Regulation of S-Adenosylmethionine Synthetase in Escherichia · VOL. 104, 1970 E-s 0 c 0. 0' 0 >-02Vr-0l-01 S-ADENOSYLMETHIONINE SYNTHETASE IN E. COLI 7 6 5 4 3 21 0 40 80 120 160

JOURNAL OF BACTERIOLOGY, Nov. 1970, p. 734-747Copyright X 1970 American Society for Microbiology

Regulation of S-Adenosylmethionine Synthetasein Escherichia coli

CAROLINE T. HOLLOWAY, RONALD C. GREENE, AND CHING-HSIANG SU

Veterans Administration Hospital, and Department of Biochemistry, Duke Medical Center,Durham, North Carolina 27705

Received for publication 10 August 1970

Addition of methionine to the growth medium of Escherichia coli K-12 leads to a

reduction in the specific activity of S-adenosylmethionine (SAM) synthetase. Thusthe enzyme appears to be repressible rather than inducible. Mutant strains (prob-ably metJ-) are constitutive for SAM synthetase as well as for the methionine bio-synthetic enzymes, suggesting that the regulatory systems for these enzymes have atleast some elements in common. Cells grown to stationary phase in complete me-

dium, which have low specific activities of the enzymes, were routinely used forderepression experiments. The lag in growth and derepression when these cells are

incubated in minimal medium is shortened by threonine. Ethionine, norleucine, anda-methylmethionine are poor substrates or nonsubstrates for SAM synthetase andare ineffective repressors. Selenomethionine, a better substrate for SAM synthetasethan methionine, is also slightly more effective at repression than methionine. Al-though SAM is considered to be a likely candidate for the corepressor in the controlof the methionine biosynthetic enzymes, addition of SAM to the growth mediumdoes not cause repression. Measurement of SAM uptake shows that too little istaken into the cells to have a significant effect, even if it were active in the controlsystem.

The control of the biosynthesis of the asparticfamily of amino acids in Escherichia coli hasbeen studied by many workers. This system,which has served as a model for the control ofbranched biosynthetic pathways, has been thesubject of several recent reviews (3, 31, 32).The enzymes of this pathway which are uniqueto methionine biosynthesis are repressed byaddition of methionine to the growth medium(16, 23, 24, 30, 33). The structural genes for theseenzymes are scattered in the genetic map of E.coli but, since apparently single gene mutationslead to loss of ability to repress any of them,they all seem to be controlled by a single reg-ulatory system (2, 13, 16, 22). None of thesereports considered control of S-adenosylmethio-nine (SAM) synthetase. We have found thatSAM synthetase is repressed by addition ofmethionine to the growth medium and haveisolated ethionine-resistant mutants of E. coliK-12 which are constitutive for SAM synthetase,cystathionine synthetase, and cystathionase. Theresults suggest that control of SAM synthetaseshares at least some regulatory elements withthat of the methionine biosynthetic enzymes.

MATERIALS AND METHODS

Materials. Glutathione, disodium adenosine tri-phosphate (ATP), and chloramphenicol were ob-tained from the Sigma Chemical Co.; L-threonine,L-norleucine, and L-cystathionine were from Calbio-chem; DL-, D-, and L-methionines and L-ethionine werefrom Mann Research Laboratories; DL-a-methyl-methionine and DL-selenomethionine were from theCyclo Chemical Corp.; reduced nicotinamide adeninedinucleotide (NADH) was from Pabst Laboratories,5, 5' dithio-bis (2-nitrobenzoic acid) from the AldrichChemical Co., and beef heart lactic dehydrogenasefrom the Worthington Biochemical Corp. DisodiumATP-8-14C was obtained from Schwarz BioResearchInc. Unlabeled SAM was prepared by a modificationof the method of Schlenk et al. (25) by using chroma-tographic purification similar to that of Shapiro andEhninger (28). Radioactive SAM was prepared byincubating ATP-8-14C and methionine with partiallypurified yeast SAM synthetase (8) followed by chro-matography on Dowex 50 X-8 (minus 400 mesh) witha gradient from 0 to 6 N HCI. The radioactive SAMpeak fractions were combined and lyophilized. Theunlabeled SAM gave a single ultraviolet-absorbing,ninhydrin-reactive spot on paper electrophoresis. Be-tween 95 and 98% of the radioactivity of the labeledSAM migrated with the unlabeled SAM. Degradation

734

Vol. 104, No. 2Printed In U.S.A.

on July 7, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

S-ADENOSYLMEIHIONINE SYNTHETASE IN E. COLI

-z

U nP( E

r-

t-)E

t (hours) t (hours)FIG. 1. Effect of methionine on derepression of wild-type cells. Wild-type cells were grown in LB broth, har-

vested, and incubated in minimal dextrose medium containing (0) no methionine, (0) 2 X 104 M L-methionineor (A) 10 Mf L-methionine.

products which accumulate on long-term storage canbe removed by rechromatography on Dowex 50 X-2 byusing 1 N HCl to wash off impurities and by eluting theSAM with 3 N HCl.

O-succinylhomoserine was synthesized by themethod of Flavin and Slaughter (7). The product hadthe reported melting point and infrared spectrum.

Bacterial strains. Wild-type E. coli K-12 (X)F+ wasobtained from Samson Gross of the BiochemistryDepartment at Duke University. Several strains whichare constitutive for SAM synthetase (exemplified hereby D8 and E31) were isolated as spontaneous ethio-iine-resistant mutants of wild-type E. coli K-12. Thegenetics of these strains will be reported in detail else-vhere. Transduction experiments have shown them to)e closely linked to the metB locus (C-H. Su, R. C.3reene, and C. T. Holloway, Bacteriol. Proc., p. 136,970), and they are probably mutations at the metJocus described by Lawrence, Smith, and Rowbury13) in Salmonella typhimurium.

Media and growth of cells. Chemically defined mediawere based on a modification of the medium of Davisand Mingioli (4) containing, per liter: 7 g of K2HPO4,3 g of KH2PO4, 1 g of (NH4)2SO4, 0.1 g of MgSO4.7H20, and 5 g of dextrose (or in some cases 5 g ofsodium lactate) with supplementation as required.Complete media were LB broth (27) containing 10 g oftryptone (Difco), 5 g of yeast extract (Difco), and 10 gof NaCl per liter, or nutrient broth (Difco). Cells wereroutinely grown at 37 C and 250 rev/min in a NewBrunswick model G-25 gyrotory incubator shaker.For derepression experiments, cells were grown over-night (-16 hr) in 1-liter Bellco shake flasks contain-ing 500 ml of medium (usually LB broth). An ade-quate volume of culture was centrifuged at 11,000rev/min for 15 min in a Sorvall RC-2B refrigeratedcentrifuge. The medium was discarded, and the cellswere suspended in sufficient minimal medium (0.5%dextrose) to give an absorbancy of 2.0 at 550 nm (-3mg of cells/ml). Portions of the cell suspension

VOL. 104, 1970 735

on July 7, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

HOLLOWAY, GREENE, AND SU

5 ' 1 , 1 : g 04~~~~'0

I-~~~~~~

Z ~ (hors tQors

E 1 2 3 4 5 6 o 1 2 3 4 s 6

C ~~~~~~~0

t (hours) t (hours)

I-u VI~~~~~~~~

-c~~~~~~~~~~~~~~~c

wE

FIG. 2. Effect of threonine on growth and derepression. Wild-type cells grown to stationary phase in LB brothwere incubated in minimal dextrose medium containing (0) no threonine or (0) S X JO3 M L-threonline.

(usually 250 ml) were added to 1-liter Bellco shakeflasks containing specified supplements. The flaskswere shaken to mix the contents and were incubatedat 37 C and 150 rev/min in a New Brunswick modelG77 Metabolyte shaking water bath. Growth was fol-lowed by dilution of 0.5-ml samples of culture to 5 mlwith water and measurement of the absorbancy at 550nm with a Gilford model 240 spectrophotometer. Atperiodic intervals, samples for enzyme assay weretaken as described below.

Harvest and toluenization of cells. Samples of culturefluid were centrifuged for 10 min at 12,000 rev/min inthe cold. The supernatant fluid was discarded, and thecells were washed by suspension in 5 ml of 0.02 Mtris(hydroxymethyl)aminomethane (Tris)-hydrochlo-ride buffer, pH 7.6 (3 C), and centrifugation. If thecells were not to be assayed immediately, the pelletwas stored in a freezer overnight. The washed cellswere suspended in sufficient 0.02 M Tris-hydrochloride,pH 7.6, to give an absorbancy of 20 at 550 nm (cal-culated from the absorbancy of the culture and thevolume of sample taken). Toluene (0.05 volume) was

added, the suspension was incubated at 37 C for 10min, and the cells were stored in ice until the enzymeassays were completed.Enzyme assays. Adenosylmethionine synthetase

activity was assayed by measuring conversion of 14C-ATP to 14C-SAM with Dowex 50 used to separate theradioactive SAM from other labeled compounds.Assay mixtures contained: 150 umoles of Tris-hydro-chloride, pH 8.5; 100 Amoles of KCl; 15 umoles ofMgCl2; 8 ;umoles of glutathione; 20 jAmoles of L-methi-onine; 10 ,umoles of ATP-8-14C (0.25 uCi); and 0.4 mlof toluenized cells, in a total volume of 1 ml. The mix-tures were incubated for 30 min at 37 C, and the reac-tions were stopped by the addition of 1 ml of 6%HC104. The precipitated material was removed bycentrifugation and 1 ml of each supernatant fluid waspipetted pnto a column (0.6 by 4 cm) of Dowex 50X-2(H) (200 to 400 mesh). Each column was washedwith three 10-ml samples of 1 N HCI followed by 5 mlof water to remove residual HCI and 9 ml of 1 N NH3to strip off the adenosylmethionine. The NH3 eluateswere acidified by the addition of 1 ml of concentrated

736 J. BACTERIOL.

on July 7, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

VOL. 104, 1970

E-

s0

c

0.

0'

0 >-

02Vl-01r-0

S-ADENOSYLMETHIONINE SYNTHETASE IN E. COLI

76

5

4

3

21

0 40 80 120 160t (min)

U,

0'

0

Ct> y-

w

I0lCJ

0 40 80 120 160 0 40 80 120 160

t (min) t (min)FIG. 3. Repression of wild-type cells by methionine in the presence of threonine. Cells grown to stationary

phase in LB broth were incubated in minimal dextrose medium containing 5 X 1JO M L-threonine and (I) no

methionine, (0) 2 X J0- M L-methionine, or (AX) 6 X 10' M L-methionine.

HCl and 4-ml samples were counted in the 2: 1 tolueneTriton X-100 counting mixture of Patterson andGreene (19). A series of incubations with differentquantities of toluenized cells gave a slightly curvedplot of enzyme activity versus cell concentration. Athigh rates of SAM synthesis (330 nmoles per incuba-tion mixture per 30 min), specific activities were about25% lower than those with low rates of SAM synthesis(54 nmoles per incubation mixture per 30 min). Inspite of this deviation from linearity, the assay is ade-quate for comparison of cells with large differences inspecific activity. In separate experiments, sonic extractsand toluenized cells gave the same SAM synthetaseactivity. More than 75% of added labeled SAM couldbe recovered after incubation with toluenized cellsunder assay conditions, indicating that extensivedegradation did not take place during the assay. Theactivities of blank incubations without cells or withoutmethionine are comparable and low even when com-pared to the activities of repressed cells (generally lessthan 1 nmole per mg of cells). Cystationine synthetasewas assayed by a modification of the procedure of

Kaplan and Flavin (11) by measurement of the a-keto.butyrate resulting from the elimination reaction ofO-succinylhomoserine. Since at certain stages of thederepression experiments the cells contain a substancethat reacts with lactic dehydrogenase and NADH togive high blanks, the toluenized cells were washed bysedimentation and suspension in an equal volume ofbuffer. Assay mixtures containing: 100 jAmoles of Tris-hydrochloride, pH 8.1; 0.125 ,umole of pyridoxalphosphate; 5 jsmoles of O-succinylhomoserine, andwashed toluenized cells (0.2 ml for wild type, 0.1 ml forconstitutive mutants) in a total volume of 0.5 ml areincubated for 20 min at 37 C. Incubation tubes areplaced in an ice bath, and 0.1 ml of6% HCl04 is addedto each tube. The acid is neutralized by addition of0.03 ml of 2 M potassium carbonate, and an appro-priate volume of supernatant (0.4 ml for wild type,0.2 ml for constitutive mutants) is added to a cuvettecontaining 0.2 jAmole of NADH and sufficient 0.05 M,

pH 7.5, potassium phosphate to give a total volume of1 ml. The absorbancy of the solution at 340 nm wasmeasured; beef heart lactic dehydrogenase (-25 pg)

737

I I I I^

,

rS

on July 7, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

HOLLOWAY, GREENE, AND SU

t (min) t (min)

WGQ2-0=4

E~~~~~~~~

0E 40 8 10 10 0 0 80 10 6

UJ~~~~~~~~~U4

00~~~~~~~~~~1'0

0 40 80 120 160 0 40 80 120 160ot (min) t(min)

FIG. 4. Effect of chloramphenicol on enzyme formation. LB broth-grown wild-type cells were incubated inminimal dextrose medium supplemented with 5 X 103 M L-threonine alone (@), or with the addition of40 ju/mlofchloramphenicol (0).

TABLE 1. Enzyme activities of wild-type andconstitutive strains of Escherichia colis

Enzyme activitiesb

StrainSAII thionine Pysta-synthetase synthetae thons

Wild type 0.037 0.015 0.1E31 0.82 1.83 4.4D8 0.83 1.88 3.2

Cells were grown to stationary phase in LBbroth, harvested, toluenized, and assayed.

b Results are expressed as 10-3 internationalenzyme units per mg of cells.

was added, and after 5 min the absorbancy at 340 nmwas read again. Experimental values were correctedfor NADH disappearance in blank incubation mix-tures without O-succinylhomoserine. For a givensample of toluenized cells, net NADH oxidation isproportional to the quantity of cells. Cells of a metB-mutant (cystathionine synthetaseless) gave no activityin the assay. In separate experiments, toluenized cellsand sonic extracts gave the same cystathionine synthe-tase activities. Cystathionase was assayed by a modifi-cation of the procedure of Flavin (6) by measurementof the rate of cystathionine-dependent reduction ofEllman's reagent [5,5' dithio-bis-(2-nitrobenzoicacid); (5)] at room temperature at 410 nm with aGilford model 2000 recording spectrophotometer.Assay mixtures contained: 1 pmole of L-cystathionine;0.125 ,zmole of MgSO4, 160 Amoles of potassium phos-phate, pH 7.4; 1 umole of Ellman's reagent, and 0.1 ml

738 J. BAcrERIOL.

on July 7, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

S-ADENOSYLMETHIONINE SYNTHETASE IN E. COLI

76

ECS

n44t

z4

4

5

24

E

tA 3'-Co -cE E'. ra .

_. l,,u

0 40 80 120 160t (min)

40 80 120 160t (min)

ofE

_ ._

en v

X E >

_0%

2 8 -B~~~~~~~~

0 40 80 120 160t (min)

t (min)FIG. 5. Effect ofmethionine on enzyme synthesis in strain D8. D8 cells grown to stationary phase on LB broth

were incubated in 5 X 10O X L-threonine-supplemented minimal dextrose medium with the addition of (0) no

methionine, (0) 10-' M L-methionine, or (A) S X O-3 M L-methionine.

of toluenized cells in a total volume of 1 ml. Eachexperimental value was corrected for a blank incuba-tion without cystathionine. The rate of cystathioninecleavage is directly proportional to the amount of agiven toluenized cell preparation added to the incuba-tion mixture. Cells of a metC( mutant (cystathionase-less) had negligible activity. When stationary-phasecells are used, the toluenized cell assay gives slightlylower cystathionase activities (70 to 90%) than is ob-tained with sonic extracts. When using the toluenizedcell assay, we noted that cystathionase activity droppedon incubation of constitutive cells in minimal medium.When these assays were repeated with sonic extracts,no drop in activity was seen, indicating that somechange which influenced the toluenized cell assay(possibly accumulation of an inhibitor which does notdiffuse out of toluenized cells) had occurred, ratherthan a loss in enzyme activity. Thus increases in cysta-thionase may be masked, and the toluenized cell assay

can be used only for qualitative comparison of cellswith relatively large differences in activity.

Uptake measurement. Log-phase cells were har-vested by centrifugation and suspended in sufficientminimal medium (0.5% dextrose) to give an absorb-ance of 10 to 12 at 550 nm. Samples (0.9 ml) of cellswere warmed at 37 C for 5 min, and 0.1 ml of a 10-4 Msolution of labeled compound was added. At indicatedtimes, 0.1-ml samples were pipetted into tubes con-taining 5 ml of minimal medium at room temperature,the contents were mixed with a Vortex mixer and werefiltered through a 25-mm membrane filter (0.45 Ampore size; Millipore Corp.) using a sampling manifold(Millipore model 3025). The tube and filter wererinsed with an additional 2 ml of medium. The firstfiltration was finished about 15 sec after taking thesample, and the wash took another 10 to 15 sec. Con-trols without cells and with boiled cells were done tomeasure nonspecific binding of label. Duplicate sam-

5[

0

VOL. 104, 1970 739

on July 7, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

HOLLOWAY, GREENE, AND SU

la"'0

cG)C _

*a E

0c

w

.-

40 80 120 160t (min)

t (min) t (in)FsG. 6. Effect of methionine on enzyme synthesis in strain E31. E31 cells grown to stationary phase in LB

broth were incubated in 5 X 1JO3 M L-threonine-supplemented minimal dextrose medium with the addition of (0)no methionine, (0) 1-3 M L-methionine, or (/\) 5 X 1O3 M L-methionine.

ples (0.02 ml) of each suspension were pipetted ontodiscs (Millipore) to measure the radioactivity in thesuspension. The discs were placed in counting vialsand dried overnight in a vacuum desiccator. Fivemilliliters of scintillation fluid [0.4% 2,5-diphenyl-oxazole; 0.01% 1,4-bis-2-(5-phenyloxazolyl) benzenein toluene] was added to each vial, and the sampleswere counted in a Packard TriCarb model 3375liquid scintillation spectrometer.

RESULTS

In our initial experiments stationary-phasecells which had been grown on minimal mediumcontaining dextrose had high levels of SAMsynthetase. When such cells were diluted intofresh minimal medium, the specific activity of theSAM synthetase dropped during the early stages

of growth and then rose as the cells entered latelog phase. In medium containing methionine,the specific activity dropped to lower values andthe subsequent rise in specific activity was de-layed. Thus it appeared that addition of methio-nine to the growth medium caused repressionrather than induction of SAM synthetase.

Since derepression can be studied best in cellswith low initial specific activities, several mediawere evaluated for growth of stationary-phasecells with acceptably low enzyme activities. Thelevel of SAM synthetase in stationary-phasecells is a function of many parameters, includingthe nature of the carbon source and the finalpH of the medium as well as amino acid supple-mentation. No attempt has been made to system-atically evaluate these parameters, but cells grown

740 J. BACrERIOL.

t (min)

* =06au

._ E

O E

00

Ift

a

on July 7, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

S-ADENOSYLMETHIONINE SYNTHETASE IN E. COLI

EC0Inin

m

4

0(0

4

t bin)

0 40 80 120 160 0 40 80 120 160t (min) t (min)

FIG. 7. Effect of ethionine on derepression. Wild-type cells grown to stationary phase on LB broth were incu-bated in threonine (S X 10O3 M) containing minimal dextrose medium with (a) no addition, (0) 2 X 10O4 ML-ethionine, or (/\) 103 M L-ethionine.

in complete media generally have lower activitythan those grown in defined media. The lowestactivity was obtained in cells grown in LB broth,so these cells were routinely used in derepressionexperiments. When it is desired to use cellsgrown on defined media, somewhat higher butacceptably low activities are obtained withcells grown in minimal medium containing0.5% sodium lactate in place of dextrose as thecarbon source and supplemented with mM L-methionine.

Cells with low specific activities obtained bygrowth to stationary phase on LB broth wereharvested, suspended in minimal medium, andincubated alone or with methionine supple-mentation. Figure 1 shows the growth andenzyme specific activities of these cells. Thegrowth behavior shows the lag that is charac-

teristic for cells grown in rich medium andtransferred to an unsupplemented medium. Inthe absence of methionine the specific activitiesof all three enzymes increase with time, but, ashas been reported for the enzymes of the methio-nine biosynthetic pathway, the increases are notcoordinate (6). Cystathionine synthetase andSAM synthetase both increase about 10- to 12-foldwhereas cystathionase shows only a small increasein activity. Both levels of methionine supple-mentation prevent any increase in enzyme ac-tivities for the first 4 hr of incubation. After thistime, the cells in the medium supplemented with2 X 10 4 M L-methionine are released from repres-sion (probably because of depletion of the exog-eneous supply of methionine) with cystathioninesynthetase showing the most rapid increase inspecific activity and cystathionase showing the

VOL. 104, 1-970 741

on July 7, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

HOLLOWAY, GREENE, AND SU

I0 40 80 120 160t (min)

t (min) t (min)FIG. 8. Effect ofnorleucine on derepression. LB broth-grown wild-type cells were incubated in minimal dextrose

medium supplemented with S X 103 M L-threonine and (I) no further addition, (0) 2 X 1J4 M L-norleuCine,or (Ax) IlP3ML-norleucine.

slowest increase. Little if any derepression isseen during the 8-hr incubation period in cellsgrown on medium supplemented with mM L-

methionine.Although these experimental conditions allow

demonstration of substantial increases in enzymeactivity, which are prevented by addition ofmethionine to the growth medium, the long lagperiod before derepression is inconvenient.Attempts to reduce the intracellular methioninepool and thus speed up derepression were madeby addition of threonine and lysine to the growthmedium, since these amino acids inhibit the ac-tivities of the aspartokinase I, homoserine dehy-drogenase I complex and aspartokinase III inE. coli K-12 (17, 18, 29). The early stages ofderepression of cells pregrown in LB broth arenot affected by addition of 5 X 10O3 M L-lysine

to the incubation medium. During the laterstages lysine does cause an increase in enzymeactivities above those of the control, but thiseffect has not been further investigated. On theother hand (Fig. 2), 5 X 10O' M L-threoninehas a marked effect on growth and derepressionof LB broth-grown cells. Unlike strain Hfr H,which is inhibited by threonine (17), the growthof the strain used in these experiments is stimu-lated by threonine. The time course of derepres-sion of the enzymes is also accelerated by additionof threonine to the growth medium, with theonset of enzyme synthesis preceding rapid growthin media either with or without threonine. Be-cause of the marked growth stimulation, ap-parently, in this strain, the ability to synthesizethreonine limits protein synthesis after transferfrom LB broth to minimal medium. Threonine

742 J. BACTERIOL.

1

6E

'-44> 3400

49

w n(A =

I EuJ-

Z UX

UJC)Z EZ %A

&O CEZ lbQN

2OiL

I)(-)

on July 7, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

S-ADENOSYLMETHIONINE SYNTHETASE IN E. COLI

76

E'-5,

m4

z

m0L0

Z D- S

Ic4

(noi

40 80t (min)

0 40 80 120 160 0 40 80 120 160t (min) t (min)

FIG. 9. Effect of a-methylmethionine on derepression. Wild-type cells grown on LB broth were incubated inS X 10OuM L-threonine supplemented minimal dextrose medium containing, (0) no a-methylmethionine, (0)2 X 10' m DL-a-methylmethionine, or (A) 6 X 10' m DL-a-methylmethionine.

has a much smaller effect on cells, pregrown inin minimal-sodium lactate-methionine medium,which do not show a lag in growth or derepres-sion. Regardless of the basis of the threonineeffect, incubation of LB broth-grown cells inmedium containing 5 x 10-3 M L-threonineprovides a convenient system for evaluation ofthe repressive effects of methionine and itsanalogues. The more rapid derepression ac-companied by relatively little growth allowsdemonstration of the repressive effects of muchlower concentrations of methionine. As shownin Fig. 3, 2 X 10-5 M L-methionine clearly causesrepression for 80 to 120 min, after which enzymesynthesis begins, probably reflecting depletion ofmethionine. Figure 4 shows that chloramphenicol(40 Ag/ml) stops growth and prevents the in-creases in enzyme activities, which indicates that

protein synthesis is required for the increasesin enzyme activity, and the phenomenon prob-ably represents derepression.

Since the development of enzyme activities arenot coordinate, there was no indication that acommon control system is involved in the regula-tion of their synthesis until several constitutivemutants with elevated activities of all threeenzymes were isolated. Table 1 shows the enzymeactivities of stationary-phase, LB broth-growncells of the wild-type and constitutive strains.The activities of all three enzymes are markedlyincreased in the mutants D8 and E31 when incu-bated in threonine containing minimal mediumalone or with the addition of methionine (Fig.5 and 6). Although there is an increase in thespecific activity of cystathionine synthetase and anapparent decrease in that of cystathionase during

120 160

VOL. 104, 1970 743

on July 7, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

HOLLOWAY, GREENE, AND SU

IEI

z

0

to0nImI

75

4

3

2

An

rn-

_n &£ 0

Om ECU,

._ ,%_ C

.24,

.S-

2,.w

0-'

I0 40 80 120 160t (min)

t (min) t (min)FIG. 10. Comparison of the repressive effects of methionine and selenomethionine. Wild-type cells grown to

stationary phase in LB broth were incubated in threonine (5 X 1O-I M) containing minimal dextrose medium with(0) no additions, (0) 4 X 1J5 M DL-selenomethionine, (AL) 4 X 105 DL-methionine, or (v) 104 M DL-methio-nine.

the incubation, these changes are not influencedby methionine at a concentration (5 X 10-3 M)250 times greater than that (2 X 10 M) whichcauses significant repression in wild-type cells.

Addition of chloramphenicol (40 ,ug/ml) tothe growth medium prevents the increase inspecific activity of cystathionine synthetase butdoes not prevent the apparent drop in activity ofcystathionase. This decrease represents somepeculiarity in the toluenized cell assay procedurerather than a loss of enzyme.

Figures 7 to 9 show the effects of ethionine,norleucine, and a-methylmethionine on growthand derepression of wild type cells. While ethio-nine and norleucine can be used at relatively highconcentrations, a-methylmethionine concentra-

tions higher than those shown inhibit growth,and the results are complicated by its inhibitionof methionyl transfer ribonucleic acid (tRNA)synthetase (26). None of these compounds hasan appreciable effect on the enzyme activitiescompared to that of methionine. Ethionine at10- M exhibits slight repression of cystathioninesynthetase, and norleucine (both at 2 X 10-4and 10- M) slightly stimulates SAM syn-thetase formation. These differences are smalland may not be significant. Selenomethionine,on the other hand, is slightly more effective atpreventing enzyme synthesis than an equal con-

centration of methionine. As shown in Fig. 10,4 x 10- M DL-selenomethionine represses aswell as 10-4 M DL-methiOnine, and both show

t (min)

80

744 J. BACTERIOL.

on July 7, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

S-ADENOSYLMETHIONINE SYNTHETASE IN E. COLI

more prolonged repression than 4 X 105 M DL-methionine.

Because of indications that the corepressor forcontrol of the methionine biosynthetic enzymesmight be SAM or one of its metabolites, thereactivity of the analogues with SAM synthetasewas measured to see whether there was any cor-relation between the repressive activity of acompound and its ability to be converted to anadenosyl derivative. As shown in Table 2, theactivity of E. coli SAM synthetase is approxi-mately the same in the presence of 0.01 and 0.02M L-methionine, indicating that the enzyme issaturated in this concentration range. A smallamount of SAM is synthesized in the reactionmixtures containing D-methionine, but, since theoptical purity of the D-methionme was not tested,the observed activity may be due to L-methioninecontamination of the material used. Regardlessof its reactivity with the enzyme, D-methioninedoes not appreciably inhibit the reaction withL-methionine. Ethionine is a relatively poorsubstrate for E. coli SAM synthetase, whereasnorleucine and a-methylmethionine do not appearto be substrates at all. In contrast to the otheranalogues, selenomethionine reacts with theenzyme more than twice as rapidly as methionine.Thus the analogues which are ineffective atrepression are not good substrates for SAMsynthetase, whereas one analogue, selenomethio-nine, which reacts with the enzyme better thanmethionine, is also slightly more effective atrepression.With this correlation in mind, the ability of

SAM to repress enzyme formation was tested.As shown in Fig. 11, addition of SAM to theincubation medium during derepression of cellspregrown in LB broth has no effect on the de-velopment of enzyme activity. Since equivalentconcentrations of methionine caused significantrepression under these conditions, these resultstended to rule out SAM as an effector. Beforesuch a conclusion could be drawn, however, itwas necessary to measure the uptake of SAM bythe cells. Figure 12 shows the results of one suchuptake experiment. Radioactivity correspondingto approximately 5 nmoles of SAM (adenine-8-14C) was concentrated per gram of cells whenthe rapid stage of uptake was completed. In acontrol experiment with the same cell suspension,the methionine transport system was shown to befunctioning. It is questionable whether anySAM was taken up, since the amount of labelbound to the cells (1.5 %) was less than theradioactive impurities in the SAM preparation(2.5%). Preliminary estimates of the SAM pool

TABLE 2. Reaction of methionine andanalogues with SAM synthetase"

Substrate Activityb

0.01 M L-Methioline. 31.50.02 M L-MethiOline. 31.90.02 M D-MethiOline ..................... 2.10.02 M DL-MethiOnine................... 30.50.04 M DL-Methionine................... 31.70.02 M L-EthiOnile...................... 7.20.02 M DL-a-Methylmethionine........... 00.02 M L-Norleucine...................... 00.02 M DL-Selenomethionine............. 65.6

- The incubation procedure described in thetext was followed except that the ATP-8-14C hadtwice the normal specific activity (0.05 ,uCi/Mmole)and each incubation mixture contained 6 mg oftoluenized E31 cells grown to stationary phasein minimal dextrose medium.

b Results expressed as nanomoles per 30 minper milligram of cells.

in wild-type cells during log phase in minimalmedium are in the range of 30 to 50 nmoles/g.Thus, even if SAM were taken up, the changein the pool size would be too small to causemeasurable repression, and no conclusions aboutthe role of SAM in enzyme regulation can bedrawn from these experiments.

DISCUSSION

SAM synthetase was reported to be inducibleby methionine in yeast (20), but its control hasnot been previously studied in E. coli. The resultspresented here clearly indicate that the enzymeis repressed by addition of methionine to thegrowth medium. The mutants used in theseexperiments cotransduce with metB markerswith high efficiency suggesting that they are inthe metJ locus, which was described as a methio-nine regulatory gene in S. typhimurium (13).Since these mutants are constitutive for cysta-thionine synthetase and cystathionase (and byinference for the other enzymes of methioninebiosynthesis) as well as being constitutive forSAM synthetase, it is reasonable to concludethat SAM synthetase is regulated by the samesystem that controls the enzymes of methioninebiosynthesis.The relative abilities of the methionine

analogues to cause repression of enzyme syn-thesis may reflect directly their reactivities withthe control system or indirectly their abilities tobe converted to compounds which in turn reactwith the control system. MetG mutants of S.typhimurium, which have been shown to have an

VOL. 104, 1970 745

on July 7, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

0 40 80 120 160t (min)

4neW E

Z -F2 ,5

;r-> o.

UK)

I0

0 40 80 120 160 0 40 80 120 160t (min) t (min)

FIG. 11. Effect ofSAM on enzyme synthesis. LB broth-grown wild-type cells were incubated in minimadextrose medium supplemented with 5 X JO-IM L-threonine and (a) no further additions, (0) 2 X 10-5M SAMor (A) 6 X 105m SAM.

6

co0)4

2

CL

0 2 3 4 5 6 7t (min)

FIG. 12. Uptake of SAM. Wild-type cells weregrown in minimal dextrose medium at 37 C to an ab-sorbancy (550 nm) of 1.8. The cells were harvested andsuspended in fresh minimal dextrose medium to give asuspension with an absorbancy (550 nm) of 12.6 (.19mg/ml). The suspension (0.9 ml) was incubated at 37 Cfor 5 min and 0.1 ml ofJ0-4 Mf SAM (8-14C) was addedat zero time. Samples were taken.

altered methionyl (tRNA) synthetase (10), arenot derepressed for the methionine biosyntheticenzymes, suggesting that methionyl tRNA isnot involved in repression. On the other hand,several groups of workers (1, 12, 14, 21) havereported the involvement of SAM in regulationof methionine synthesis, mostly at the level offeedback inhibition. Kerr and Flavin (12) re-cently showed that an ethionine-resistant mutantof Neurospora crassa, originally isolated byMetzenberg (15), has low SAM synthetase andoverproduces methionine. They propose thatin this strain, the methionine overproductionresults from release of feedback inhibition ratherthan derepression. In contrast, we have reported(9) that low SAM synthetase mutants of E. coliK-12 have elevated levels of cystathionine syn-thetase and cystathionase (and presumably ofthe other methionine biosynthetic enzymes).The behavior of these strains as well as the resultswith analogues of methionine are consistent withthe hypothesis that the effector in repression ofthe methionine biosynthetic enzymes in E. coli

746

76

E 5C

In-4-3Uz4

0

4

-

='en

w

ZtA>_ ._"

fnDz EZ >'

VI -

ofin,

0A

A

08

on July 7, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

S-ADENOSYLMETHIONINE SYNTHETASE IN E. COLI

is SAM or one of its metabolites. Unfortunately,the impermeability of E. coli K-12 to SAMprevents a direct test of this hypothesis.

ACKNOWLEDGMENTS

The competent technical assistance of D. J. Peninger, EllenMack, and Alice Estes is gratefully acknowledged.

This investigation was supported by Public Health Serviceresearch grant GM 10317 from the National Institute of GeneralMedical Sciences, and by a fellowship to C.T.H. (grant MH08394) from the National Institute of Mental Health.

LITERATURE CITED

1. Cherest, H., F. Eichler, and H. de Robichon-Szulmajster.1969. Genetic and regulatory aspects of methionine bio-synthesis in Saccharomyces cerevisiae. J. Bacteriol. 97:328-336.

2. Cohen, G. N., and F. Jacob. 1959. Sur la repression de lasynthdse des enzymes intervenant dans la formation dutryptophan chez Escherichia colt. Compt. Rend. H. Acad.Sci. 248:3490-3492.

3. Datta, P. 1969. Regulation of branched biosynthetic path-ways in bacteria. Science 165:556-562.

4. Davis, B. D., and E. S. Mingioli. 1950. Mutants of Escherichiacoil requiring methionine or vitamin Bis. J. Bacteriol. 60:17-28.

5. Ellman, G. L. 1959. Tissue sulfhydryl groups. Arch. Biochem.Biophys. 82:70-77.

6. Flavin, M. 1962. Microbial transsulfuration: the mechanismof an enzymatic disulfide elimination reaction. J. Biol.Chem. 237:768-777.

7. Flavin, M., and C. Slaughter. 1965. Synthesis of the succinicester of homoserine, a new intermediate in the bacterialbiosynthesis of methionine. Biochemistry 4:1370-1375.

8. Greene, R. C. 1969. Kinetic studies of the mechanism ofS-adenosylmethionine synthetase fromi yeast. Biochemistry8:2255-2265.

9. Greene, R. C., C-H Su, and C. T. Holioway. 1-970. S-adeno-sylmethionine synthetase deficient mutants of Escherichiacoll K-12 with impaired control of methionine biosynthe-sis. Biochem. Biophys. Res. Commun. 38:1120-1126.

10. Gross, T. S., and R. J. Rowbury. 1969. Methionyl transferRNA synthetase mutants of Salmonella typhimurium whichhave normal control of the methionine biosynthetic en-

zymes. Biochim. Biophys. Acta 184: 233-236.11. Kaplan, M. M., and M. Flavin. 1966. Cystathionine y-syn-

thetase of salmonella. J. Biol. Chem. 241:4463-4471.12. Kerr, D., and M. Flavin. 1970. The regulation of methionine

synthesis and the nature of cystathionine y-synthase inneurospora. J. Biol. Chem. 241:1842-1855.

13. Lawrence, D. A., D. A. Smith, and R. J. Rowbury. 1968.Regulation of methionine synthesis in Salmonella typhi-murium: mutants resistant to inhibition by analogues ofmethionine. Genetics 58-:473-492.

14. Lee, L.-W., J. M. Ravel, and W. Shive. 1966. Multimetabo-lite control of a biosynthetic pathway by sequential metabo-lites. J. Biol. Chem. 241:5479-5480.

15. Metzenberg, R. L. 1968. Repair of multiple defects of a

747

regulatory mutant of Neurospora by high osmotic pressureand by reversion. Arch. Biochem. Biophys. 125:532-541.

16. Patte, J.-C., G. LeBras, and G. N. Cohen. 1967. Regulationby methionine of the synthesis of a third aspartokinase andof a second homoserine dehydrogenase in Escherichia coliK12. Biochim. Biophys. Acta 136: 245-257.

17. Patte, J.-C., G. LeBras, T. Loviny, and G. N. Cohen. 1963.Retro-inbibition et repression de I'homoserine deshydro-genase d'Escherichia coli. Biochim. Biophys. Acta 67:16-30.

18. Patte, J.-C., P. Truffa-Bachi, and G. N. Cohen. 1966. Thethreonine-sensitive homoserine dehydrogenase and asparto-kinase activities of Escherichia coli. Biochim. Biophys.Acta 128:426-438.

19. Patterson, M. S., and R. C. Greene. 1965. Measurement oflow energy beta-emitters in aqueous solution by liquidscintillation counting of emulsions. Anal. Chem. 37:854-857.

20. Pigg, C. J., W. A. Sorsoli, and L. W. Parks. 1964. Inductionof the methionine-activating enzyme in Saccharomycescerevisiae. J. Bacteriol. 87:920-923.

21. de Robichon-Szulmajster, H., and H. Cherest. 1967. Regula-tion of homoserine 0-transacetylase, first step in methio-nine biosynthesis in Saccharomyces cerevisiae. Biochem.Biophys. Res. Commun. 28:256-262.

22. Rowbury, R. J. 1965. Resistance to norleucine and control ofmethionine synthesis in Escherichia coli. Nature (London)206:962-963.

23. Rowbury, R. J., and D. D. Woods. 1961. Further studies on

the repression of methionine synthesis in Escherichia coli.J. Gen. Microbiol. 24:129-144.

24. Rowbury, R. J., and D. D. Woods. 1966. The regulation ofcystathionine formation in Escherichia coli. J. Gen. Mi-crobiol. 42:155-163.

25. Schlenk, F., J. L. Dainko, and S. M. Stanford. 1959. Im-proved procedure for the isolation of S-adenosylmethio-nine and S-adenosylethionine. Arch. Biochem. Biophys.83:28-34.

26. Schlesinger, S. 1967. Inhibition of growth of Escherichia coliand of homoserine O-transsuccinylase by a-methylmethio-nine. J. Bacteriol. 94:327-332.

27. Scott, J. R. 1968. Genetic studies on bacteriophage P1. Vir-ology 36:564-574.

28. Shapiro, S. K., and D. J. Ehninger. 1966. Methods for theanalysis and preparation of adenosylmethionine andadenosylhomocysteine. Anal. Biochem. 15:323-333.

29. Stadtman, E. R. 1961. Feed-back inhibition and repressionof aspartokinase activity in Escherichia coil and Saccharo-myces cerevisiae. J. Biol. Chem. 236:2033-2038.

30. Taylor, R. T., H. Dickerman, and H. Weissbach. 1966. Con-trol of one carbon metabolism in a methionine-Bi2 auxo-

troph of Escherichia coli. Arch. Biochem. Biophys. 117:405-412.

31. Truffa-Bachi, P., and G. N. Cohen. 1968. Some aspects ofamino acid biosynthesis in microorganisms. Annu. Rev.Biochem. 37:79-108.

32. Umbarger, H. E. 1969. Regulation of amino acid metabolism.Annu. Rev. Biochem. 38:323-370.

33. Wijesundra, S., and D. D. Woods. 1960. Suppression ofmethionine synthesis in Escherichia coil by growth in thepresence of this amino acid. J. Gen. Microbiol. 22:229-241.

VOL. 104, 1970

on July 7, 2020 by guesthttp://jb.asm

.org/D

ownloaded from