Phenotypic Suppression of a Fructose-1.,6-Diphosphate

JOURNAL OF BACTERIOLOGY, JUly 1973, p. 268-276 Vol. 115, No. 1 Copyright ( 1973 American Society for Microbiology Printed in U.S.A.

Phenotypic Suppression of a Fructose-1.,6-Diphosphate Aldolase Mutation in Escherichia coli RENATE SCHREYER AND AUGUST BOCK Fachbereich fur Biologie der Universitat Regensburg, 84 Regensburg, Germany Received for publication 13 April 1973

Strain NP 315 of Escherichia coli possesses a thermolabile fructose-1, 6-diphosphate (FDP) aldolase; its growth on carbohydrate substrates is inhibited probably as a consequence of the accumulation of high intracellular levels of FDP. Studies of one class of phenotypic revertants of strain NP 315 which have regained their ability to grow on C. substrates at 40 C showed that in these strains the buildup of the inhibitory FDP pool is prevented by additional mutations in enzymes catalyzing the conversion of the substrate offered in the medium to FDP. For example, mutations affecting 6-phosphogluconate dehydrogenase activity (gnd-) may be selected in great number without any mutagenesis and enrichment simply by isolating revertants of strain NP 315 able to grow on gluconate at 40 C. Similarly, an additional mutation in phosphoglucose isomerase (pgi-) restores the ability of these fda- gnd- strains to grow on glucose at 40 C. Glucose metabolism of these fda- gnd- pgi- strains was investigated. The enzymes of the Entner-Doudoroff pathway are induced to an appreciable extent upon growth of these mutants on glucose medium; further evidence for glucose degradation via this route (which normally is induced only in the presence of gluconate) was provided by following the fate of the Cl label of radioactive glucose in L-alanine. Predominant labeling of the carboxyl-carbon of L-alanine was observed, inciating a major contribution of the Entner-Doudoroff path to pyruvate formation from glucose. Chromatographic analysis of the intermediates of glucose metabolism showed further that glucose apparently is at least partly metabolized via a bypass consisting of the accumulation of extracellular gluconic acid which arises by dephosphorylation of 6-phosphogluconolactone and possibly of 6-phosphogluconate. This extracellular gluconate is then taken up and metabolized in the normal manner via the Entner-Doudor- off enzymes.

The inhibitory effects of high sugar phosphate of the medium to FDP, thereby preventing ac- concentrations under conditions of any iinbal- cumulation of FDP. For example, growth inhibi- ance in carbohydrate catabolism is well known tion on gluconate as carbon source is reversed (1) and is usually ascribed to a general toxicity by mutations (gnd-) affecting activity of gluco- of these components. Thus, growth of a mutant nate-6-phosphate dehydrogenase (EC 1.1.1.44); of Escherichia coli with a defective fruc- similarly, glucose-supported growth was some- tose-1, 6-diphosphate (FDP) aldolase (EC what restored by mutations in gnd and phos- 4.1.2.13) is severely inhibited under conditions phoglucose isomerase (EC 5.3.1.9) (pgi-). One where high intracellular concentrations of FDP of these apparent fda- gnd- pgi- mutants was accumulate (1, 2). used in the investigation of the path of glucose In an attempt to more closely characterize degradation because carbon flow through gly- this growth inhibition, we have isolated revert- colysis and the hexosemonophosphate shunt ants of the original temperature-sensitive FDP should be prevented by the mutational blocks aldolase strain. One class of revertants could be and the Entner-Doudoroff-path (5) (Fig. 1) identified as strains possessing mutations in should be repressed upon growth on glucose enzymes leading from the C6 or C5 substrate (10). The results show that in these mutants the 26E VOL. 115, 1973 GLUCOSE METABOLISM IN E. COLI MUTANTS 269 Entner-Doudoroff path contributes to glucose Glucose degradation. I MATERIALS AND METHODS Glucose-6-P b 6-phosphog uconolactone The following strains were used: K-10, a proto- I pgl g6 - Gluconate pgi 6-phosphogluconate = trophic Hfr strain (rel-); strain NP 315 (fda-), a / edd ~~ i4 DPG derivative of strain K-10 with a temperature-sensitive fructose-1,6-diphosphate aldolase (1); strain NP 8 >/ R~~~~~~nd 3151 (fda- gnd-) is a spontaneous revertant of strain eds NP 315 scored for growth at 40 C on gluconate mini- pfkJ mal plates. Strain NP 31515 (fda- gnd- pgi-) is a derivative of strain NP 3151; it was selected for growth Fructose-1,6-diP Tricese-P + Pyruvate at 40 C on glucose minimal medium plates. j l fds The minimal medium used was a modified salt 2 Triose-P solution P of Fraenkel and Neidhardt (9). Only ~Ir one-half of the original phosphate concentration was Glycolysis Pentose-phosphats cycle Entner-Doudoroff- employed, and it was supplemented with 0.2% Pathway (NH4)2SO4 and 0.4% of the indicated carbon source. FIG. 1. Pathways ofglucose and gluconate dissimi- Plates contained 1.5% agar (Serva). Cultures were lation by Escherichia coli (17). Gene designations are grown in gyratory water bath shakers; the bacterial fda, fructose-1,6-diphosphate aldolase; pgi, phospho- growth was measured by following the optical density glucose isomerase; gnd, 6-phosphogluconate dehy- (OD) increase at 420 or 405 nm in a Zei,B PMQ II or an drogenase; pgl, 6-phosphogluconolactonase; edd, 6- Eppendorf photometer. phosphogluconate dehydrase; eda, 2-keto-3-deoxy- Preparation of cell-free extracts and enzyme 6-phosphogluconate aldolase; pfk, phosphofructoki- assays. For preparation of cell-free extracts, the nase; glk, gluconokinase. Dashed arrows indicate cultures were grown to an OD420 of about 1.5 and inducible enzymes. harvested by centrifugation. The sedimented cells were washed once with the respective extraction kinase and gluconate-6-phosphate dehydrogenase by buffer and subsequently broken by sonic treatment using the assay system of Moellering and Bergmeyer with a Branson sonic oscillator for three 1-min periods (19). It was found that high levels of fructose-1,6- at setting no. 3. The extracts were cleared by centrifu- diphosphate greatly interfered with the quantitative gation for 20 min at 20,000 x g, and their protein determination of gluconate and gluconate-6-phos- concentration was determined quantitatively by the phate by inhibiting gluconate-6-phosphate dehy- method of Lowry et al. (18) with bovine serum drogenase activity (3). Prior to these determinations, albumin as standard protein. The activities of phos- fructose-1,6-diphosphate was, therefore, removed by phoglucose isomerase, glucose-6-phosphate dehy- incubation in the presence of FDP-aldolase, triose- drogenase, gluconate-6-phosphate dehydrogenase, phosphate isomerase, a-glycerophosphate dehydro- and gluconokinase were assayed in the reaction mix- genase, and NADH. All the enzymes employed in tures by the method of Fraenkel and Levisohn (10); these determinations were obtained from Boehringer the activities of 6-phosphogluconate dehydrase and and Sons, Mannheim. 2-keto-3-deoxy-6-phosphogluconate aldolase were de- Labeling pattern of L,alanine in cultures grown termined by employing the procedure of Kovachevich on [1-14C]glucose or [1-'4C]gluconate. The contribu- and Wood (16) as modified by Fradkin and Fraenkel tion of the Entner-Doudoroff path to the degradation (6). of glucose or gluconate was assessed by following the Analysis of sugar phosphates in the soluble pool labeling pattern of L-alanine in the protein of cells and in the medium. The amount of sugar phosphates grown on [1-'4C]glucose or [1-'4C]gluconate as sub- accumulated in the cellular pool and excreted into the strate (10). The isolation and degradation of alanine medium was determined as described by Fraenkel (8). was carried out by the method of Fraenkel and The cells from 2 ml of the cultures were separated Levisohn (10) except that alanine was separated from from the medium by filtration through Sartorius the other amino acids of the protein hydrolysate by membrane filters (pore size 0.45 1sm), extracted by three successive descending chromatographies on boiling in 2 ml of water for 4 min, and cleared by Whatman no. 1 paper. The solvent systems used in centrifugation. The sugar phosphate contents of this the given sequence (20 h each, 20 C) were isopropanol- extract and of the medium were measured by the water-glacial acetic acid (75: 15: 10), a-picoline- procedures of Bergmeyer. Thus, fructose-1,6-diphos- water-ammonia (78:20:2), and phenol-water (88: 12). phate was assayed quantitatively by the aldolase-, Chromatography. Separation of sugar phosphates triosephosphate isomerase-, and a-glycerophosphate was achieved by chromatography on Whatman no. dehydrogenase-mediated oxidation of reduced nico- 3MM paper, using a solvent system of 1 M ammo- tinamide adenine dinucleotide (NADH) (4); glucose- nium acetate (pH 5.0), 95% ethanol, and 0.1 M 6-phosphate was measured by glucose-6-phosphate ethylenediaminetetraacetate disodium salt (30:70: 1, dehydrogenase-coupled NADP-reduction (13); simi- vol/vol/vol), descending at 30 C (20, 17). Separation of larly, gluconate-6-phosphate was measured with the gluconolactone from other intermediates was accom- aid of 6-phosphogluconate dehydrogenase (14). Glu- plished by chromatography on Merck cellulose thin- conic acid was determined quantitatively via glucono- layer plates under the conditions given by Grassetti et 270 SCHREYER AND BOCK J. BACTERIOL. al. (12). The location of radioactivity on chromato- lular accumulation of FDP in the pfk- strains grams was detected by means of a Berthold chromato- whereas high levels of this intermediate can be gram scanner, model LB 280, for the paper chromato- demonstrated in strain NP 315, it is likely that grams and with a Berthold scanner II, model LB 2723, FDP accumulation, and not that of any other for thin-layer plates. Standards were run on the same sugar phosphate, is responsible for the growth chromatogram and detected with the periodate benzi- inhibition observed. dine spray (11). Quantitative evaluation of the radio- of revertants. To gain more infor- activity was done by cutting out or scraping off the Isolation respective area from the paper or plates and by mation about the site of growth inhibition by counting in a toluene-0.5% diphenyloxazole scintilla- FDP, revertants of strain NP 315 (fda-) were tion mixture using a BF 5001 liquid scintillation isolated. At first, strains were selected which spectrometer. regained the ability to grow on gluconate minimal medium plates at 40 C. Spontaneous RESULTS AND DISCUSSION revertants to temperature-resistant growth on Growth studies. It has been shown previ- gluconate arose quite frequently. About 50 colo- ously that blocking carbon flow through glycol- nies per 108 cells plated appeared at 40 C. ysis by a mutational alteration of FDP aldolase Fifty-six of these 40 C revertants were analyzed (Fig. 1) results in the complete and immediate for the enzymes of gluconate metabolism and cessation of growth on C6 carbon sources includ- for inhibition of gluconate-6-phosphate dehy- ing gluconate (1, 2). On the other hand, Korn- drogenase by FDP (3). Fifty-four of them turned berg and Smith (15) reported that mutants of E. out to possess a drastically reduced activity of coli lacking phosphofructokinase activity are gluconate-6-phosphate dehydrogenase (Table able to utilize the hexosemonophosphate path- 1). No change of the pattern of inhibition of way as a degradative route during glucose- gluconate-6-phosphate dehydrogenase by FDP supported growth if a sufficient supply of phos- could be found in the two residual strains. phoenolpyruvate (PEP) is provided for the up- These apparent gnd- strains were used to se- take of glucose via the PEP phosphotransferase lect spontaneous revertants able to grow at 40 system. In contrast to these pfk- strains, the C on glucose minimal medium plates. A high FDP aldolase mutant grows neither at the temperature dependency of appearance of restrictive condition with glucose-6-phosphate phenotypic glucose-positive revertants could as substrate nor under conditions where cells be observed. At 35, 38, and 40 C about 8,000, were pregrown on pyruvate which results in the 600, and 340 temperature-resistant colonies induction of PEP synthase (Fig. 2). Cessation of appeared per 3 x 108 cells plated, respectively. growth of strain NP 315 at 40 C, therefore, Most of the low-temperature revertants seems not to be due to shortage of PEP. Since showed improved FDP aldolase activities, the main difference between pfk- and fda- whereas enzymatic analysis of a total of nine strains should reside in the lack of the intracel- 40 C revertants revealed that, in addition to the fda and gnd mutations, seven of them showed a drastically reduced activity of phosphoglucose isomerase (Table 1). Growth on fructose of 0,75-

400C 0,50 C TABLE 1. Specific activities of 6-phosphogluconate 300 C 0- 0 dehydrogenase and phosphoglucose isomerase in cell-free extracts ofstrains NP 315, NP 3151, and NP 31515 grown on gluconate minimal medium at 30 C 6L 0,25- /o O (30) Pyr 6-Phosphogluco- Phosphoglucose L- Gluc (400 C) Strain nate dehydrog- isomerase enase sp acta sp acta

0,1 NP 315 (fda- 116 660 NP 3151 (fda- <4 646 gnd-) -1 -2 0 1 2 3 NP 31515 (fda < 1 < 1 TIME (hrs) gnd- pgi-) FIG. 2. Growth of mutant NP 315 in glucose- 6-phosphate minimal medium at 30 C and after a a Specific activities are expressed in nanomoles of shift to 40 C, 0; the effect ofpregrowth of mutant NP substrate converted per minute per milligram protein 315 in pyruvate minimal medium at 30 C on subse- at 20 to 22 C. All extracts were devoid of fructose-1, 6- quent growth with glucose as carbon source at 40 C, diphosphate aldolase activity and showed wild-type- . like glucose-6-phosphate dehydrogenase activity. VOL. 115, 1973 GLUCOSE METABOLISM IN E. COLI MUTANTS 271 these apparent fda- gnd- pgi- strains was still TABLE 2. Mean doubling times ofstrains NP315, NP strictly temperature-sensitive. Table 2 lists the 3151, and NP 31515 on glucose and gluconate minimal growth rates of some of these revertants with medium glucose and gluconate as substrate. It shows that gluconate-supported growth is near nor- StrnGrowth Doubling times (min) mal; this confirms the result of Fraenkel (7) that temp (C) Glucose Gluconate the gnd function is dispensable during gluconate- supported growth. Doubling times of strain NP NP 315 30 165 135 NG 31515 (fda- gnd- pgi-) on glucose were found to NP 315 40 NGa NP 3151 30 180 150 be variable within the indicated range in differ- NP 3151 40 NG 100 ent experiments. Colony formation on glucose NP 31515 30 210-270 150 minimal medium was first visible within 3 days NP 31515 40 240-330 105 at 40 C. These results show that strain NP 315 fda-, aNo growth. therefore, provides a convenient means for rap- idly selecting mutants in the gnd and pgi loci this purpose at 30 C to preclude any interfer- without any mutagenesis and enrichment proce- ence of aldolase inactivation with any possible dure. This technique should work for selection induction of the enzymes of gluconate metabo- of mutants in any other enzyme of the Cf- or lism. Gluconokinase levels of strains NP 3151 C5-carbohydrate metabolism because release of and NP 31515, upon growth on glucose at 30 C, growth inhibition seems to be a result of the nearly reach those which are observed in crude prevention of FDP accumulation, so any muta- extracts of gluconate cells (Table 4). Strain NP tion blocking carbon flow from the substrate 315 possesses an appreciably induced level, too, offerred in the medium to FDP should release which is in accordance with the evidence for growth inhibition. As expected, the gnd muta- gluconate excretion during glucose-supported tion of strain NP 3151 (fda- gnd-) prevents the growth of this strain (1). 6-Phosphogluconate build-up of a detectable FDP pool on gluconate dehydrase and 2-keto-3-deoxy-6-phosphoglu- medium and the gnd and pgi lesions in NP conate aldolase activities, which were deter- 31515 (fda- gnd- pgi-) the accumulation of mined in a coupled assay, are significantly in- FDP both on glucose and gluconate medium duced in glucose cells of strains NP 3151 and (Table 3). Fructose, which at 40 C cannot be NP 31515, whereas in the parent strain NP 315 metabolized via the hexosemonophosphate their level is almost completely repressed. route nor by glycolysis, does not give rise to high Direct evidence for the use of the Entner- FDP values in strain NP 31515 (fda- gnd- pgi-) Doudoroff pathway for glucose degradation probably because of lack of energy for uptake comes from the labeling pattern of L-alanine and phosphorylation. from cells grown at the expense of [1- Route of glucose degradation in fda- gnd- "4C]glucose or [1- 4C]gluconate, respectively pgi- strains. Under the assumption that the in (10). The wild strain K-10 almost exclusively vitro-determined enzyme activities reflect the degrades glucose via glycolysis and the hexose- in vivo conditions, the lesions in the fda- gnd- monophosphate route (Table 5). The slightly pgi- strains separate the glycolytic and the increased labeling of the carboxyl group of pentose phosphate sequence from a hypotheti- L-alanine derived from protein of glucose-grown cally possible route for glucose degradation via cells of strain NP 315 indicates a minor contri- glucose-6-phosphate 6-phosphogluconolac- bution of the Entner-Doudoroff path to pyru- tone gluconate-6-phosphate and the Entner- vate formation. In contrast, pyruvate is synthe- Doudoroff enzymes. The assumption of such a sized in strains NP 3151 and NP 31515 to a con- pathway, however, does not agree with the siderable extent via the Entner-Doudoroff behavior of gnd- pgi- double mutants isolated pathway. by Kupor and Fraenkel (17; and D. G. Fraenkel, The induced level of gluconokinase observed personal communication) which are completely in strains NP 3151 (fda- gnd-) and NP 31515 unable to grow on glucose. The elucidation of (fda- gnd- pgi-) suggests gluconic acid-the the pathway for glucose dissimilation in our inducer of this enzyme-to be an intermediate strains, therefore, seemed desirable. In a first of glucose degradation by these cells. Such a step the specific activities of gluconokinase role of gluconate in glucose metabolism has (glk) and of 2-keto-3-deoxy-6-phosphogluconate been described by Kupor and Fraenkel (17) for aldolase (eda) were determined in order to phosphogluconolactonase mutants of E. coli assess any contribution of these enzymes to which excrete gluconolactone and rephosphory- glucose degradation. The cells were grown for late its extracellular hydrolysis-product glu- 272 SCHREYER AND BOCK J. BACTERIOL. TABLE 3. Intracellular FDP concentration in strains K-10, NP315, NP3151, and NP31515 at different growth conditionsa nmol of FDP/mg of cell protein Strain Glucose, 30 C Glucose, 40 C Gluconate, 30 C Gluconate, 40 C K-10 8 (<2)" <2 (<2) 5 (5) 2 (3) NP 315 49 (16) 260 (55) 31 (5) 236 (62) NP 3151 39 (14) 244 (43) 4 (<2) 2 (8) NP 31515 7 (4) <2 (8) 14 (<2) <2 (<2) a Cells were grown in minimal medium with 0.4% carbon source; in the case of of the 40 C values the cultures were exposed to the high temperature for 60 min. b Numbers in parentheses represent nanomoles of FDP leaked into the medium from an amount of cells equivalent to 1 mg of cellular protein.

TABLE 4. Specific activities of gluconokinase and of 6-phosphogluconate dehydrase and KDPG aldolase in cell-free extracts of strains K 10, NP 315, NP 3151, and NP 31515 grown at 30 C. Sp act of glu- Sp act of Entner- conokinasea ~~~Doudoroff enzymes" Strain conokinasea (edd and eda) Glucose Gluconate Glucose Gluconate cells cells cells cells K-10 10 72 9 132 NP 315 (fda- 45 118 14 88 NP 3151 (fda- gnd) 68 133 39 123 NP 31515 (fda- gnd- pgi-) 112 118 51 110 a Expressed as nanomoles of gluconate phosphorylated per minute x milligrams of protein. b Expressed as nanomoles pyruvate formed per minute x milligram of protein.

TABLE 5. Distribution of label from [1-_4C]glucose in Again, an accumulation of gluconate and also of Cl and C3 of L-alanine in the protein ofstrains K 10, glucose-6-phosphate and 6-phosphogluconate NP 315, NP 3151, and NP 31515 grown at 30 C can be observed. No gluconolactone could be dpm in the carboxyl detected on the thin-layer plates in this experi- Strain Substrate group of L-alanine ment. 30 C cells of strain NP 31515 (not shown) (%) essentially yield the same result. Figure 4B more directly demonstrates the K-10 Glucose 8.9 intermediate role of gluconic acid in glucose K-10 Gluconate 95.6 metabolism by strain NP 31515. In this experi- NP 315 Glucose 16.8 were to NP 315 Gluconate 96.3 ment, glycerol-grown cells transferred NP 3151 Glucose 35.9 [1- 4C]glucose in the presence of chloram- NP 3151 Gluconate 96.8 phenicol. This treatment prevents the induction NP 31515 Glucose 70.3 of gluconokinase (which is absent in cell-free NP 31515 Gluconate 96.1 extracts of glycerol cells) and also causes gluconate to accumulate to a much higher level than that observed in the absence of chloram- conic acid. The existence of a similar bypass in phenicol. The experiments presented therefore strains NP 3151 (fda- gnd-) and NP 31515 show that a bypass sequence (fda- gnd- pgi-) was tested by growth on Pi [1- '4C]glucose and chromatographic analysis of glucose -. glucose-6-phosphate -. I -; any detectable intermediates (17). Figure 3 shows the results for strain NP 3151. At 30 C adenosine 5'-triphosphate (Fig. 3A) levels of glucose-6-phosphate, 6-phos- gluconate-*u 6-phosphogluconate phogluconate, gluconolactone, and gluconate is followed in cells from strain NP 31515 (fda- are small, but detectable. At 40 C (Fig. 3B) a gnd- pgi-). It is not yet clear, however, from preferential accumulation of gluconate takes these data if the intermediate I which is dephos- place. Figure 4A shows the results of the same phorylated is 6-phosphogluconolactone or 6- experiment done with strain NP 31515 at 40 C. phosphogluconate. VOL. 115, 1973 GLUCOSE METABOLISM IN E. COLI MUTANTS 273

30 40 t TIME (min) FIG. 3. Chromatographic separation of intermediates from glucose metabolism. A culture of strain NP 3151 was grown in glycerol minimal medium to an OD40, of 1.18, washed with basal salts solution, and incubated without carbon source for 30 min at 30 C. A 0.4-ml amount of this culture was incubated with 0.01 ml of [12Cjglucose (2.6 mg/ml) and 0.01 ml [1-14CJglucose (57 ,gCi4smol) at 30(A) or 40 C (B). Samples of 0.02 ml were taken at the indicated times into 0.02 ml of 0 C acetone and chromatographed in the systems described. Values for glucose (0), gluconate (0), glucose-6-phosphate (A), and 6-phosphogluconate (0) are from the paper chromatograms; those for gluconolactone (A) are from the cellulose thin-layerplates. Values in A forgluconate, gluconolactone, 6-phosphogluconate, and glucose-6-phosphate are multiplied with a factor of 2.

Dephosphorylation of 6-phosphogluconolac- phogluconolactone which is indicated by the tone would lead to increased levels of 6- maltose-blue phenotype should also be cor- gluconolactone which has been shown to be related with a decreased concentration of 6- responsible for conferrring to lactonase-deficient phosphogluconate in the cellular pool. Indeed, cells the "maltose-blue" phenotype (17). Strain strain NP 31515 accumulates less 6-phosphoglu- NP 31515 was checked for this property and conate than glucose-6-phosphate in the soluble showed the maltose-blue characteristic on mal- pool, which could (amongst other possible ex- tose-glycerol plates (17) both at 30 and 40 C. planations) be interpreted as a preferential The preferential dephosphorylation of 6-phos- dephosphorylation of 6-phosphogluconolactone 274 SCHREYER AND BOCK J. BACTERIOL.

u ~~0 Q 3.10

2.10'~~~~~~~~~

10'

6.104' El

5.10'-

0.~2.10'2~ ~ -°/

10*' 0 o

10 20 310 L 50 60 70 60 m 10 TIME (min)

FIG. 4. A, Intermediates of glucose metabolism of strain NP 31515 at 40 C. A 0.8-ml amount of a glycerol-grown culture (OD 1.22) of strain NP 31515 (washed and preincubated as described in the legend to Fig. 3) was mixed with 0.02 ml of [1-14Clglucose (57,uCi/,umol) and 0.02 ml ofglucose (2.6 mg/ml) and incubated at 40 C. Samples (0.02 ml) were taken, mixed with 0.02 ml of acetone, and chromatographed (17). No gluconolactone could be detected on thin-layer plates. Symbols: 0, glucose; *, gluconate; A, glucose-6-phosphate; 0, 6-phosphogluconate. B, The experiment was done as described under A except that chloramphenicol was given together with the labeled glucose to cells of strain NP 31515 in a final concentration of 100 ,g/ml. Samples (0.04 ml) were taken, the cells were removed from it by centrifugation, and 0.02 ml of the supernatant fluid were spotted directly onto the chromatograms. Symbols: 0, glucose; 0, gluconate; 0, 6-phosphoglaco- nate. No gluconolactone could be detected on cellulose thin-layer plates.

(Table 6). The high level of 6-phosphogluconate for the observation that the fda- gnd- pgi- in the medium which could also be detected as strain metabolizes glucose via an apparently metabolite in the experiments illustrated by similar route as 6-phosphogluconolactonase- Fig. 3 and 4, however, provides additional deficient strains (17). (i) Either some metabo- evidence for 6-phosphogluconolactone excretion lite which accumulates in these mutants could without any preceding dephosphorylation. inhibit the activity of the lactonase (ii) Strain Several possible explanations could account NP 31515 (fda- gnd- pgi-) also could carry a VOL. 115, 1973 GLUCOSE METABOLISM IN E. COLI MUTANTS 275 TABLE 6. Intracellular pools of glucose-6-phosphate and 6-phosphogluconate in strains NP 3151 (fda- gnd-) and NP 31515 (fda- gnd- pgi-) Glucose-6- 6-phospho- Strain Substrate phosphate Temp (C) (nmol/mg of (nmol/mggluconateof protein) protein) NP 3151 Glucose 30 11 (24)b 37 (31) NP 3151 Glucose 30/40a 140 (36) 100 (36) NP 3151 Gluconate 30 13 ( <2) 26 (15) NP 3151 Gluconate 30/40 < 2 (20) 18 ( <2) NP 31515 Glucose 30 89 (255) 6 (200) NP 31515 Glucose 30/40 100(258) 10 (199) a Pools measured 60 min after a shift from 30 to 40 C. b Numbers in parentheses represent nanomoles of sugar phosphates in the medium from an amount of cells equivalent to 1 mg of cellular protein. lesion in the pgl structural gene; this possibility, 3. Brown, A. T., and Ch. L. Wittenberger. 1971. Mechanism however, seems improbable because the gluco- for regulating the distribution of glucose carbon between the Embden-Meyerhof and hexose-monophos- nate and glucose positive revertants arose spon- phate pathways in Streptococcus faecalis. J. Bacteriol. taneously with a frequency incompatible with 106:456-467. the assumption of a double mutation; or (iii) the 4. Bflcher, Th., and H. J. Hohorst. 1970. D-Fructose-1,6- gnd mutation causes all metabolites prior to the diphosphat, Dihydroxyacetonphosphat und D- lesion Glyzerinaldehyd-3-phosphat, p. 1283-1288. In H. U. to accumulate, and the phosphatase Bergmeyer (ed.), Methoden der enzymatischen Ana- involved in dephosphorylation preferentially lyse. Verlag Chemie, Weinheim. acts on 6-phosphogluconolactone. 5. Entner, N., and M. Doudoroff. 1952. Glucose and glu- It still needs to be discussed why the pgi- conic acid oxidation of Pseudomonas saccharophila. J. Biol. Chem. 196:853-862. gnd- strains of E. coli isolated by Kupor and 6. Fradkin, J. E., and D. G. Fraenkel. 1971. 2-Keto-3-deoxy- Fraenkel (17) by selection for inability to form gluconate 6-phosphate aldolase mutants of Escherichia colonies on glucose plates (17) do not exhibit the coli. J. Bacteriol. 108:1277-1283. same pattern of glucose metabolism described 7. Fraenkel, D. G. 1968. Selection of Escherichia coli mutants lacking glucose-6-phosphate dehydrogenase or for our fda- gnd- pgi- mutants (which were gluconate-6-phosphate dehydrogenase. J. Bacteriol. selected for ability for glucose supported 95:1267-1271. growth). At the moment, no definite answer is 8. Fraenkel, D. G. 1968. The accumulation of glucose- possible. In vivo leakiness of the mutational 6-phosphate from glucose and its effect in an Esche- richia coli mutant lacking phosphoglucose isomerase blocks of our strains which could facilitate and glucose-6-phosphate dehydrogenase. J. Biol. induction of the Entner-Doudoroff enzymes by Chem. 243:6451-6457. the extracellular gluconate might be responsible 9. Fraenkel, D. G., and F. C. Neidhardt. 1961. Use of for the different behaviour. The assumption of chloramphenicol to study control of RNA synthesis in bacteria. Biochim. Biophys. Acta 53:96-110. physiological leakiness is supported by the ex- 10. Fraenkel, D. G., and S. R. Levisohn. 1967. Glucose and periment of Fig. 4B. Conversion of glucose to gluconate metabolism in a mutant of Escherichia coli gluconate in the absence of a functioning Ent- lacking gluconate-6-phosphate dehydrase. J. Bacteriol. ner-Doudoroff path may only be explained if 93:1579-1581. 11. Gordon, H. T., W. Thornburg, and L. N. Werum. 1956. some phosphoenolpyruvate necessary for glu- Rapid paper chromatography of carbohydrates and cose uptake may be formed either via glycolysis related compounds. Anal. Chem. 28:849-855. or the hexosemonophosphate route. 12. Grassetti, D. R., J. F. Murray, and J. L. Wellings. 1965. Thin-layer chromatography of phosphorylated glycol- ACKNOWLEDGMENTS ysis intermediates. J. Chromatogr. 18:612-614. 13. Hohorst, H. J. 1970. D-Glucose-6-phosphat und D-Fruc- We are very much indebted to Dan G. Fraenkel for tose-6-phosphat p. 1200-1204. In H. U. Bergmeyer valuable suggestions and for reading the manuscript. (ed.), Methoden der enzymatischen Analyse. Verlag Chemie, Weinheim. LITERATURE CITED 14. Hohorst, H. J. 1970. D-Gluconat-6-phosphat p. 1210-1212. In H. U. Bergmeyer (ed.), Methoden der 1. Bock, A., and F. C. Neidhardt. 1966. Properties of a enzymatischen Analyse. Verlag Chemie Weinheim. mutant of Escherichia coli with a temperature-sensi- 15. Kornberg, H. L., and J. Smith. 1970. Role of phospho- tive fructose-1,6-diphosphate aldolase. J. Bacteriol. fructokinase in the utilization of glucose by Escherichia 92:470-476. coli. Nature (London) 227:44-46. 2. Bock, A., and F. C. Neidhardt. 1966. Isolation of a 16. Kovachevich, R., and W. A. Wood. 1955. Carbohydrate mutant of Escherichia coli with a temperature-sensi- metabolism by Pseudomonas fluorescens. III. Purifica- tive fructose-1,6-diphosphate aldolase activity. J. Bac- tion and properties of 6-phosphogluconate dehydrase. teriol. 92: 464-469. J. Biol. Chem. 213:745-756. 276 SCHREYER AND BOCK J. BACTERIOL.

17. Kupor, S. R., and D. G. Fraenkel. 1972. Glucose metabo- nates, p. 1205-1209. In H. U. Bergmeyer (ed.), Metho- lism in 6-Phosphogluconolactonase mutants of Esche- den der enzymatischen Analyse. Verlag Chemie, Wein- richia coli. J. Biol. Chem. 247:1904-1910. heim. 18. Lowry, 0. H., N. J. Rosebrough, A. L. Farr, and R. J. 20. Wawskiewicz, E. J. 1961. A two-dimensional system of Randall. 1951. Protein measurement with the Folin paper chromatography of some sugar phosphates. Ana- phenol reagent. J. Biol. Chem. 193:265-275. lyt. Chem. 33:252-259. 19. Moellering, H., und H, U. Bergmeyer. 1970. D-Gluco-