluent volume.

Blue dextran 2000 was covalently bound to Sepha-rose 4B by the method of Thompson et al. (27). Thepooled G-75 fractions were added to a column of bluedextran-Sepharose (1.6 by 38 cm) previously equili-brated with 50 mM Tris-chloride until the Am, of theeluent was zero. The enzyme was eluted with a 0 to0.5 mM linear gradient of adenylates (0.25 mM AMPplus 0.25 mM ATP). The peak of the adenylate kinaseactivity eluted at approximately the 0.4 mM adenylateconcentration. The peak adenylate kinase fractions(12 ml) were pooled and added to a column (1.3 by 15cm) of DEAE-cellulose (Whatman DE 52) equili-brated with 50mM Tris-chloride (pH 7.4). The columnwas eluted with a 0 to 0.2 M linear KCl gradient.Adenylate kinase activity appeared in the eluent atapproximately 0.15 M KCI. Fractions containing thehighest activity were pooled and stored in aliquots at-20°C in the presence of 1 mg of bovine serum albu-min per ml. The protein concentration before thebovine serum albumin addition was 20,ug/ml.Enzyme purity was estimated with the use of disc

electrophoresis in 7% polyacrylamide gels by the pro-cedure of Gabriel (11). Gels were stained with 0.5%amido black in 7% acetic acid for 60 to 120 min atroom temperature and later destained with 7% aceticacid for 24 to 48 h. For purity estimations, 120 ,ug ofprotein was layered on each gel. Parallel gels werestained for enzyme activity with nitroblue tetrazolium

and phenazine methosulfate as terminal electron ac-ceptor dyes (12). For comparison, a standard of com-mercially available muscle adenylate kinase was runon an identical gel.

RESULTSIf adenylate kinase were partially inactivated,

the conversion of AMP to ADP would presum-ably be impaired. Changes in the relative con-centrations of the adenine nucleotides mightadversely affect regulation of the energy metab-olism of the cell. Further, a decrease in theconcentration of ADP might become a limitingfactor in oxidative phosphorylation and, hence,in growth. Growth of CR341T28 cells at partiallypermissive temperatures that cause partial in-activation of adenylate kinase thus seemed to bea promising approach to study of the correlationbetween alterations in energy charge values oradenylate ratios and functional capacities of thecell such as growth and viability.Among a population ofE. coli CR341T28 cells,

over 90% form colonies on nutrient agar at 34°C,but only 10% do so at 380C (Fig. 1). There is noobvious explanation for this range of tempera-ture sensitivities, since the heat lability of theadenylate kinase is apparently caused by a pointmutation. The distribution may be due to mi-croheterogeneity in culture caused by physiolog-ical individuality among the cells. Alternatively,the probability that any given cell will form acolony when plated may decrease with temper-

801

60F0

U

J

>L

0

201

40F

30 34 38

TEMPERATURE (°C)

FIG. 1. Colony-forming ability ofE. coli CR341T28incubated inpourplates at various temperatures. Foreach temperature, cells were grown in liquid cultureat 30°C to a density of 5 x 108 cells per ml asdetermined by particle counts. Then a sample wasserially diluted, plated in nutrient agar, and incu-bated at the indicated temperature for 2 days.

loop I I I

VOL. 145, 1981


ature even though the cells may be essentiallyidentical. In liquid medium, the cut-off temper-ature was slightly lower and appeared to besharper, occurring at about 340C.When the mutant strain was shifted from a

permissive (300C) to a nonpermissive (360C)temperature, the viable-cell count declined inagreement with previous observations (6,8). Thechange was coordinant with the decrease in ad-enylate energy charge (Fig. 2). The cells couldnot be revived by shifting back to 300C after 1.5h at 360C. The specific activity of adenylate

Time (hours)

C' 0.4E0)

n.0.3

, 0.2

c

~0.4

E 0.2

=v 0.1

co

-0

0.04

0

4 5 6 7

0 2 4 6 8Time (hours)

FIG. 2. Effect on the adenylate energy charge, ad-enylate kinase activity, and viability of E. coliCR341728 of a shift from a permissive to a nonper-missive temperature. (A) Cells were grown in liquidculture at 30°C and then shifted to 37°C. The specificactivity of adenylate kinase was determined in di-alyzed cell-free extracts according to the proceduregiven in the text for the reverse reaction. (B) Cellswere grown in liquid culture for 2 h at 30°C and thenshifted to 36°C. The viability was estimated by abilityto form colonies at 30°C and the energy charge by theluciferase method.

0.40

0.3 X >

0

0.2 C. I

3

0.15

0.1

1.0 m

0.8 e

0.60.4 m0.2 (O e

kinase also decreased rapidly following the shiftin temperature.

Exponential growth of E. coli CR341T28 wasobserved at temperatures up to 340C. At 340Cand below, the viable colony counts and particlecounts correlated very well (e.g., Fig. 1). Attemperatures below 300C, the growth rate of themutant strain was the same as that of the nearlyisogenic strain CR34 (Fig. 3). This indicated thatat these temperatures, the adenylate kinase ac-tivity in strain CR341T28 was probably notgrowth-rate limiting. Above 300C, strainCR341T28 grew exponentially but at rates sub-stantially lower than those of CR34 cells (Fig.3). In this temperature range the ratio of thegrowth rate constants of the mutant and wild-type strains decreased with temperature in anessentially linear manner (Fig. 4). At 34.50C orhigher, colony counts and particle counts corre-lated poorly for strain CR341T28, as shown inFig. 1.The relative specific activities of adenylate

kinase in the two strains provided further ap-parent correlation between the lowered growthrates of strain CR341T28 at temperatures above300C and the adenylate kinase lesion (Fig. 4).The specific activities for the mutant and parentstrains differed by a factor of about two, even atpermissive temperatures (Fig. 5). This differencein specific activities was reported by Cousin et

1401-

Q

.E

70t

24 28 32 36

Tempera ture ('C)

FIG. 3. Growth rates of the parental (CR34) (A)and the temperature-sensitive (CR341 T28) (0) strainsof E. coli at various temperatures. The growth con-stant is defined as 0.69/tg,,, where tgen is the timerequired to double the number of cells. The percent-age of cells able to form colonies was monitored bycomparing particle counts with colony counts andwas always at or near 100%.

- . _ .~ A -

I I

30-'-37° :.A. , . '-N ^~~~~~~

% -*0-

c_ -w o

/ A-30°_-36°

A__

A

/A

I/

J. BACTERIOL.

F

u.uzi


O _

*= C4jO.5

4 I.-

< nAliL

1.0-

ko0.8

'-'0.6

* *\*~~~\0~~~~~~

-\_9_9-J~~

%O0

3024 28 32Temperature (OC)

FIG. 4. The ratios of the growth constant IL (0)and ofadenylate kinse specific activity (0) for E. coliCR341T28 to the corresponding parameters for E.coli CR34 as a function of temperature.

0.61

0,

E

0

._

0

c

c1

4 0.2-

24 28 32 36

Temperature (°C)

FIG. 5. The specific activities of adenylate kinasein E. coli CR341T28 (0) and E. coli CR34 (A) grownat various temperatures. Samples were taken from a

liquid culture at the indicated temnperature at a den-sity of about 9 x 108 cells per ml. The samples were

sonicated and assayed as described in the text.

al. (8). The Km of adenylate kinase for ADP is0.14 mM for strain CR34 and 0.56mM for strainCR341T28 (see below). The experiments thatled to Fig. 5 were perforned at an ADP concen-

tration of 4 mM; thus, the wild-type enzyme wasabout 90% saturated, and the mutant enzyme,about 75%. This difference thus accounts forabout 20% of the observed twofold difference inrate. The difference in specific activities at per-

missive temperatures may be due to a lowerturnover rate of catalysis (V) for the enzymefrom the mutant strain, a lower molar level ofenzyme in the cell, or differential loss of activityduring purification. In an attempt to test the lastpossibility, we found that, for each strain, theextracts obtained by sonication and by use ofthe French pressure cell had equal adenylatekinase specific activities.The adenylate energy charge was initially de-

termined by the luciferase method for bothstrains at permissive and partially permissivetemperatures. In the luciferase assay, errors inADP and especially AMP values were largesince these compounds must be estimated bydifference. Energy charge can be determinedwith comparatively little error because this pa-rameter is relatively insensitive to changes inthe concentration of AMP, but ratios betweenthe adenylates are subject to large error. Theenergy charge at 300C and below was about 0.90in both strains and remained at this value athigher temperatures in the CR34 cells, but de-clined in strain CR341T28 to a value of 0.82 at34C. The close correlation in the mutant strainbetween the decreases at temperatures above300C in specific activity of adenylate kinase, inthe energy charge, and in the growth rate furthersupports the suggestion that the lowered enzymeactivity limits the growth rate at these temper-atures.The [3H]adenosine incorporation technique

(see above) permitted determination of adeninenucleotide levels, and therefore ratios and en-ergy charge values, with substantially less errorthan in the luciferase method. Since the turn-over time of an ATP molecule in E. coli is about0.1 s (17), it was considered that after 90 s theadenylate pool was uniformly labeled. The en-ergy charge values obtained with this methodwere slightly higher than those obtained by theluciferase technique. The energy charge declinedin strain CR341T28 with increasing temperaturefrom 0.90 at 300C to 0.88 at 340C (Fig. 6A).Energy charge values near 0.9 have been ob-served consistently in actively metabolizing cells(4, 5). The lowered growth rate observed whenthe energy charge fell below 0.90 in the mutantstrain suggests that regulatory interactions in-volving the adenine nucleotides may be limitingthe growth rate. The energy charge in the pa-rental strain remained relatively high, 0.93,throughout the entire temperature range.The ATP/ADP and ATP/AMP ratios, as de-

termined by the isotope method (Fig. 6B and60), increased as temperature decreased below30°C in both strains. This has also been observedin E. coli B (C.C.G., unpublished data). Thereason for this increase is not clear. The ATP/

A

*0

0

0

VOL. 145, 1981

V;.*rt


0O9

0S

<2

a.

FIG. 6.

determineAdenylateATP/AMJtemperatu,were dete,Cultures uextracted,rated bychromatog

ADP rati300C, bulstrain ab25% to 34

ADP/AM50% of thgrowth teratio wasCR341T2about 300tended toperature.CR341T2creased fAMP ratiof the valTherm

nase. Inthe speci

.0 . . . . A mined in the cell-free extract remained constant,at approximately 0.4 ,umol/min per mg ofprotein

?5 . when assayed at 300C, during the exponentialgrowth phase (Fig. 2). When the growth temper-

P0. ature of the culture was raised to 370C, the cellsPO

P stopped growing after 30 min. Coincident with10 the cessation of growth, the specific activity of*>_B adenylate kinase decreased rapidly. A shift-up

9 : S in temperature for this mutant strain of E. coliAA*A<s_-_has previously been shown to result in a rapid

8 decline in the intracellular ATP concentration7 (7, 13) in accordance with the adenylate kinase

% inactivation.50 Adenylate kinase from the temperature-sen-^C sitive mutant grown at 300C is thermolabile in

40 . A...A.... cell-free extracts (8, 13). The temperature-sen-10 sitive adenylate kinase was 95% inactivated

when a crude extract (3.2 mg/ml of protein) was20 incubated at 370C for 30 min. This observation_______________ agrees with previous results of Cousin et al. (8),

sA whereas Glaser et al. (13) showed little or no

D adenylate kinase inactivation in the E. coli4 CR341T28 extract under similar conditions. The. ^ purified temperature-sensitive enzyme, even in

3 > a much more dilute solution (51 jug/ml of pro-~ . . tein), was more resistant to thermal inactivation,

2* ..F .~w--* | gretaining 31% of initial activity after 30 min at24 28 32 36 370C. Adenylate kinase from the parent strain

Temperature (0C) was fully active after the same treatment.Adenine nucleotide concentration ratios as Purification of adenylate kinase. Thed by the incorporation of [3H]adenosine. method used to purify adenylate kinase was aenergy charge (A), ATP/ADP ratio (BJ, modification of that used by Holmes and SingerP ratio (C), andADP/AMP ratio (D) of the (16).:re-sensitive (0) and parental (A) strains Blue dextran-Sepharose chromatography (27)rrmined at various growth temperatures. was employed because it is a rapid procedureverepulse-labeled with [3H]adenosine and that results in considerable purification withand the adenine nucleotides were sepa- relatively little inactivation of the unstable mu-poly(ethylene)imine-cellulose thin-layer tant adenylate kinase. Since adenylate kinase

graphy. requires both ATP and AMP as substrates, itwas hoped that the use of low concentrations of

io was constant in strain CR34 above both adenine nucleotides would be effective int declined noticeably in the mutant eluting this enzyme, rather than others that bind-ove this temperature and was about either ATP or AMP.0% lower at 340C than at 300C. The Table 1 summarizes the purification, whichIP ratio in the mutant strain was about resulted in a 278-fold purification of the mutanttat in the parental strain at nearly all enzyme and a final specific activity of 50 umol/,mperatures (Fig. 6D). The ATP/AMP min per mg of protein at 300C. The details ofalso lower at all temperatures in strain the purification are given above. Because the'8 than in strain CR34 (Fig. 60). At enzyme was needed only for kinetic compari-IC, the ATP/AMP ratio in strain CR34 sons, extensive efforts to maximize yield did notbecome constant with respect to tem- seem necessary. The same procedure, applied toThe ATP/AMP ratio in strain adenylate kinase from the parent strain, resulted

!8 declined as the temperature in- in a 475-fold purification and a final specificrom 26 to 340C. At 340C, the ATP/ activity of 976 pmol/min per mg of protein atio in the mutant strain was about 60% 300C. The difference between the specific activ-lue observed at 300C. ities of the mutant and the parent enzymesial inactivation of adenylate ki- increased duriffg the course of the purification.a culture of E. coli CR341T28 at 300C, The similr degree of apparent purity observedfic activity of adenylate kinase deter- in polyacrylamide gel electrophoresis of the two

1.

.

0m 0aI-

V

a.0

I.-

1

5

J. BACTERIOL.


TABLE 1. Purification of adenylate kinase from atemperature-sensitive mutant, E. coli CR341T28

Sp. Act

Total Yield PmofPuoficaF'raction activ- (%)l ATP/min tion fac-

ity per mg of torprotein at300)

Streptomycin-S04 130 100 0.18 1ppta

(NH4)2SO4 ppt 145 112 0.42 2Sephadex G-75 48 40 1.68 9Blue dextran- 15 12 27.3 152

SepharoseDEAE-cellulose 7 5 50.0 278

appt, Precipitate.

purified enzymes suggests that the mutant en-

zyme was partially inactivated during the puri-fication procedure, and that the major band mayhave contained both active and inactive mutantadenylate kinase protein.Polyacrylamide gel electrophoresis. Be-

fore bovine serum albumin was added for stabi-lization, 100 jig (5 ml) of the purified adenylatekinase from each strain was concentrated byultrafiltration and subjected to electrophoresison a 7% acrylamide gel. Duplicate gels were run;one was stained for protein and the other forenzyme activity. A similar pattern of one majorand two minor protein bands was observed withboth enzymes. The mobility of the major band(estimated as approximately 90% of total pro-tein) corresponded to that of the adenylate ki-nase activity in the enzyme-stained gel. Themobility of the major protein band also matchedthat of commercial muscle adenylate kinasewhen run as a standard in the same gel systemand stained for either protein or enzyme activity.Kinetic properties of mutant and parent

adenylate kinase at 300C. Table 2 summarizesthe kinetic parameters observed for the adenyl-ate kinase reaction when using dialyzed, par-

tially purified enzymes from the mutant andparent strains. The assay temperature (300C)corresponded to a permissive growth tempera-ture for the mutant strain, and the highest tem-perature at which maximum adenylate kinaseactivity was observed in the mutant celLs (Fig.5). At this temperature, the parent adenylatekinase had slightly higher affinities than themutant enzyme for all three substrates, withaffinity ratios ranking from 1.5 for ATP to 4 forADP. The Hill coefficients, approximately 1 forATP and AMP and 2 for ADP, were the same

for both enzymes. When adenylate kinase wasisolated from the mutant strain grown at 340C,where the enzyme level has been shown to bereduced by 40%, the specific activity in extractswas reduced, as expected, but the Km values for

TABLE 2. Kinetic parameters of adenylate kinasefrom E. coli CR34 and from the temperature-

sensitive mutant E. coli CR341228 assayed at 30°CATP1 AMPa ADP

StrainK. n K,. n K. n

CR34 0.17 0.09 0.08 1.0 0.14 1.8CR341T28 0.25 0.9 0.17 1.0 0.56 1.8

a At saturating concentration of cosubstrate: 1 mMfor the parental enzyme, and 2 mM for the tempera-ture-sensitive enzyme.

all substrates were the same as those for theenzyme isolated from the mutant strain grownat 30°C. Variation of the concentration of one ofthe cosubstrates, AMP or ATP, over a 10-foldrange (0.05mM to 0.5 mM) produced a less thantwofold variation in the affinity of either enzymefor the remaining substrate.The substrate inhibition observed at high

AMP concentrations for the parent adenylatekinase was not observed for the mutant enzymeeven at much higher AMP concentrations (Fig.7).Kinetic properties of adenylate kinase as

a function oftemperature. Cultures of E. coliCR341T28 show a gradually decreasing level ofadenylate kinase activity with an increase in thegrowth temperature above 300C, reaching com-plete inactivation (and concomitant cell death)around 370C. The affinities of the partially pu-rified mutant and parent adenylate kinases forthe three substrates were determined as a func-tion oftemperature over the same range (Fig. 8).The reaction mixtures were incubated in cu-vettes at the indicated temperatures for 10 min,and the reactions were initiated by the additionof adenylate kinase. Control experimentsshowed that in the presence of substrate thetemperature-sensitive adenylate kinase was sta-ble for the duration of the assay even at 400C.The affinities of adenylate kinase from the par-ent strain for all three nucleotides were essen-tially unchanged over the temperature range 30to 400C.

In contrast, the affinities of the mutant ade-nylate kinase for all three substrates decreasedsubstantially as the assay temperature wasraised from 300C to 400C. The change was great-est in the Km value for ATP, which increasedfrom 0.25 mM at 300C to 0.70 mM at 400C. TheKm values for AMP and ADP increased abouttwofold over the same temperature range. Acontrol experiment showed that the tempera-ture-sensitive adenylate kinase, when incubatedat 400C for 10 min (in the presence of 1 mMADP to reduce inactivation) and subsequentlyassayed at 300C, exhibited the same affinity for

VOL. 145, 1981


0

3.0

I

Q5 1.0~i' 1.5 2 3 4(AMP] mM

FIG. 7. Reaction rate as a function ofAMP concentration for partially purified adenylate kinase from E.coli CR34 (parent) and E. coli CR341T28 (mutant) at 30°C. The concentrations ofATP were 4 mM and 1 mMfor the mutant and the parent enzymes respectively.

1.0

Q9

aE

Q8

0.7

s6

05

04

0.3

02

0.1

,_,.OATP

----------

,-" __~~~~~~~~AAIF

>Mutant

Ian

30 32 34 36 38 40

Tep (C)FIG. 8. Michaelis constants for adenylate kinase from E. coli CR34 (parent, bottom 3 lines) and E. coli

CR341728 (mutant, top 3 lines) as functions of temperature. Cosubstrate concentration (AMP when ATP wasvaried; ATP whenAMP was varied) was constant at2 mM. Partiallypurified adenylate kinases from the twostrains grown at 30'C were assayed according to the procedures described in the text.

ADP as did the enzyme assayed at 300C withoutany prior incubation at elevated temperature.The Hill coefficients for the three substrates

were the same at 30°C and at 400C (1.8 for ADP,1.0 for AMP, and 0.9 for ATP) for the mutantenzyme. However, when the log of the maximumvelocity of the reaction was plotted against theinverse ofthe temperature, the resulting Arrhen-ius plot for the parent enzyme showed the ex-

pected straight line relationship, whereas therewas a deviation from linearity for the mutantenzyme at elevated (37.5 to 42.50C) tempera-tures (Fig. 9).

DISCUSSIONAt temperatures below 300C, the adenylate

ratios were altered in the mutant strain in a waythat is consistent with a less active adenylate

J. BACTERIOL.

_ >~


0

0 6.0E2 5.0E 4.0- fParent

4.0

E_ 3.0\

Mutant

X 2.0C0 V

0 400 350 30

3.1 32 3.3103/T

FIG. 9. Arrhenius plots ofthe maximum velocity ofthe adenylate kinase reaction for the enzymes par-tially purified from E. coli CR34 (parent) and E. coliCR341728 (mutant). Activation energies of 11.9 and11.3 kcal/mol were calculated from the slope of thecurve for the parental enzyme and from the linearpart of the curve for the mutant enzyme, respectively.

kinase. The value of (ADP)2/[(ATP)(AMP)] av-eraged 0.54 in the parent strain and 0.31 in themutant, and did not vaiy with temperature ineither strain over the range 26°C to 340C. TheATP/AMP and ADP/AMP ratios were alsolower in the mutant in this permissive temper-ature range (Fig. 6C and 6D). However, at 260Cto 300C, growth of the mutant strain was essen-tially identical to that of the parent. It may besignificant that the ATP/ADP ratio in the mu-tant was equal to that of the parental strain atthese temperatures. Although no definitive con-clusions in this regard are possible, these resultssuggest that the ATP/ADP ratio may, undersome conditions, be more important than theATP/AMP ratio in relation to growth.The pattern of nucleotide concentrations in

the mutant strain at temperatures above 30 or320C is puzzling. Although the specific activityof adenylate kinase in extracts of the mutantwas lower than in the same strain at lowertemperatures, the mass action ratio was as highas at lower temperatures; as expected, the ADP/AMP ratio was lower and ATP/AMP ratiomuch lower than in the 260C to 30°C range.However, the ATP/ADP ratio was also lower.This effect balanced the decrease in ADP/AMP,with the result that the mass action ratio, which

may be expressed as the product of ADP/AMPand ADP/ATP, remained essentially constant.The reason for the decrease in the ATP/ADP

ratio in the mutant at higher temperatures isnot clear. It may perhaps not be related directlyto the heat lability of adenylate kinase. Theglycerol-3-phosphate acyltransferase activity ofthis strain is also heat labile (8), apparently as aconsequence of the same mutation that affectsadenylate kinase (9). It is possible that thisdefect in phospholipid biosynthesis may affectmembrane functioi in such a way as to limitelectron transport phosphorylation. Thus, theanalytical results are consistent with the sugges-tion of simultaneous decreases in vivo at highertemperatures in the activity of adenylate kinase(hence, the decreases in the ATP/AMP andADP/AMP ratios) and in the capacity for oxi-dative phosphorylation (hence, the decrease inthe ATP/ADP ratio).The close correlation between growth rate and

specific activity of adenylate kinase shown inFig. 4 is also puzzling in view of the fact that themass action ratio, expected to be the most sen-sitive indicator of enzyme function in vivo, didnot decrease in the temperature range where theactivity of the enzyme, as measured in extracts,fell sharply with increasing temperature. Theresults indicate that a value of the ATP/AMPratio nearly 50% below its normal range is com-patible with normal growth (mutant strain at300C; Fig. 4). This observation is surprising,since several enzymes have been shown to re-spond to the ATP/AMP ratio. At temperaturesabove 300C, decreases in the adenylate energycharge and in the ATP/ADP ratio accompanieda further decrease in the ATP/AMP ratio, sothat it is not possible to attribute the sharpdecline in growth rate relative to the parentalstrain with confidence to any one factor.The altered primary structure of the mutant

enzyme might lead to lower specific activity ofadenylate kinase in mutant cells growing at per-missive temperatures either by causing a reduc-tion in the turnover number for the enzyme orby increasing the susceptibility of the mutantenzyme to hydrolysis by intracellular proteases.Previous studies (8, 13) have ruled out an in-creased level of general proteolytic activity inthe temperature-sensitive strain, but it is knownthat defective enzymes may be more susceptiblethan normal enzymes to proteolysis (14). Thesusceptibility might become more pronouncedat higher temperatures, and might thus contrib-ute to the gradual decrease in adenylate kinaseactivity in the mutant as a function of growthtemperature above 300C. The inactivation of themutant adenylate kinase in vivo at 360C to 370Cmight also reflect thermal denaturation of the

VOL. 145, 1981


mutant enzyme in addition to or instead of pro-teolytic cleavage. Substrates and ammoniumsulfate offer protection against thermal inacti-vation (8); however, this could equally affecteither mechanism of inactivation. The degree ofinactivation that occurs during incubations ofthe mutant enzyme, at different stages of puri-fication, at different temperatures and other in-cubation conditions, varies somewhat both inour own experiments and according to previousreports (8, 13). In general, however, the mutantenzyme appears to be less sensitive to tempera-ture increase in vitro than in vivo. Furthermore,the enzyme appears to become more heat stableduring the purification procedure. This wouldargue that thermnal inactivation may be at leastin part proteolytic.At the permissive temperature (300C), the

kinetic properties of the temperature-sensitiveadenylate kinase are quite similar to those of theparent adenylate kinase or to those previouslyreported for adenylate kinase from wild-type E.coli (16) except that affinities for AMP, ADP,and ATP are somewhat reduced.The kinetic properties of the parent and the

mutant adenylate kinases are affected differ-ently by changes in the assay temperature. Themaximal velocity of the reaction catalyzed bythe parent enzyme yielded a linear Arrheniusplot over the temperature range studied (30 to400C). The Km values for the three substratesremained nearly constant over the same tem-perature range. In contrast, the maxmum veloc-ity of the reaction catalyzed by the mutant ad-enylate kinase gave a linear Arrhenius plot onlybelow approximately 370C. Above this temper-ature, the experimental points diverged down-ward from linearity, suggesting structural alter-ations. Although nominally suggesting a de-crease in activation energy, this curve probablyreally reflects inactivation of the enzyme. Thereis a progressive twofold or larger decrease in theaffinity of the mutant enzyme for all three sub-strates as the assay temperature is increasedbetween 30 and 400C, again suggesting confor-mational alterations in the mutant enzyme inthis critical temperature range. The reactionremained second order with respect to ADP(Hill slope = 1.8) between 36 and 400C, but thispresumably merely reflects the participation oftwo molecules of ADP in the reaction, ratherthan any regulatory interactions between sites.

In summary, both a reduction in the level ofadenylate kinase and a reduction in the affinityof the enzyme for its substrates probably con-tribute to the decreased adenylate kinase activ-ity and the decreased rate of cell growth ob-served between 30 and 340C in E. coliCR341T28.

ACKNOWVLEDGMENTSWe thank Jack Barber for helpful assistance.This work was supported by Public Health Service grant

AM-09863 from the National Institute of Arthritis, Metabo-lism, and Digestive Diseases and National Science Foundationgrant PCM-76-18321. C.C.G. was supported by Public HealthService Research training grant GM-07185.

LITERATURE CITED1. Atkinson, D. E. 1968. Regulation of enzyme function.

Annu. Rev. Microbiol. 23:47-68.2. Atkinson, D. E. 1977. Cellular energy metabolism and its

regulation, p. 85. Academic Press, Inc., New York.3. Ball, W. J., Jr., and D. E. Atkinson. 1975. Adenylate

energy charge in Saccharomyces cerevisiae during star-vation. J. Bacteriol. 121:975-982.

4. Chapman, A. G., and D. E. Atkinson. 1977. Adeninenucleotide concentrations and turnover rates. Theircorrelation with biological activity in bacteria and yeast.Adv. Microbial Physiol. 15:253-306.

5. Chapman, A. G., L Fall, and D. E. Atkinson, 1971.Adenylate energy charge in Escherichia coli duringgrowth and starvation. J. Bacteriol. 108:1072-1086.

6. Cousin, D. 1967. Mutants thermosensibles d'Escherichiacoli K12. fl. Etude d'une mutation letale affectant uner6action g6n6ratrice d'energie. Ann. Inst. Pasteur(Paris) 113:309-325.

7. Cousin, D., and J. P. Belaich. 1966. Sur une mutationthermosensible d'Escherichia cdi affectant une fonc-tion energ6tique. C. R. Acad. Sci 263:886-888.

8. Cousin, D., G. Buttin, and D. Margarita. 1969. Mutantsthermosensibles d'Escherichia coli K12. m. Une mu-tation letale d'E. coli affectant l'activite de l'adenylatekinase. Ann. Inst. Pasteur (Paris) 117:612-630.

9. Cronan, J. E., Jr. 1978. Molecular biology of bacterialmembrane lipids. Annu. Rev. Biochem. 47:163-189.

10. Cronan, J. E., Jr., T. K. Ray, and P. R. Vagelos. 1970.Selection and characterization of an E. coli mutantdefective in membrane phospholipid biosynthesis. Proc.Natl. Acad. Sci. U.S.A. 65:737-744.

11. Gabriel, 0. 1971. Analytical disc gel electrophoresis.Methods Enzymol. 22:565-577.

12. Gabriel, 0. 1971. Locating enzymes on gels. MethodsEnzymol. 22:5788.

13. Glaser, M., W. Nulty, and P. R. Vagelosa 1975. Role ofadenylate kinase in the regulation of macromolecularbiosynthesis in a putative mutant of Escherichia colidefective in membrane phospholipid biosynthesis. J.Bacteriol. 123:128-136.

14. Goldberg, A. L., and A. C. St. John. 1976. Intracellularprotein degradation in mammalian and bacterial cells:Part 2. Annu. Rev. Biochem. 45:747-803.

15. Hochstadt-Ozer, J., and M. CasheL 1972. The regula-tion of purine utilization in bacteria. V. Inhibition ofpurine phosphoribosyltransferase activities and purineuptake in isolated membrane vesicles by guanosinetetraphosphate. J. Biol. Chem. 247:7067-7072.

16. Holmes, R. K., and AL F. Singer. 1973. Purification andcharacterization of adenylate kinase as an apparentadenosine triphosphate-dependent inhibitor of ribonu-clease II in Escherichia coli. J. Biolk Chem. 248:2014-2021.

17. Holms, W. H., L. D. Hamilton, and A. G. Robertson.1972. The rate ofturnover ofthe adenosine triphosphatepool of Escherichia coli growing aerobically in simpledefined media. Arch. Mikrobiol. 83:95-109.

18. Kohiyama, ML, D. Cousin, A. Ryter, and F. Jacob.1966. Mutants thermosensibles d'Escherichia coli K12.1. Isolement et caract4risa%ion rapid. Ann. Inst. Pasteur(Paris) 110:465-486.

19. Luria, S. E., J. L Suit, and C. A. Plate. 1975. Initiationof transcription is temperature-dependent in an E. coli

J. BACTERIOL.


mutant with a ta adenylate kinase. Biochem. Biophys.Res. Commun. 67:353-358.

20. Noda, L. 1973. Adenylate kinase, p. 279-305. In P. D.Boyer (ed.), The enzymes, 3rd ed., vol. 8. AcademicPress, Inc., New York.

21. Oliver, I. T., and J. L Peel. 1956. Myokinase activity inmicroorganisms. Biochim. Biophys. Acta 20:390-392.

22. Randerath, K., and E. Randerath. 1967. Thin-layerseparation methods for nucleic acid derivatives. Meth-ods Enzymol. 12:323-347.

23. Schaffner, W., and C. Weissmann. 1973. A rapid, sen-

sitive, and specific method for the determination ofprotein in dilute solution. Anal. Biochem. 56:502-514.

24. Strehler, B. L 1963. Bioluminescence assay: principles

and practice, p. 99-181. In D. Glick (ed.), Methods ofbiochemical analysis, vol. 16. John Wiley & Sons, Inc.,New York.

25. Swedes, J. S., M. E. Dial, and C. S. McLaughlin. 1979.Regulation of protein synthesis during energy limitationofSaccharomyces cerevisiae. J. Bacteriol. 138:162-170.

26. Swedes, J. S., R. J. Sedo, and D. E. Atkinson. 1975.Relation of growth and protein synthesis to the adenyl-ate energy charge in an adenine-requiring mutant ofEscherichia coli. J. Biol. Chem. 260:6930-6938.

27. Thompson, S. T., K H. Cass, and E. Stellwagen. 1975.Blue dextran-sepharose: An affinity column for dinu-cleotide fold proteins. Proc. Natl. Acad. Sci. U.S.A. 72:669-672.

VOL. 145, 1981

Documents

Adenylate EnergyCharge Escherichia CR341T28 Properties Heat-Sensitive … · E. COLIWITHHEAT-SENSITIVE ADENYLATE KINASE 1375 avalueofabout0.90(3-5),andproteinsynthesis appears to