Vol. 141, No. 2. JouRNAL OF BACTERIOLOGY, Feb. 1980, p. 751-757. 0021-9193/80/02-0751/07$02.00/0. Mutation in the crp Gene of Salmonella typhimurium ...
Vol. 141, No. 2
JouRNAL OF BACTERIOLOGY, Feb. 1980, p. 751-757 0021-9193/80/02-0751/07$02.00/0
Mutation in the crp Gene of Salmonella typhimurium Which Interferes with Inducer Exclusion B. J. SCHOLTE AND P. W. POSTMA* Laboratory of Biochemistry, B. C. P. Jansen Institute, University of Amsterdam, 1018 TVAmsterdam, The Netherlands
A mutation in the crp gene of Salmonella typhimurium is described which overcame the defects of ptsHI deletion mutants for growth on a number of nonphosphotransferase system compounds. This mutation abolished inducer exclusion in a leaky ptsI mutant. The possible implications for the mechanism of inducer exclusion are discussed.
The gram-negative bacteria Escherichia coli and Sabnonella typhimurium contain a phosphoenolpyruvate-dependent sugar phosphotransferase system (PTS) which is active in the translocation and concomitant phosphorylation of a number of sugars (15). The PTS consists of a number of protein components which catalyze the transfer of phosphoryl groups from phosphoenolpyruvate to the sugar (Fig. 1). In addition to their role in transport, some of these proteins have also been implicated in the regulation of cell metabolism. In particular, glucose-specific factor m (I1G1) or a closely related regulatory protein is thought to play a central role (15, 17, 20). The regulatory role of the PTS is illustrated by the pleiotropic nature of mutants defective in one or more components of the PTS. E. coli or S. typhimurium mutant strains which lack enzyme I or HPr or both do not grow on a number of non-PTS sugars, including glycerol, melibiose, maltose, and lactose (see Table 2; for a review, see reference 15). In E. coli this phenomenon has been attributed mainly to a low rate of cyclic AMP (cAMP) synthesis. There is evidence that the PTS is involved in regulation of adenylate cyclase activity, as suggested by the inhibition of adenylate cyclase activity by PTS sugars (7). Phosphorylated enzyme I and a phosphorylated regulatory protein have been proposed as activators of adenylate cyclase (13, 17). For S. typhimurium, on the other hand, the inability of pts mutants to grow on many nonPTS carbon sources has been explained mainly on the basis of inducer exclusion (15, 18-20). The non-phosphorylated form of a protein component of the PTS is thought to inhibit the transport systems for glycerol, melibiose, and maltose, preventing the entry ofthese substrates (inducers). It has been proposed that the primary effector of this inhibition is mGlc, a protein component of the PTS involved in transport of glucose via enzyme Hl-BGc (Fig. 1). Support for
this proposal has been provided by the isolation of crr mutants. The crr mutation lowers the activity of IIIGiC in the cell and at the same time restores growth of pts mutants on several nonPTS carbon sources (19; see also Table 2). The concept of this type of inducer exclusion predicts that inhibition of the entry of inducer prevents induction and subsequent growth, irrespective of the presence of cAMP. In other words, if inducer exclusion is involved in the growth inhibition of pts mutants on melibiose, glycerol, or maltose, one would not expect that cAMP would restore growth of these pts mutants. In this paper we report that either cAMP or a mutation in the cAMP binding protein (crp*) is able to overcome the effect of pts mutations in S. typhimurium with respect to growth on some non-PTS compounds. In particular, our results show that inducer exclusion in ptsI mutants is abolished either by the addition of cAMP or by the introduction of a crp* mutation. The consequences for the current concept of inducer exclusion are discussed.
MATERIAlJS AND METHODS Bacterial strains. The strains of S. tyhunurium used in this study are listed in Table 1. The phenotypic characteristics of representative strains are described in Table 2. Media and growth conditions. Cells were grown at 370C on a rotary shaker in liquid medium A [containing, per liter of distilled water: (NH4)2SO4, 1 g; K2HPO4, 10.5 g, KH2PO4, 4.5 g, MgSO4, 0.1 g] supplemented with 20 pg of tryptophan per ml and a carbon source (0.2%). Media were solidified with agar (1.5%; Difco Laboratories, Detroit, Mich.). The effect of cAMP was tested in two ways: (i) by adding 5 mM cAMP to the liquid growth medium and following growth by measuring the increase in optical density at 600 nm; (ii) by placing a sterile filter with 5 pmol of cAMP in the middle of an agar plate and recording the diameter of the growth zone after 24 to 48 h. Transductions involving TnlO were performed on nutrient agar plates (0.8% nutrient broth, 0.5% NaCl, and 751
SCHOLTE AND POSTMA
FIG. 1. The
enzyme I and HPr
of sugar-specific proteins. Here are msconsisting of Il-A /II exclusion is indicated
(-) of non-phosphotransfera-se transport systems (S represents melibiose, maltose-or Activation glycerol) by non-phosphorylated is of adenylate cyclase by phosphorylated
tetracycline per ml. (46 mCi/mmol), [U(184 mCi/mmol), (2,8-3H]-
agar) containing 25jpg
[U_,4C]glUCOSe MCi! (284
mmol) were obtained f3rom The Radiochemical Centre, Amersham, England. cAMP was purchased from Sigma Chemical Co., St. Louis, Mo.
Preparation of cell extracts. Cell were grown nutrient broth. After being washed with overnight 0.9% NaCl, the cells were suspended in a buffer conta.ining 25 mM potassium phosphate, 10 mM MgC12, 0. 1 mM EDTA, and 0.5 mM dithioerythritol (final pH, 7.5) and were broken by passage through an Aminco French pressure cell at 1,200 kg/cm2. The homogenate was centrifuged for 20 mmn at 12,000 x g and 49C to
(dry weight) at 200C. Oxygen consumption was measured with a Clark-type electrode in medium A (final volume, 1.6 ml-). Substrates were added at the concentrations indicated below. The oxidation velocity is expressed as nanoatoms of oxygen consumed per minute per milligram (dry weight) at 25°C. Isolation of mutants and genetic methods. Strains containing only a crr mutation were constructed as follows. A AptsHI41 crr double mutant was transduced with phage P22 grown on strain SB3687 [A(pt8I-crr)166, see reference 4], with selection for pt8+ recombinants (growth on mannitol). Since in the middle of the pt8l gene (5) and Apt8HI41 ends starts at the end of the pt8I gene (4), A(ptsI-crr)166 pts8 recombinants can be obtained which still contain the crr mutation. The properties of these crr strains will be reported elsewhere. Briefly, they do not grow on a number of non-PTS compounds, including xylose, citrate, succinate, and malate (Table 2). Revertants of the crr mutants which regained the Crr+ phenotype could be obtained on minimal plates containing xylose or citrate as the carbon source, either spontaneously or upon addition of diethyl sulfate. One class of revertants contained reversions that mapped near cysA and represented true crr+ reversions. A second class of revertants contained a mutation cotransducible with cysG which was subsequently named crp*. A strain containing only this crp* mutation was constructed by replacing the crr-306 mutation in a crp crr strain by the ptsHI-crr-49 deletion. This deletion extends beyond the crr-306 mutation. PP869 was transduced with phage P22 grown on PP866, with selection for both tetracycline resistance and inability to grow on mannitol. Transduction of the resulting strain with wild-type phage to growth on mannitol in the absence of cysteine yielded the crp strain PP914. In an analogous way, crp ptsIl 7 mutants were isolated. Preparation of P22 transducing lysates and transduction with phage P22 were performed as described previo"isly (2).
RESULTS Effect of cAMP onpts mutants. The model of inducer exclusion as described above predicts that strains carrying a ptsHI deletion mutation intact cells and debris. The resultming cell cannot grow on non-PTS carbon sources such as and centrifged for 120 mmn at 100,000 melibiose, or maltose as long as 1G1c i glycerol, clear supernatant which dialyzed yielded in the non-phosphorylated form. Under present Protein buffer.t three changes of night against these conditions, the transport of these carbon al. (9). determiined by, the method of Lowry Determinati'on of bning activity frcAMP. sources is inhibited. We have found, however, that cAMP, added externally, stimulated the indig oH]cAMP to the cAMP binding protein growth of ptsHI deletion mutants of S. typhiwas determined in the dialyzed supernatant essentially murium on glycerol and melibiose (but not on al. (11), with the following described by Pastan modifications. The incubation mixture contained maltose) (Table 2). For instance, PP642 mM AMP, 10 mM potassium phosphate buffer (AptsHI41) doubled on glycerol every 75 min of [3H]cAMP (specific-activity, 7.5), various 5 mM cAMP was added to the medium. when ml. of protein 2,300 cpm/pmol), and 10 tof20 absence of cAMP, no growth was obIn the for added to [4C]glucose (10 mM) served. Since in these strains GIIPc is supposed bound wH included in the pellet. to be completely in the non-phosphorylated Tranport and oxidation studies. Growth of form due to the absence of enzyme I and HPr, and tranport of labeled compounds were as desfribed one might argue that an alternative route for is expred previously (14). Trasmport miir per minute phosphorylation of IlIGIC, independent of enzyme moles of substrate taken ex-
crp MUTATION IN S. TYPHIMURIUM
VOL. 141, 1980
TABLE 1. Origin and genotype of Salmonella strainsa Strain Stram
Isolation procedure, parental
Suc ~~~~~ ~~~strain ~ ~ Source
Relevant genotype Relevantgenotype
SB3507 trpB223 SB2309 trpB223 A (cysK-ptsHI)41 PP386 crr-306 A (cysK-ptsHI)41 trpB223 PP776 crr-306 trpB223 PP782 crr-306 PP825 crp*-771 crr-306 PP838 cysG1510:.:Tn10 crr-306 PP869 crp*-771 crr-306 PP914 crp*-771 A (cysK-ptsHI)41 PP642 PP866 A(cysK-ptsHI-crr)49 cysA1539.:Tn1O trpB223 PP712 cysG1510.:Tn10 A(cysK-ptsHI)41 PP886 cysG1510 :TnlO A(cysK-ptsHl)41 trpB223 PP889 crp*-771 A(cysK-ptsHI)41 trpB223 SB1476 ptsIl7 crp*-771 ptsIl 7 SB1786 cya-502 PP958 cya-502 crp*-771 cysA20 SB3687 A(ptsI-crr)167 tpB223 TT172 NK186 TA3335
cysG1510:.:TnlO cysA1539-.:TnlO crp
Glp+ SB2309, spontaneous PP386 x P22(SB3687) cysA20 x P22(PP776) Xyl+ PP782, DES PP782 x P22(TT172) PP838 x P22(PP825) cysA20 x P22(SB2309) PP642 x P22(TT172) PP886 x P22(PP825)
E. Balbinder J. C. Cordaro This study This study This study This study This study This study This study This study This study This study This study This study P. E. Hartman This study P. E. Hartman This study P. E. Hartman P. E. Hartman, ref 4 J. Roth J. Roth B. N. Ames, ref. 1
Genetic nomenclature according to Sanderson and Hartman (21). Glp, Glycerol; Xyl, xylose; DES, diethyl sulfate; P22, phage P22. a
TABLE 2. Phenotype of strains used in this studya Genotype
presentb Mtld + -
Glp Mel Malt Xyl
+ + +
crr crp *
AptsHI crp* AptsHI crr crr crr
aMtl, Mannitol; Glp, glycerol; Mel, melibiose; Malt,
maltose; Xyl, xylose. b
If cAMP was present, 5 ,umol was added. 'Growth was monitored on chemically defined media containing 0.2% (wt/vol) of the carbon source, and fermentation was tested on eosin methylene blue plates containing 1% sugar. +, Growth and fermentation after 48 h of incubation at 37°C; -, no growth and no fermentation under these conditions. d Phenotype.
I and HPr, is created when these mutants are grown in the presence of cAMP. This would allow transport of glycerol and melibiose and subsequent growth on these carbon sources. The experiments shown in Fig. 2 eliminated this possibility. Figure 2A shows an example of inducer
exclusion as described earlier by Saier and Roseman (20). The PTS sugar methyl a-glucoside inhibits glycerol oxidation in a leaky ptsI mutant. The rapid inhibition is explained by the
mechanism of inducer exclusion: all of the phos-
pho-LIIGlc present in the cell is dephosphorylated
rapidly by methyl a-glucoside via its interaction with sugar-specific enzyme II-BGlc (see also Fig. 1). Figure 2B shows that glycerol oxidiation by a ptsHI deletion mutant, grown on glycerol in the presence of cAMP, is not sensitive to inhibition by methyl a-glucoside. This indicates that in ptsHI deletion strains, grown under these conditions, inducer exclusion was not effective, since any phospho-IIIGlc formed by the alternative pathway would be dephosphorylated by methyl a-glucoside via enzyme II-BGlc. Mutations which modify the cAMP binding protein. We found other evidence indicating that cAMP is in some way involved in inducer exclusion. As described by Saier and Roseman (19), growth defects of pts mutants on various non-PTS compounds can be suppressed by a crr mutation. In these pts crr double mutants, the crr mutation results in a lowered IIIG1c level. Mutants containing only a crr mutation were constructed as described in Materials and Methods. These crr mutants did not grow on a number of non-PTS compounds, including xylose,
SCHOLTE AND POSTMA ©
FIG. 2. Influence of methyl a-gluco8ide (aMG) on glycerol oxidation. Cells were grown on minimal salts medium and the carbon source indicated below. 02 consumption was measured as described in the text. The reaction wa8started by the addition of 6 mM glycerol (at the first arrow). At the second arrow, 1 mM methyl a-glucoside was added. The dotted line in A represents oxidation in the absence of methyl a-glucoside. (A) SB1476ptsIl7grown on 0.2% glycerol (1.8 mg dry weight per ml). (B) PP642 A(cysK-ptsHI)41 grown on 0.2% glycerol plus 5 mM cAMP (0.5 mg dry weight per ml). (C) ptsIl7 crp*-771 grown on 0.2% glycerol (0.6 mg dry weight per ml).
citrate, succinate, and malate, but grew nornally on all PTS sugars or on the non-PTS sugars glycerol, maltose, and melibiose (Table 2). Lack of growth on xylose, citrate, or succinate could be overcome by the addition of external cAMP (Table 2). For instance, PP782 (crr-306) doubled every 60 min on xylose in the presence of 5 mM cAMP, whereas no growth occurred in the absence of cAMP. Starting with these crr strains, we isolated suppressors of the crr mutation which allowed, in addition, pts mutants to grow on a number of non-NTS carbon sources. Revertants of crr strains were sought which regained the ability to grow on xylose, citrate, and succinate at the same time. Some of these revertants carried a mutation which mapped close to cysA and were shown to have regained iGic activity. These stains presumably acquired reversions in the crr gene, correcting the original crr mutation. A second class of revertants also exhibited the Crr+ phenotype, but the mutation did not map near cysA. Subsequent mapping showed that these mutations were localized close to cysG, being 15 to 30% cotransducible with cysGl5l1Y.:TnlO (a strain containing a TnlO transposon inserted in cysG). For instance, when PP838 (crr-306 cysG1510:.:Tn1O) was transduced with phage P22 grown on PP825 (crp*-771) on minimal plates lacking cysteine, the resulting cysteine+ transductants (43 colonies) fell into two classes: 14% of the transductants grew on citrate, succinate, and xylose, whereas 86% were unable to do so. Both classes tained lowered IfflGc activity, indicating that the original crr mutation was still present. Several lines of evidence suggest that this suppressor mutation was
probably in the crp gene, which codes for the cAMP binding protein. (i) Hong et aL (8) have shown that crp is 16% cotransducible with cysG439. (ii) The suppressor mutation can partially restore growth defects of cya mutants, i.e., mutants that lack adenylate cyclase and consequently are unable to grow on many carbon sources (10, 12). Strains containing both cya and the suppressor mutation (for instance, PP958 cya-502 crp*-771) are able to grow again on, for instance, mannitol or melibiose, carbon sources on which the cya strain does not grow (see also reference 22). (iii) Figure 3 shows that this suppressor mutation caused a decrease in the binding affinity for cAMP as measured in a cell extract. The strain used was constructed by transducing the suppressor mutation into a cysG strain, eliminating the possibility that more than one mutation was involved. From these data we conclude that the suppressor mutation was in c-AMP concentration ( jM)
.8 A u
0 2 1/c-AMP concentration (jM) FIG. 3. Binding affinity for cAMP in crude extracts. The points in the figures represent the mean of three determinations with the same preparation. (A) Binding of cAMP at different concentrations of cAMP; values are calculated as picomoles per milligram of protein. -(B) Double-reciprocal plot of the data of (A). The values for crp * and crp are corrected for aspecific binding by subtracting the values for the crp strain. Symbols: 0, crp' (SB3507); *, crp (PP914); A, crp (TA3335).
VOL. 141, 1980
crp MUTATION IN S. TYPHIMURIUM
the crp gene; it will be designated crp* hereafter. Strains carrying only the crp* mutation, for instance PP914, showed nonnal growth on all carbon sources tested (Table 2), indicating that the cAMP binding protein was still functional although it had altered binding properties. Strains with a defective cAMP binding protein instead do not grow on any of these carbon sources (6, 10). Revertants have been isolated from several crr strains. Until now, in all such cases, the suppressor mutation was cotransducible with either cysG or cysA. Effect of crp* on pts mutants. In view of the observed effects of cAMP on pts mutants, we decided to test whether the crp* mutation also affected the growth of pts mutants. The following results were found. (i) A crp* AptsHI double mutant was able to grow on glycerol or melibiose but not on maltose (Table 2). These results resemble those from experiments with externally added cAMP. Oxidation of glycerol or melibiose was not inhibited by methyl a-glucoside in these strains, results similar to those obtained with these strains grown in the presence of cAMP. (ii) A strain carrying both the leaky ptsI17 mutation and a crp* mutation showed no inducer exclusion as measured by the effect of methyl a-glucoside on both glycerol oxidation (Fig. 2C) and glycerol transport (Fig. 4). Similar results were obtained when thiomethyl galactoside transport via the melibiose transport system was measured (data not shown).
The crp* mutation did not abolish inducer exclusion by interfering with the synthesis of IIIGI,. Transport of methyl a-glucoside, a specific substrate of the IIlGc/II-BGlc system, was normal or even elevated in crp* strains compared with that in the crp+ parent (Fig. 5). Finally, the crp* mutation did not result in constitutive synthesis of the transport system for glycerol or melibiose, as illustrated for glycerol in Fig. 6. Cells grown in minimal medium with galactose as a carbon source did not take up glycerol, whereas growth on glycerol induced normal uptake. This means that there was a normal requirement for the inducer in the crp* strain.
DISCUSSION Regulation of bacterial metabolism by the phosphoenolpyruvate-dependent PTS is a complex phenomenon. It has been proposed that IIIGkc or a closely connected regulatory protein interacts with a number of non-sugar phosphotransferase transport systems and adenylate cyclase (15,17). Non-phosphorylated 111Gkc inhibits these transport systems (inducer exclusion), whereas phosphorylated IIIGk activates adenylate cyclase. Although this scheme can account for a number of observations, two difficulties become clear upon closer inspection of this model. (i) crr mutations abolish inducer exclusion inpts mutants by eliminating non-phosphorylated IIIc, and allow growth of pts crr double mutants on non-PTS compounds such as melibiose, maltose, glyceroL and lactose. However, since growth on these carbon sources requires in L-
2 0 Time (min)
FIG. 4. Inducer exclusion in a crp strain. Cells were grown on a minimal salts medium containing 0.2% glycerol. Transport of 1 mM [U- 14CJglycerol (specific activity, 110 cpm/nmol) was measured as described in the text. (A) SB1476: *, no methyl aglucoside; 0, 1 mM methyl a-glucoside. (B) crp *771 ptsII7: *, no methyl a-glucoside; 0, 1 mM methyl a-
60 30 Time (sec)
FIG. 5. Transport of methyl a-glucoside. Cells were grown on minimal salts medium containing 0.2% glucose and harvested in the mid-exponential phase. Transport of 0.25 mM methyl a-[U- 14C]glucoside (specific activity, 1,400 cpm/nmol) was measured as described in the text. Symbols: 0, crp+ (SB3507); , crp* (PP914).
SCHOLTE AND POSTMA
for activity or none at all. However, in both cases the mechanism by which cAMP and the crp* gene product influenced inducer exclusion is as yet unclear. The prompt inhibition of glycerol oxidation or transport in ptsIl7 strains by a PTS substrate (Fig. 2A and 4) cannot be explained by a mechanism acting only at the level of operon transcription. These results urge us to combine inducer exclusion and regulation by cAMP and crp* in one model. Apparently, the crpj mutation and cAMP influence one of the conditions necesary for the occurrence of inducer exclusion. The most obvious explanation, namely that the crp* mutation interferes with the synthesis of the PTS proteins, in particular IlHc, can be rejected on the basis of the evidence presented in Table 2 and Fig. 5. There are several other possibilities which have to be investigated. One of these is direct involvement of cAMP in inducer exclusion. Another is modulation of the hypothetical
'i- E U
Time (sec) FIG. 6. .Induction of glycerol uptake. Cells were growl; azud the uptake of I mM [U. "4Cjglycerol (spe. cific activiity, 110 cpm/nmol) was measured as described in the text. Symbols: 0, crp+ strain (SB3507) grown on (02% glycerol; 0, crpi strain grown on 0.2% galactose; A, crp* strain (PP914) grown on 0.2% glycerol; A c,crp* strain (PP914) grown on 0.2%galactose.
regulatory activity of IIiGi by cAMP. Possibly, cAMP and the crp* mutation inhibit during cell growth the formation of a protein other than mGic that is essential to inducer exclusion. Fi-
nally, an important parameter which is possibly influenced by the crp* mutation might be the relative amounts of the various non-phosphotransferase transport systems which are affected by inducer exclusion, and IVllc. This might deall cases cAMP (cya and crp mutants exhibit 198 it. terminie the extent of the inhibition. Ri1V,0 1xJp, %n-scarn rg nn thaa Amfiv-+,ua UMzsgVzL, t11 W6llUll UXItSLlVE f*h is surprising that these pts crr mutants grow at ACKNOWLEDGMEDGS all on these carbon sources since they lack phosWe thank J. Roth (University of Utah) and B. N. Ames pho.IIlG and thus are unable to activate ade- (University of California) for the generous gift of strains. This study was supported by a grant from The Netherlands nylate cyclase. (ii) Conversely, the earlier finding for the Advancement of Pure Research (Z.W.O.) that growth of pts mutants of E. coli on these Organization the auspices of The Netherlands Foundation for Chemnon-PTS compounds can sometimes be stimu- under ical Research (S.O.N.). lated by cAMP (for a review, see reference 15) LITERATURE CITED is equally surprising. Although in this case the low intracellular cAMP concentration, due to 1. Alper, M. D., and B. N. Ames. 1978. Transport of antibiotics and metabolite analogs by systems under the absence of phospho-EIGk, is corrected, incyclic AMP control: positive selection of Salmonella ducer exclusion is still existent. typhimurium cya and crp mutants. J. Bacteriol. 133: In this paper we present data that shed some 149-157. light on these difficulties. A mutation, crp*, is 2. Blume, A. J., and E. Balbinder. 1966. The tryptophan described that allowed pts mutants to grow on operon of Salmonella typhimurium. Fine-structure analysis by deletion mapping and abortive transduction. these non-PTS sugars. Surprisingly, the crp * Genetics 53:577-592. mutation was isolated as a suppressor of a crr 3. Botsford, J. L., and M. Drexler. 1978. The cyclic 3',5'mutation which, by itself, is a suppressor of pts adenosine monophosphate receptor protein and regumutations. The results led us to the conclusion lation of cyclic 3',5'-adenosine monophosphate synthesis in Escherichia coli. Mol. Gen. Genet. 165:47456. that the crp* mutation interfered with inducer J. C., T. Melton, J. P. Stratic, M. Atagiin, C. exclusion in some way. The effect of the crp* 4. Cordaro, Gladding, P. E. Hartman, and S. Roseman. 1976. mutation may be explained by assuming that it Fosfomycin resistance: selection method for intemal stimulated the adenylate cyclase activity. The and extended deletions of the phosphoenolpyruvate: sugar phosphotransferase genes of Salnonella typhiinvolvement of the crp gene in the regulation of mwium. J. Bacteriol. 128:785-793. cAMP production and adenylate cyclase syntheJ. C., and S. Roseman. 1972. Deletion mapsis has been reported (3, 16). Alternatively, the 5. Cordaro, ping of the genes coding for HPr and enzyme I of the crp* mutation could alter the cAMP binding phosphoenolpyruvate:sugar phosphotransferase system in Salmonella typhimurium. J. Bacteriol. 112:17-29. protein in such a way that it needed less cAMP it,
VOL. 141, 1980 6. Emmer, M., B. de Crombrugghe, I. Pastan, and R. Perlman. 1970. Cyclic AMP receptor protein of E. coli: its role in the synthesis of inducible enzymes. Proc. Natl. Acad. Sci. U.S.A. 66:480-487. 7. Harwood, J. P., C. Gazdar, C. Prasad, A. Peterkofsky, S. J. Curtiss, and W. Epstein. 1976. Involvement of the glucose Enzymes II in the regulation of adenylate cyclase by glucose in Escherichia coli. J. Biol. Chem. 251:2462-2468. 8. Hong, J. S., G. R. Smith, and B. N. Ames. 1971. Adenosine 3':5'-cyclic monophosphate concentration in the bacterial host regulates the viral decision between lysogeny and lysis. Proc. Natl. Acad. Sci. U.S.A. 61: 2258-2262. 9. Lowry, 0. H., N. J. Rosebrough, A. L. Farr, and R. J. Randall. 1951. Protein determination with the Folin phenol reagent. J. Biol. Chem. 193:265-275. 10. Pastan, I., and S. Adhya. 1976. Cyclic adenosine 5'monophosphate in Escherichia coli. Bacteriol. Rev. 40: 527-551. 11. Pastan, I., M. Gallo, and W. B. Anderson. 1974. The purification and analysis of mechanism of action of a cyclic AMP-acceptor protein from Escherichia coli. Methods Enzymol. 38:367-376. 12. Perlnan, R. L., and L. Pastan. 1969 Pleiotropic deficiency of carbohydrate utilization in an adenyl cyclase deficient mutant of Escherichia coli. Biochem. Biophys. Res. Commun. 37:151-157. 13. Peterkofsky, A., and C. Gazdar. 1975. Interaction of Enzyme I of the phosphoenolpyruvate:sugar phosphotransferase system with adenylate cyclase of Escherichia coli. Proc. Natl. Acad. Sci. U.S.A. 72: 2920-2924. 14. Postma, P. W. 1977. Galactose transport in Salmonella
crp MUTATION IN S. TYPHIMURIUM
typhimurium. J. Bacteriol. 129:630-639. 15. Postma, P. W., and S. Roseman 1976. The bacterial phosphoenolpyruvate:sugar phosphotransferase system. Biochim. Biophys. Acta 457:213-257. 16. Potter, K., G. Chaloner-Larsson, and H. Yamazaki. 1974. Abnormally high rate of cyclic AMP excretion from an Escherichia coli mutant deficient in cyclic AMP receptor protein. Biochem. Biophys. Res. Commun. 57:379-385. 17. Saier, M. H. 1977. Bacterial phosphoenolpyruvate:sugar phosphotransferase systems: structural, functional, and evolutionary interrelationships. Bacteriol. Rev. 41:856871. 18. Sauer, M. H., and S. Roseman. 1972. Inducer exclusion and repression of enzyme synthesis in mutants of Salmonella typhimurium defective in Enzyme I of the phosphoenolpyruvate:sugar phosphotransferase system. J. Biol. Chem. 247:972-975. 19. Saier, M. H., and S. Roseman. 1976. Sugar transport. The crr mutation: its effect on repression of enzyme synthesis. J. Biol. Chem. 251:6598-6605. 20. Saier, M. H., and S. Roseman. 1976. Sugar transport. Inducer exclusion and regulation of the melibiose, maltose, glycerol, and lactose transport systems by the phosphoenolpyruvate:sugar phosphotransferase system. J. Biol. Chem. 251:6606-6615. 21. Sanderson, K. E., and P. E. Hartman. 1978. Linkage map of Salmonella typhimurium, edition V. Microbiol. Rev. 42:471-519. 22. Takebe, Y., M. Shibuya, and Y. Kaziro. 1978. A new extragenic suppressor of cya mutation. Mutant cyclic AMP receptor protein with an increased affinity for cyclic AMP. J. Biochem. 83:1615-1623.