Phosphotransferase SystemSugars by Escherichia coli - Journal of

Phosphotransferase SystemSugars by Escherichia coli - Journal of

JOURNAL OF BACTERIOLOGY, Mar. 1983, p. 1187-1195 Vol. 153, No. 3 0021-9193/83/031187-09$02.00/0 Copyright C 1983, American Society for Microbiology ...

1MB Sizes 0 Downloads 0 Views

Recommend Documents

Escherichia coli
Fakultativ pathogener Keim, Sapprophyt der intestinalen Normalflora des Menschen. □ ca 107 E. coli/g Fäzes ~ 0,01% de

Escherichia coli - The Journal of Immunology
Alain P. Gobert, Marjolaine Vareille, Anne-Lise Glasser, ...... Madrid, L. V., C. Y. Wang, D. C. Guttridge, A. J. Schott

Nitric Oxide Formation by Escherichia coli - The Journal of Biological
Mar 31, 2003 - Hughes, M. N., Scott, C., Thomson, A. J., Green, J., and. Poole, R. K. (2002) EMBO J. 21, 3235–3244). F

Escherichia coli - CDC
tions in Guatemala or remaining at least 14 days in Antigua in Guatemala, or in Cuernavaca, Mexico, were included in the

Transcriptional Response of Escherichia coli to TPEN - Journal of
May 12, 2006 - Carbonic anhydrase. 60. B1053. yceE. 2.5. 1.3E-03. Drug resistance protein. 65. B0313. betI. 2.4. 5.9E-03

Escherichia coli - ernährungs umschau
Ursprünglich trugen die Bakterien den. Namen „Bacterium coli commune“. Der. Gattungsname Bacterium deutet auf die m

Heat stable enterotoxin produced by Escherichia coli - Archives of
Nov 29, 1988 - Hospital 'Di Cristina', Palermo, and §Paediatric Division, Hospital of Mantua, Italy. SUMMARY .... Paler

Suppression of the Escherichia coli rpoH Opal Mutation by Ribosomes
reversions have been classified into several groups according to their map ..... whether suhD affects the level or prope

Allosteric control of Escherichia coli rRNA promoter complexes by DksA
synthesis in Escherichia coli (Paul et al. 2004). DksA concentrations are relatively constant (Rutherford et al. 2007),

Plastic Produced by Escherichia coli - Chinese Academy of Sciences
monomers of polybutylene succinate (PBS) and polylactic acid (PLA) both biodegradable plastics, through fermentation. Be

JOURNAL OF BACTERIOLOGY, Mar. 1983, p. 1187-1195

Vol. 153, No. 3

0021-9193/83/031187-09$02.00/0 Copyright C 1983, American Society for Microbiology

Adenylate Cyclase Is Required for Chemotaxis to Phosphotransferase System Sugars by Escherichia coli ROY A. BLACK,t* ANN C. HOBSON,f AND JULIUS ADLER Department of Biochemistry and Department of Genetics, College of Agricultural and Life University of Wisconsin-Madison, Madison, Wisconsin 53706

Sciences,

Received 17 August 1982/Accepted 27 November 1982

system sugars.

Most chemotactic responses of Escherichia coli are mediated by one of three methyl-accepting chemotaxis proteins (39), but one apparent exception is the response to sugars transported by the phosphotransferase system (PTS sugars) (25). We report here that chemotaxis to these sugars (PTS chemotaxis) appears to be mediated by adenylate cyclase, the nucleotide cyclase linked to the phosphotransferase system. The phosphotransferase system concomitantly transports and phosphorylates a variety of sugars (11, 30). Although the components vary depending on the sugar, most commonly phosphate is transferred from phosphoenolpyruvate to a soluble component designated enzyme I and from enzyme I to another soluble component called HPr. Finally, the phosphate is transferred from HPr to the sugar being transported in a process catalyzed by one of a number of membrane-bound enzymes II, each of which is relatively specific for one sugar. The enzymes II also serve as receptors for chemotaxis to PTS sugars; binding of a given sugar to its enzyme II is required for chemotaxis t Present address: Department of Physiology, University of California School of Medicine, San Francisco, CA 94143. t Present address: Department of Biochemistry, University of Califomia, Berkeley, CA 94720.

toward that sugar via the phosphotransferase system (3, 21). Metabolism of the compound (beyond the phosphorylation step) is not required for chemotaxis toward it (3, 21). In addition to an appropriate enzyme II, enzyme I and HPr are required for all PTS chemotactic responses (3, 21), and PTS chemotaxis appears to be obligatorily coupled to transport (21). Therefore, it has been suggested that an alteration in the level of phosphorylation of some component of the phosphotransferase system during translocation of a sugar triggers the chemotactic signal (21). The adenylate cyclase of E. coli, which converts ATP to cyclic AMP (cAMP) (27), is also coupled to the phosphotransferase system. PTS sugars inhibit the synthesis of cAMP by this enzyme (16, 28, 34) under conditions similar to the conditions required for chemotaxis to these compounds; the appropriate enzyme II is required for a sugar to affect adenylate cyclase activity (15, 35), metabolism of the compound is not required (28), and mutations in enzyme I and HPr lower adenylate cyclase activity or result in abnormal regulation of the activity by PTS sugars (10, 14, 29, 34). Just as for chemotaxis, it has been suggested that the level of phosphorylation of a component of the phosphotransferase sys-

1187

Downloaded from http://jb.asm.org/ on October 27, 2018 by guest

We report that in Escherichia coli, chemotaxis to sugars transported by the phosphotransferase system is mediated by adenylate cyclase, the nucleotide cyclase linked to the phosphotransferase system. We conclude that adenylate cyclase is required in this chemotaxis pathway because mutations in the cyclase gene (cya) eliminate or impair the response to phosphotransferase system sugars, even though other components of the phosphotransferase system known to be required for the detection of these sugars are relatively unaffected by such mutations. Moreover, merely supplying the mutant bacteria with the products of this enzyme, cyclic AMP and cyclic GMP, does not restore the chemotactic response. Because a residual chemotactic response is observed in certain strains with residual cyclic GMP synthesis but no cyclic AMP synthesis, it appears that the guanylate cyclase activity rather than the adenylate cyclase activity of the enzyme may be required for chemotaxis to sugars transported by the phosphotransferase system. Mutations in the cyclic nucleotide phosphodiesterase gene, which increase the level of both cyclic AMP and cyclic GMP, also reduce chemotaxis to these sugars. Therefore, it appears that control of the level of a cyclic nucleotide is critical for the chemotactic response to phosphotransferase

1188

BLACK, HOBSON, AND ADLER

tem is the essential variable in determining ade-

nylate cyclase activity (29, 34). There is evidence that the adenylate cyclase synthesizes cyclic GMP (cGMP) as well as cAMP (36, 40). In view of our previous studies implicating a role for cGMP in chemotaxis (7, 8, 17), we considered the possibility that an alteration in the guanylate cyclase activity of the adenylate cyclase might generate the chemotaxis signal for PTS sugars. Our results suggest that indeed a cyclic nucleotide is involved in the chemotactic response to these attractants.

supplemented with 5 mM cAMP were considered probable Acya derivatives; cAMP is required for the synthesis of flagella (43) and chemotaxis proteins (37, 38). Such transductants were further tested for the ability to grow on minimal agar plates containing maltose as the sole carbon source and the required amino acids; cAMP is required for the utilization of maltose (27). Those transductants which neither swarmed without cAMP nor grew on the maltosecontaining plates were considered to be Acya. The cAMP phosphodiesterase activity in E. coli also hydrolyzes cGMP, and the gene coding for this activity is cotransducible with metC (unpublished data), as it is in Salmonella typhimurium (6). We refer to the gene coding for this activity as cpd, the designation used in S. typhimurium (6). Strains AW723 and AW724 are spontaneous cpd mutants that were selected by the procedure of Alper and Ames (6). Briefly, this procedure is based on the fast growth rate of such mutants on minimal agar plates containing succinate and ammonia; colonies of these mutants appear on such plates before a lawn arises. Cpd- strains grow poorly on minimal agar plates containing glycerol and cAMP (6), and this test was used to screen putative mutants obtained from the selection procedure. Those strains which also appeared to be Cpd- on the basis of this test were assayed for cGMP phosphodiesterase activity (41). For strain AW723, this assay was carried out with bacteria permeabilized by osmotic shock (32) and also with crude extracts prepared by ultrasonic disruption and by use of a French pressure cell; for

TABLE 1. Bacterial strains Strain

Genotype

Source

Reference

Comments

CA-8000

HfrH thi

J. Beckwith

9

Parent of CA-8306, CA-8404, 20-2, AW723, and AW724

CA-8306

HfrH thi Acya

J. Beckwith

9

This strain carries a deletion in cya, the gene for adenylate cyclase

CA-8404

HfrH thi Acya crp* Smr

J. Beckwith

33

crp* designates a cAMP-independent cAMP receptor protein

20-2

HfrH thi cya

J. Beckwith

AW723

HfrH thi cpd

This paper Independent spontaneous cyclic nucleotide phosphodiesterase mutant derived from CA-8000 (see text)

AW724

HfrH thi cpd

This paper Independent spontaneous cyclic nucleotide phosphodiesterase mutant derived from CA-8000 (see text)

AW725

HfrH thi cpd Acya

This paper AW723 made Acya (see text)

RP487

thr his leu met lac xyl ara This laboratory 26 Agal eda rpsL tonB

AW729

RP487 Acya (see RP487)

ZSC103agl glk ptsM

A strain commonly used in chemotaxis studies

This paper See text W. Epstein

Curtisa

S.J. Curtis, Ph. D. thesis, University of Chicago, Chicago, Ill., 1973.

Downloaded from http://jb.asm.org/ on October 27, 2018 by guest

MATERIALS AND METHODS Bacteria. All of the strains used are E. coli K-12 derivatives and are described in Table 1. The deletion mutation in the adenylate cyclase gene (Acya) was transferred from strain CA-8306 to strains CA-8000, AW723, and RP487 as follows: an isoleucine- and valine-requiring derivative of the recipient strain was obtained by ampicillin enrichment (24) after transduction with P1 kc phage (24) propagated on an HfrH llv- strain selected by E. N. Kort (unpublished data); this derivative was then infected with P1 kc phage grown on strain CA-8306 (Acya), and llv+ transductants were screened for the cya marker. (ilv is 67% cotransducible with cya [9, 36].) Transductants which failed to form rings on a standard tryptone swarm plate (1) but did form rings if the plate was

J. BACTERIOL.

VOL. 153, 1983

1189

above, except that in all experiments comparing cya+ and cya strains the tryptone broth was supplemented with the sugar to be used as attractant at a concentration of 10 mM and 5 mM cAMP. Growth with the sugar induced components required for chemotaxis (4, 21, 31). The cAMP was required for transcription of flagellar proteins (43) and chemotaxis-specific proteins (37, 38) in cya strains, and it prevented catabolite repression by the inducing sugar in cya+ strains (12). In most cases preinduction is not actually required for this chemotaxis assay because of either partial constitutivity or the occurrence of induction during the assay (3, 4). However, we chose to preinduce in experiments with cya strains, because induction of at least some phosphotransferase system components requires cAMP (31), which is not normally included in the chemotaxis assay. After growth, bacteria were harvested by centrifugation at 4,000 x g for 2 to 5 min at room temperature and washed two times at room temperature by suspension in 10 mM potassium phosphate (pH 7.0)-0.1 mM EDTA-0.1 mM L-methionine-50 mM D-galactose and centrifugation. The pellet was then resuspended in the same medium to an optical density at 590 nm of 0.005 (3.5 x 106 cells per ml). The number of bacteria accumulated in 1 h at 30°C inside a capillary tube containing the attractant compound in the same medium was determined by a previously described procedure (2). Each point in the figures represents the average of duplicates; the separate values are indicated by error bars. All experiments were performed at least twice, and all results were similar to those reported below. D-Galactose was included in the medium (both in the capillary and in the bacterial suspension) to block the chemotaxis which is known to occur to some PTS sugars via the galactose-binding protein (4). That this procedure successfully blocked the alternative pathway for D-mannose chemotaxis was demonstrated with strain ZSC103agl; in the absence of Dgalactose, this strain was attracted to D-mannose even though it is missing mannose enzyme ll, but inclusion of D-galactose totally prevented accumulation. Sugar transport assay. Bacteria were grown, harvested, and washed as described above for the chemotaxis assay, except that the wash medium contained 10 mM potassium phosphate (pH 7.0), 0.1 mM EDTA, 0.1 mM L-methionine, and 10 mM sodium DL-lactate. After the final wash the bacteria were suspended to a density of about 5 x 109 cells per ml and placed on ice. For each assay, a sample of the suspension was diluted to a density of 3.5 x 108 cells per ml in 1 ml of the wash medium supplemented with 50 mM D-galactose (see below), and the resulting suspension was incubated for 5 min at 30°C. Then the radioactively labeled sugar (in 10 p.1) was added to a final concentration of 18 p.M for D-mannose (2.7 mCi/mmol), 15 p.M for D-glucose (7 mCi/mmol), or 5 p.M for D-mannitol (20 mCi/mmol), and 0.2-mI samples were withdrawn for filtering at 15, 30, 45, and 60 s after the addition. Each sample was pipetted into a filtration apparatus in which a 25-mm membrane filter (pore size, 0.6 p.m; polycarbonate; Nuclepore Corp.) was covered with 2 ml of ice-cold wash medium containing the sugar being tested at a concentration of 0.4 M. Suction was applied as the sample was added, and the filters were washed twice with 2 ml of the same ice-cold solution, dried, and counted by liquid scintillation spectrometry. In all

Downloaded from http://jb.asm.org/ on October 27, 2018 by guest

strain AW724, only a crude extract prepared by ultrasonic disruption was tested. No activity was detected in either strain. The cpd mutation from strain AW723 was transduced into strain CA-8000 as follows: strain CA-8000 was made metC by ampicillin enrichment (24) after transduction with P1 kc phage (24) propagated on an HfrH metC strain selected by E. N. Kort (unpublished data); this derivative was then infected with P1 kc phage grown on strain AW723, and Met' transductants were selected. c pd transductants were identified by fast growth on plates containing succinate and ammonia and slow growth on plates containing glycerol and cAMP (6). and the cGMP phosphodiesterase defect was confirmed by an enzyme assay (41). Growth of bacteria. Except where otherwise indicated, the bacteria that were used for behavioral or biochemical assays were grown in tryptone broth (1% tryptone [Difco Laboratories], 0.5% NaCI) at 35°C with rotary shaking to an optical density at 590 nm of 0.50 to 0.60 (3.5 x 108 to 4.2 x 108 cells per ml). Minimal agar plates for genetic procedures (other than selection and screening for cpd mutants) contained Vogel-Bonner salts (42), the required amino acids at concentrations of 1 mM, a carbon source at a concentration of 0.2%, and 0.001% thiamine (when required). Assays for cAMP and cGMP production. Assays for cAMP and cGMP production were performed essentially as described by Black et al. (7). Briefly, the method used involves suspending freshly washed bacteria in buffer, treating samples with trichloroacetic acid at different times, removing the trichloroacetic acid, by ether extraction, and assaying for the cyclic nucleotide by a radioimmunoassay. Our procedures were different from the previously described procedures in the following ways: (i) in the experiment with strain AW725, 2-ml samples rather than 1-ml samples were withdrawn for each time point; (ii) in all assays for cAMP, 20 j.al of the ether-extracted sample was diluted to 665 ,ul, 100 ,u1 of which was used for the assay; and (iii) in the experiment comparing production by strains CA-8000 and AW723, lactate was omitted from the buffer. Assay for guanylate cyclase. Bacteria were harvested by centrifugation, washed three times at room temperature by suspension in 50 mM Tris-hydrochloride (pH 7.8)-10 mM sodium DL-lactate-0.1 mM L-methionine and centrifugation, and finally resuspended in the same medium to an optical density at 590 nm of 30 (2.1 x 1o"' cells per ml). The resulting suspensions were subjected to two 10-s bursts of ultrasonic disruption with a Branson model W185 Sonifier Cell Disruptor fitted with a microtip. The resulting crude extracts were used immediately for the assay of guanylate cyclase activity. This assay measured the conversion of [a-32P]GTP to [32P]cGMP and was essentially the assay described by Macchia et al. (22), except that the GTP concentration in the reaction mixture was 15 p.M, 1.5 x 106 to 1.8 x 106 cpm of [a-32P]GTP was used, and 1 mM cAMP was added (in addition to the 3 mM cGMP) to reduce hydrolysis of the [32P]cGMP by cyclic nucleotide phosphodiesterase. For each experiment the assay mixture contained 0.27 ml of crude extract in a final volume of 0.30 ml; 0.07-mI samples were removed immediately after the reaction was started (zero time) and at 6-min intervals thereafter. Chemotaxis assay. Bacteria were grown as described

CHEMOTAXIS TO PTS SUGARS

BLACK, HOBSON, AND ADLER

1190 0

8 A

from New England Nuclear Corp. [8-3H]cGMP was obtained from ICN. D-[U-14C]glucose and D-[U14C]mannose were obtained from Amersham/Searle; D-[1-'4C]mannitol was from New England Nuclear Corp. D-Galactose and lactose were from Sigma Chemical Co. L-Arabinose, maltose, and D-mannose were from Calbiochem. D-Glucose was from MCB (associate of E. Merck AG, Darmstadt, Germany). DMannitol was from Eastman Kodak.

a)

0 .0

g

6

cya cpd

0 4

E0l. 1-

_

Acyo c:pd

0 O

3

I

B

Cy+

u

c) 0 (.0

sis and cGMP synthesis. We first confirmed the cAMP deficiency caused by the deletion (Acya) in the adenylate cyclase gene in strain CA-8306 (9). We found that strain CA-8306 is totally nonmotile when it is grown without cAMP and is unable to grow on L-arabinose, lactose, or malt-

1

I

2

4

cya cpd

Q

ose. Furthermore, a strain carrying the Acya mutation, as well as a mutation in the cyclic

0

0.

yo cpd

0)

o

0

|

6

12

a

18

24

30

Time, min FIG. 1. Rates of production of cAMP (A) and cGMP (B) by strains AW723 (cya'+ cpd) (O) and AW725 (Acya cpd) (0). The bacteria excrete both cAMP and cGMP into the medium, a nd the assay (7) measured the combined intra- and ext racellular cAMP or cGMP level. The bacteria were gr.own in tryptone broth supplemented with 0.2% gluco grow poorly on tryptone broth alone were washed and suspended in inc ubation medium (zero time). Samples of the suspen sion were withdrawn at different times and treated w(ith trichloroacetic acid. cAMP and cGMP levels in the acid-soluble portion were determined by radioimmiunoassays. The cAMP values for strain AW725 werre less than 0.1 pmoU/107 cells for all time points. The residual production by strain AW725 shown in (B) wass still observed if only the cGMP fraction from an AG1 -X8 column (19) was assayed.

)ea(ndathemnuthets

nucleotide phosphodiesterase gene (cpd), was totally deficient in cAMP production in vivo (Fig. IA). The cpd strain was used for the production assay to eliminate variations due to differences in the rate of degradation rather than differences in the rate of synthesis of cAMP. In agreement with Shibuya et al. (36), we

found that the Acya mutation also significantly reduced the production of cGMP in vivo, to 31% of the level in the isogenic cya+ strain under our conditions (Fig. 1B); values of 22 and 28% were obtained in two additional experiments. To determine more definitively whether adenylate cyclase is involved in the synthesis of cGMP, we tested crude extracts of strains CA-8000 (wild type) and CA-8404 (Acya crp*) for the ability to convert [o&2P]GTP to [32P]cGMP. (Because of the crp* mutation, strain CA-8404 has a cAMP receptor protein which activates transcription in the absence of cAMP [33]. Therefore, the use of this strain takes account of the possibility that 30

wild

typ

.

o a 20_

cases, linear rates of uptake were obseXrved for at least s. The figures reported below ar,*e based on the average of two time courses. As with ithe chemotaxis assay, D-galactose was included to b

",

systems other than the phosphotran sferase system.

°

60

lock uptake by

The

of this procedure for D-mannose was demonstrated with strain ZSC103agl; i n the absence of D-galactose this strain accumulated ra(dioactivity from labeled Domannose at a significant ratte even though it is missing the mannose enzyme II, bui t inclusion of Dgalactose virtually eliminated accumul lation. Previous work has indicated that this procedture should also block D-glucose uptake by non-pho wsphotransferase systems (3). There are no alternative D -mannitol transport systems (20). Chemicals. Cyclic nucleotide radlioimmunoassay kits and [a-32P]GTP (10 to 50 Ci/mmol ) were obtained success

/

C

E

/ 4

o

Acya crp* o

0

I Q0

i

0 °

6

1

1

12 8 Time, min FIG. 2. Guanylate cyclase activity in crude extracts of strains CA-8000 (wild type) () and CA-8404 (Acya crp*) (0). Essentially the same results were obtained in two additional experiments.

Downloaded from http://jb.asm.org/ on October 27, 2018 by guest

w

RESULTS Role of adenylate cyclase in both cAMP synthe

2

x a) 0

J. BACTERIOL.

VOL. 153, 1983 12,000

CHEMOTAXIS TO PTS SUGARS

1191

vivo cGMP production by the Acya strain described above (and also observed with strain CA-8404) could have been due to an impaired wild cya gene product that was able to function type 8,000 _ partially in vivo but not at all in vitro, or it could have been due to another enzyme which was not detected in the in vitro assay. However, it is clear from our evidence and previous work (36, 4,000 _ 40) that mutations in cya affect cGMP synthesis as well as cAMP synthesis. Requirement of adenylate cyclase for chemotaxis to PTS sugars. To test whether adenylate cyclase might be involved in chemotaxis to PTS sugars, we compared the levels of accumulation .a of strains CA-8000 (wild type) and CA-8306 (Acya) in capillary assays with D-glucose, Dmannose, or D-mannitol as the attractant (Fig. 3). In all three cases, strain CA-8306 showed virtually no accumulation above the background value (no PTS sugar in the capillary tube), whereas with strain CA-8000 there was an accumulation above the background level of several 8000Fc. Mannitol thousand bacteria. This defect was specific for PTS sugars; L-aspartate was used as a positive wild control in each experiment, and strain CA-8306 clearly responded to this amino acid in all cases 4,000[ type (see the legend to Fig. 3). Strains CA-8000 and CA-8306 were also comparable in their responsAcya es to L-serine (112,970 cells for strain CA-8000 LI l-4 1-2 loand 133,320 cells for strain CA-8306 at 10-2 M Lserine in one experiment) and to D-galactose Molority (6,270 cells for strain CA-8000 and 6,025 cells for FIG. 3. Chemotaxis toward PTS sugars by strains strain CA-8306 at 10-4 M D-galactose in one A. Glucose

0

C

0

0

-3

CA-8000 (wild type) (0) and CA-8306 (Acya) (0). (A) D-Glucose chemotaxis. (B) D-Mannose chemotaxis. (C) D-Mannitol chemotaxis. In each experiment, 1 mM L-aspartate was used as a positive control. The levels of accumulation with aspartate were 66,165 cells for strain CA-8000 and 131,670 cells for strain CA-8306 (A), 87,780 cells for strain CA-8000 and 184,140 cells for strain CA-8306 (B), and 90,530 cells for strain CA8000 and 31,130 cells for strain CA-8306 (C). (The data plotted are not normalized to the aspartate response in this figure or Fig. 4 through 7. For all of the chemotaxis assays in this experiment and the experiments shown in Fig. 4 through 7, 50 mM D-galactose was included in both the bacterial suspensions and the capillary tubes to block chemotaxis via systems other than the phosphotransferase system; galactose was also included in both the bacterial suspensions and the capillary tubes for the aspartate controls. The inclusion of galactose slightly increased the levels of accumulation with aspartate, perhaps because galactose serves as an energy source (2).

experiment).

To ensure that the PTS chemotaxis deficiency was a result of the cya mutation, this marker was transduced from strain CA-8306 into strain CA; 8000, and the resulting Acya transductant was tested for chemotaxis to D-mannose. This strain showed no chemotactic response to the sugar. To test the possibility that the chemotaxis deficiency in the Acya strains was due not to the cyclase defect but to a defect in another protein affected by the deletion, we determined the effect of a cya point mutation in strain 20-2 on chemotaxis to D-mannose. Even though this mutation resulted in only a leaky Cya- phenotype (some utilization of L-arabinose, lactose, and maltose and only partial dependence on cAMP for formation of rings on tryptone swarm plates), it too markedly reduced the chemotactic response (Fig. 4). Thus, there is a correlation between a defective adenylate cyclase and imcGMP may be synthesized by a protein which paired chemotaxis to PTS sugars. We then considered whether it was the cyrequires cAMP for transcription and is absent in clase defect per se that was impeding PTS a cya mutant for this reason.) No activity was detected in the strain CA-8404 extract (Fig. 2). It chemotaxis or whether this defect might be causing a deficiency in some other component of was previously reported (40) that a partially purified adenylate cyclase from E. coli convert- the phosphotransferase system that is essential ed GTP to cGMP, but there was no genetic for chemotaxis. To test the latter possibility, we identification of this enzyme. The residual in assayed the sugar uptake abilities of cya+ and

Downloaded from http://jb.asm.org/ on October 27, 2018 by guest

0

J. BACTERIOL.

BLACK, HOBSON, AND ADLER

1192

12,000r o

8,000 -

TABLE 2. Uptake of sugars by cya strains relative to an isogenic cya+ straina Strain

wild type

CA-8000 (wild type)

0

a)

4,000

0 =

~~~~~~~~~J

lo4

-

10-l

100

100

Monnose molarity FIG. 4. Chemotaxis toward D-mannose by strains CA-8000 (wild type) (D) and 20-2 (cya) (0). The levels of accumulation in the aspartate controls were 109,065 cells for strain CA-8000 and 81,345 cells for strain 20-2. D-Galactose at a concentration of 50 mM was included in both the bacterial suspensions and the capillary tubes (see the legend to Fig. 3).

tion as strain CA-8306 allowed us to rule out the possibility that cAMP synthesis is required. These strains were strains CA-8404 (Acya crp*) and AW729 (a Acya transductant of strain RP487, a strain unrelated to strain CA-8000). We found that both of these strains were totally deficient in cAMP production, as expected, but showed substantial chemotaxis to D-mannose cya strains grown under the conditions de- (Fig. 5). These strains did have at least as much scribed above for the chemotaxis assays. We residual cGMP production as shown in Fig. 1B. found no substantial deficiency in the sugar Possible explanations for the difference in PTS transport systems of cya strains (Table 2), indi- chemotaxis between these two strains and strain cating that the required enzymes II and other CA-8306 (Fig. 3) are discussed below. components necessary for transport were present and functional. (A similar result has been 12,000 F A i obtained with cya mutants of S. typhimurium [311.) The extent to which the cya strains are wild 8,000 reduced in transport cannot account for their type defectiveness in chemotaxis, since substantial PTS chemotactic responses are observed even 4,000 in cases where transport is greatly reduced (21; unpublished data). 0 We also considered the possibility that it is not the cyclase itself that is required for PTS chemo0 16,000r B taxis but merely the presence of one of its products, cAMP or cGMP. This possibility was wild 0 12,000_ ruled out by the finding that inclusion of either 5 type mM cAMP or 5 mM cGMP in the chemotaxis a) assay (in both bacterial suspensions and capil0 CID 8,0001lary tubes) did not restore PTS chemotaxis to strain CA-8306 (Acya). The addition of exogenous cyclic nucleotides does raise the intracellu4,000_ lar concentrations of these compounds; exogenous cAMP at a concentration of 5 mM supports a normal level of transcription of cAMP-depen0'-21 IC-I 0 I10-55 10''-4 10'-33 0C? dent operons in cya mutants (27; unpublished Monnose molarity data with strain CA-8306), and 5 mM cGMP similarly stimulates transcription in Acya strains FIG. 5. Chemotaxis toward D-mannose by strains with mutant cAMP receptor proteins which are CA-8000 (wild type) (0) and CA-8404 (Acya crp*) (0) (A) and by strains RP487 (wild type) (0) and AW729 activated by cGMP (5; unpublished data). The data presented above do not indicate (Acya) (0) (B). The levels of accumulation in the controls were 108,570 cells for strain CAwhether it is the cAMP synthetic activity or the aspartate and 8000 79,090 cells for strain CA-8404 (A) and cGMP synthetic activity of the adenylate cy- 155,760 cells for strain RP487 and 149,930 cells for clase that is required for PITS chemotaxis. strain AW729 (B). For all of the experiments, D(There could, of course, be a third synthetic galactose at a concentration of 50 mM was included in activity of the enzyme.) An analysis of two both the bacterial suspensions and the capillary tubes additional strains carrying the same Acya muta- (see the legend to Fig. 3). _

Downloaded from http://jb.asm.org/ on October 27, 2018 by guest

103 1002

_C^5 0

100 77 ± 3

83 ± 16 58 ± 1 ND NDb 73 ± 2 a Assays were carried out with 50 mM D-galactose present to block uptake by systems other than the phosphotransferase system (see text). b ND, Not determined.

CA-8306 (Acya) 20-2 (cya)

u

% Uptake of: D-Glucose D-Mannose D-Mannitol

CHEMOTAXIS TO PTS SUGARS

VOL. 153, 1983 4,000

a 2

C

12,000r

B.

U 0

c3

8,000 _

4,000

F

cpd t *flt o~If10.' I,10 10C, 10 I10| 16

Molarity FIG. 6. (A) Chemotaxis toward D-mannose by strains CA-8000 (wild type) (0), AW723 (cpd) (A), and AW724 (cpd) (O). (B) Chemotaxis toward D-glucose by strains CA-8000 (wild type) (0) and AW723 (cpd) (A). The levels of accumulation in the aspartate controls were 152,900 cells for strain CA-8000, 102,630 cells for strain AW723, and 92,180 cells for strain AW724 (A) and 217,910 cells for strain CA-8000 and 117,590 cells for strain AW723 (B). For all of the experiments, D-galactose at a concentration of 50 mM was included in both the bacterial suspensions and the capillary tubes (see the legend to Fig. 3). The response of strain CA-8000 to D-mannose (A) was lower than the response in Fig. 3B because for the experiment shown in this figure, the bacteria were not grown with D-mannose.

DISCUSSION The data reported here indicate that PTS chemotaxis requires adenylate cyclase and cyclic nucleotide phosphodiesterase in addition to the phosphotransferase system. Our principal findings are as follows: (i) that either a deletion or a point mutation in the gene coding for adenylate cyclase (cya) impairs PTS chemotaxis, whereas transport of PTS sugars is relatively unaffected by such a mutation; (ii) that merely

3

A

0

.0

D..

cpd

2

3

0 x

a1)

'C

wild

E

type

07

Effect of cyclic nucleotide phosphodiesterase mutations (cpd) on PTS chemotaxis. cpd mutants of E. coli are defective in the hydrolysis of both cAMP and cGMP (unpublished data). If the function of adenylate cyclase during the chemotactic response is to alter the level of cGMP or a related compound, a cpd mutation might be expected to inhibit PTS chemotaxis by impairing the regulation of cyclic nucleotide levels. Therefore, we tested two independently isolated cpd mutants, strains AW723 and AW724, for chemotaxis toward D-mannose; both of these mutants were found to be defective (Fig. 6A). Strain AW723 was also tested for its response to Dglucose, and a similar result was obtained (Fig. 6B). A strain constructed by transducing the cpd mutation from strain AW723 into strain CA-8000 (wild type) (see above) was found to be as defective in chemotaxis to D-mannose as the original strain; the transductant showed an accumulation of 665 cells compared with 2,330 cells

0: 0 u

I

I.

1.0i B

0 .

0o

cpd

9

CA/t

_

x

n a?

0.5 _ *

0

6

12

wild type

18

24

30

Time, min FIG. 7. Rates of production of cAMP (A) and cGMP (B) by strains CA-8000 (wild type) (0) and AW723 (cpd) (0). The points represent the averages of two experiments, and the error bars indicate the ranges.

Downloaded from http://jb.asm.org/ on October 27, 2018 by guest

for the wild type at 10-3 M D-mannose. This result indicates that the PTS chemotaxis defect can.indeed be attributed to the cpd mutation. Chemotaxis to L-aspartate was relatively unaffected by the cpd mutations (see the legend to Fig. 6). To determine whether the PTS chemotaxis defect of cpd mutants could in fact be due to the elevated level of a cyclic nucleotide, we measured the rates of production of both cAMP and cGMP by strain AW723. We found that the rate of cAMP production was 345% of the rate in the parent, strain CA-8000, and that the rate of cGMP production was 205% of the rate in the parent (Fig. 7).

rA

2,000 -

0

1193

1194

BLACK, HOBSON, AND ADLER

chemotaxis? It appears more likely that cGMP or a related compound, rather than cAMP. is involved since certain strains with residual cGMP production but no cAMP production have a residual P1TS chemotactic response. Our previous studies on amino acid chemotaxis also indicated no role for cAMP but a central role for cGMP or a related compound (7, 8, 17). Thus far we have been unable to demonstrate an increase in cGMP level during stimulation of bacteria by a PTS sugar. This result may mean that cGMP is not the critical compound, or it may simply reflect the brevity and minor extent of the change expected for the PTS chemotactic signal. (PTS sugars elicit a very weak response compared with amino acid attractants [4, 23].) Indeed, even after stimulation of the bacteria with amino acid attractants, we no longer consistently observed the increase in intracellular cGMP level reported previously (7). To measure the intracellular concentration of cGMP, it is necessary to remove extracellular cGMP by filtering the bacteria and then washing them on the filter, and much of the chemotactic signal may be lost during this procedure. Furthermore, the experiment is complicated by irreproducible filtering of the dense bacterial suspensions which are required to collect enough cGMP to measure. Clearly an improved technique for measuring intracellular cyclic nucleotides will be required to detect any alteration in the levels of such compounds during P1'S chemotaxis. What are the implications of our findings concerning P1S chemotaxis for other chemotaxis pathways? cya mutants retain responses to sugars detected by periplasmic binding proteins and to amino acids, so it is clear that adenylate cyclase is not involved in producing the intracellular signal in these cases. If the non-PTS chemotaxis pathways also use cGMP or a related compound as an intracellular regulator, they must be linked to their own cyclases. In conclusion, our model for chemotaxis to P1S sugars is that concomitant with transport and phosphorylation of a sugar, there is an increase in the guanylate cyclase activity (or in a third, as-yet-unidentified activity) of the adenylate cyclase, leading to a change in swimming behavior. The control of this activity could be mediated by the level of phosphorylation of a component of the phosphotransferase system, as has been suggested for the modulation of adenylate cyclase activity (29, 34). ACKNOWLEDGMENTS We thank Terilee Norene for excellent technical assistance in the performance of the chemotaxis assays. This work was supported by a Public Health Service grant from the National Institute of Allergy and Infectious Diseases, by a grant from the National Science Foundation, and by a grant from the Muscular Dystrophy Association. R.A.B.

Downloaded from http://jb.asm.org/ on October 27, 2018 by guest

supplying the bacteria with the products of this enzyme, cAMP and cGMP, does not restore the chemotactic response; (iii) that certain Acya strains with residual cGMP synthesis, but no cAMP synthesis, have a residual chemotactic response to PITS sugars; and (iv) that mutations in the cyclic nucleotide phosphodiesterase gene (cpd) severely reduce PT'S chemotaxis and increase the rate of production of both cAMP and cGMP. Previous work (36, 40) indicated that adenylate cyclase synthesizes cGMP as well as cAMP, and our findings confirm this conclusion. We extended the previous studies by using a cpd background (to eliminate apparent differences in synthesis due to differences in cyclic nucleotide phosphodiesterase activity) and by measuring the in vitro conversion of GTP to cGMP by extracts from cya+ and cya strains. We do not know whether the in vivo cGMP production which remains in Acya strains is due to residual activity of the cya gene product or to another cyclase. Although the mutation was determined to be a deletion on the basis of non-revertibility and an inability to give recombinants with other mutations (9), the extent of the deletion is not known. We are not certain why the Acya mutation eliminates PTS chemotaxis in strain CA-8306 whereas a residual response remains in strains CA-8404 (Acya crp*) and AW729 (a Acya transductant of strain RP487). If the deletion mutation does not abolish all guanylate cyclase activity of the cya gene product, one possibility is that there is more residual guanylate cyclase activity in strains CA-8404 and AW729. In the case of strain CA-8404, another possibility is that the altered crp gene product interacts with the defective adenylate cyclase and partially restores its function in chemotaxis; this suggestion arises from a report that the wild-type crp gene product does interact with adenylate cyclase (18). Another possible explanation for the presence of a residual response in strains CA-8404 and AW729 but not in strain CA-8306 is that there is an alternative guanylate cyclase activity or an alternative chemotaxis pathway which is expressed to a greater extent under the influence of crp* and in the strain RP487 background than it is in strain CA-8306. The observation that not only cya mutations but also mutations in the gene for cyclic nucleotide phosphodiesterase (cpd) impair chemotaxis to PTS sugars suggests that modulation of the level of a cyclic nucleotide is critical in PTS chemotaxis. In fact, both cya and cpd mutants have normal motility and respond to non-PTS chemotaxis attractants, indicating that the mutations do not significantly affect the basal level of any central mediator in chemotaxis. Which cyclic nucleotide plays a role in PTS

J. BACTERIOL.

VOL. 153, 1983

received support from a Public Health Service predoctoral training grant from the National Institutes of Health.

124:26-38.

1195

21. Lengeler, J., A.-M. Auburger, R. Mayer, and A. Pecher. 1981. The phosphoenolpyruvate-dependent carbohydrate: phosphotransferase system enzymes II as chemoreceptors in chemotaxis of Escherichia coli K-12. Mol. Gen. Genet. 183:163-170. 22. Macchia, V., S. Varrone, H. Weissbach, D. L. Miller, and I. Pastan. 1975. Guanylate cyclase in Escherichia coli. J. Biol. Chem. 250:6214-6217. 23. Mesibov, R., and J. Adler. 1972. Chemotaxis toward amino acids in Escherichia coli. J. Bacteriol. 112:315-326. 24. Miller, J. H. 1972. Experiments in molecular genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. 25. Niwano, M., and B. L. Taylor. 1982. Novel sensory adaptation mechanism in bacterial chemotaxis to oxygen and phosphotransferase substrates. Proc. NatI. Acad. Sci. U.S.A. 79:11-15. 26. Parkinson, J. S., and P. T. Revello. 1978. Sensory adaptation mutants of E. coli. Cell 15:1221-1230. 27. Perlman, R. L., and I. Pastan. 1969. Pleiotropic deficiency of carbohydrate utilization in an adenyl cyclase deficient mutant of Escherichia coli. Biochem. Biophys. Res. Commun. 37:151-157. 28. Peterkofsky, A., and C. Gazdar. 1974. Glucose inhibition of adenylate cyclase in intact cells of Escherichia coli B. Proc. Natl. Acad. Sci. U.S.A. 71:2324-2328. 29. Peterkofsky, A., and C. Gazdar. 1975. Interaction of enzyme I of the phosphoenolpyruvate:sugar phosphotransferase system with adenylate cyclase of Escherichia coli. Proc. NatI. Acad. Sci. U.S.A. 72:2920-2924. 30. Postma, P. W., and S. Roseman. 1976. The bacterial

phosphoenolpyruvate:sugar phosphotransferase system. Biochim. Biophys. Acta 457:213-257. 31. Rephaeli, A. W., and M. H. Saier, Jr. 1980. Regulation of genes coding for enzyme constituents of the bacterial phosphotransferase system. J. Bacteriol. 141:658-663. 32. RoUlins, C. M., and F. W. Dahiquist. 1980. Methylation of chemotaxis-specific proteins in Escherichia coli cells permeable to S-adenosylmethionine. Biochemistry 19:4627-

4632. 33. Sabourin, D., and J. Beckwith. 1975. Deletion of the Escherichia coli crp gene. J. Bacteriol. 122:338-340. 34. Saier, M. H., Jr., and B. U. Feucht. 1975. Coordinate regulation of adenylate cyclase and carbohydrate permeases by the phosphoenolpyruvate:sugar phosphotransferase system in Salmonella typhimurium. J. Biol. Chem. 250:7078-7080. 35. Saier, M. H., Jr., B. U. Feucht, and L. J. Hofstadter. 1976. Regulation of carbohydrate uptake and adenylate cyclase activity mediated by the enzymes II of the phosphoenolpyruvate:sugar phosphotransferase system in Escherichia coli. J. Biol. Chem. 251:883-892. 36. Shibuya, M., Y. Takebe, and Y. Kaziro. 1977. A possible involvement of cya gene in the synthesis of cyclic guanosine 3':5'-monophosphate in E. coli. Cell 12:521-528. 37. Silverman, M., and M. Simon. 1974. Characterization of Escherichia coli flagellar mutants that are insensitive to catabolite repression. J. Bacteriol. 120:1196-1203. 38. Silverman, M., and M. Simon. 1977. Identification of polypeptides necessary for chemotaxis in Escherichia coli. J. Bacteriol. 130:1317-1325. 39. Springer, M. S., M. F. Goy, and J. Adler. 1979. Protein methylation in behavioural control mechanisms and in signal transduction. Nature (London) 280:279-284. 40. Tao, M., and A. Huberman. 1970. Some properties of Escherichia coli adenyl cyclase. Arch. Biochem. Biophys. 141:236-240. 41. Thompson, W. J., G. Brooker, and M. M. Appleman. 1974. Assay of cyclic nucleotide phosphodiesterases with radioactive substrates. Methods Enzymol. 38:205-212. 42. Vogel, H. J., and D. M. Bonner. 1956. Acetylornithinase of Escherichia coli: partial purification and some properties. J. Biol. Chem. 218:97-106. 43. Yokota, T., and J. S. Gots. 1970. Requirement of adenosine 3',5'-cyclic phosphate for flagella formation in Escherichia coli and Salmonella typhimurium. J. Bacteriol. 103:513-516.

Downloaded from http://jb.asm.org/ on October 27, 2018 by guest

LITERATURE CITED 1. Adler, J. 1966. Chemotaxis in bacteria. Science 153:708716. 2. Adler, J. 1973. A method for measuring chemotaxis and use of the method to determine optimum conditions for chemotaxis by Escherichia coli. J. Gen. Microbiol. 74:7791. 3. Adler, J., and W. Epstein. 1974. Phosphotransferasesystem enzymes as chemoreceptors for certain sugars in Escherichia coli chemotaxis. Proc. Nat]. Acad. Sci. U.S.A. 71:2895-2899. 4. Adler, J., G. L. Hazelbauer, and M. M. Dahl. 1973. Chemotaxis toward sugars in Escherichia coli. J. Bacteriol. 115:824-847. 5. Alexander, J. K. 1980. Suppression of defects in cyclic adenosine 3',5'-monophosphate metabolism in Escherichia coli. J. Bacteriol. 144:205-209. 6. Alper, M. D., and B. N. Ames. 1975. Cyclic 3',5'-adenosine monophosphate phosphodiesterase mutants of Salmonella typhimurium. J. Bacteriol. 122:1081-1090. 7. Black, R. A., A. C. Hobson, and J. Adler. 1980. Involvement of cyclic GMP in intracellular signaling in the chemotactic response of Escherichia coli. Proc. NatI. Acad. Sci. U.S.A. 77:3879-3883. 8. Black, R. A., A. C. Hobson, and J. Adler. 1982. Control of the methylation level of a protein involved in bacterial chemotaxis, p. 91-98. In E. Usdin, R. T. Borchardt, and C. R. Creveling (ed.), Biochemistry of S-adenosylmethionine and related compounds. Macmillan Press, Basingstoke, England. 9. Brickman, E., L. Soll, and J. Beckwith. 1973. Genetic characterization of mutations which affect catabolite-sensitive operons in Escherichia coli, including deletions of the gene for adenyl cyclase. J. Bacteriol. 116:582-587. 10. Castro, L., B. U. Feucht, M. L. Morse, and M. H. Saier, Jr. 1976. Regulation of carbohydrate permeases and adenylate cyclase in Escherichia coli. J. Biol. Chem. 251:55225527. 11. Dills, S. S., A. Apperson, M. R. Schmidt, and M. H. Saier, Jr. 1980. Carbohydrate transport in bacteria. Microbiol. Rev. 44:385-418. 12. Dobrogosz, W. J., and P. B. Hamilton. 1971. The role of cyclic AMP in chemotaxis in Escherichia coli. Biochem. Biophys. Res. Commun. 42:202-207. 13. Epstein, W., L. B. Rothman-Denes, and J. Hesse. 1975. Adenosine 3':5'-cyclic monophosphate as mediator of catabolite repression in Escherichia coli. Proc. NatI. Acad. Sci. U.S.A. 72:2300-2304. 14. Feucht, B. U., and M. H. Saier, Jr. 1980. Fine control of adenylate cyclase by the phosphoenolpyruvate:sugar phosphotransferase systems in Escherichia coli and Salmonella typhimurium. J. Bacteriol. 141:603-610. 15. Harwood, J. P., C. Gazdar, C. Prasad, A. Peterkofsky, S. J. Curtis, and W. Epstein. 1976. Involvement of the glucose enzymes II of the sugar phosphotransferase system in the regulation of adenylate cyclase by glucose in Escherichia coli. J. Biol. Chem. 251:2462-2468. 16. Harwood, J. P., and A. Peterkofsky. 1975. Glucose-sensitive adenylate cyclase in toluene-treated cells of Escherichia coli B. J. Biol. Chem. 250:4656-4662. 17. Hobson, A. C., R. A. Black, and J. Adler. 1982. Control of bacterial motility in chemotaxis. Symp. Soc. Exp. Biol. 35:105-121. 18. Joseph, E., C. Bernsley, N. Guiso, and A. Ullmann. 1982. Multiple regulation of the activity of adenylate cyclase in Escherichia coli. Mol. Gen. Genet. 185:262-268. 19. Krlshnan, N., and G. Krishna. 1976. A simple and sensitive assay for guanylate cyclase. Anal. Biochem. 70:1831. 20. Lengeler, J. 1975. Mutations affecting transport of the hexitols D-mannitol, D-glucitol, and galactitol in Escherichia coli K-12; isolation and mapping. J. Bacteriol.

CHEMOTAXIS TO PTS SUGARS