Isolation and Properties of a Bacillus subtilis Mutant Unable to

Vol. 145, No. 2

JouRNAL OF BACTERIOLOGY, Feb. 1981, p. 760-767 0021-9193/81/020760-08$02.00/0

Isolation and Properties of a Bacillus subtilis Mutant Unable to Produce Fructose-Bisphosphatase YASUTARO FUJITA' AND ERNST FREESE`* Department of Biochemistry, Hamamatsu University, School ofMedicine, 36(0 Handacho, Hamamatsu, Japan 431-31,1 and Laboratory ofMolecular Biology, National Institute of Neurological and Communicative Disorders and Stroke, Bethesda, Maryland 202052

A Bacillus subtilis mutation (gene symbol fdpAl), producing a deficiency of D-fructose-1,6-bisphosphate 1-phosphohydrolase (EC 3.1.3.11, fructose-bisphosphatase), was isolated and genetically purified. An fdpAl-containing mutant did not produce cross-reacting material. It grew on any carbon source that allowed growth of the standard strain except myo-inositol and D-gluconate. Because the mutant could grow on D-fructose, glycerol, or L-malate as the sole carbon source, B. subtilis can produce fructose-6-phosphate and the derived cell wall precursors from these carbon sources in the absence of fructose-bisphosphatase. In other words, during gluconeogenesis B. subtilis must be able to bypass this reaction. Fructose-bisphosphatase is also not needed for the sporulation of B. subtilis. The fdpAl mutation has the pleiotropic consequence that mutants carrying it cannot produce inositol dehydrogenase (EC 1.1.1.18) and gluconate kinase (EC 2.7.1.12) under conditions that normally induce these enzymes.

Fructose-1,6-bisphosphate 1-phosphohydrolase (EC 3.1.3.11, fructose-bisphosphatase) catalyzes the dephosphorylation of fructose-1,6-bisphosphate (FbP) to yield fructose-6-phosphate (fructose-6-P) and inorganic phosphate. The fructose-bisphosphatases of Bacillus licheniformis and B. subtilis are similar in their molecular weight, manganese requirement for function and stability, activation by phosphoenolpyruvate (Penolpyruvate), and inhibition by AMP (7, 13). Inhibition of the B. subtilis enzyme by AMP is overcome by P-enolpyruvate, whose concentration increases during gluconeogenic growth, and the enzyme activity is inhibited by highly phosphorylated nucleotides such as guanosine 5'-diphosphate 3'-diphosphate, etc. The enzyme of mammals, plants, and Escherichia coli differs from the Bacillus enzyme in almost all of these respects (10, 14). In enterobacteria such as E. coli or Salmonella typhimurium, fructose-bisphosphatase activity is necessary for gluconeogenic growth on carbon sources such as L-malate, succinate, or glycerol; mutants lacking this enzyme activity (gene symbol fdp) cannot grow on these carbon sources because they cannot produce fructose-6P, which is needed at least for glucosamine and thus mucopeptide synthesis (2, 3, 18). Fructosebisphosphatase has therefore been assumed to be the only enzyme by which fructose-6-P can be produced rapidly enough to allow the normal rate of gluconeogenesis in most microorganisms and cells of higher organisms (phosphofructoki-

nase functions only slowly in the reverse direction because the free-energy change of the reaction, -3.4 kcal/mol (14.2 J/mol), greatly favors the forward direction). Only a few microorganisms contain a pyrophosphate-dependent phosphofructokinase which can efficiently convert FbP to fructose-6-P if the cells contain pyrophosphate (17). We were therefore surprised to find that a fructose-bisphosphatasedeficient mutant of B. subtilis could grow at a high rate and sporulate well in media containing only gluconeogenic carbon sources. The enzymatic and physiological properties of the mutant are here investigated in detail. MATERIALS AND METHODS Bacterial strains. Strain 60015 (metC7 t?pC2), a derivative of B. subtilis strain 168, is our standard strain. Strain 61656 (fdpl hisAl leu-8 metB5 trpC2) is an ade+ transductant of strain 61509 (adeA16 hisAl leu-8 metB5 trpC2) by PBS1 phage grown in strain 60866 (fdpAl trpC2 metC7 Not; "Not" stands for one or more additional mutations). Strain YF001 (hisAl leu-8 trpC2 metB5) is a sister transductant of 61656. Media and growth conditions. Nutrient sporulation medium (NSMP) was prepared by autoclaving 8 g of nutrient broth (Difco Laboratories) in 950 ml of distilled water. When this had cooled, the following sterile solutions were added: 5 ml of 2 M potassium phosphate buffer (pH 6.5), 5 ml of a metal mixture (0.14 M CaCl2, 0.01 M MnCl2, and 0.2 M MgCl2), and 0.5 ml of FeC13 (2 mM in 0.01 M HCI). The synthetic salt mixture (S6) used for growth in liquid minial medium contained: 100 mM potassium morpholinopropanesulfonate (pH 7.0), 10 mM

760

VOL. 145, 1981

B. SUBTILIS FRUCTOSE-BISPHOSPHATASE ENZYME

(NH4)2SO4, 5 mM K2HPO4-KH2PO4 (pH 7.0), 1 mM MgCl2, 0.7 mM CaCl2, 50 ,LM MnCl2, 1 MuM ZnCl2, and 5 MuM FeCl3. As the carbon source, 25 mM hexose, 30 mM pentose, 12.5 mM disaccharide, 50 mM glycerol, or 40 mM potassium L-malate (all final concentrations) was added as indicated in the text. Amino acids were added to the above media to give L-tryptophan (25 Lig/ml) and L-methionine (10 ,ug/ml) for strains 60015 and 60866 and, in addition, L-histidine and L-leucine (50 ,ug/nml each) for strains YFOO1 and 61656. Cells were grown for 15 to 20 h on plates containing 33 g of tryptose blood agar base (TBAB; Difco) and inoculated at an initial optical density at 600 am (OD6w) of 0.05 into the growth medium. The Erlenmeyer flasks containing the medium (not more than 20% of the flask volume) were shaken at 37°C to give maximal aeration in a reciprocating water bath shaker, growth was followed by measuring the OD600. Preparation of cell extracts. Ten to twenty OD6w units of cells were harvested (chloramphenicol was added to give 100 Lg/ml), centrifuged (3,000 x g, 10 min), and washed in 10 ml of 100 mM potassium phosphate buffer (pH 7) and 1 mM MgCl2 (PM) at room temperature. The cells were suspended in 1 ml of 0.1 M Tris-chloride (pH 8) at 4°C and lysed with lysozyme (100 yg/ml) for 30 min at 4°C and then for 10 min at 37°C. After the cell debris was removed by centrifugation (27,000 x g, 20 min), the supernatant solution was used to assay all enzymes except fructosebisphosphatase. To assay the latter, cells were suspended in 10 mM of 4(2-hydroxyethyl)-1-piperazine ethanesulfonate (adjusted to pH 8 by KOH), 5 mM MnCi2, 100 mM NaCl, and 5 mM P-enolpyruvate (HEPES-Mn-Na-PEP) and lysed with lysozyme as above. Enzyme assays. Fructose-bisphosphatase was assayed by the method of Fujita and Freese (7). The reaction mixture (1 ml) contained 100 ,umol of Trischloride (pH 8), 0.1 pnol of MnCi2, 1 Mmol of FbP, 0.2 Mmol of P-enolpyruvate, 2 Mumol of NADP+, 2 U of Dglucose-6-phosphate (glucose-6-P) dehydrogenase (Oriental Co., Tokyo), 2 U of phosphoglucose isomerase (Sigma Chemical Co.), and 100 pl (approximately 100 Mg of protein) of cell extract. Initial rates (nanomoles per minute per milliliter at 25°C) of NADPH formation were determined from the increase in absorbance at 340 nm (A340), using the molar extinction coefficient of 6,200 (NADPH). Glucose and glucose-6-P dehydrogenases (EC 1.1.1.47 and EC 1.1.1.49) were assayed according to Fujita et al. (9) with slight modifications. For glucose dehydrogenase, the reaction mixture (1 ml) contained 200 ,mol of Tris-chloride (pH 8), 800 pmol of KCI, 100 Mmol of D-glucose, 2 Mmol of NAD', and 100 pl of the extract. For glucose-6-P dehydrogenase, the reaction mixture contained 25 Mumol of D-glucose-6-P, 100 Mumol of Tris-chloride (pH 8), 5 Mumol of MgCl2, 0.5 Mumol of NADP+, and 100 id of extract. Initial rates (nanomoles per minute per milliliter at 25°C) were measured from the change of Au0 (molar extinction coefficient of NADH = 6,220 and of NADPH = 6,200). Inositol dehydrogenase was assayed by the method of Ramaley et al. (15). The reaction mixture (1 ml)

761

contained 100 Mumol of Tris-chloride (pH 9), 50 Mmol of myo-inositol, 0.5 umol of NAD+, and 100 MlI of cell extract. Initial rates (nanomoles per minute per mililiter at 25°C) of NADH formation were measured for the increase of Am0. Gluconate kinase was assayed according to Bergmeyer et al. (1). The reaction mixture (1 ml) contained 100 Mmol of Tris-chloride (pH 8), 3.3 umol of MgCl2, 3.2 Mmol of ATP, 0.3 Mmol of sodium gluconate, 0.4 ,umol of NADP+, 0.5 U of gluconate-6-phosphate dehydrogenase (Sigma), and 200 Al of cell extract. The rate (nanomoles per minute per milliliter) of NADPH formation was measured when the rate of increase in Au0 was constant. Sucrase (EC 3.2.1.26) was assayed by the method of Fujita et al. (8) with some modification. The reaction mixture (100 pl) contained 5 pmol of sodium phosphate buffer (pH 6.5), 3 Mumol of sucrose, 5 ,umol of NaCl, and 10 Ml of cell extract. After incubation at 37°C for 15 min, the reaction tubes were placed in boiling water for 3 min. Glucose was determined by glucose oxidase (Boehringer Mannheim Corp.), peroxidase (Boehringer) and o-dianisidine. Two milliliters of reagent (containing 120 mM sodium phosphate [pH 7.4], 1.5 U of peroxidase per ml, 9 U of glucose oxidase per ml, and 50 Mg of o-dianisidine-HCl per ml) was added to the reaction mixture. After incubation (30 min, 25°C), the amount of oxidized o-dianisidine was determined by measuring the A470 and reading the nanomolar amount from a calibration curve prepared with glucose standards. Specific activities (nanomoles per minute per milligram of protein) were calculated by protein measurements according to Lowry et al. (11). Preparation of antibody against fructose-bisphosphatase. A New Zealand white rabbit (about 4 kg) received a subcutaneous injection (on the ventral side) of 1.5 ml of an emulsion containing 1 ml of complete Freund adjuvant (Difco) and 0.5 ml (7 U) of partially purified fructose-bisphosphatase solution in HEPES-Mn-Na-PEP containing 5% glycerol (obtained by AH-Sepharose 4B column chromatography [7]). After 4 weeks, 2 ml of the same emulsion was injected again subcutaneously. The animal was bled 1 week later, and the serum obtained was stored at -20°C. The antibody activity remained essentially constant for at least 16 months. Test for the presence of cross-reacting material in cell extracts. Cells of strains 60015 and 61656 were grown in 400 ml of NSMP supplemented with the required amino acids. At OD600 = 1.5, the cells were harvested, washed, and disrupted in a French pressure cell (at 1,400 kg/cm2 [20,000 lb/in2]). The lysates were centrifuged at 40,000 x g for 20 min, and the superaatant solution was used for the cross-reaction test (18 mg of protein/ml for strain 60015, 17 mg of protein/ml for strain 61656). To tubes containing different amounts (5, 10, 20, and 40 l) of serum, 500-fold diluted in HEPES-MnNa-PEP, a fixed amount of cell extract (10 Ml) was added. The mixtures were incubated at 37°C for 1 h. The resulting precipitates were removed by centrifugation (27,000 x g, 20 min), and the supernatant solutions were assayed for fructose-biphosphatase ac-

762

J. BACTERIOL.

FUJITA AND FREESE

tivity; this showed that the antibody contained in 1 ml of serum precipitated 20 mg of the enzyme, corresponding to an activity of 2.6 mmol of fructose-6-P produced per min at 25°C. To determine the amount of antibody remaining, 10 pl of fructose-bisphosphatase (Sepharose 6B fraction of [7]) was added to 100 1p of each supernatant solution. The mixtures were incubated at 4°C for 24 h, and the resulting precipitates were removed by centrifugation (27,000 x g, 20 min). The supernatant solutions were assayed for enzyme activity. Ouchterlony test. For the preparation of extracts, cells were grown in 2 liters of NSMP containing 10 mM D-glucose and the amino acids required by the particular mutant. They were harvested at ODooo = 1.7, washed in PM, suspended in 30 ml of HEPES-MnNa-PEP, and lysed with 110 jig of lysozyme per ml at 37°C for 10 min. The lysate was centrifuged at 120,000 x g for 4 h. To the supernatant solution, (NH4)2SO4 was added to give 80% saturation. The resulting precipitate (after 1 h at 4°C) was collected by centrifugation (27,000 x g, 20 min) and dissolved in 4 ml of HEPES-Mn-Na-PEP; the undissolved material was removed by centrifugation at 27,000 x g for 20 min. The extract was fivefold concentrated by vacuum dialysis in a collodion bag (Schleicher & Schueli Co.) while dialyzing against HEPES-Mn-Na-PEP. The final protein concentrations were 330 mg/ml (strain 60015) and 290 mg/ml (strain 61656). For partial purification of the antibody, 0.025 ml of the serum was mixed with 0.2 ml of the concentrated extract of the fdp mutant (61656), and the mixture was incubated for 24 h at 4°C. The precipitate was removed by centrifugation (4,500 x g, 10 min). The supernatant solution was used for the Ouchterlony test.

For the Ouchterlony test, 20 pi of the concentrated extracts and the partially purified serum was applied to separate wells (0.4-cm diameter, 0.15-cm depth, and 0.7-cm distance between wells) in 1% agar (Nakarai Co., Kyoto) containing sodium barbiturate buffer (pH 8.6). The applied agar plates were incubated at 20°C for 40 h in a moist box. Pictures of patterns of precipitin lines were taken with a Nikon F2 camera.

RESULTS Isolation of a mutant lacking fructosebisphosphatase activity. Many sporulation mutants of B. subtilis isolated earlier (5) were examined for their growth on different carbon sources. One mutant (60866) was unable to grow on D-gluconate or myo-inositol as the sole carbon source. When this strain was grown in NSMP, the OD600 increase of the culture ceased earlier than for the standard strain (60015); filaments developed and the cells eventually lysed (Fig. 1). The OD increased much higher when the medium was supplemented with 5 mM D-glucose or D-mannose, whereas it remained even lower with D-fructose. Addition of 10 mM glycerol to NSMP did not affect the maximal OD but prevented lysis. Eventually, the cells adapted to grow on

I

I

*

I

I

4

5

ose

*Mnr ose

NSMPoniy ceroI

I

V

I

I

ET1 2 3

INCUBATION TIME(hr) FIG. 1. Growth of the pleiotropic fdp strain 60866 in NSMP plus different carbon sources. Cells grown on TBAB plates were inoculated at OD600 = 0.05 into 30 ml of NSMP plus 5 mM D-glucose, D-mannose, glycerol, D-fructose, or no addition. Cultures of the standard strain reached ODes values larger than 2 with all of these carbon sources.

carbon source. This adaptation occurred slowly in the presence of fructose, faster in glycerol, and fastest in L-malate (especially when its concentration was higher than 10 mM). The observation that the cells became filamentous when their growth stopped in NSMP suggested that the mutant was (transiently) unable to synthesize cell wall precursors (glucosamine-6-phosphate, etc.). Because only glucose and mannosb initially allowed growth to a higher OD600, it appeared likely that the strain lacked fructosebisphosphatase activity which should supply fructose-6-P during gluconeogenesis. Actually, strain 60866 lacked this enzyme activity during growth on glucose (glycolysis), malate (gluconeogenesis), or other carbon sources (Table 1). In our standard strain, the fructose-bisphosphatase seemed to be produced constitutively (7), but careful determinations revealed that the specific activity of crude extracts of cells grown in synthetic medium with D-glucose as the sole carbon source was 40% lower than that of cells grown with L-malate as the sole carbon source any

B. SUBTILIS FRUCTOSE-BISPHOSPHATASE ENZYME

VOL. 145, 1981

TABLE 1. Specific activity of fructosebisphosphatase in extracts of various strainsa Fructosebisphosphatase

Strain

Medium (carbon source)

(nmol/

min per mg of protein) 25

S6 + glucose S6 + malate 39 NSMP + 5 mM glucose 42 60866 S6 + glucose