crystallopoietes and a morphogenetic mutant - journal of bacteriology

JOURNAL OF BACTERIOLOGY, June 1978, p. 1064-10730021-9193/78/0134-1064$02.00/0Copyright © 1978 American Society for Microbiology

Vol. 134, No. 3

Printed in U.S.A.

Regulation of Cyclic AMP Levels in Arthrobactercrystallopoietes and a Morphogenetic Mutant

R. W. HAMILTONt AND P. E. KOLENBRANDER* *

Department ofMicrobiology and Cell Biology, The Pennsylvania State University, University Park,Pennsylvania 16802

Received for publication 16 December 1977

The extracellular levels of cyclic AMP (cAMP), cAMP phosphodiesteraseactivity, and adenylate cyclase activity were measured at various intervals duringgrowth and morphogenesis of Arthrobacter crystallopoietes. There was a signif-icant rise in the extracellular cAMP level at the onset of stationary phase, andthis rise coincided with a decrease in intracellular cAMP. The phosphodiesteraseactivity measured in vitro increased in the early exponential phase of growth asintracellular cAMP decreased, and, conversely, prior to the onset of stationaryphase the phosphodiesterase activity decreased as the intracellular cAMP levelsincreased. Adenylate cyclase activity was greater in cell extracts prepared fromcells grown in a medium where morphogenesis was observed. Pyruvate stimulatedadenylate cyclase activity in vitro. A morphogenetic mutant, able to grow only asspheres in all mnedia tested, was shown to have altered adenylate cyclase activity,whereas no significant difference compared to the parent strain was detectable ineither the phosphodiesterase activity or the levels of extracellular cAMP. Theroles of the two enzymes, adenylate cyclase and phosphodiesterase, and excretionofcAMP are discussed with regard to regulation of intracellular cAMP levels andmorphogenesis.

Ad n i characteristic of members ofthe genus Arthrobacter is their ability to un-dergo a unique morphogenetic cycle. Stationary-phase cells are spherical, but upon inoculationinto a rich organic medium, the spheres elongateinto rod-shaped cells and continue to grow anddivide as rods throughout exponential growth.At the onset of the stationary phase, the rod-shaped cells return to a spherical form by reduc-tive cell division or fragmentation of the rods.Arthrobacter crystalpoietes offers an addi-

tional advantage for study in that the cellularmorphology can be nutritionally controlled (1).CelLs growing in a glucose-based minimal me-dium grow only as spheres and fail to undergothe normal morphogenetic cycle. However, ad-dition of one ofa number of specific amino acidsor organic acids such as asparagine, succinate, orbutyrate results in an increased growth rate andinduction of sphere-to-rod morphogenesis. Theinducer compound is the preferred substrate forgrowth, and its depletion from the medium re-sults in resumption of the slower growth rateand a return to spherical-shaped celLs.

t Present address: Department of Microbiology, M.S. Her-shey Medical Center, The Pennsylvania State University,Hershey, PA 17033.

t Present address: Microbiology Section, Laboratory ofMicrobiology and Immunology, National Institute of DentalResearch, Bethesda, MD 20014.

In a previously published report (3) we dem-onstrated that cyclicAMP (cAMP) was involvedin morphogenesis of A. crystallopoietes. In thisreport it was found that the intracellular concen-tration of cAMP remained relatively constantduring growth of A. crystallopoietes as spheresin glucose-based medium (see Fig. 1A). How-ever, after inoculation ofglucose-grown cells intosuccinate-based medium, there was an immedi-ate 30-fold rise in intracellular levels of cAMP,which then rapidly fell to a rather stable levelthroughout exponential growth (see Fig. 1B).This level was four- to fivefold higher than thatfound in glucose-grown cells. This initial peak ofintracellular cAMP just preceded the morpho-logical change of spherical- to rod-shaped cells.At the onset of the stationary phase, there wasa second peak of intracellular cAMP, which rap-idly dropped to a stable low concentration instationary-phase cells. Again, this peak of intra-cellular cAMP just preceded a point of morpho-logical change, i.e., the change from rod- tospherical-shaped cells.An investigation of A. crystallpoietes, Mph-

3, a morphogenetic mutant ofA. crystalopojeteswhich is unable to undergo morphogenesis andgrows solely as spheres in both succinate- andglucose-based media, revealed the absence ofthe two peaks of intracellular cAMP (3). More-over, the intracellular cAMP level in spherical

.064

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

05

Dec

embe

r 20

21 b

y 17

3.24

7.13

1.13

7.

INTRACELLULAR CYCLIC AMP IN ARTHROBACTER 1065

cells growing in succinate-based medium wasnearly identical to the parent strain growing asspheres in glucose-based medium (see Fig. 10).This report is concerned with the mechanisms

involved in the control of cAMP levels duringgrowth and morphogenesis of A. crystallo-poietes. These mechanisms include the extrac-tion of cAMP, hydrolysis of cAMP by phospho-diesterase, and formation ofcAMP by adenylatecyclase. The regulation of adenylate cyclase ac-tivity and the relationship ofcAMP to cataboliterepression in A. crystallopoietes compared tothat reported in other procaryotic systems arediscussed.

MATERIALS AND METHODSStrains and culture conditions. A culture of A.

crystallopoietes (ATCC 15481) was obtained from J.C. Ensign, Department of Bacteriology, University ofWisconsin. The morphological mutant of A. crystal-lopoietes, Mph-3, was obtained as previously described(3). Cultures were grown in a mineral salts phosphatemedium as described previously (3). Carbon sources,glucose and sodium succinate, were filter sterilizedseparately as concentrated solutions and added afterautoclaving to the mineral salts-phosphate growthmedium.

For measurement of extracellular cAMP, cultureswere grown at 30°C with shaking aeration in 1,500-micapacity Erlenmeyer flasks containing 200 ml of me-dium. Bacterial growth was monitored by use of aKlett-Summerson colorimeter with a red filter (660nm).

Cells used to prepare cell extracts for assay ofphosphodiesterase and adenylate cyclase were grownin a 20-liter fermentor (Microferm Fermenter modelMF-1285, New Brunswick Scientific Co., New Bruns-wick, N.J.). Incubation was at 30°C, with aeration at40 liters/min and agitation set at 350 rpm.Assay of cAMP. The protein-binding assay of Gil-

man (2) was used to measure cAMP levels as previ-ously described (3). Concentrations of cAMP in bac-terial extracts were determined from a standard curve,which was linear in the range between 1 and 20 pmolof cAMP.

Extracellular cAMP was measured by filtering 10ml of culture fluid through a membrane filter (Milli-pore Corp., 0.45-eum porosity, 25 mm in diameter) intoa screw-capped test tube which was then capped andheated for 10 min at 95°C. Samples of the filteredculture medium were assayed in duplicate for cAMPby the protein-binding assay by use of test kits sup-plied by Boehringer Mannheim Corp.

Preparation of cell extract for enzyme assays.Samples of 500 to 2,000 ml were removed from thefermentor and immediately cooled to 4°C with ice.The bacterial cells were pelleted by centrifugation at10,000 x g for 20 min at 4°C; the supernatant wasremoved and discarded. Cell pellets were suspended in5 to 10 ml of 10 mM tris(hydroxymethyl)amino-methane-hydrochloride (pH 8.0), and the bacterialcells were disrupted by three passages through anAminco French pressure cell at 20,000 lb/in2. Theresulting crude extract was clarified by centrifugation

at 15,000 x g for 15 min. The clarified supernatantcontaining the enzyme activities was removed andstored at -55°C until assayed. Protein concentrationof the clarified crude extracts was determined by themethod of Lowry et al. (9) using bovine serum albuminas the standard.

Phosphodiesterase assay. The activity of cAMPphosphodiesterase was assayed with slight modifica-tion of the procedure described by Thompson andAppleman (24). The complete reaction mixture con-tained (final concentration in 250 ul): 32 mMtris (hydroxymethyl) aminomethane -maleate - hydro-chloride (pH 7.0), 2 mM dithiothreitol, 6.25 mMMgCl2, 0.5 mM cAMP, and enough [3H]cAMP (20Ci/mmol) to give approximately 200,000 cpm per assaytube. The reaction was started by the addition of 100Id of crude extract (containing 0.2 mg of protein). Thereaction was incubated for 30 min at 30°C and thenterminated by heating for 2.5 min at 95°C. Aftercooling to 30°C, 50 jig (in 50,l of distilled water) ofsnake venom (Crotalus atrox) was added, and themixture was incubated for 15 min. The reaction vessels(10- by 75-mm borosilicate test tubes) were placed inan ice bath and cooled to 4°C, and 1 ml of 1:3 (wt/vol)15% ethanolic slurry of Bio-Rad resin (AG2-X8, 200 to400 mesh) was added, blended in a Vortex mixer, andallowed to equilibrate for 15 min. The tubes were thenblended in a Vortex mixer and centrifuged for 5 min atmaximum speed in a swinging-bucket rotor in a clinicalcentrifuge. The resin formed a pellet at the bottom ofthe test tube, and 0.5 ml of the resulting supernatantwas drawn off and added to 10 ml of Hydromix (York-town Research, Hackensack, N.J.) for scintillationcounting. Crude extract added to the reaction mixtureand immediately boiled was used as a control. Underthese conditions, the formation of product in the phos-phodiesterase reaction mixture was linear with respectto time for 60 min.Adenylate cyclase assay. The adenylate cyclase

assay followed the two-column method of Salomon etal. (18) with minor modification. The reaction mixturecontained 25mM Tris-hydrochloride (pH 7.9), 5.0 mMMgCl2, 20 mM creatine phosphate, 100 U of creatinephosphokinase per ml, 0.75 mM cAMP, 5 mM ATP,1,500,000 cpm of a-[32P]ATP (11 Ci/mmol), 1 mg ofbovine serum albumin per ml, 2.5 mM dithiothreitol,and, if indicated, 75 mM sodium pyruvate. The reac-tion was started by the addition of 20 Ml of crudeextract (40,ug of protein), bringing the total incubationmixture to 50 ul. The reaction mixture was incubatedat 30°C for 45 min, and the reaction was terminatedby heating for 2 min at 95°C. [3H]cAMP (approxi-mately 30,000 cpm) in 50 I1 of distilled water wasadded to monitor cAMP recovery in the followingchromatographic procedures.

After addition of 0.8 ml of water to each reactiontube, the tube contents were mixed and decanted intoa column (2.5-ml Sarpette pipette tips, Walter SarstedtInc., Princeton, N.J.) containing 1 ml of Bio-Rad cat-ion exchange resin (AG 50W-X8, 200-400 mesh, hydro-gen form). The eluants from this and two successive1-ml glass-distilled water washes were discarded. A 3-ml volume of glass-distilled water was then added tothe column; the eluant passed directly into a columncontaining 0.6 g of neutral alumina (which had beenpreviously washed with 8.0 ml of 0.1 M imidazole), and

VOL. 134, 1978

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

05

Dec

embe

r 20

21 b

y 17

3.24

7.13

1.13

7.

1066 HAMILTON AND KOLENBRANDER

the subsequent eluant was discarded. A 4-ml volumeof 0.1 M imidazole was then added to the column, andthe eluant was collected directly into a scintillationvial. A 10-ml volume of Hydromix (Yorktown Re-search) was added for scintillation counting of theaqueous sample.The columns can be recycled for further use by

addition of 5 ml of 0.1 N HCI to the Bio-Rad AG 5OW-X8 cation exchange columns, followed by a 10-ml,glass-distilled water wash. The neutral alumina col-umns can be recycled by washing with 10 ml of 0.1 Mimidazole.

In A. crystalopoietes, the rate of reaction of ade-nylate cyclase was maximal at pH 7.9 using tris-(hydroxymethyl)aminomethane-hydrochloride buffer.A divalent cation was essential for the reaction, andaddition of 5.0 mM Mg2" resulted in optimal activity.It was found that Mn2" at 5.0mM concentration couldreplace Mg2e, and was equally effective as Mg2. Often,cell-free preparations of adenylate cyclase are unsta-ble, and although this is more of a problem in partic-ulate preparations, e.g., plasma membranes, it can alsobe a problem in soluble preparations (20). It wasnecessary to stabilize the adenylate cyclase ofA. crys-tallopoietes with 1 mg of bovine serum albumin perml to achieve proportional product formation over an

extended period of time.Many crude preparations used for adenylate cyclase

assays contain high levels of ATP-degrading enzymes.Therefore, the ATP-regenerating system was added tothe reaction mixture. Creatine phosphate and creatinephosphokinase were used for this purpose as the phos-phoenolpyruvate and pyruvate kinase-regeneratingsystem has been demonstrated to be capable of acti-vating or inhibiting adenylate cyclase (20). The 0.75mM cAMP which was added to the reaction mixtureserved a dual purpose; it reduced interference due tophosphodiesterase activity and it could prevent anaccelerated formation of product during boiling, whichwas used to terminate the reaction (20).The amount of cyclic nucleotide (product) degra-

dation in the assay was examined using A. crystallo-poietes extract from succinate-grown cells containing8 U of phosphodiesterase activity (the most activepreparation tested) as measured in the phosphodies-terase assay. This phosphodiesterase preparation iscatalytically capable of degrading at least 50% of thecAMP present in the adenylate cyclase reaction mix-ture. However, under the conditions employed in theadenylate cyclase assay, the phosphodiesterase activ-ity was much lower and capable of degrading only2.5% of the cAMP (Table 1). This low value of activityprobably resulted because the higher pH that wasrequired for the adenylate cyclase activity (pH 7.9)was well above the optimal pH for phosphodiesteraseactivity (pH 7.0). The product of the adenylate cyclaseassay was identified as cAMP (Table 2) by using thin-layer chromatography as outlined below. The forma-tion of cAMP during the course of the adenylatecyclase reaction was linear for at least 45 min assayedin the presence or absence of pyruvate.

Thin-layer chromatography. The thin-layerchromatographic analysis employed was a modifiedprocedure of Tao (22). Using a capillary tubing, theradioactive sample was applied to the cellulose thin-

TABLE 1. Phosphodiesterase activity detectable inthe reaction mixture used to measure adenylate

cyclase activityActivity (cpm) recovered from:

Percent

Expt Cc Remain- cpm re-CacAMP spot der of the coveredgram chromat- as cAMP

ogram

1 1,560,000 1,521,000 39,000 97.52 1,125,000 1,095,000 30,000 97.4

a The reaction was identical to the standard reactionof adenylate cyclase, except for the addition to thereaction mixture of 2,000,000 cpm of [3H]cAMP (20Ci/mmol) to monitor phosphodiesterase activity.After 45 min of incubation at 30'C, the reaction wasterminated by heating, and the reaction tubes werecentrifuged in a clinical centrifuge to pellet denaturedprotein. A 40-pl volume of the 50-1 reaction mixturewas spotted and developed on the cellulose chromat-ogram.

TABLE 2. Identification ofproduct of adenylatecyclase assay as cAMP

Approx cpm of :]P recovered from: Percent

Expta cpm in as- ATP, Columns identi-say mix- APan as CAMP fled as

tue APsoAMP spot cAMP b

1 2,000,000 1,819,810 15,380 15,300 1012 22,000,000 1,805,830 13,937 14,450 97

a The reaction was the standard reaction for ade-nylate cyclase as described in Materials and Methods.After terminating the reaction by heating, two sampleswere processed by the two-column method, and twosamples were centrifuged in a clinical centrifuge topellet denatured protein. A 5-jl volume of each super-natant was spotted onto a cellulose thin-layer sheet,which was developed as described in Materials andMethods.

b Percent identified as cAMP = [(counts per minuteof cAMP from columns)/(counts per minute fromcAMP spot)] x 100.

layer sheet (Eastman Chromogram Sheet 6064, with-out fluorescent indicator) and dried under a stream ofcool air, and 5 IAI of a 5-mg/ml nonradioactive cAMPsolution was added as carrier. The sheet was placed inan Eastman Chromogram developing apparatus(model 6071) and developed at room temperature witha solvent system composed of 1.0 M ammonium ace-tate-95% ethanol (30:75, vol/vol). After the solventfront had migrated to the top of the thin-layer sheet,the chromatogram was removed and allowed to dry.The nucleotides were visualized under short-wave UVlight (254 nm). Each of the sample lanes was cut into8-mm strips, and each strip was placed in 7 ml ofscintillation fluid consisting of 6.4 g of 2,5-diphenylox-azole dissolved in 800 ml of toluene and 300 ml of 2-methoxyethanol. Radioactivity was measured in aBeckman model LS 3155T scintillation counter.The cAMP, with an Rf of 0.44, was well separated

from ATP, ADP, and AMP, which have Rf values of

J. BACTrERIOL.

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

05

Dec

embe

r 20

21 b

y 17

3.24

7.13

1.13

7.


0.02, 0.04, and 0.11, respectively, and from adenosineand adenine, which have respective R( values of 0.63and 0.58 (22).

Materials. [8-3H]cAMP (specific activity, 20Ci/mmol) was purchased from Schwarz/Mann (Or-angeburg, N.Y.), a-[32P]ATP (11 Ci/mmol) was sup-plied by Amersham/Searle (Arlington Heights, IM.).Cyclic nucleotides, their dibutyryl derivatives, phos-phorylated nucleic acid bases, dithiothreitol, imidaz-ole, creatine phosphate, creatine phosphokinase,snake venom, and neutral alumina (WN-3) were pur-chased from Sigma Chemical Co. (St. Louis, Mo.). 2-Methoxyethanol was supplied by Eastman Kodak Co.(Washington, D.C.). All other chemicals were obtainedcommercially and were of the highest purity available.

RESULTSExtraceliular concentration of cAMP.

The concentration ofcAMP found in the growthmedium for A. crystallopoietes growing in glu-cose-based medium is illustrated in Fig. 1A. NocAMP was immediately detectable, but a smallamount (0.5 x 10-9 M) was found at mid-expo-nential growth and continued to increase at aslow rate until the onset of the stationary phase.At this point there was a significant rise in theexternal cAMP concentration (Fig. 1A, dottedline). After this rise, the level of cAMP in themedium remained relatively constant through-out the stationary phase at 20 x 10-9 M.

In succinate-based medium the extracellularcAMP could be detected early in the exponentialgrowth phase (Fig. 1B). The level rose relativelyslowly throughout this growth phase despite theelevated level of internal cAMP during exponen-tial growth. It was demonstrated previously thatduring growth in succinate there is a suddenincrease and ensuing decrease in internal cAMPat the onset of stationary phase, coincident withthe change from rod- to spherical-shaped cells(3). It should be noted that, just prior to thehighest point in this peak of internal cAMP, theexternal cAMP level increased and ,continued toincrease as the internal level decreased. It ispossible that the internal level after the secondpeak was regulated by excretion of the cyclicnucleotide. During stationary phase the externallevel continued to increase but at a markedlyslower rate and reached levels of approximately90 x i0-9 M.The external cAMP levels in the morphoge-

netically altered mutant, Mph-3, growing on suc-cinate nearly paralleled that of the parent strain,but reached slightly lower levels (about 75 x10-9 M) during stationary phase (Fig. 1C). Onceagain the rise in external cAMP concentrationoccurred at the point of decrease in intemalcAMP. However, the change in the internal levelwas slight when compared to the parent strain,

whereas the external levels ofcAMP were nearlyidentical in the two strains. It is possible that inMph-3 any cAMP synthesized during the onsetof stationary phase is immediately excreted intothe medium and not allowed to accumulate inthe cell.The pattern of cAMP excretion (as well as

internal cAMP levels) by Mph-3 growing in glu-cose-based medium was nearly identical to theparent strain growing in this medium (Fig. 1A).However, the levels of external cAMP onlyreached 15 x 10'- M during stationary phase(data not shown).Phosphodiesterase activity during

growth. The phosphodiesterase activity mea-sured using cells of the parent strain grown inglucose-based medium is shown in Fig. 2A. Mid-exponential glucose-grown cells were transferredto fresh glucose medium, and cells were har-vested and assayed for phosphodiesterase activ-ity at various intervals during growth. The levelof phosphodiesterase activity remained rela-tively constant at 2.0 U throughout exponentialgrowth. I early stationary phase there was anincrease in activity which soon leveled off at 4.0U of activity. The pattern of phosphodiesteraseactivity in the mutant strain, Mph-3, grown inglucose-based medium was similar to the parentas demonstrated in Fig. 2D, although the expo-nential values of activity were slightly higher at3.0 U, and the decrease in activity during expo-nential phase began earlier in the mutant thanin the parent strain.An examination of the parent strain growing

in succinate-based medium gave a different pro-file of activity (Fig. 2B). The levels of phospho-diesterase activity during exponential growthrose to a much higher value than found in glu-cose-grown cells. The activity of phosphodies-terase continued to rise, from 3.5 to 8.0 U, as theintemal cAMP level decreased from its initialpeak value (cf. Fig. 1B). This suggests that theintemal concentration ofcAMP during exponen-tial growth may be controlled by the phospho-diesterase activity since no rise in extemalcAMP was found at this point, and the level ofinternal cAMP falls much too rapidly to be dueto dilution of cyclic AMP by cell growth. In lateexponential phase there was a drop in phospho-diesterase activity, and this drop just precedesthe second rise in the intemal cAMP level (Fig.2B, dotted line). During the stationary phase thephosphodiesterase activity remained relativelyconstant at 4 U, which is similar to the levelobserved in glucose-grown cells. Mph-3 growingin succinate-based medium showed a pattem ofphosphodiesterase activity which was nearlyidentical to wild-type cells on that medium, ex-cept for a slight increase in activity of phospho-

VOL. 134, 1978

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

05

Dec

embe

r 20

21 b

y 17

3.24

7.13

1.13

7.


C SUCCINATE *. .1020-200

a-150 a7I-

4U

50.0 -3E

10 20HOURS

FIG. 1. Extracellular and intracellular levels of cAMP and cell growth of A. crystallopoietes in glucosemedium (A) or succinate medium (B), and ofA. crystallopoietes, Mph-3, in succinate medium (C). Dotted lineis a reference marker and will be at the same position in subsequent figures. Symbols: 0, bacterial growth;A, intracelular cAMP; *, extracellular cAMP.

diesterase during early exponential growth,where there was a peak value of 12.5 U of activity(Fig. 2C).Adenylate cyclase activity during

growth. A comparison of the internal levels ofcAMP in Mph-3 and parent strain, each growingon succinate, indicated an alteration in the reg-ulation of cAMP metabolism in the mutant.However, an investigation of the external cAMP

levels and phosphodiesterase activities of themutant showed they were nearly identical tothose observed in wild-type cells. The absenceof significant differences in the mutant and par-ent strains regarding the external cAMP levelsand phosphodiesterase activity led to an inves-tigation of adenylate cyclase, the enzyme re-

sponsible for the formation of cAMP.The activity of adenylate cyclase in the parent

E0 200(100,-zm 40

w20-J

r 200D. 100I-zm 40

20w

r- 20010

cn

z:D 40

20w-Jy

900

50t0

30o.4

lo U

J. BACTERIOL.

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

05

Dec

embe

r 20

21 b

y 17

3.24

7.13

1.13

7.

INTRACELLULAR CYCLIC -AMP IN ARTHROBACTER 1069

HOURS HOURSFIG. 2. Phosphodiesterase activity and ceUgrowth FIG. 3. Adenylate cyclase activity and cell growth

of A. crystallopoietes growing in glucose (A) or suc- in A. crystallopoietes during growth in glucose (A) orcinate (B); and A. crystallopoietes, Mph-3, growing succinate (B); and A. crystallopoietes, Mph-3, grow-in succinate (C) or glucose medium (D). Phosphodi- ing in succinate (C) or glucose (D). Adenylate cyclaseesterase activity is plotted as units of activity, and I activity is plotted as units of activity, and I U equalsU of activity equals 1 nmol of cAMP converted to I nmol ofcAMP formed from ATPper min per mg ofAMP per min per mg of protein. Dotted lines are protein. Dotted line is a reference line for comparisonreference markers and refer to Fig. 1. Symbols: 0, with other figures. Symbols: 0, cell growth; U, unitsbacterialgrowth; A, units ofphosphodiesterase activ- of adenylate cyclase activity.ity.

strain during growth in glucose-based medium isshown in Fig. 3A. As expected from the internalcAMP data, the adenylate cyclase activity re-

mained low throughout exponential growth (lessthan 3 U). There was a slight increase in activityprior to the onset of stationary phase whichcoincided with the excretion of cAMP into the

VOL. 134, 1978

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

05

Dec

embe

r 20

21 b

y 17

3.24

7.13

1.13

7.


medium. The activity continued to increase untillate stationary phase, where the activity reached15U.When glucose-grown cells (100 Klett units,

adenylate cyclase activity, 0.7 U) were trans-ferred to succinate-based medium, there was animmediate increase of adenylate cyclase activityto approximately 15 U, or a 20-fold rise in activ-ity (Fig. 3B). The pattern of adenylate cyclaseactivity fluctuated during exponential growthwith an increase to 27 U of activity early ingrowth and decreasing to 8 U of activity at mid-logarithmic growth. There was a small increasein activity to 15 U before stationary phase, anda rise to 50 U of activity in stationary-phasecells.The mutant, Mph-3, showed an altered pat-

tern of activity when grown in succinate-basedmedium (Fig. 3C). There appeared to be littlechange in the adenylate cyclase activity, whichremained fairly constant around 2 to 4 Uthroughout exponential growth. In general theactivity appeared to be four- to fivefold less thanthe parent strain, which is consistent with thedecreased levels of intracellular cAMP found inMph-3.The pattern of adenylate cyclase activity in

Mph-3 grown in glucose-based medium.was alsoaltered from that of the wild type (Fig. 3D).However, in this medium the activity wasgreater than that of the parent strain. The sig-nificance of this finding is not certain, since theinternal levels of cAMP in Mph-3 parallel thelevels of the wild-type strain on glucose. Whencompared to the parent strain, the difference inadenylate cyclase activity in the mutant duringgrowth in either succinate or glucose medium isevident and suggests that the adenylate cyclaseis altered in Mph-3.Adenylate cyclase activity in the pres-

ence ofpyruvate. In many procaryotic systemsthe activity of adenylate cyclase is affected bysmall molecules (5, 6, 12, 23, 26). Pyruvate andother keto-group compounds have been shownto stimulate adenylate cyclase activity of Brev-ibacteriwn liquefaciens both in vivo and in vitro(4, 10). Pyruvate was also found to stimulateadenylate cyclase activity in A. crystallopoietes.An exination of the adenylate cyclase ac-

tivity of glucose-grown A. crystalopoietes in thepresence of 75mM pyruvate is shown in Fig. 4A.The activity appeared to remain relatively con-stant throughout growth at approximately 55 Uof activity, except for a slight decrease to 43 Uat the onset of stationary phase. This activity ismore than 20 times the activity observed withno pyruvate. When mid-exponential glucose-grown cells (activity 60 U) were transferred tosuccinate, there was an immediate decrease in

FIG. 4. Adenylate cyclase activity measured in thepresence ofpyruvate and cell growth ofA. crystallo-poietes in glucose (A) or succinate (B); and A. crys-tallopoietes, Mph-3, growing in succinate (C) or glu-cose (D). Adenylgte cyclase activity isplotted as unitsof activity and I U equals I nmol of cAMP fonnedfrom A7TPper min per mg ofprottin. Dotted line is areference line for comparison with otherfigures. Sym-bolk: 0, bacterial growth; U, units of adenylate cy-clase activity.

adenylate cyclase activity to about 20U of activ-ity (Fig. 4B). Tis value rose to 35 U in 3 h. The

J. BACTERIOL.

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

05

Dec

embe

r 20

21 b

y 17

3.24

7.13

1.13

7.


activity then dropped until the cells reachedabout 80 Klett units of turbidity. Just prior tothe second peak of internal cAMP (Fig. 4B,dotted line), there was a peak in the adenylatecyclase activity (50 U), and the activity remainedrelatively high throughout stationary phase.Note that the adenylate cyclase of glucose-grown cells was capable of great stimulation bypyruvate, whereas the succinate-grown cells ex-hibited a decreased stimulation. Hence, in thepresence of pyruvate the activity of adenylatecyclase in glucose-grown cells was actuallygreater than succinate-grown cells, which is justopposite of the results obtained by assaying ade-nylate cyclase activity without pyruvate in thereaction mixture. It may be significant that thepeak of adenylate cyclase activity in succinate-grown wild-type cells just preceded, but paral-leled, the second peak of internal cAMP.The pattern of adenylate cyclase activity was

altered in the mutant with respect to the parentstrain. In succinate-based medium the activitypeaked initially at 15 U, but then remained lowat approximately 5 U throughout the exponen-tial phase of growth (Fig. 40). This activity isthree- to fourfold lower than in the wild-typestrain. The decreased activity of adenylate cy-clase in the mutant growing in succinate, withrespect to the parent, correlates well with thelack of elevated internal cAMP in the mutant,and suggests that the failure to produce cAMPin normal amounts is due to an alteration inadenylate cyclase. In glucose-based medium theinitial activity was slightly higher than the par-ent strain (Fig. 4D), but the activity continuedto decrease throughout exponential phase. Theactivity remained constant during stationaryphase at 25.0 U. These data suggest that smallmolecules, e.g., pyruvate, may be involved in theregulation of adenylate cyclase activity, in vivo,in A. crystallopojetes.

DISCUSSIONSaier et al. (17) showed that in Escherichia

coli and Salmonella typhimurium extraction ofcAMP was an important mechanism in the reg-ulation of internal cAMP levels. Since the dryweight of A. crystallopoietes growing in succi-nate, at the point where the external level ofcAMP began to rise (dotted line, Fig. 1B), wasequal to 0.77 mg (dry weight) per ml of culture,and the peak of internal cAMP was equal to 100pmol/mg (dry weight), it can be calculated thatthe intracellular concentration of cAMP wasabout 3 x 10-5 M. To determine if all of theintracellular cAMP was excreted into the culturemedium or if additional cAMP was synthesizedat this time and immediately excreted, the intra-cellular and extracellular levels of cAMP were

compared by expressing both as picomoles ofcAMP per milliliter of culture. When expressedin this way, the intracellular concentration atthe point in growth described above (Fig. 1B,dotted line) was 77 pmol/ml of culture and fellto 5.0 pmol/ml of culture as the cells entered thestationary phase of growth. Note that this intra-cellular change of 72 pmol/ml ofculture is nearlyidentical to the rise in extracellular cAMP of 68pmol/ml of culture. Similar calculations for glu-cose-grown cells give a cellular concentrationchange of 8.7 pmol/ml of culture, whereas theextracellular concentration changes by 17pmol/ml of culture. Thus, it appears that cAMPis excreted into the medium, resulting in a low-ering of intemal cAMP level at the onset of thestationary phase in A. crystallopoietes. WhenMph-3 was grown in succinate-based medium, achange in extemal cAMP level of 65 pmol/ml ofculture was observed at the onset of the station-ary phase. This value was nearly identical fothat found in the parent strain. However, theintracellular level was calculated to change only3 pmol/ml of culture. A possible explanation forthis phenomenon is that, at the onset of thestationary phase, Mph-3 synthesizes nearly anequivalent amount of cAMP to that in the par-ent strain, but that all cAMP made is immedi-ately excreted, which prevents an intracellularaccumulation of cAMP.The profile of phosphodiesterase activity in

the parent strain, growing in succinate-basedmedium, suggests that phosphodiesterase activ-ity is involved in the regulation ofinternal cAMPlevels during exponential growth. At the begin-ning of exponential growth, the phosphodiester-ase activity continues to rise as the intracellularcAMP level decreases from its initial peak value.An opposite pattern is found before the secondpeak of cAMP (Fig. 2B, dotted line), i.e., thephosphodiesterase activity progressively de-creases prior to the second increase in internalcAMP. The fact that little extracellular cAMPcan be detected during the decrease of the initialpeak of intracellular cAMP (cf. Fig. 1B) furthersubstantiates the role of the phosphodiesteraseenzyme in the regulation of internal cAMP con-centration during exponential growth. The pat-tern of phosphodiesterase activity in Mph-3 dur-ing growth in succinate- or glucose-based me-dium does not appear appreciably altered fromthat of the parent strain, which suggests that analtered phosphodiesterase activity is not respon-sible for the lower intracellular cAMP level inMph-3 growing in succinate-based medium.

Opposite to the findings during exponentialgrowth, the excretion of cAMP appears to beinvolved in the regulation of internal cAMPlevels during the stationary phase of growth.

VOL. 134, 1978

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

05

Dec

embe

r 20

21 b

y 17

3.24

7.13

1.13

7.


Only during the stationary phase do both suc-cinate- and glucose-grown cells have equivalentphosphodiesterase activities and nearly equiva-lent levels of intracellular cAMP. However,there is an increase in adenylate cyclase activityduring the stationary phase of growth (Fig. 3).Therefore, it is likely that regulation of internalcAMP concentration during the stationaryphase involves the concerted action of adenylatecyclase, phosphodiesterase, and excretion ofcAMP.The activity of adenylate cyclase in A. crys-

tallopoietes measured in the absence ofpyruvatereflects the activity expected from the intracel-lular cAMP data. There is little adenylate cy-clase activity measurable in glucose-grown cells,but upon inoculation of mid-exponential glu-cose-grown cells into succinate-based medium,there is an immediate 20-fold increase in ade-nylate cyclase specific activity, which correlateswell with the finding of an initial peak of intra-cellular cAMP. Also, as predicted by the externalcAMP and phosphodiesterase activity data, theactivity of adenylate cyclase in Mph-3 is muchlower than in the parent strain, remaining atleast four- to fivefold lower throughout exponen-tial growth.

Similar to B. liquefaciens (10), the activity ofadenylate cyclase in A. crystallopoietes is capa-ble of stimulation by pyruvate. The finding ofgreater adenylate cyclase activity in glucose-grown cells, compared to succinate-grown cells,in the presence of pyruvate is not consistentwith the low internal cAMP level found in cellsgrown in this medium. Perhaps the adenylatecyclase activity in the presence of pyruvate is ameasure of the maximum possible rate ofcAMPsynthesis, and is not a true measure of the invivo activity. Alternatively, the activity of suc-cinate-grown cells may already be stimulated bysmall molecules which were present in vivo,which prevent further stimulation by pyruvatein vitro. The observation of a peak in adenylatecyclase activity in succinate-grown wild-typecells just preceding, but paralleling, the secondpeak of intracellular cAMP may indicate thatsmall molecules are regulatory during this laterstage of growth, but are not effective at earlierstages. This is further substantiated by the ab-sence ofa pyruvate-stimulated peak ofadenylatecyclase activity paralleling the first peak of in-tracellular cAMP. Although the interpretationof these findings is not clear, the observation ofincreased adenylate cyclase activity by pyruvatedoes suggest that metabolic intermediates mayplay a role in vivo in the regulation of adenylatecyclase activity in A. crystallopoietes.The activity ofpyruvate-stimulated adenylate

cyclase in Mph-3 growing in succinate-based

medium is four- to fivefold less than that of theparent strain grown in the same medium. More-over, a similar reduction in activity in the mu-tant compared to the parent strain was observedwhen adenylate cyclase was assayed in the ab-sence of pyruvate. These findings suggest thatan alteration in the adenylate cyclase activity inMph-3 may be a major cause of the reducedintracellular level of cAMP which is involved inthe sphere-to-rod-to-sphere morphogenesis inthe parent strain.The mechanism of regulation of the cAMP

levels in E. coli has recently been examined andinvolves interaction of the adenylate cyclase en-zyme with the phosphoenolpyruvate-dependentphosphotransferase system (13). The model in-cludes the concept that adenylate cyclase is nor-mally complexed with enzyme I of the phospho-transferase system. Adenylate cyclase can ex-press a high level of activity, when enzyme Iexists in the phosphorylated form, but its activ-ity is low when enzyme I is dephosphorylated.The high-activity state is favored by the pres-ence of phosphoenolpyruvate and the absence ofglucose, whereas the opposite conditions favor alow-activity state. Hence, the observation of lowlevels of cAMP during growth of E. coli inglucose-based medium (14) is due to the de-phosphorylation ofenzyme I and the subsequentinactivation of adenylate cyclase.However, there is no evidence that cAMP is

regulated in A. crystallopoietes in the samemanner as in E. coli, or that cAMP is involvedin catabolite repression. A. crystallopoietes is anobligate aerobe with an oxidative physiology, incontrast to the facultative anaerobes (e.g., E.coli), and A. crystallopoietes utilizes organicacids preferentially to glucose and other carbo-hydrates. Krulwich and Ensign (7) studied thediauxic growth of A. crystallopoietes. Unlike E.coli, glucose transport was found to involve ac-tive transport with a glucose-specific permease.This system was inducible by glucose, but in thepresence of succinate the synthesis of the per-mease was repressed, and the activity was in-hibited. This led them to the conclusion thatglucose utilization was directly inhibited by theorganic acid (i.e., succinate). Schechter et al.(19) investigated enzyme induction and repres-sion in A. crystallopoietes. Histidase inductionby histidine was reduced by incubation of cellswith either glucose or succinate. Each requiredan extended period of time for maximum effect(greater than 100 min). Also, succinate or glu-cose inhibited the transport of histidine. Induc-tion of L-serine dehydratase by glycine was se-verely and permanently repressed by glucose,but unaffected by succinate. Isocitrate lyase wasseverely repressed by succinate or fumarate, but

J. BACTrERIOL.

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

05

Dec

embe

r 20

21 b

y 17

3.24

7.13

1.13

7.


glucose had no effect. In each case, exogenouslyadded cAMP had no effect on enzyme produc-tion or repression. Schechter et al. (19) suggestedthat, in A. crystallopoietes, the extent to whicha compound is involved in repression has noth-ing to do with its ready utilization as a growthsubstrate, but may be related to the pathway bywhich it is utilized. Thus, it appears unlikelythat, in A. crystallopoietes, cAMP plays a sig-nificant role in catabolite repression. Regulationof adenylate cyclase in A. crystallopoietes mustalso be different from E. coli, since no phospho-transferase system for glucose or other sugars(T. A. Krulwich, personal communication) hasbeen found to exist in A. crystallopoietes.This conclusion is supported by recent find-

ings in Pseudomonas aeruginosa. P. aeruginosais also an obligate aerobe which has no apparentphosphotransferase system, which utilizes or-ganic acids preferentially to glucose, and inwhich synthesis of inducible enzymes for carbo-hydrate utilization is strongly repressed by suc-cinate (8, 11, 15, 16, 25). Siegel et al. (21) inves-tigated the role ofcAMP in catabolite repressionin P. aeruginosa. The intracellular cAMP re-mained at a constant value regardless of thecarbon source, and exogenous cAMP failed toreverse catabolite repression. They concludedthat cAMP is unlikely to have a role in cataboliterepression in P. aeruginosa. It is possible thatboth P. aeruginosa and A. crystallopoietes,which exhibit oxidative metabolism and prefer-ential growth on organic acids, fail to utilizecAMP as a controlling element in cataboliterepression.

ACKNOWLEDGMENTSThis study was supported by a Public Health Service

Biomedical Supports grant. R. W. Hamilton was supported bya Public Health Service training grant in microbiology,GM00512-15, from the National Institute of General MedicalSciences.

LITERATURE CITED1. Ensign, J. C., and R. S. Wolfe. 1964. Nutritional control

of morphogenesis in Arthrobacter crystallopoietes. J.Bacteriol. 87:924-932.

2. Gilman, A. G. 1970. A protein binding assay for adenosine3',5'-cyclic monophosphate. Proc. Natl. Acad. Sci.U.S.A. 67:305-312.

3. Hamilton, R. W., E. C. Achberger, and P. E. Kolen-brander. 1977. Cyclic AMP and morphogenesis in Ar-throbacter crystallopoietes: effect of cyclic adenosine3',5'-monophosphate. J. Bacteriol. 129:874-879.

4. Hirata, M., and 0. Hayaishi. 1965. Pyruvate dependentadenyl cyclase activity of Brevibacterium liquefaciens.Biochem. Biophys. Res. Commun. 21:361-365.

5. Ide, M. 1969. Adenyl cyclase of E. coli. Biochem. BiophysRes. Commun. 36:42-46.

6. Khandelwal, R. L., and I. R. Hamilton. 1972. Effectorsof purified adenyl cyclase from Streptococcus salivar-ius. J. Biol. Chem. 246:3297-3304.

7. Krulwich, T. A., and J. C. Ensign. 1969. Alteration ofglucose metabolism ofArthrobacter crystallopoietes bycompounds which induce sphere-to-rod morphogenesis.J. Bacteriol. 97:526-534.

8. Lessie, T. G., and F. Neidhardt. 1967. Adenosine tri-phosphate linked control of Pseudomonas aeruginosaglucose-6-phosphate dehydrogenase. J. Bacteriol.93:1337-1345.

9. Lowry, 0. H., N. J. Rosebrough, A. L Farr, and R. J.Randall. 1951. Protein measurements with the Folinphenol reagent. J. Biol. Chem. 193:265-275.

10. Lynch, T. J., E. A. Tallant, and W. Y. Cheung. 1975.Brevibacterium liquefaciens and its in vivo stimulationby pyruvate. J. Bacteriol. 124:1106-1112.

11. Ng, F., and E. Dawes. 1973. Chemostat studies on theregulation of glucose metabolism in Pseudomonasaeruginosa by citrate. Biochem. J. 132:129-140.

12. Peterkofsky, A. 1976. Cyclic nucleotides in bacteria, p.1-48. In P. Greengard and G. A. Robinson (ed.), Ad-vances in cyclic nucleotide research, vol. 7. AcademicPress Inc., New York.

13. Peterkofsky, A. 1977. Regulation of E. coli adenylatecyclase by phosphorylation-dephosphorylation. TrendsBiol. Sci. 2:12-14.

14. Peterkofsky, A., and C. Gazdar. 1974. Glucose inhibi-tion of adenylate cyclase in intact cells of E. coli B.Proc. Natl. Acad. Sci. U.S.A. 71:2324-2328.

15. Phibbs, P. V., Jr., and R. G. Eagon. 1970. Transportand phosphorylation of glucose, fructose, and mannitolby Pseudomonas aeruginosa. Arch. Biochem. Biophys.138:476-482.

16. Romano, A. H., J. J. Eberhard, S. L. Dingly, and T.D. McDowell. 1970. Distribution of the phosphoenol-pyruvate: glucose phosphotransferase system in bacte-ria. J. Bacteriol. 104:808-813.

17. Saier, M. H., Jr., B. U. Feucht, and M. T. McCaman.1975. Regulation of intracellular adenosine cyclic 3',5'-monophosphate levels in Escherichia coli and Salmo-nella typhimurium: evidence for energy-dependent ex-cretion of the cyclic nucleotide. J. Biol. Chem.250:7593-7601.

18. Salomon, Y., C. Londos, and M. Rodbell. 1974. Ahighly sensitive adenylate cyclase assay. Anal. Biochem.58:541-548.

19. Schechter, S. L, Z. Gold, and T. A. Krulwich. 1972.Enzyme induction and repression in A. crystallopoietes.Arch. Microbiol. 85:280-293.

20. Schultz, G. 1974. General principles of assays for adenyl-ate cyclase and guanylate cyclase assays. Methods En-zymol. 38:115-125.

21. Siegel, L. S., P. B. Hylemon, and P. V. Phibbs, Jr.1977. Cyclic adenosine 3',5'-monophosphate levels andactivities of adenylate cyclase and cyclic adenosine 3',5'-monophosphate phosphodiesterase in Pseudomonasand Bacteroides. J. Bacteriol. 129:87-96.

22. Tao, M. 1974. Preparation and properties of adenylatecyclase from Escherichia coli. Methods Enzymol.38:155-160.

23. Tao, M., and F. Lipmann. 1969. Isolation of adenylcyclase from E. coli. Proc. Natl. Acad. Sci. U.S.A.63:86-92.

24. Thompson, W. J., and M. M. Appleman. 1971. Multiplecyclic nucleotide phosphodiesterase activities from ratbrain. Biochemistry 10:311-316.

25. Tiwari, N., and J. Campbell. 1969. Enzymatic controlof the metabolic activity of Pseudomonas aeruginosagrown in glucose or succinate media. Biochim. Biophys.Acta 192:395-401.

26. Umezawa, K., K. Takai, S. Tsuji, Y. Kurashina, and0. Hayaishi. 1974. Adenyl cyclase from Brevibacter-ium liquefaciens. III. In situ stimulation by pyruvate.Proc. Natl. Acad. Sci. U.S.A. 71:4598-4601.

VOL. 134, 1978

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

05

Dec

embe

r 20

21 b

y 17

3.24

7.13

1.13

7.

crystallopoietes and a morphogenetic mutant - journal of bacteriology

Documents